maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/lib/AsmParser/LLParser.cpp

4044 lines
138 KiB
C++
Raw Normal View History

Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
//===-- LLParser.cpp - Parser Class ---------------------------------------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file defines the parser class for .ll files.
//
//===----------------------------------------------------------------------===//
#include "LLParser.h"
#include "llvm/AutoUpgrade.h"
#include "llvm/CallingConv.h"
#include "llvm/Constants.h"
#include "llvm/DerivedTypes.h"
#include "llvm/InlineAsm.h"
#include "llvm/Instructions.h"
#include "llvm/Module.h"
Assembly and Bitcode support for unsigned/signed overflow flags and exact sdiv flags. llvm-svn: 76475
2009-07-21 05:19:07 +08:00
#include "llvm/Operator.h"
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
#include "llvm/ValueSymbolTable.h"
#include "llvm/ADT/SmallPtrSet.h"
assert(0) -> LLVM_UNREACHABLE. Make llvm_unreachable take an optional string, thus moving the cerr<< out of line. LLVM_UNREACHABLE is now a simple wrapper that makes the message go away for NDEBUG builds. llvm-svn: 75379
2009-07-12 04:10:48 +08:00
#include "llvm/Support/ErrorHandling.h"
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
#include "llvm/Support/raw_ostream.h"
using namespace llvm;
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
static std::string getTypeString(Type *T) {
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
std::string Result;
raw_string_ostream Tmp(Result);
Tmp << *T;
return Tmp.str();
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
/// Run: module ::= toplevelentity*
Refactor some parser interfaces to fix PR3278 and a FIXME: ParseAssemblyString with a specified module would not parse into the module, it would create and return a new one. llvm-svn: 61635
2009-01-05 04:44:11 +08:00
bool LLParser::Run() {
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
// Prime the lexer.
Lex.Lex();
Refactor some parser interfaces to fix PR3278 and a FIXME: ParseAssemblyString with a specified module would not parse into the module, it would create and return a new one. llvm-svn: 61635
2009-01-05 04:44:11 +08:00
return ParseTopLevelEntities() ||
ValidateEndOfModule();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ValidateEndOfModule - Do final validity and sanity checks at the end of the
/// module.
bool LLParser::ValidateEndOfModule() {
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
// Handle any instruction metadata forward references.
if (!ForwardRefInstMetadata.empty()) {
for (DenseMap<Instruction*, std::vector<MDRef> >::iterator
I = ForwardRefInstMetadata.begin(), E = ForwardRefInstMetadata.end();
I != E; ++I) {
Instruction *Inst = I->first;
const std::vector<MDRef> &MDList = I->second;
for (unsigned i = 0, e = MDList.size(); i != e; ++i) {
unsigned SlotNo = MDList[i].MDSlot;
if (SlotNo >= NumberedMetadata.size() || NumberedMetadata[SlotNo] == 0)
return Error(MDList[i].Loc, "use of undefined metadata '!" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(SlotNo) + "'");
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
Inst->setMetadata(MDList[i].MDKind, NumberedMetadata[SlotNo]);
}
}
ForwardRefInstMetadata.clear();
}
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
// If there are entries in ForwardRefBlockAddresses at this point, they are
// references after the function was defined. Resolve those now.
while (!ForwardRefBlockAddresses.empty()) {
// Okay, we are referencing an already-parsed function, resolve them now.
Function *TheFn = 0;
const ValID &Fn = ForwardRefBlockAddresses.begin()->first;
if (Fn.Kind == ValID::t_GlobalName)
TheFn = M->getFunction(Fn.StrVal);
else if (Fn.UIntVal < NumberedVals.size())
TheFn = dyn_cast<Function>(NumberedVals[Fn.UIntVal]);
if (TheFn == 0)
return Error(Fn.Loc, "unknown function referenced by blockaddress");
// Resolve all these references.
if (ResolveForwardRefBlockAddresses(TheFn,
ForwardRefBlockAddresses.begin()->second,
0))
return true;
ForwardRefBlockAddresses.erase(ForwardRefBlockAddresses.begin());
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
for (unsigned i = 0, e = NumberedTypes.size(); i != e; ++i)
if (NumberedTypes[i].second.isValid())
return Error(NumberedTypes[i].second,
"use of undefined type '%" + Twine(i) + "'");
for (StringMap<std::pair<Type*, LocTy> >::iterator I =
NamedTypes.begin(), E = NamedTypes.end(); I != E; ++I)
if (I->second.second.isValid())
return Error(I->second.second,
"use of undefined type named '" + I->getKey() + "'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!ForwardRefVals.empty())
return Error(ForwardRefVals.begin()->second.second,
"use of undefined value '@" + ForwardRefVals.begin()->first +
"'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!ForwardRefValIDs.empty())
return Error(ForwardRefValIDs.begin()->second.second,
"use of undefined value '@" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(ForwardRefValIDs.begin()->first) + "'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Support MDNode forward reference. llvm-svn: 75031
2009-07-09 03:23:54 +08:00
if (!ForwardRefMDNodes.empty())
return Error(ForwardRefMDNodes.begin()->second.second,
"use of undefined metadata '!" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(ForwardRefMDNodes.begin()->first) + "'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Support MDNode forward reference. llvm-svn: 75031
2009-07-09 03:23:54 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Look for intrinsic functions and CallInst that need to be upgraded
for (Module::iterator FI = M->begin(), FE = M->end(); FI != FE; )
UpgradeCallsToIntrinsic(FI++); // must be post-increment, as we remove
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
bool LLParser::ResolveForwardRefBlockAddresses(Function *TheFn,
std::vector<std::pair<ValID, GlobalValue*> > &Refs,
PerFunctionState *PFS) {
// Loop over all the references, resolving them.
for (unsigned i = 0, e = Refs.size(); i != e; ++i) {
BasicBlock *Res;
Revert 85678/85680. The decision is to stay with the current form of indirectbr, thus we don't need "blockaddr(@func, null)". Eliminate it for simplicity. llvm-svn: 85699
2009-11-01 09:27:45 +08:00
if (PFS) {
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
if (Refs[i].first.Kind == ValID::t_LocalName)
Res = PFS->GetBB(Refs[i].first.StrVal, Refs[i].first.Loc);
Revert 85678/85680. The decision is to stay with the current form of indirectbr, thus we don't need "blockaddr(@func, null)". Eliminate it for simplicity. llvm-svn: 85699
2009-11-01 09:27:45 +08:00
else
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
Res = PFS->GetBB(Refs[i].first.UIntVal, Refs[i].first.Loc);
} else if (Refs[i].first.Kind == ValID::t_LocalID) {
return Error(Refs[i].first.Loc,
typo llvm-svn: 85812
2009-11-03 02:28:45 +08:00
"cannot take address of numeric label after the function is defined");
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
} else {
Res = dyn_cast_or_null<BasicBlock>(
TheFn->getValueSymbolTable().lookup(Refs[i].first.StrVal));
}
Revert 85678/85680. The decision is to stay with the current form of indirectbr, thus we don't need "blockaddr(@func, null)". Eliminate it for simplicity. llvm-svn: 85699
2009-11-01 09:27:45 +08:00
if (Res == 0)
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
return Error(Refs[i].first.Loc,
"referenced value is not a basic block");
// Get the BlockAddress for this and update references to use it.
BlockAddress *BA = BlockAddress::get(TheFn, Res);
Refs[i].second->replaceAllUsesWith(BA);
Refs[i].second->eraseFromParent();
}
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
//===----------------------------------------------------------------------===//
// Top-Level Entities
//===----------------------------------------------------------------------===//
bool LLParser::ParseTopLevelEntities() {
while (1) {
switch (Lex.getKind()) {
default: return TokError("expected top-level entity");
case lltok::Eof: return false;
case lltok::kw_declare: if (ParseDeclare()) return true; break;
case lltok::kw_define: if (ParseDefine()) return true; break;
case lltok::kw_module: if (ParseModuleAsm()) return true; break;
case lltok::kw_target: if (ParseTargetDefinition()) return true; break;
case lltok::kw_deplibs: if (ParseDepLibs()) return true; break;
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
case lltok::LocalVarID: if (ParseUnnamedType()) return true; break;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::LocalVar: if (ParseNamedType()) return true; break;
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
case lltok::GlobalID: if (ParseUnnamedGlobal()) return true; break;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::GlobalVar: if (ParseNamedGlobal()) return true; break;
rename lltok::Metadata -> lltok::exclaim. We name tokens after their syntactic form, not their semantic form. llvm-svn: 92294
2009-12-30 12:56:59 +08:00
case lltok::exclaim: if (ParseStandaloneMetadata()) return true; break;
rename NamedOrCustomMD -> MetadataVar to follow conventions of all the rest of the code. llvm-svn: 92295
2009-12-30 13:02:06 +08:00
case lltok::MetadataVar: if (ParseNamedMetadata()) return true; break;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// The Global variable production with no name can have many different
// optional leading prefixes, the production is:
// GlobalVar ::= OptionalLinkage OptionalVisibility OptionalThreadLocal
First step in fixing PR8927: Add a unnamed_addr bit to global variables and functions. This will be used to indicate that the address is not significant and therefore the constant or function can be merged with others. If an optimization pass can show that an address is not used, it can set this. Examples of things that can have this set by the FE are globals created to hold string literals and C++ constructors. Adding unnamed_addr to a non-const global should have no effect unless an optimization can transform that global into a constant. Aliases are not allowed to have unnamed_addr since I couldn't figure out any use for it. llvm-svn: 123063
2011-01-09 00:42:36 +08:00
// OptionalAddrSpace OptionalUnNammedAddr
// ('constant'|'global') ...
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
case lltok::kw_private: // OptionalLinkage
case lltok::kw_linker_private: // OptionalLinkage
case lltok::kw_linker_private_weak: // OptionalLinkage
Change the `linker_private_weak_def_auto' linkage to `linkonce_odr_auto_hide' to make it more consistent with its intended semantics. The `linker_private_weak_def_auto' linkage type was meant to automatically hide globals which never had their addresses taken. It has nothing to do with the `linker_private' linkage type, which outputs the symbols with a `l' (ell) prefix among other things. The intended semantic is more like the `linkonce_odr' linkage type. Change the name of the linkage type to `linkonce_odr_auto_hide'. And therefore changing the semantics so that it produces the correct output for the linker. Note: The old linkage name `linker_private_weak_def_auto' will still parse but is not a synonym for `linkonce_odr_auto_hide'. This should be removed in 4.0. <rdar://problem/11754934> llvm-svn: 162114
2012-08-18 02:33:14 +08:00
case lltok::kw_linker_private_weak_def_auto: // FIXME: backwards compat.
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
case lltok::kw_internal: // OptionalLinkage
case lltok::kw_weak: // OptionalLinkage
case lltok::kw_weak_odr: // OptionalLinkage
case lltok::kw_linkonce: // OptionalLinkage
case lltok::kw_linkonce_odr: // OptionalLinkage
Change the `linker_private_weak_def_auto' linkage to `linkonce_odr_auto_hide' to make it more consistent with its intended semantics. The `linker_private_weak_def_auto' linkage type was meant to automatically hide globals which never had their addresses taken. It has nothing to do with the `linker_private' linkage type, which outputs the symbols with a `l' (ell) prefix among other things. The intended semantic is more like the `linkonce_odr' linkage type. Change the name of the linkage type to `linkonce_odr_auto_hide'. And therefore changing the semantics so that it produces the correct output for the linker. Note: The old linkage name `linker_private_weak_def_auto' will still parse but is not a synonym for `linkonce_odr_auto_hide'. This should be removed in 4.0. <rdar://problem/11754934> llvm-svn: 162114
2012-08-18 02:33:14 +08:00
case lltok::kw_linkonce_odr_auto_hide: // OptionalLinkage
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
case lltok::kw_appending: // OptionalLinkage
case lltok::kw_dllexport: // OptionalLinkage
case lltok::kw_common: // OptionalLinkage
case lltok::kw_dllimport: // OptionalLinkage
case lltok::kw_extern_weak: // OptionalLinkage
case lltok::kw_external: { // OptionalLinkage
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
unsigned Linkage, Visibility;
if (ParseOptionalLinkage(Linkage) ||
ParseOptionalVisibility(Visibility) ||
switch the .ll parser to use SourceMgr. llvm-svn: 74735
2009-07-03 07:08:13 +08:00
ParseGlobal("", SMLoc(), Linkage, true, Visibility))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
break;
}
case lltok::kw_default: // OptionalVisibility
case lltok::kw_hidden: // OptionalVisibility
case lltok::kw_protected: { // OptionalVisibility
unsigned Visibility;
if (ParseOptionalVisibility(Visibility) ||
switch the .ll parser to use SourceMgr. llvm-svn: 74735
2009-07-03 07:08:13 +08:00
ParseGlobal("", SMLoc(), 0, false, Visibility))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
break;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_thread_local: // OptionalThreadLocal
case lltok::kw_addrspace: // OptionalAddrSpace
case lltok::kw_constant: // GlobalType
case lltok::kw_global: // GlobalType
switch the .ll parser to use SourceMgr. llvm-svn: 74735
2009-07-03 07:08:13 +08:00
if (ParseGlobal("", SMLoc(), 0, false, 0)) return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
break;
}
}
}
/// toplevelentity
/// ::= 'module' 'asm' STRINGCONSTANT
bool LLParser::ParseModuleAsm() {
assert(Lex.getKind() == lltok::kw_module);
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
std::string AsmStr;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(lltok::kw_asm, "expected 'module asm'") ||
ParseStringConstant(AsmStr)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Add a special streamer to libLTO that just records symbols definitions and uses. The result produced by the streamer is used to give the linker more accurate information and to add to llvm.compiler.used. The second improvement removes the need for the user to add __attribute__((used)) to functions only used in inline asm. The first one lets us build firefox with LTO on Darwin :-) llvm-svn: 126830
2011-03-02 12:14:42 +08:00
M->appendModuleInlineAsm(AsmStr);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
/// toplevelentity
/// ::= 'target' 'triple' '=' STRINGCONSTANT
/// ::= 'target' 'datalayout' '=' STRINGCONSTANT
bool LLParser::ParseTargetDefinition() {
assert(Lex.getKind() == lltok::kw_target);
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
std::string Str;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
switch (Lex.Lex()) {
default: return TokError("unknown target property");
case lltok::kw_triple:
Lex.Lex();
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(lltok::equal, "expected '=' after target triple") ||
ParseStringConstant(Str))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
M->setTargetTriple(Str);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
case lltok::kw_datalayout:
Lex.Lex();
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(lltok::equal, "expected '=' after target datalayout") ||
ParseStringConstant(Str))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
M->setDataLayout(Str);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
}
/// toplevelentity
/// ::= 'deplibs' '=' '[' ']'
/// ::= 'deplibs' '=' '[' STRINGCONSTANT (',' STRINGCONSTANT)* ']'
bool LLParser::ParseDepLibs() {
assert(Lex.getKind() == lltok::kw_deplibs);
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Lex.Lex();
if (ParseToken(lltok::equal, "expected '=' after deplibs") ||
ParseToken(lltok::lsquare, "expected '=' after deplibs"))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (EatIfPresent(lltok::rsquare))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
std::string Str;
if (ParseStringConstant(Str)) return true;
M->addLibrary(Str);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
while (EatIfPresent(lltok::comma)) {
if (ParseStringConstant(Str)) return true;
M->addLibrary(Str);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
return ParseToken(lltok::rsquare, "expected ']' at end of list");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
/// ParseUnnamedType:
/// ::= LocalVarID '=' 'type' type
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool LLParser::ParseUnnamedType() {
Remove support for parsing the "type i32" syntax for defining a numbered top level type without a specified number. This syntax isn't documented and blocks forward progress. llvm-svn: 133371
2011-06-19 08:03:46 +08:00
LocTy TypeLoc = Lex.getLoc();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
unsigned TypeID = Lex.getUIntVal();
Remove support for parsing the "type i32" syntax for defining a numbered top level type without a specified number. This syntax isn't documented and blocks forward progress. llvm-svn: 133371
2011-06-19 08:03:46 +08:00
Lex.Lex(); // eat LocalVarID;
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
Remove support for parsing the "type i32" syntax for defining a numbered top level type without a specified number. This syntax isn't documented and blocks forward progress. llvm-svn: 133371
2011-06-19 08:03:46 +08:00
if (ParseToken(lltok::equal, "expected '=' after name") ||
ParseToken(lltok::kw_type, "expected 'type' after '='"))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (TypeID >= NumberedTypes.size())
NumberedTypes.resize(TypeID+1);
Type *Result = 0;
if (ParseStructDefinition(TypeLoc, "",
NumberedTypes[TypeID], Result)) return true;
if (!isa<StructType>(Result)) {
std::pair<Type*, LocTy> &Entry = NumberedTypes[TypeID];
if (Entry.first)
return Error(TypeLoc, "non-struct types may not be recursive");
Entry.first = Result;
Entry.second = SMLoc();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// toplevelentity
/// ::= LocalVar '=' 'type' type
bool LLParser::ParseNamedType() {
std::string Name = Lex.getStrVal();
LocTy NameLoc = Lex.getLoc();
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Lex.Lex(); // eat LocalVar.
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(lltok::equal, "expected '=' after name") ||
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
ParseToken(lltok::kw_type, "expected 'type' after name"))
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
return true;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Result = 0;
if (ParseStructDefinition(NameLoc, Name,
NamedTypes[Name], Result)) return true;
if (!isa<StructType>(Result)) {
std::pair<Type*, LocTy> &Entry = NamedTypes[Name];
if (Entry.first)
return Error(NameLoc, "non-struct types may not be recursive");
Entry.first = Result;
Entry.second = SMLoc();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// toplevelentity
/// ::= 'declare' FunctionHeader
bool LLParser::ParseDeclare() {
assert(Lex.getKind() == lltok::kw_declare);
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Function *F;
return ParseFunctionHeader(F, false);
}
/// toplevelentity
/// ::= 'define' FunctionHeader '{' ...
bool LLParser::ParseDefine() {
assert(Lex.getKind() == lltok::kw_define);
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Function *F;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
return ParseFunctionHeader(F, true) ||
ParseFunctionBody(*F);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
/// ParseGlobalType
/// ::= 'constant'
/// ::= 'global'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool LLParser::ParseGlobalType(bool &IsConstant) {
if (Lex.getKind() == lltok::kw_constant)
IsConstant = true;
else if (Lex.getKind() == lltok::kw_global)
IsConstant = false;
Pacify gcc-4.3 (thinks IsConstant may be used uninitialized). llvm-svn: 64220
2009-02-11 00:24:55 +08:00
else {
IsConstant = false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return TokError("expected 'global' or 'constant'");
Pacify gcc-4.3 (thinks IsConstant may be used uninitialized). llvm-svn: 64220
2009-02-11 00:24:55 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
return false;
}
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
/// ParseUnnamedGlobal:
/// OptionalVisibility ALIAS ...
/// OptionalLinkage OptionalVisibility ... -> global variable
/// GlobalID '=' OptionalVisibility ALIAS ...
/// GlobalID '=' OptionalLinkage OptionalVisibility ... -> global variable
bool LLParser::ParseUnnamedGlobal() {
unsigned VarID = NumberedVals.size();
std::string Name;
LocTy NameLoc = Lex.getLoc();
// Handle the GlobalID form.
if (Lex.getKind() == lltok::GlobalID) {
if (Lex.getUIntVal() != VarID)
return Error(Lex.getLoc(), "variable expected to be numbered '%" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(VarID) + "'");
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
Lex.Lex(); // eat GlobalID;
if (ParseToken(lltok::equal, "expected '=' after name"))
return true;
}
bool HasLinkage;
unsigned Linkage, Visibility;
if (ParseOptionalLinkage(Linkage, HasLinkage) ||
ParseOptionalVisibility(Visibility))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Extend the AsmWriter to print unnamed numbered types as "%0 = type ..." and unnamed numbered global variables as "@0 = global ...". Extend the AsmParser to recognize these forms. llvm-svn: 78859
2009-08-13 07:32:33 +08:00
if (HasLinkage || Lex.getKind() != lltok::kw_alias)
return ParseGlobal(Name, NameLoc, Linkage, HasLinkage, Visibility);
return ParseAlias(Name, NameLoc, Visibility);
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseNamedGlobal:
/// GlobalVar '=' OptionalVisibility ALIAS ...
/// GlobalVar '=' OptionalLinkage OptionalVisibility ... -> global variable
bool LLParser::ParseNamedGlobal() {
assert(Lex.getKind() == lltok::GlobalVar);
LocTy NameLoc = Lex.getLoc();
std::string Name = Lex.getStrVal();
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool HasLinkage;
unsigned Linkage, Visibility;
if (ParseToken(lltok::equal, "expected '=' in global variable") ||
ParseOptionalLinkage(Linkage, HasLinkage) ||
ParseOptionalVisibility(Visibility))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (HasLinkage || Lex.getKind() != lltok::kw_alias)
return ParseGlobal(Name, NameLoc, Linkage, HasLinkage, Visibility);
return ParseAlias(Name, NameLoc, Visibility);
}
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
// MDString:
// ::= '!' STRINGCONSTANT
change ParseMDString and ParseMDNode to take arguments of the right type. This exposed a raft of other problems, which I'll deal with in subsequent patches. llvm-svn: 92273
2009-12-30 05:53:55 +08:00
bool LLParser::ParseMDString(MDString *&Result) {
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
std::string Str;
if (ParseStringConstant(Str)) return true;
change ParseMDString and ParseMDNode to take arguments of the right type. This exposed a raft of other problems, which I'll deal with in subsequent patches. llvm-svn: 92273
2009-12-30 05:53:55 +08:00
Result = MDString::get(Context, Str);
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
return false;
}
// MDNode:
// ::= '!' MDNodeNumber
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
//
/// This version of ParseMDNodeID returns the slot number and null in the case
/// of a forward reference.
bool LLParser::ParseMDNodeID(MDNode *&Result, unsigned &SlotNo) {
// !{ ..., !42, ... }
if (ParseUInt32(SlotNo)) return true;
// Check existing MDNode.
if (SlotNo < NumberedMetadata.size() && NumberedMetadata[SlotNo] != 0)
Result = NumberedMetadata[SlotNo];
else
Result = 0;
return false;
}
rename ParseMDNode -> ParseMDNodeID, since it parses !42, not !{... } as you'd expect. llvm-svn: 92291
2009-12-30 12:15:23 +08:00
bool LLParser::ParseMDNodeID(MDNode *&Result) {
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
// !{ ..., !42, ... }
unsigned MID = 0;
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
if (ParseMDNodeID(Result, MID)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
// If not a forward reference, just return it now.
if (Result) return false;
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
// Otherwise, create MDNode forward reference.
PR9214: Convert Metadata API to use ArrayRef. llvm-svn: 129932
2011-04-22 03:59:31 +08:00
MDNode *FwdNode = MDNode::getTemporary(Context, ArrayRef<Value*>());
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
ForwardRefMDNodes[MID] = std::make_pair(FwdNode, Lex.getLoc());
rename MetadataCache -> NumberedMetadata to follow the convention used by other things. Convert it to a vector since it is a dense numbering. llvm-svn: 92293
2009-12-30 12:51:58 +08:00
if (NumberedMetadata.size() <= MID)
NumberedMetadata.resize(MID+1);
NumberedMetadata[MID] = FwdNode;
change ParseMDString and ParseMDNode to take arguments of the right type. This exposed a raft of other problems, which I'll deal with in subsequent patches. llvm-svn: 92273
2009-12-30 05:53:55 +08:00
Result = FwdNode;
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
return false;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
}
Refactor metadata parsing routines into separate functions. llvm-svn: 76455
2009-07-21 03:00:08 +08:00
simplify some code and unbreak the build by not consuming an extra token. llvm-svn: 92279
2009-12-30 06:35:39 +08:00
/// ParseNamedMetadata:
Parse named metadata. llvm-svn: 77410
2009-07-29 08:34:02 +08:00
/// !foo = !{ !1, !2 }
bool LLParser::ParseNamedMetadata() {
rename NamedOrCustomMD -> MetadataVar to follow conventions of all the rest of the code. llvm-svn: 92295
2009-12-30 13:02:06 +08:00
assert(Lex.getKind() == lltok::MetadataVar);
Parse named metadata. llvm-svn: 77410
2009-07-29 08:34:02 +08:00
std::string Name = Lex.getStrVal();
rename NamedOrCustomMD -> MetadataVar to follow conventions of all the rest of the code. llvm-svn: 92295
2009-12-30 13:02:06 +08:00
Lex.Lex();
Parse named metadata. llvm-svn: 77410
2009-07-29 08:34:02 +08:00
simplify some code and unbreak the build by not consuming an extra token. llvm-svn: 92279
2009-12-30 06:35:39 +08:00
if (ParseToken(lltok::equal, "expected '=' here") ||
rename lltok::Metadata -> lltok::exclaim. We name tokens after their syntactic form, not their semantic form. llvm-svn: 92294
2009-12-30 12:56:59 +08:00
ParseToken(lltok::exclaim, "Expected '!' here") ||
simplify some code and unbreak the build by not consuming an extra token. llvm-svn: 92279
2009-12-30 06:35:39 +08:00
ParseToken(lltok::lbrace, "Expected '{' here"))
Parse named metadata. llvm-svn: 77410
2009-07-29 08:34:02 +08:00
return true;
Make NamedMDNode not be a subclass of Value, and simplify the interface for creating and populating NamedMDNodes. llvm-svn: 109061
2010-07-22 07:38:33 +08:00
NamedMDNode *NMD = M->getOrInsertNamedMetadata(Name);
Add support for empty named metadata too. This isn't particularly useful, but it is nice for consistency. llvm-svn: 108262
2010-07-14 03:42:44 +08:00
if (Lex.getKind() != lltok::rbrace)
do {
if (ParseToken(lltok::exclaim, "Expected '!' here"))
return true;
change ParseMDString and ParseMDNode to take arguments of the right type. This exposed a raft of other problems, which I'll deal with in subsequent patches. llvm-svn: 92273
2009-12-30 05:53:55 +08:00
Add support for empty named metadata too. This isn't particularly useful, but it is nice for consistency. llvm-svn: 108262
2010-07-14 03:42:44 +08:00
MDNode *N = 0;
if (ParseMDNodeID(N)) return true;
Make NamedMDNode not be a subclass of Value, and simplify the interface for creating and populating NamedMDNodes. llvm-svn: 109061
2010-07-22 07:38:33 +08:00
NMD->addOperand(N);
Add support for empty named metadata too. This isn't particularly useful, but it is nice for consistency. llvm-svn: 108262
2010-07-14 03:42:44 +08:00
} while (EatIfPresent(lltok::comma));
Parse named metadata. llvm-svn: 77410
2009-07-29 08:34:02 +08:00
if (ParseToken(lltok::rbrace, "expected end of metadata node"))
return true;
return false;
}
Support stand alone metadata syntax. !0 = constant metadata !{i32 21, i32 22} @llvm.blah = constant metadata !{i32 1000, i16 200, metadata !0} llvm-svn: 74630
2009-07-02 03:21:12 +08:00
/// ParseStandaloneMetadata:
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
/// !42 = !{...}
Support stand alone metadata syntax. !0 = constant metadata !{i32 21, i32 22} @llvm.blah = constant metadata !{i32 1000, i16 200, metadata !0} llvm-svn: 74630
2009-07-02 03:21:12 +08:00
bool LLParser::ParseStandaloneMetadata() {
rename lltok::Metadata -> lltok::exclaim. We name tokens after their syntactic form, not their semantic form. llvm-svn: 92294
2009-12-30 12:56:59 +08:00
assert(Lex.getKind() == lltok::exclaim);
Support stand alone metadata syntax. !0 = constant metadata !{i32 21, i32 22} @llvm.blah = constant metadata !{i32 1000, i16 200, metadata !0} llvm-svn: 74630
2009-07-02 03:21:12 +08:00
Lex.Lex();
unsigned MetadataID = 0;
LocTy TyLoc;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Ty = 0;
Derive MDNode from MetadataBase instead of Constant. Emit MDNodes into METADATA_BLOCK in bitcode file. llvm-svn: 76834
2009-07-23 09:07:34 +08:00
SmallVector<Value *, 16> Elts;
factor code even more. llvm-svn: 92280
2009-12-30 06:40:21 +08:00
if (ParseUInt32(MetadataID) ||
ParseToken(lltok::equal, "expected '=' here") ||
ParseType(Ty, TyLoc) ||
rename lltok::Metadata -> lltok::exclaim. We name tokens after their syntactic form, not their semantic form. llvm-svn: 92294
2009-12-30 12:56:59 +08:00
ParseToken(lltok::exclaim, "Expected '!' here") ||
factor code even more. llvm-svn: 92280
2009-12-30 06:40:21 +08:00
ParseToken(lltok::lbrace, "Expected '{' here") ||
Compute isFunctionLocal in MDNode ctor or via argument in new function getWhenValsUnresolved(). Document PFS argument to ParseValID() and ConvertGlobalOrMetadataValIDToValue(). llvm-svn: 93108
2010-01-10 15:14:18 +08:00
ParseMDNodeVector(Elts, NULL) ||
factor code even more. llvm-svn: 92280
2009-12-30 06:40:21 +08:00
ParseToken(lltok::rbrace, "expected end of metadata node"))
Derive MDNode from MetadataBase instead of Constant. Emit MDNodes into METADATA_BLOCK in bitcode file. llvm-svn: 76834
2009-07-23 09:07:34 +08:00
return true;
PR9214: Convert Metadata API to use ArrayRef. llvm-svn: 129932
2011-04-22 03:59:31 +08:00
MDNode *Init = MDNode::get(Context, Elts);
rename MetadataCache -> NumberedMetadata to follow the convention used by other things. Convert it to a vector since it is a dense numbering. llvm-svn: 92293
2009-12-30 12:51:58 +08:00
// See if this was forward referenced, if so, handle it.
switch to TrackingVH instead of WeakVH, since these can never be RAUW'd and go to null. This also gets us some sorely lacking type safety. llvm-svn: 92272
2009-12-30 05:43:58 +08:00
std::map<unsigned, std::pair<TrackingVH<MDNode>, LocTy> >::iterator
Support MDNode forward reference. llvm-svn: 75031
2009-07-09 03:23:54 +08:00
FI = ForwardRefMDNodes.find(MetadataID);
if (FI != ForwardRefMDNodes.end()) {
Introduce a new temporary MDNode concept. Temporary MDNodes are not part of the IR, are not uniqued, and may be safely RAUW'd. This replaces a variety of alternate mechanisms for achieving the same effect. llvm-svn: 111681
2010-08-21 06:02:26 +08:00
MDNode *Temp = FI->second.first;
Temp->replaceAllUsesWith(Init);
MDNode::deleteTemporary(Temp);
Support MDNode forward reference. llvm-svn: 75031
2009-07-09 03:23:54 +08:00
ForwardRefMDNodes.erase(FI);
rename MetadataCache -> NumberedMetadata to follow the convention used by other things. Convert it to a vector since it is a dense numbering. llvm-svn: 92293
2009-12-30 12:51:58 +08:00
assert(NumberedMetadata[MetadataID] == Init && "Tracking VH didn't work");
} else {
if (MetadataID >= NumberedMetadata.size())
NumberedMetadata.resize(MetadataID+1);
if (NumberedMetadata[MetadataID] != 0)
return TokError("Metadata id is already used");
NumberedMetadata[MetadataID] = Init;
Support MDNode forward reference. llvm-svn: 75031
2009-07-09 03:23:54 +08:00
}
Support stand alone metadata syntax. !0 = constant metadata !{i32 21, i32 22} @llvm.blah = constant metadata !{i32 1000, i16 200, metadata !0} llvm-svn: 74630
2009-07-02 03:21:12 +08:00
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseAlias:
/// ::= GlobalVar '=' OptionalVisibility 'alias' OptionalLinkage Aliasee
/// Aliasee
testcase and asmparser fix for PR4066 llvm-svn: 70080
2009-04-26 05:26:00 +08:00
/// ::= TypeAndValue
/// ::= 'bitcast' '(' TypeAndValue 'to' Type ')'
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
/// ::= 'getelementptr' 'inbounds'? '(' ... ')'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
///
/// Everything through visibility has already been parsed.
///
bool LLParser::ParseAlias(const std::string &Name, LocTy NameLoc,
unsigned Visibility) {
assert(Lex.getKind() == lltok::kw_alias);
Lex.Lex();
unsigned Linkage;
LocTy LinkageLoc = Lex.getLoc();
if (ParseOptionalLinkage(Linkage))
return true;
if (Linkage != GlobalValue::ExternalLinkage &&
Introduce new linkage types linkonce_odr, weak_odr, common_odr and extern_weak_odr. These are the same as the non-odr versions, except that they indicate that the global will only be overridden by an *equivalent* global. In C, a function with weak linkage can be overridden by a function which behaves completely differently. This means that IP passes have to skip weak functions, since any deductions made from the function definition might be wrong, since the definition could be replaced by something completely different at link time. This is not allowed in C++, thanks to the ODR (One-Definition-Rule): if a function is replaced by another at link-time, then the new function must be the same as the original function. If a language knows that a function or other global can only be overridden by an equivalent global, it can give it the weak_odr linkage type, and the optimizers will understand that it is alright to make deductions based on the function body. The code generators on the other hand map weak and weak_odr linkage to the same thing. llvm-svn: 66339
2009-03-07 23:45:40 +08:00
Linkage != GlobalValue::WeakAnyLinkage &&
Linkage != GlobalValue::WeakODRLinkage &&
Add the private linkage. llvm-svn: 62279
2009-01-16 04:18:42 +08:00
Linkage != GlobalValue::InternalLinkage &&
Add plumbing for the `linker_private' linkage type. This type is meant for "private" symbols which the assember shouldn't strip, but which the linker may remove after evaluation. This is mostly useful for Objective-C metadata. This is plumbing, so we don't have a use of it yet. More to come, etc. llvm-svn: 76385
2009-07-20 09:03:30 +08:00
Linkage != GlobalValue::PrivateLinkage &&
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
Linkage != GlobalValue::LinkerPrivateLinkage &&
Change the `linker_private_weak_def_auto' linkage to `linkonce_odr_auto_hide' to make it more consistent with its intended semantics. The `linker_private_weak_def_auto' linkage type was meant to automatically hide globals which never had their addresses taken. It has nothing to do with the `linker_private' linkage type, which outputs the symbols with a `l' (ell) prefix among other things. The intended semantic is more like the `linkonce_odr' linkage type. Change the name of the linkage type to `linkonce_odr_auto_hide'. And therefore changing the semantics so that it produces the correct output for the linker. Note: The old linkage name `linker_private_weak_def_auto' will still parse but is not a synonym for `linkonce_odr_auto_hide'. This should be removed in 4.0. <rdar://problem/11754934> llvm-svn: 162114
2012-08-18 02:33:14 +08:00
Linkage != GlobalValue::LinkerPrivateWeakLinkage)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(LinkageLoc, "invalid linkage type for alias");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Constant *Aliasee;
LocTy AliaseeLoc = Lex.getLoc();
testcase and asmparser fix for PR4066 llvm-svn: 70080
2009-04-26 05:26:00 +08:00
if (Lex.getKind() != lltok::kw_bitcast &&
Lex.getKind() != lltok::kw_getelementptr) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseGlobalTypeAndValue(Aliasee)) return true;
} else {
// The bitcast dest type is not present, it is implied by the dest type.
ValID ID;
if (ParseValID(ID)) return true;
if (ID.Kind != ValID::t_Constant)
return Error(AliaseeLoc, "invalid aliasee");
Aliasee = ID.ConstantVal;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Aliasee->getType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(AliaseeLoc, "alias must have pointer type");
// Okay, create the alias but do not insert it into the module yet.
GlobalAlias* GA = new GlobalAlias(Aliasee->getType(),
(GlobalValue::LinkageTypes)Linkage, Name,
Aliasee);
GA->setVisibility((GlobalValue::VisibilityTypes)Visibility);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// See if this value already exists in the symbol table. If so, it is either
// a redefinition or a definition of a forward reference.
fix PR5295 where the .ll parser didn't reject a function after a global or global after a function with conflicting names. Update some testcases that were accidentally depending on this behavior. llvm-svn: 85081
2009-10-26 07:22:50 +08:00
if (GlobalValue *Val = M->getNamedValue(Name)) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// See if this was a redefinition. If so, there is no entry in
// ForwardRefVals.
std::map<std::string, std::pair<GlobalValue*, LocTy> >::iterator
I = ForwardRefVals.find(Name);
if (I == ForwardRefVals.end())
return Error(NameLoc, "redefinition of global named '@" + Name + "'");
// Otherwise, this was a definition of forward ref. Verify that types
// agree.
if (Val->getType() != GA->getType())
return Error(NameLoc,
"forward reference and definition of alias have different types");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If they agree, just RAUW the old value with the alias and remove the
// forward ref info.
Val->replaceAllUsesWith(GA);
Val->eraseFromParent();
ForwardRefVals.erase(I);
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Insert into the module, we know its name won't collide now.
M->getAliasList().push_back(GA);
Eliminate some calls to Value::getNameStr. llvm-svn: 116670
2010-10-16 19:28:23 +08:00
assert(GA->getName() == Name && "Should not be a name conflict!");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
/// ParseGlobal
/// ::= GlobalVar '=' OptionalLinkage OptionalVisibility OptionalThreadLocal
First step in fixing PR8927: Add a unnamed_addr bit to global variables and functions. This will be used to indicate that the address is not significant and therefore the constant or function can be merged with others. If an optimization pass can show that an address is not used, it can set this. Examples of things that can have this set by the FE are globals created to hold string literals and C++ constructors. Adding unnamed_addr to a non-const global should have no effect unless an optimization can transform that global into a constant. Aliases are not allowed to have unnamed_addr since I couldn't figure out any use for it. llvm-svn: 123063
2011-01-09 00:42:36 +08:00
/// OptionalAddrSpace OptionalUnNammedAddr GlobalType Type Const
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= OptionalLinkage OptionalVisibility OptionalThreadLocal
First step in fixing PR8927: Add a unnamed_addr bit to global variables and functions. This will be used to indicate that the address is not significant and therefore the constant or function can be merged with others. If an optimization pass can show that an address is not used, it can set this. Examples of things that can have this set by the FE are globals created to hold string literals and C++ constructors. Adding unnamed_addr to a non-const global should have no effect unless an optimization can transform that global into a constant. Aliases are not allowed to have unnamed_addr since I couldn't figure out any use for it. llvm-svn: 123063
2011-01-09 00:42:36 +08:00
/// OptionalAddrSpace OptionalUnNammedAddr GlobalType Type Const
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
///
/// Everything through visibility has been parsed already.
///
bool LLParser::ParseGlobal(const std::string &Name, LocTy NameLoc,
unsigned Linkage, bool HasLinkage,
unsigned Visibility) {
unsigned AddrSpace;
Extend the IL for selecting TLS models (PR9788) This allows the user/front-end to specify a model that is better than what LLVM would choose by default. For example, a variable might be declared as @x = thread_local(initialexec) global i32 42 if it will not be used in a shared library that is dlopen'ed. If the specified model isn't supported by the target, or if LLVM can make a better choice, a different model may be used. llvm-svn: 159077
2012-06-23 19:37:03 +08:00
bool IsConstant, UnnamedAddr;
GlobalVariable::ThreadLocalMode TLM;
Reject uses of unnamed_addr in declarations. llvm-svn: 123358
2011-01-13 09:30:30 +08:00
LocTy UnnamedAddrLoc;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy TyLoc;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Ty = 0;
Extend the IL for selecting TLS models (PR9788) This allows the user/front-end to specify a model that is better than what LLVM would choose by default. For example, a variable might be declared as @x = thread_local(initialexec) global i32 42 if it will not be used in a shared library that is dlopen'ed. If the specified model isn't supported by the target, or if LLVM can make a better choice, a different model may be used. llvm-svn: 159077
2012-06-23 19:37:03 +08:00
if (ParseOptionalThreadLocal(TLM) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseOptionalAddrSpace(AddrSpace) ||
Reject uses of unnamed_addr in declarations. llvm-svn: 123358
2011-01-13 09:30:30 +08:00
ParseOptionalToken(lltok::kw_unnamed_addr, UnnamedAddr,
&UnnamedAddrLoc) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseGlobalType(IsConstant) ||
ParseType(Ty, TyLoc))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If the linkage is specified and is external, then no initializer is
// present.
Constant *Init = 0;
if (!HasLinkage || (Linkage != GlobalValue::DLLImportLinkage &&
Remove the one-definition-rule version of extern_weak linkage: this linkage type only applies to declarations, but ODR is only relevant to globals with definitions. llvm-svn: 66650
2009-03-11 16:08:06 +08:00
Linkage != GlobalValue::ExternalWeakLinkage &&
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Linkage != GlobalValue::ExternalLinkage)) {
if (ParseGlobalValue(Ty, Init))
return true;
}
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (Ty->isFunctionTy() || Ty->isLabelTy())
fix typo llvm-svn: 64091
2009-02-09 04:00:15 +08:00
return Error(TyLoc, "invalid type for global variable");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
GlobalVariable *GV = 0;
// See if the global was forward referenced, if so, use the global.
Fix PR3372 llvm-svn: 63501
2009-02-02 15:24:28 +08:00
if (!Name.empty()) {
fix PR5295 where the .ll parser didn't reject a function after a global or global after a function with conflicting names. Update some testcases that were accidentally depending on this behavior. llvm-svn: 85081
2009-10-26 07:22:50 +08:00
if (GlobalValue *GVal = M->getNamedValue(Name)) {
if (!ForwardRefVals.erase(Name) || !isa<GlobalValue>(GVal))
return Error(NameLoc, "redefinition of global '@" + Name + "'");
GV = cast<GlobalVariable>(GVal);
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else {
std::map<unsigned, std::pair<GlobalValue*, LocTy> >::iterator
I = ForwardRefValIDs.find(NumberedVals.size());
if (I != ForwardRefValIDs.end()) {
GV = cast<GlobalVariable>(I->second.first);
ForwardRefValIDs.erase(I);
}
}
if (GV == 0) {
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
GV = new GlobalVariable(*M, Ty, false, GlobalValue::ExternalLinkage, 0,
Extend the IL for selecting TLS models (PR9788) This allows the user/front-end to specify a model that is better than what LLVM would choose by default. For example, a variable might be declared as @x = thread_local(initialexec) global i32 42 if it will not be used in a shared library that is dlopen'ed. If the specified model isn't supported by the target, or if LLVM can make a better choice, a different model may be used. llvm-svn: 159077
2012-06-23 19:37:03 +08:00
Name, 0, GlobalVariable::NotThreadLocal,
AddrSpace);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else {
if (GV->getType()->getElementType() != Ty)
return Error(TyLoc,
"forward reference and definition of global have different types");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Move the forward-reference to the correct spot in the module.
M->getGlobalList().splice(M->global_end(), M->getGlobalList(), GV);
}
if (Name.empty())
NumberedVals.push_back(GV);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Set the parsed properties on the global.
if (Init)
GV->setInitializer(Init);
GV->setConstant(IsConstant);
GV->setLinkage((GlobalValue::LinkageTypes)Linkage);
GV->setVisibility((GlobalValue::VisibilityTypes)Visibility);
Extend the IL for selecting TLS models (PR9788) This allows the user/front-end to specify a model that is better than what LLVM would choose by default. For example, a variable might be declared as @x = thread_local(initialexec) global i32 42 if it will not be used in a shared library that is dlopen'ed. If the specified model isn't supported by the target, or if LLVM can make a better choice, a different model may be used. llvm-svn: 159077
2012-06-23 19:37:03 +08:00
GV->setThreadLocalMode(TLM);
First step in fixing PR8927: Add a unnamed_addr bit to global variables and functions. This will be used to indicate that the address is not significant and therefore the constant or function can be merged with others. If an optimization pass can show that an address is not used, it can set this. Examples of things that can have this set by the FE are globals created to hold string literals and C++ constructors. Adding unnamed_addr to a non-const global should have no effect unless an optimization can transform that global into a constant. Aliases are not allowed to have unnamed_addr since I couldn't figure out any use for it. llvm-svn: 123063
2011-01-09 00:42:36 +08:00
GV->setUnnamedAddr(UnnamedAddr);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Parse attributes on the global.
while (Lex.getKind() == lltok::comma) {
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Lex.getKind() == lltok::kw_section) {
Lex.Lex();
GV->setSection(Lex.getStrVal());
if (ParseToken(lltok::StringConstant, "expected global section string"))
return true;
} else if (Lex.getKind() == lltok::kw_align) {
unsigned Alignment;
if (ParseOptionalAlignment(Alignment)) return true;
GV->setAlignment(Alignment);
} else {
TokError("unknown global variable property!");
}
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
//===----------------------------------------------------------------------===//
// GlobalValue Reference/Resolution Routines.
//===----------------------------------------------------------------------===//
/// GetGlobalVal - Get a value with the specified name or ID, creating a
/// forward reference record if needed. This can return null if the value
/// exists but does not have the right type.
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
GlobalValue *LLParser::GetGlobalVal(const std::string &Name, Type *Ty,
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy Loc) {
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
PointerType *PTy = dyn_cast<PointerType>(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (PTy == 0) {
Error(Loc, "global variable reference must have pointer type");
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Look this name up in the normal function symbol table.
GlobalValue *Val =
cast_or_null<GlobalValue>(M->getValueSymbolTable().lookup(Name));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If this is a forward reference for the value, see if we already created a
// forward ref record.
if (Val == 0) {
std::map<std::string, std::pair<GlobalValue*, LocTy> >::iterator
I = ForwardRefVals.find(Name);
if (I != ForwardRefVals.end())
Val = I->second.first;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If we have the value in the symbol table or fwd-ref table, return it.
if (Val) {
if (Val->getType() == Ty) return Val;
Error(Loc, "'@" + Name + "' defined with type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Val->getType()) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Otherwise, create a new forward reference for this value and remember it.
GlobalValue *FwdVal;
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
if (FunctionType *FT = dyn_cast<FunctionType>(PTy->getElementType()))
Remove the one-definition-rule version of extern_weak linkage: this linkage type only applies to declarations, but ODR is only relevant to globals with definitions. llvm-svn: 66650
2009-03-11 16:08:06 +08:00
FwdVal = Function::Create(FT, GlobalValue::ExternalWeakLinkage, Name, M);
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
else
Switch GlobalVariable ctors to a sane API, where *either* a context or a module is required. llvm-svn: 75025
2009-07-09 03:03:57 +08:00
FwdVal = new GlobalVariable(*M, PTy->getElementType(), false,
GlobalValue::ExternalWeakLinkage, 0, Name);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefVals[Name] = std::make_pair(FwdVal, Loc);
return FwdVal;
}
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
GlobalValue *LLParser::GetGlobalVal(unsigned ID, Type *Ty, LocTy Loc) {
PointerType *PTy = dyn_cast<PointerType>(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (PTy == 0) {
Error(Loc, "global variable reference must have pointer type");
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
GlobalValue *Val = ID < NumberedVals.size() ? NumberedVals[ID] : 0;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If this is a forward reference for the value, see if we already created a
// forward ref record.
if (Val == 0) {
std::map<unsigned, std::pair<GlobalValue*, LocTy> >::iterator
I = ForwardRefValIDs.find(ID);
if (I != ForwardRefValIDs.end())
Val = I->second.first;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If we have the value in the symbol table or fwd-ref table, return it.
if (Val) {
if (Val->getType() == Ty) return Val;
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Error(Loc, "'@" + Twine(ID) + "' defined with type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Val->getType()) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Otherwise, create a new forward reference for this value and remember it.
GlobalValue *FwdVal;
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
if (FunctionType *FT = dyn_cast<FunctionType>(PTy->getElementType()))
Remove the one-definition-rule version of extern_weak linkage: this linkage type only applies to declarations, but ODR is only relevant to globals with definitions. llvm-svn: 66650
2009-03-11 16:08:06 +08:00
FwdVal = Function::Create(FT, GlobalValue::ExternalWeakLinkage, "", M);
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
else
Switch GlobalVariable ctors to a sane API, where *either* a context or a module is required. llvm-svn: 75025
2009-07-09 03:03:57 +08:00
FwdVal = new GlobalVariable(*M, PTy->getElementType(), false,
GlobalValue::ExternalWeakLinkage, 0, "");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefValIDs[ID] = std::make_pair(FwdVal, Loc);
return FwdVal;
}
//===----------------------------------------------------------------------===//
// Helper Routines.
//===----------------------------------------------------------------------===//
/// ParseToken - If the current token has the specified kind, eat it and return
/// success. Otherwise, emit the specified error and return failure.
bool LLParser::ParseToken(lltok::Kind T, const char *ErrMsg) {
if (Lex.getKind() != T)
return TokError(ErrMsg);
Lex.Lex();
return false;
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
/// ParseStringConstant
/// ::= StringConstant
bool LLParser::ParseStringConstant(std::string &Result) {
if (Lex.getKind() != lltok::StringConstant)
return TokError("expected string constant");
Result = Lex.getStrVal();
Lex.Lex();
return false;
}
/// ParseUInt32
/// ::= uint32
bool LLParser::ParseUInt32(unsigned &Val) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Lex.getKind() != lltok::APSInt || Lex.getAPSIntVal().isSigned())
return TokError("expected integer");
uint64_t Val64 = Lex.getAPSIntVal().getLimitedValue(0xFFFFFFFFULL+1);
if (Val64 != unsigned(Val64))
return TokError("expected 32-bit integer (too large)");
Val = Val64;
Lex.Lex();
return false;
}
Extend the IL for selecting TLS models (PR9788) This allows the user/front-end to specify a model that is better than what LLVM would choose by default. For example, a variable might be declared as @x = thread_local(initialexec) global i32 42 if it will not be used in a shared library that is dlopen'ed. If the specified model isn't supported by the target, or if LLVM can make a better choice, a different model may be used. llvm-svn: 159077
2012-06-23 19:37:03 +08:00
/// ParseTLSModel
/// := 'localdynamic'
/// := 'initialexec'
/// := 'localexec'
bool LLParser::ParseTLSModel(GlobalVariable::ThreadLocalMode &TLM) {
switch (Lex.getKind()) {
default:
return TokError("expected localdynamic, initialexec or localexec");
case lltok::kw_localdynamic:
TLM = GlobalVariable::LocalDynamicTLSModel;
break;
case lltok::kw_initialexec:
TLM = GlobalVariable::InitialExecTLSModel;
break;
case lltok::kw_localexec:
TLM = GlobalVariable::LocalExecTLSModel;
break;
}
Lex.Lex();
return false;
}
/// ParseOptionalThreadLocal
/// := /*empty*/
/// := 'thread_local'
/// := 'thread_local' '(' tlsmodel ')'
bool LLParser::ParseOptionalThreadLocal(GlobalVariable::ThreadLocalMode &TLM) {
TLM = GlobalVariable::NotThreadLocal;
if (!EatIfPresent(lltok::kw_thread_local))
return false;
TLM = GlobalVariable::GeneralDynamicTLSModel;
if (Lex.getKind() == lltok::lparen) {
Lex.Lex();
return ParseTLSModel(TLM) ||
ParseToken(lltok::rparen, "expected ')' after thread local model");
}
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseOptionalAddrSpace
/// := /*empty*/
/// := 'addrspace' '(' uint32 ')'
bool LLParser::ParseOptionalAddrSpace(unsigned &AddrSpace) {
AddrSpace = 0;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (!EatIfPresent(lltok::kw_addrspace))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
return ParseToken(lltok::lparen, "expected '(' in address space") ||
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
ParseUInt32(AddrSpace) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::rparen, "expected ')' in address space");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseOptionalAttrs - Parse a potentially empty attribute list. AttrKind
/// indicates what kind of attribute list this is: 0: function arg, 1: result,
/// 2: function attr.
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
bool LLParser::ParseOptionalAttrs(Attributes &Attrs, unsigned AttrKind) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Attrs = Attribute::None;
LocTy AttrLoc = Lex.getLoc();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
while (1) {
switch (Lex.getKind()) {
default: // End of attributes.
if (AttrKind != 2 && (Attrs & Attribute::FunctionOnly))
return Error(AttrLoc, "invalid use of function-only attribute");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Remove old backwards compatibility support from the parser for autoupgrading the old malloc/free instructions, and for 'sext' and 'zext' as function attributes (they are spelled signext/zeroext now), and support for result value attributes being specified after a function. Additionally, diagnose invalid attributes on functions with an error message instead of an abort in the verifier. llvm-svn: 133229
2011-06-17 11:16:47 +08:00
// As a hack, we allow "align 2" on functions as a synonym for
// "alignstack 2".
if (AttrKind == 2 &&
(Attrs & ~(Attribute::FunctionOnly | Attribute::Alignment)))
return Error(AttrLoc, "invalid use of attribute on a function");
if (AttrKind != 0 && (Attrs & Attribute::ParameterOnly))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(AttrLoc, "invalid use of parameter-only attribute");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
Add new function attribute - noimplicitfloat Update code generator to use this attribute and remove NoImplicitFloat target option. Update llc to set this attribute when -no-implicit-float command line option is used. llvm-svn: 72959
2009-06-06 05:57:13 +08:00
case lltok::kw_zeroext: Attrs |= Attribute::ZExt; break;
case lltok::kw_signext: Attrs |= Attribute::SExt; break;
case lltok::kw_inreg: Attrs |= Attribute::InReg; break;
case lltok::kw_sret: Attrs |= Attribute::StructRet; break;
case lltok::kw_noalias: Attrs |= Attribute::NoAlias; break;
case lltok::kw_nocapture: Attrs |= Attribute::NoCapture; break;
case lltok::kw_byval: Attrs |= Attribute::ByVal; break;
case lltok::kw_nest: Attrs |= Attribute::Nest; break;
case lltok::kw_noreturn: Attrs |= Attribute::NoReturn; break;
case lltok::kw_nounwind: Attrs |= Attribute::NoUnwind; break;
Replace the -unwind-tables option with a per function flag. This is more LTO friendly as we can now correctly merge files compiled with or without -fasynchronous-unwind-tables. llvm-svn: 132033
2011-05-25 11:44:17 +08:00
case lltok::kw_uwtable: Attrs |= Attribute::UWTable; break;
Add the returns_twice attribute to LLVM. llvm-svn: 141001
2011-10-03 22:45:37 +08:00
case lltok::kw_returns_twice: Attrs |= Attribute::ReturnsTwice; break;
Add new function attribute - noimplicitfloat Update code generator to use this attribute and remove NoImplicitFloat target option. Update llc to set this attribute when -no-implicit-float command line option is used. llvm-svn: 72959
2009-06-06 05:57:13 +08:00
case lltok::kw_noinline: Attrs |= Attribute::NoInline; break;
case lltok::kw_readnone: Attrs |= Attribute::ReadNone; break;
case lltok::kw_readonly: Attrs |= Attribute::ReadOnly; break;
Reintroduce the InlineHint function attribute. This time it's for real! I am going to hook this up in the frontends as well. The inliner has some experimental heuristics for dealing with the inline hint. When given a -respect-inlinehint option, functions marked with the inline keyword are given a threshold just above the default for -O3. We need some experiments to determine if that is the right thing to do. llvm-svn: 95466
2010-02-06 09:16:28 +08:00
case lltok::kw_inlinehint: Attrs |= Attribute::InlineHint; break;
Add new function attribute - noimplicitfloat Update code generator to use this attribute and remove NoImplicitFloat target option. Update llc to set this attribute when -no-implicit-float command line option is used. llvm-svn: 72959
2009-06-06 05:57:13 +08:00
case lltok::kw_alwaysinline: Attrs |= Attribute::AlwaysInline; break;
case lltok::kw_optsize: Attrs |= Attribute::OptimizeForSize; break;
case lltok::kw_ssp: Attrs |= Attribute::StackProtect; break;
case lltok::kw_sspreq: Attrs |= Attribute::StackProtectReq; break;
case lltok::kw_noredzone: Attrs |= Attribute::NoRedZone; break;
case lltok::kw_noimplicitfloat: Attrs |= Attribute::NoImplicitFloat; break;
Add support for naked functions llvm-svn: 76198
2009-07-18 02:07:26 +08:00
case lltok::kw_naked: Attrs |= Attribute::Naked; break;
Add a new function attribute, nonlazybind, which inhibits lazy-loading optimizations when emitting calls to the function; instead those calls may use faster relocations which require the function to be immediately resolved upon loading the dynamic object featuring the call. This is useful when it is known that the function will be called frequently and pervasively and therefore there is no merit in delaying binding of the function. Currently only implemented for x86-64, where it turns into a call through the global offset table. Patch by Dan Gohman, who assures me that he's going to add LangRef documentation for this once it's committed. llvm-svn: 133080
2011-06-16 04:36:13 +08:00
case lltok::kw_nonlazybind: Attrs |= Attribute::NonLazyBind; break;
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
case lltok::kw_address_safety: Attrs |= Attribute::AddressSafety; break;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Add a new function attribute, 'alignstack'. It will indicate (when the backends implement support for it) that the stack should be forcibly realigned in the prologue (and the process reversed in the epilogue). llvm-svn: 95945
2010-02-12 08:31:15 +08:00
case lltok::kw_alignstack: {
unsigned Alignment;
if (ParseOptionalStackAlignment(Alignment))
return true;
Attrs |= Attribute::constructStackAlignmentFromInt(Alignment);
continue;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_align: {
unsigned Alignment;
if (ParseOptionalAlignment(Alignment))
return true;
Attrs |= Attribute::constructAlignmentFromInt(Alignment);
continue;
}
Add a new function attribute, 'alignstack'. It will indicate (when the backends implement support for it) that the stack should be forcibly realigned in the prologue (and the process reversed in the epilogue). llvm-svn: 95945
2010-02-12 08:31:15 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Lex.Lex();
}
}
/// ParseOptionalLinkage
/// ::= /*empty*/
Add the private linkage. llvm-svn: 62279
2009-01-16 04:18:42 +08:00
/// ::= 'private'
Add plumbing for the `linker_private' linkage type. This type is meant for "private" symbols which the assember shouldn't strip, but which the linker may remove after evaluation. This is mostly useful for Objective-C metadata. This is plumbing, so we don't have a use of it yet. More to come, etc. llvm-svn: 76385
2009-07-20 09:03:30 +08:00
/// ::= 'linker_private'
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
/// ::= 'linker_private_weak'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= 'internal'
/// ::= 'weak'
Introduce new linkage types linkonce_odr, weak_odr, common_odr and extern_weak_odr. These are the same as the non-odr versions, except that they indicate that the global will only be overridden by an *equivalent* global. In C, a function with weak linkage can be overridden by a function which behaves completely differently. This means that IP passes have to skip weak functions, since any deductions made from the function definition might be wrong, since the definition could be replaced by something completely different at link time. This is not allowed in C++, thanks to the ODR (One-Definition-Rule): if a function is replaced by another at link-time, then the new function must be the same as the original function. If a language knows that a function or other global can only be overridden by an equivalent global, it can give it the weak_odr linkage type, and the optimizers will understand that it is alright to make deductions based on the function body. The code generators on the other hand map weak and weak_odr linkage to the same thing. llvm-svn: 66339
2009-03-07 23:45:40 +08:00
/// ::= 'weak_odr'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= 'linkonce'
Introduce new linkage types linkonce_odr, weak_odr, common_odr and extern_weak_odr. These are the same as the non-odr versions, except that they indicate that the global will only be overridden by an *equivalent* global. In C, a function with weak linkage can be overridden by a function which behaves completely differently. This means that IP passes have to skip weak functions, since any deductions made from the function definition might be wrong, since the definition could be replaced by something completely different at link time. This is not allowed in C++, thanks to the ODR (One-Definition-Rule): if a function is replaced by another at link-time, then the new function must be the same as the original function. If a language knows that a function or other global can only be overridden by an equivalent global, it can give it the weak_odr linkage type, and the optimizers will understand that it is alright to make deductions based on the function body. The code generators on the other hand map weak and weak_odr linkage to the same thing. llvm-svn: 66339
2009-03-07 23:45:40 +08:00
/// ::= 'linkonce_odr'
Change the `linker_private_weak_def_auto' linkage to `linkonce_odr_auto_hide' to make it more consistent with its intended semantics. The `linker_private_weak_def_auto' linkage type was meant to automatically hide globals which never had their addresses taken. It has nothing to do with the `linker_private' linkage type, which outputs the symbols with a `l' (ell) prefix among other things. The intended semantic is more like the `linkonce_odr' linkage type. Change the name of the linkage type to `linkonce_odr_auto_hide'. And therefore changing the semantics so that it produces the correct output for the linker. Note: The old linkage name `linker_private_weak_def_auto' will still parse but is not a synonym for `linkonce_odr_auto_hide'. This should be removed in 4.0. <rdar://problem/11754934> llvm-svn: 162114
2012-08-18 02:33:14 +08:00
/// ::= 'linkonce_odr_auto_hide'
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
/// ::= 'available_externally'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= 'appending'
/// ::= 'dllexport'
/// ::= 'common'
/// ::= 'dllimport'
/// ::= 'extern_weak'
/// ::= 'external'
bool LLParser::ParseOptionalLinkage(unsigned &Res, bool &HasLinkage) {
HasLinkage = false;
switch (Lex.getKind()) {
Add plumbing for the `linker_private' linkage type. This type is meant for "private" symbols which the assember shouldn't strip, but which the linker may remove after evaluation. This is mostly useful for Objective-C metadata. This is plumbing, so we don't have a use of it yet. More to come, etc. llvm-svn: 76385
2009-07-20 09:03:30 +08:00
default: Res=GlobalValue::ExternalLinkage; return false;
case lltok::kw_private: Res = GlobalValue::PrivateLinkage; break;
case lltok::kw_linker_private: Res = GlobalValue::LinkerPrivateLinkage; break;
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
case lltok::kw_linker_private_weak:
Res = GlobalValue::LinkerPrivateWeakLinkage;
break;
Add plumbing for the `linker_private' linkage type. This type is meant for "private" symbols which the assember shouldn't strip, but which the linker may remove after evaluation. This is mostly useful for Objective-C metadata. This is plumbing, so we don't have a use of it yet. More to come, etc. llvm-svn: 76385
2009-07-20 09:03:30 +08:00
case lltok::kw_internal: Res = GlobalValue::InternalLinkage; break;
case lltok::kw_weak: Res = GlobalValue::WeakAnyLinkage; break;
case lltok::kw_weak_odr: Res = GlobalValue::WeakODRLinkage; break;
case lltok::kw_linkonce: Res = GlobalValue::LinkOnceAnyLinkage; break;
case lltok::kw_linkonce_odr: Res = GlobalValue::LinkOnceODRLinkage; break;
Change the `linker_private_weak_def_auto' linkage to `linkonce_odr_auto_hide' to make it more consistent with its intended semantics. The `linker_private_weak_def_auto' linkage type was meant to automatically hide globals which never had their addresses taken. It has nothing to do with the `linker_private' linkage type, which outputs the symbols with a `l' (ell) prefix among other things. The intended semantic is more like the `linkonce_odr' linkage type. Change the name of the linkage type to `linkonce_odr_auto_hide'. And therefore changing the semantics so that it produces the correct output for the linker. Note: The old linkage name `linker_private_weak_def_auto' will still parse but is not a synonym for `linkonce_odr_auto_hide'. This should be removed in 4.0. <rdar://problem/11754934> llvm-svn: 162114
2012-08-18 02:33:14 +08:00
case lltok::kw_linkonce_odr_auto_hide:
case lltok::kw_linker_private_weak_def_auto: // FIXME: For backwards compat.
Res = GlobalValue::LinkOnceODRAutoHideLinkage;
break;
Add a new "available_externally" linkage type. This is intended to support C99 inline, GNU extern inline, etc. Related bugzilla's include PR3517, PR3100, & PR2933. Nothing uses this yet, but it appears to work. llvm-svn: 68940
2009-04-13 13:44:34 +08:00
case lltok::kw_available_externally:
Res = GlobalValue::AvailableExternallyLinkage;
break;
Add plumbing for the `linker_private' linkage type. This type is meant for "private" symbols which the assember shouldn't strip, but which the linker may remove after evaluation. This is mostly useful for Objective-C metadata. This is plumbing, so we don't have a use of it yet. More to come, etc. llvm-svn: 76385
2009-07-20 09:03:30 +08:00
case lltok::kw_appending: Res = GlobalValue::AppendingLinkage; break;
case lltok::kw_dllexport: Res = GlobalValue::DLLExportLinkage; break;
case lltok::kw_common: Res = GlobalValue::CommonLinkage; break;
case lltok::kw_dllimport: Res = GlobalValue::DLLImportLinkage; break;
case lltok::kw_extern_weak: Res = GlobalValue::ExternalWeakLinkage; break;
case lltok::kw_external: Res = GlobalValue::ExternalLinkage; break;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Lex.Lex();
HasLinkage = true;
return false;
}
/// ParseOptionalVisibility
/// ::= /*empty*/
/// ::= 'default'
/// ::= 'hidden'
/// ::= 'protected'
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
///
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool LLParser::ParseOptionalVisibility(unsigned &Res) {
switch (Lex.getKind()) {
default: Res = GlobalValue::DefaultVisibility; return false;
case lltok::kw_default: Res = GlobalValue::DefaultVisibility; break;
case lltok::kw_hidden: Res = GlobalValue::HiddenVisibility; break;
case lltok::kw_protected: Res = GlobalValue::ProtectedVisibility; break;
}
Lex.Lex();
return false;
}
/// ParseOptionalCallingConv
/// ::= /*empty*/
/// ::= 'ccc'
/// ::= 'fastcc'
/// ::= 'coldcc'
/// ::= 'x86_stdcallcc'
/// ::= 'x86_fastcallcc'
Add support for thiscall calling convention. Patch by Charles Davis and Steven Watanabe! llvm-svn: 103902
2010-05-16 17:08:45 +08:00
/// ::= 'x86_thiscallcc'
Address review comments: add 3 ARM calling conventions. Dispatch C calling conv. to one of these conventions based on target triple and subtarget features. llvm-svn: 73530
2009-06-17 02:50:49 +08:00
/// ::= 'arm_apcscc'
/// ::= 'arm_aapcscc'
/// ::= 'arm_aapcs_vfpcc'
Add MSP430 interrupt calling conv. No functionality change yet. llvm-svn: 90738
2009-12-07 10:27:35 +08:00
/// ::= 'msp430_intrcc'
Add ret instruction to PTX backend llvm-svn: 114788
2010-09-25 15:46:17 +08:00
/// ::= 'ptx_kernel'
/// ::= 'ptx_device'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= 'cc' UINT
Address review comments: add 3 ARM calling conventions. Dispatch C calling conv. to one of these conventions based on target triple and subtarget features. llvm-svn: 73530
2009-06-17 02:50:49 +08:00
///
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
bool LLParser::ParseOptionalCallingConv(CallingConv::ID &CC) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
switch (Lex.getKind()) {
default: CC = CallingConv::C; return false;
case lltok::kw_ccc: CC = CallingConv::C; break;
case lltok::kw_fastcc: CC = CallingConv::Fast; break;
case lltok::kw_coldcc: CC = CallingConv::Cold; break;
case lltok::kw_x86_stdcallcc: CC = CallingConv::X86_StdCall; break;
case lltok::kw_x86_fastcallcc: CC = CallingConv::X86_FastCall; break;
Add support for thiscall calling convention. Patch by Charles Davis and Steven Watanabe! llvm-svn: 103902
2010-05-16 17:08:45 +08:00
case lltok::kw_x86_thiscallcc: CC = CallingConv::X86_ThisCall; break;
Address review comments: add 3 ARM calling conventions. Dispatch C calling conv. to one of these conventions based on target triple and subtarget features. llvm-svn: 73530
2009-06-17 02:50:49 +08:00
case lltok::kw_arm_apcscc: CC = CallingConv::ARM_APCS; break;
case lltok::kw_arm_aapcscc: CC = CallingConv::ARM_AAPCS; break;
case lltok::kw_arm_aapcs_vfpcc:CC = CallingConv::ARM_AAPCS_VFP; break;
Add MSP430 interrupt calling conv. No functionality change yet. llvm-svn: 90738
2009-12-07 10:27:35 +08:00
case lltok::kw_msp430_intrcc: CC = CallingConv::MSP430_INTR; break;
Add ret instruction to PTX backend llvm-svn: 114788
2010-09-25 15:46:17 +08:00
case lltok::kw_ptx_kernel: CC = CallingConv::PTX_Kernel; break;
case lltok::kw_ptx_device: CC = CallingConv::PTX_Device; break;
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
case lltok::kw_cc: {
unsigned ArbitraryCC;
Lex.Lex();
More dead code removal (using -Wunreachable-code) llvm-svn: 148578
2012-01-21 05:51:11 +08:00
if (ParseUInt32(ArbitraryCC))
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
return true;
More dead code removal (using -Wunreachable-code) llvm-svn: 148578
2012-01-21 05:51:11 +08:00
CC = static_cast<CallingConv::ID>(ArbitraryCC);
return false;
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
return false;
}
rename ParseOptionalCustomMetadata -> ParseInstructionMetadata, and make it non-optional. This fixes the bug where we'd accept and ignore a spurious comma after some instructions. llvm-svn: 92300
2009-12-30 13:31:19 +08:00
/// ParseInstructionMetadata
Each instruction is allowed to have *multiple* different metadata objects on them. Though the entire compiler supports this, the asmparser didn't. llvm-svn: 92270
2009-12-30 05:25:40 +08:00
/// ::= !dbg !42 (',' !dbg !57)*
Give ParseInstructionMetadata access to the PerFunctionState object. This is in preparation for generalizing its parsing of function-local values. llvm-svn: 111893
2010-08-24 10:05:17 +08:00
bool LLParser::ParseInstructionMetadata(Instruction *Inst,
PerFunctionState *PFS) {
rename ParseOptionalCustomMetadata -> ParseInstructionMetadata, and make it non-optional. This fixes the bug where we'd accept and ignore a spurious comma after some instructions. llvm-svn: 92300
2009-12-30 13:31:19 +08:00
do {
if (Lex.getKind() != lltok::MetadataVar)
return TokError("expected metadata after comma");
Parse custom metadata attached with an instruction. llvm-svn: 83033
2009-09-29 08:01:14 +08:00
Each instruction is allowed to have *multiple* different metadata objects on them. Though the entire compiler supports this, the asmparser didn't. llvm-svn: 92270
2009-12-30 05:25:40 +08:00
std::string Name = Lex.getStrVal();
Push StringRefs through the metadata interface. llvm-svn: 145934
2011-12-06 19:50:26 +08:00
unsigned MDK = M->getMDKindID(Name);
Each instruction is allowed to have *multiple* different metadata objects on them. Though the entire compiler supports this, the asmparser didn't. llvm-svn: 92270
2009-12-30 05:25:40 +08:00
Lex.Lex();
clean up after metadata changes. llvm-svn: 84463
2009-10-19 13:31:10 +08:00
change ParseMDString and ParseMDNode to take arguments of the right type. This exposed a raft of other problems, which I'll deal with in subsequent patches. llvm-svn: 92273
2009-12-30 05:53:55 +08:00
MDNode *Node;
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
SMLoc Loc = Lex.getLoc();
Extend function-local metadata to be usable as attachments. llvm-svn: 111895
2010-08-24 10:24:03 +08:00
if (ParseToken(lltok::exclaim, "expected '!' here"))
rename lltok::Metadata -> lltok::exclaim. We name tokens after their syntactic form, not their semantic form. llvm-svn: 92294
2009-12-30 12:56:59 +08:00
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Add a comment explaining why this code doesn't just call ParseMetadataValue. llvm-svn: 111914
2010-08-24 22:35:45 +08:00
// This code is similar to that of ParseMetadataValue, however it needs to
// have special-case code for a forward reference; see the comments on
// ForwardRefInstMetadata for details. Also, MDStrings are not supported
// at the top level here.
Extend function-local metadata to be usable as attachments. llvm-svn: 111895
2010-08-24 10:24:03 +08:00
if (Lex.getKind() == lltok::lbrace) {
ValID ID;
if (ParseMetadataListValue(ID, PFS))
return true;
assert(ID.Kind == ValID::t_MDNode);
Inst->setMetadata(MDK, ID.MDNodeVal);
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
} else {
Pacify a noisy compiler, and sink this variable declaration closer to its uses. llvm-svn: 115206
2010-10-01 05:04:13 +08:00
unsigned NodeID = 0;
Extend function-local metadata to be usable as attachments. llvm-svn: 111895
2010-08-24 10:24:03 +08:00
if (ParseMDNodeID(Node, NodeID))
return true;
if (Node) {
// If we got the node, add it to the instruction.
Inst->setMetadata(MDK, Node);
} else {
MDRef R = { Loc, MDK, NodeID };
// Otherwise, remember that this should be resolved later.
ForwardRefInstMetadata[Inst].push_back(R);
}
rewrite handling of forward ref'd instruction metadata to used deferred resolution instead of creating a temporary node + rauw. There is no reason to create the temporary mdnode, then do rauw, then destroy it. llvm-svn: 100086
2010-04-01 13:14:45 +08:00
}
Each instruction is allowed to have *multiple* different metadata objects on them. Though the entire compiler supports this, the asmparser didn't. llvm-svn: 92270
2009-12-30 05:25:40 +08:00
// If this is the end of the list, we're done.
rename ParseOptionalCustomMetadata -> ParseInstructionMetadata, and make it non-optional. This fixes the bug where we'd accept and ignore a spurious comma after some instructions. llvm-svn: 92300
2009-12-30 13:31:19 +08:00
} while (EatIfPresent(lltok::comma));
return false;
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseOptionalAlignment
/// ::= /* empty */
/// ::= 'align' 4
bool LLParser::ParseOptionalAlignment(unsigned &Alignment) {
Alignment = 0;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (!EatIfPresent(lltok::kw_align))
return false;
PR3281:crash00.ll: produce this diagnostic instead of crashing: @t = global i8 0, align 3 ^ llvm-svn: 61675
2009-01-05 15:46:05 +08:00
LocTy AlignLoc = Lex.getLoc();
if (ParseUInt32(Alignment)) return true;
if (!isPowerOf2_32(Alignment))
return Error(AlignLoc, "alignment is not a power of two");
Move MaximumAlignment to be a member of the Value class. llvm-svn: 109891
2010-07-31 05:07:05 +08:00
if (Alignment > Value::MaximumAlignment)
Define a maximum supported alignment value for load, store, and alloca instructions (constrained by their internal encoding), and add error checking for it. Fix an instcombine bug which generated huge alignment values (null is infinitely aligned). This fixes undefined behavior noticed by John Regehr. llvm-svn: 109643
2010-07-29 04:12:04 +08:00
return Error(AlignLoc, "huge alignments are not supported yet");
PR3281:crash00.ll: produce this diagnostic instead of crashing: @t = global i8 0, align 3 ^ llvm-svn: 61675
2009-01-05 15:46:05 +08:00
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
/// ParseOptionalCommaAlign
/// ::=
/// ::= ',' align 4
///
/// This returns with AteExtraComma set to true if it ate an excess comma at the
/// end.
bool LLParser::ParseOptionalCommaAlign(unsigned &Alignment,
bool &AteExtraComma) {
AteExtraComma = false;
while (EatIfPresent(lltok::comma)) {
// Metadata at the end is an early exit.
rename NamedOrCustomMD -> MetadataVar to follow conventions of all the rest of the code. llvm-svn: 92295
2009-12-30 13:02:06 +08:00
if (Lex.getKind() == lltok::MetadataVar) {
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
AteExtraComma = true;
return false;
}
reject invalid comma stuff with a message. We reject the case in PR6888 with: $ llvm-as t.ll llvm-as: t.ll:2:29: error: expected metadata or 'align' store <3 x i32> %x, i32 1, i32 1>, <3 x i32>* %p ^ instead of: $ llvm-as t.ll llvm-as: llvm-svn: 102154
2010-04-23 08:50:50 +08:00
if (Lex.getKind() != lltok::kw_align)
return Error(Lex.getLoc(), "expected metadata or 'align'");
AlignLoc is never used for anything - zap it (gcc-4.6 warning). llvm-svn: 117025
2010-10-22 00:07:10 +08:00
reject invalid comma stuff with a message. We reject the case in PR6888 with: $ llvm-as t.ll llvm-as: t.ll:2:29: error: expected metadata or 'align' store <3 x i32> %x, i32 1, i32 1>, <3 x i32>* %p ^ instead of: $ llvm-as t.ll llvm-as: llvm-svn: 102154
2010-04-23 08:50:50 +08:00
if (ParseOptionalAlignment(Alignment)) return true;
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Initial implementation of 'fence' instruction, the new C++0x-style replacement for llvm.memory.barrier. This is just a LangRef entry and reading/writing/memory representation; optimizer+codegen support coming soon. llvm-svn: 136009
2011-07-26 07:16:38 +08:00
/// ParseScopeAndOrdering
/// if isAtomic: ::= 'singlethread'? AtomicOrdering
/// else: ::=
///
/// This sets Scope and Ordering to the parsed values.
bool LLParser::ParseScopeAndOrdering(bool isAtomic, SynchronizationScope &Scope,
AtomicOrdering &Ordering) {
if (!isAtomic)
return false;
Scope = CrossThread;
if (EatIfPresent(lltok::kw_singlethread))
Scope = SingleThread;
switch (Lex.getKind()) {
default: return TokError("Expected ordering on atomic instruction");
case lltok::kw_unordered: Ordering = Unordered; break;
case lltok::kw_monotonic: Ordering = Monotonic; break;
case lltok::kw_acquire: Ordering = Acquire; break;
case lltok::kw_release: Ordering = Release; break;
case lltok::kw_acq_rel: Ordering = AcquireRelease; break;
case lltok::kw_seq_cst: Ordering = SequentiallyConsistent; break;
}
Lex.Lex();
return false;
}
Add a new function attribute, 'alignstack'. It will indicate (when the backends implement support for it) that the stack should be forcibly realigned in the prologue (and the process reversed in the epilogue). llvm-svn: 95945
2010-02-12 08:31:15 +08:00
/// ParseOptionalStackAlignment
/// ::= /* empty */
/// ::= 'alignstack' '(' 4 ')'
bool LLParser::ParseOptionalStackAlignment(unsigned &Alignment) {
Alignment = 0;
if (!EatIfPresent(lltok::kw_alignstack))
return false;
LocTy ParenLoc = Lex.getLoc();
if (!EatIfPresent(lltok::lparen))
return Error(ParenLoc, "expected '('");
LocTy AlignLoc = Lex.getLoc();
if (ParseUInt32(Alignment)) return true;
ParenLoc = Lex.getLoc();
if (!EatIfPresent(lltok::rparen))
return Error(ParenLoc, "expected ')'");
if (!isPowerOf2_32(Alignment))
return Error(AlignLoc, "stack alignment is not a power of two");
return false;
}
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
reimplement insertvalue/extractvalue metadata handling to not blindly accept invalid input. Actually add a testcase. llvm-svn: 92297
2009-12-30 13:14:00 +08:00
/// ParseIndexList - This parses the index list for an insert/extractvalue
/// instruction. This sets AteExtraComma in the case where we eat an extra
/// comma at the end of the line and find that it is followed by metadata.
/// Clients that don't allow metadata can call the version of this function that
/// only takes one argument.
///
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseIndexList
/// ::= (',' uint32)+
reimplement insertvalue/extractvalue metadata handling to not blindly accept invalid input. Actually add a testcase. llvm-svn: 92297
2009-12-30 13:14:00 +08:00
///
bool LLParser::ParseIndexList(SmallVectorImpl<unsigned> &Indices,
bool &AteExtraComma) {
AteExtraComma = false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Lex.getKind() != lltok::comma)
return TokError("expected ',' as start of index list");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
while (EatIfPresent(lltok::comma)) {
reimplement insertvalue/extractvalue metadata handling to not blindly accept invalid input. Actually add a testcase. llvm-svn: 92297
2009-12-30 13:14:00 +08:00
if (Lex.getKind() == lltok::MetadataVar) {
AteExtraComma = true;
return false;
}
Silence compiler warning. llvm-svn: 115107
2010-09-30 07:32:20 +08:00
unsigned Idx = 0;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseUInt32(Idx)) return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Indices.push_back(Idx);
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
//===----------------------------------------------------------------------===//
// Type Parsing.
//===----------------------------------------------------------------------===//
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// ParseType - Parse a type.
bool LLParser::ParseType(Type *&Result, bool AllowVoid) {
SMLoc TypeLoc = Lex.getLoc();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
switch (Lex.getKind()) {
default:
return TokError("expected type");
case lltok::Type:
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Type ::= 'float' | 'void' (etc)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Result = Lex.getTyVal();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Lex.Lex();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
break;
case lltok::lbrace:
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Type ::= StructType
if (ParseAnonStructType(Result, false))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
break;
case lltok::lsquare:
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Type ::= '[' ... ']'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex(); // eat the lsquare.
if (ParseArrayVectorType(Result, false))
return true;
break;
case lltok::less: // Either vector or packed struct.
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Type ::= '<' ... '>'
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Lex.Lex();
if (Lex.getKind() == lltok::lbrace) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseAnonStructType(Result, true) ||
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
ParseToken(lltok::greater, "expected '>' at end of packed struct"))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
} else if (ParseArrayVectorType(Result, true))
return true;
break;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
case lltok::LocalVar: {
// Type ::= %foo
std::pair<Type*, LocTy> &Entry = NamedTypes[Lex.getStrVal()];
// If the type hasn't been defined yet, create a forward definition and
// remember where that forward def'n was seen (in case it never is defined).
if (Entry.first == 0) {
switch to use the new api for structtypes. llvm-svn: 137480
2011-08-13 02:06:37 +08:00
Entry.first = StructType::create(Context, Lex.getStrVal());
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Entry.second = Lex.getLoc();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Result = Entry.first;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
break;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
case lltok::LocalVarID: {
// Type ::= %4
if (Lex.getUIntVal() >= NumberedTypes.size())
NumberedTypes.resize(Lex.getUIntVal()+1);
std::pair<Type*, LocTy> &Entry = NumberedTypes[Lex.getUIntVal()];
// If the type hasn't been defined yet, create a forward definition and
// remember where that forward def'n was seen (in case it never is defined).
if (Entry.first == 0) {
switch to use the new api for structtypes. llvm-svn: 137480
2011-08-13 02:06:37 +08:00
Entry.first = StructType::create(Context);
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Entry.second = Lex.getLoc();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Result = Entry.first;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
break;
}
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
// Parse the type suffixes.
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
while (1) {
switch (Lex.getKind()) {
// End of type.
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
default:
if (!AllowVoid && Result->isVoidTy())
return Error(TypeLoc, "void type only allowed for function results");
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Type ::= Type '*'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::star:
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (Result->isLabelTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return TokError("basic block pointers are invalid");
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (Result->isVoidTy())
return TokError("pointers to void are invalid - use i8* instead");
if (!PointerType::isValidElementType(Result))
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
return TokError("pointer to this type is invalid");
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Result = PointerType::getUnqual(Result);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
break;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Type ::= Type 'addrspace' '(' uint32 ')' '*'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_addrspace: {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (Result->isLabelTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return TokError("basic block pointers are invalid");
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (Result->isVoidTy())
Use slightly more correct grammar in error messages. llvm-svn: 64148
2009-02-10 01:41:21 +08:00
return TokError("pointers to void are invalid; use i8* instead");
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!PointerType::isValidElementType(Result))
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
return TokError("pointer to this type is invalid");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
unsigned AddrSpace;
if (ParseOptionalAddrSpace(AddrSpace) ||
ParseToken(lltok::star, "expected '*' in address space"))
return true;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Result = PointerType::get(Result, AddrSpace);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
break;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// Types '(' ArgTypeListI ')' OptFuncAttrs
case lltok::lparen:
if (ParseFunctionType(Result))
return true;
break;
}
}
}
/// ParseParameterList
/// ::= '(' ')'
/// ::= '(' Arg (',' Arg)* ')'
/// Arg
/// ::= Type OptionalAttributes Value OptionalAttributes
bool LLParser::ParseParameterList(SmallVectorImpl<ParamInfo> &ArgList,
PerFunctionState &PFS) {
if (ParseToken(lltok::lparen, "expected '(' in call"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
while (Lex.getKind() != lltok::rparen) {
// If this isn't the first argument, we need a comma.
if (!ArgList.empty() &&
ParseToken(lltok::comma, "expected ',' in argument list"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Parse the argument.
LocTy ArgLoc;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *ArgTy = 0;
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
Attributes ArgAttrs1;
Attributes ArgAttrs2;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *V;
Add ParseInlineMetadata() which can parses metadata that refers to an instruction. Extend ParseParameterList() to use this new function so that calls to llvm.dbg.declare can pass inline metadata llvm-svn: 90497
2009-12-04 07:40:58 +08:00
if (ParseType(ArgTy, ArgLoc))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Add ParseInlineMetadata() which can parses metadata that refers to an instruction. Extend ParseParameterList() to use this new function so that calls to llvm.dbg.declare can pass inline metadata llvm-svn: 90497
2009-12-04 07:40:58 +08:00
split t_Metadata into t_MDNode and t_MDString, eliminating some unsafe casting. llvm-svn: 92287
2009-12-30 10:11:14 +08:00
// Otherwise, handle normal operands.
Remove old backwards compatibility support from the parser for autoupgrading the old malloc/free instructions, and for 'sext' and 'zext' as function attributes (they are spelled signext/zeroext now), and support for result value attributes being specified after a function. Additionally, diagnose invalid attributes on functions with an error message instead of an abort in the verifier. llvm-svn: 133229
2011-06-17 11:16:47 +08:00
if (ParseOptionalAttrs(ArgAttrs1, 0) || ParseValue(ArgTy, V, PFS))
split t_Metadata into t_MDNode and t_MDString, eliminating some unsafe casting. llvm-svn: 92287
2009-12-30 10:11:14 +08:00
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ArgList.push_back(ParamInfo(ArgLoc, V, ArgAttrs1|ArgAttrs2));
}
Lex.Lex(); // Lex the ')'.
return false;
}
Reject PR3281:accepted03.ll with: llvm-as: accepted03.ll:1:35: invalid unresolved type up reference declare void @r({ \7, opaque, \10 } %su) ^ llvm-svn: 61725
2009-01-06 02:34:07 +08:00
/// ParseArgumentList - Parse the argument list for a function type or function
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// prototype.
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= '(' ArgTypeListI ')'
/// ArgTypeListI
/// ::= /*empty*/
/// ::= '...'
/// ::= ArgTypeList ',' '...'
/// ::= ArgType (',' ArgType)*
Reject PR3281:accepted03.ll with: llvm-as: accepted03.ll:1:35: invalid unresolved type up reference declare void @r({ \7, opaque, \10 } %su) ^ llvm-svn: 61725
2009-01-06 02:34:07 +08:00
///
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
bool LLParser::ParseArgumentList(SmallVectorImpl<ArgInfo> &ArgList,
bool &isVarArg){
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
isVarArg = false;
assert(Lex.getKind() == lltok::lparen);
Lex.Lex(); // eat the (.
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Lex.getKind() == lltok::rparen) {
// empty
} else if (Lex.getKind() == lltok::dotdotdot) {
isVarArg = true;
Lex.Lex();
} else {
LocTy TypeLoc = Lex.getLoc();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *ArgTy = 0;
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
Attributes Attrs;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
std::string Name;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseType(ArgTy) ||
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
ParseOptionalAttrs(Attrs, 0)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (ArgTy->isVoidTy())
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
return Error(TypeLoc, "argument can not have void type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Remove support for using "foo" as symbols instead of %"foo". This is ancient syntax and has been long obsolete. As usual, updating the tests is the nasty part of this. llvm-svn: 133242
2011-06-17 14:36:20 +08:00
if (Lex.getKind() == lltok::LocalVar) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Name = Lex.getStrVal();
Lex.Lex();
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
if (!FunctionType::isValidArgumentType(ArgTy))
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
return Error(TypeLoc, "invalid type for function argument");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ArgList.push_back(ArgInfo(TypeLoc, ArgTy, Attrs, Name));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
while (EatIfPresent(lltok::comma)) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Handle ... at end of arg list.
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (EatIfPresent(lltok::dotdotdot)) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
isVarArg = true;
break;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Otherwise must be an argument type.
TypeLoc = Lex.getLoc();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseType(ArgTy) || ParseOptionalAttrs(Attrs, 0)) return true;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (ArgTy->isVoidTy())
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
return Error(TypeLoc, "argument can not have void type");
Remove support for using "foo" as symbols instead of %"foo". This is ancient syntax and has been long obsolete. As usual, updating the tests is the nasty part of this. llvm-svn: 133242
2011-06-17 14:36:20 +08:00
if (Lex.getKind() == lltok::LocalVar) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Name = Lex.getStrVal();
Lex.Lex();
} else {
Name = "";
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!ArgTy->isFirstClassType())
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
return Error(TypeLoc, "invalid type for function argument");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ArgList.push_back(ArgInfo(TypeLoc, ArgTy, Attrs, Name));
}
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
return ParseToken(lltok::rparen, "expected ')' at end of argument list");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseFunctionType
/// ::= Type ArgumentList OptionalAttrs
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
bool LLParser::ParseFunctionType(Type *&Result) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
assert(Lex.getKind() == lltok::lparen);
diagnose PR3281:crash02.ll with: llvm-as: crash02.ll:1:62: invalid function return type declare { <{ <{}>, void ([1898 x { void ()* }], opaque, ...) (), fp128 * }>, opaque } @t () ^ llvm-svn: 61679
2009-01-05 16:04:33 +08:00
if (!FunctionType::isValidReturnType(Result))
return TokError("invalid function return type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
SmallVector<ArgInfo, 8> ArgList;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool isVarArg;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseArgumentList(ArgList, isVarArg))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Reject names on the arguments lists.
for (unsigned i = 0, e = ArgList.size(); i != e; ++i) {
if (!ArgList[i].Name.empty())
return Error(ArgList[i].Loc, "argument name invalid in function type");
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
if (ArgList[i].Attrs)
Stop accepting and ignoring attributes in function types. Attributes are applied to functions and call/invokes, not to types. llvm-svn: 133266
2011-06-18 01:37:13 +08:00
return Error(ArgList[i].Loc,
"argument attributes invalid in function type");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Second attempt at de-constifying LLVM Types in FunctionType::get(), StructType::get() and TargetData::getIntPtrType(). llvm-svn: 134982
2011-07-12 22:06:48 +08:00
SmallVector<Type*, 16> ArgListTy;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
for (unsigned i = 0, e = ArgList.size(); i != e; ++i)
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
ArgListTy.push_back(ArgList[i].Ty);
Result = FunctionType::get(Result, ArgListTy, isVarArg);
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// ParseAnonStructType - Parse an anonymous struct type, which is inlined into
/// other structs.
bool LLParser::ParseAnonStructType(Type *&Result, bool Packed) {
SmallVector<Type*, 8> Elts;
if (ParseStructBody(Elts)) return true;
Result = StructType::get(Context, Elts, Packed);
return false;
}
/// ParseStructDefinition - Parse a struct in a 'type' definition.
bool LLParser::ParseStructDefinition(SMLoc TypeLoc, StringRef Name,
std::pair<Type*, LocTy> &Entry,
Type *&ResultTy) {
// If the type was already defined, diagnose the redefinition.
if (Entry.first && !Entry.second.isValid())
return Error(TypeLoc, "redefinition of type");
// If we have opaque, just return without filling in the definition for the
// struct. This counts as a definition as far as the .ll file goes.
if (EatIfPresent(lltok::kw_opaque)) {
// This type is being defined, so clear the location to indicate this.
Entry.second = SMLoc();
// If this type number has never been uttered, create it.
if (Entry.first == 0)
switch to use the new api for structtypes. llvm-svn: 137480
2011-08-13 02:06:37 +08:00
Entry.first = StructType::create(Context, Name);
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
ResultTy = Entry.first;
return false;
}
// If the type starts with '<', then it is either a packed struct or a vector.
bool isPacked = EatIfPresent(lltok::less);
// If we don't have a struct, then we have a random type alias, which we
// accept for compatibility with old files. These types are not allowed to be
// forward referenced and not allowed to be recursive.
if (Lex.getKind() != lltok::lbrace) {
if (Entry.first)
return Error(TypeLoc, "forward references to non-struct type");
ResultTy = 0;
if (isPacked)
return ParseArrayVectorType(ResultTy, true);
return ParseType(ResultTy);
}
// This type is being defined, so clear the location to indicate this.
Entry.second = SMLoc();
// If this type number has never been uttered, create it.
if (Entry.first == 0)
switch to use the new api for structtypes. llvm-svn: 137480
2011-08-13 02:06:37 +08:00
Entry.first = StructType::create(Context, Name);
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
StructType *STy = cast<StructType>(Entry.first);
SmallVector<Type*, 8> Body;
if (ParseStructBody(Body) ||
(isPacked && ParseToken(lltok::greater, "expected '>' in packed struct")))
return true;
STy->setBody(Body, isPacked);
ResultTy = STy;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseStructType: Handles packed and unpacked types. </> parsed elsewhere.
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// StructType
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= '{' '}'
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// ::= '{' Type (',' Type)* '}'
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= '<' '{' '}' '>'
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// ::= '<' '{' Type (',' Type)* '}' '>'
bool LLParser::ParseStructBody(SmallVectorImpl<Type*> &Body) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
assert(Lex.getKind() == lltok::lbrace);
Lex.Lex(); // Consume the '{'
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
// Handle the empty struct.
if (EatIfPresent(lltok::rbrace))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
LocTy EltTyLoc = Lex.getLoc();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Ty = 0;
if (ParseType(Ty)) return true;
Body.push_back(Ty);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!StructType::isValidElementType(Ty))
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
return Error(EltTyLoc, "invalid element type for struct");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
while (EatIfPresent(lltok::comma)) {
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
EltTyLoc = Lex.getLoc();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseType(Ty)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!StructType::isValidElementType(Ty))
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
return Error(EltTyLoc, "invalid element type for struct");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Body.push_back(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
return ParseToken(lltok::rbrace, "expected '}' at end of struct");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseArrayVectorType - Parse an array or vector type, assuming the first
/// token has already been consumed.
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
/// Type
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ::= '[' APSINTVAL 'x' Types ']'
/// ::= '<' APSINTVAL 'x' Types '>'
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
bool LLParser::ParseArrayVectorType(Type *&Result, bool isVector) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Lex.getKind() != lltok::APSInt || Lex.getAPSIntVal().isSigned() ||
Lex.getAPSIntVal().getBitWidth() > 64)
return TokError("expected number in address space");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy SizeLoc = Lex.getLoc();
uint64_t Size = Lex.getAPSIntVal().getZExtValue();
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(lltok::kw_x, "expected 'x' after element count"))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy TypeLoc = Lex.getLoc();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *EltTy = 0;
if (ParseType(EltTy)) return true;
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(isVector ? lltok::greater : lltok::rsquare,
"expected end of sequential type"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (isVector) {
reject 0 element vectors with: @a = internal constant void bitcast(<0 x i8> <> to void) ^ Fixes PR3685 llvm-svn: 65698
2009-03-01 02:12:41 +08:00
if (Size == 0)
return Error(SizeLoc, "zero element vector is illegal");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if ((unsigned)Size != Size)
return Error(SizeLoc, "size too large for vector");
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
if (!VectorType::isValidElementType(EltTy))
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
return Error(TypeLoc,
"vector element type must be fp, integer or a pointer to these types");
Move types back to the 2.5 API. llvm-svn: 77516
2009-07-30 06:17:13 +08:00
Result = VectorType::get(EltTy, unsigned(Size));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else {
Create FunctionType::isValidArgumentType to go along with isValidReturnType. Also create isValidElementType for ArrayType, PointerType, StructType and VectorType. Make LLParser use them. This closes up some holes like an assertion failure on: %x = type {label} but largely doesn't change any semantics. The only thing we accept now which we didn't before is vectors of opaque type such as "<4 x opaque>". The opaque can be resolved to an int or float when linking. llvm-svn: 73016
2009-06-07 15:26:46 +08:00
if (!ArrayType::isValidElementType(EltTy))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(TypeLoc, "invalid array element type");
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Result = ArrayType::get(EltTy, Size);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
return false;
}
//===----------------------------------------------------------------------===//
// Function Semantic Analysis.
//===----------------------------------------------------------------------===//
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
LLParser::PerFunctionState::PerFunctionState(LLParser &p, Function &f,
int functionNumber)
: P(p), F(f), FunctionNumber(functionNumber) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Insert unnamed arguments into the NumberedVals list.
for (Function::arg_iterator AI = F.arg_begin(), E = F.arg_end();
AI != E; ++AI)
if (!AI->hasName())
NumberedVals.push_back(AI);
}
LLParser::PerFunctionState::~PerFunctionState() {
// If there were any forward referenced non-basicblock values, delete them.
for (std::map<std::string, std::pair<Value*, LocTy> >::iterator
I = ForwardRefVals.begin(), E = ForwardRefVals.end(); I != E; ++I)
if (!isa<BasicBlock>(I->second.first)) {
Use LLVMContext for generating UndefValue constants too! llvm-svn: 74703
2009-07-03 01:04:01 +08:00
I->second.first->replaceAllUsesWith(
Move more code back to 2.5 APIs. llvm-svn: 77635
2009-07-31 07:03:37 +08:00
UndefValue::get(I->second.first->getType()));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
delete I->second.first;
I->second.first = 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
for (std::map<unsigned, std::pair<Value*, LocTy> >::iterator
I = ForwardRefValIDs.begin(), E = ForwardRefValIDs.end(); I != E; ++I)
if (!isa<BasicBlock>(I->second.first)) {
Use LLVMContext for generating UndefValue constants too! llvm-svn: 74703
2009-07-03 01:04:01 +08:00
I->second.first->replaceAllUsesWith(
Move more code back to 2.5 APIs. llvm-svn: 77635
2009-07-31 07:03:37 +08:00
UndefValue::get(I->second.first->getType()));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
delete I->second.first;
I->second.first = 0;
}
}
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
bool LLParser::PerFunctionState::FinishFunction() {
// Check to see if someone took the address of labels in this block.
if (!P.ForwardRefBlockAddresses.empty()) {
ValID FunctionID;
if (!F.getName().empty()) {
FunctionID.Kind = ValID::t_GlobalName;
FunctionID.StrVal = F.getName();
} else {
FunctionID.Kind = ValID::t_GlobalID;
FunctionID.UIntVal = FunctionNumber;
}
std::map<ValID, std::vector<std::pair<ValID, GlobalValue*> > >::iterator
FRBAI = P.ForwardRefBlockAddresses.find(FunctionID);
if (FRBAI != P.ForwardRefBlockAddresses.end()) {
// Resolve all these references.
if (P.ResolveForwardRefBlockAddresses(&F, FRBAI->second, this))
return true;
P.ForwardRefBlockAddresses.erase(FRBAI);
}
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!ForwardRefVals.empty())
return P.Error(ForwardRefVals.begin()->second.second,
"use of undefined value '%" + ForwardRefVals.begin()->first +
"'");
if (!ForwardRefValIDs.empty())
return P.Error(ForwardRefValIDs.begin()->second.second,
"use of undefined value '%" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(ForwardRefValIDs.begin()->first) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
/// GetVal - Get a value with the specified name or ID, creating a
/// forward reference record if needed. This can return null if the value
/// exists but does not have the right type.
Value *LLParser::PerFunctionState::GetVal(const std::string &Name,
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
Type *Ty, LocTy Loc) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Look this name up in the normal function symbol table.
Value *Val = F.getValueSymbolTable().lookup(Name);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If this is a forward reference for the value, see if we already created a
// forward ref record.
if (Val == 0) {
std::map<std::string, std::pair<Value*, LocTy> >::iterator
I = ForwardRefVals.find(Name);
if (I != ForwardRefVals.end())
Val = I->second.first;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If we have the value in the symbol table or fwd-ref table, return it.
if (Val) {
if (Val->getType() == Ty) return Val;
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (Ty->isLabelTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
P.Error(Loc, "'%" + Name + "' is not a basic block");
else
P.Error(Loc, "'%" + Name + "' defined with type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Val->getType()) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Don't make placeholders with invalid type.
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!Ty->isFirstClassType() && !Ty->isLabelTy()) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
P.Error(Loc, "invalid use of a non-first-class type");
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Otherwise, create a new forward reference for this value and remember it.
Value *FwdVal;
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (Ty->isLabelTy())
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
FwdVal = BasicBlock::Create(F.getContext(), Name, &F);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
else
FwdVal = new Argument(Ty, Name);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefVals[Name] = std::make_pair(FwdVal, Loc);
return FwdVal;
}
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
Value *LLParser::PerFunctionState::GetVal(unsigned ID, Type *Ty,
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy Loc) {
// Look this name up in the normal function symbol table.
Value *Val = ID < NumberedVals.size() ? NumberedVals[ID] : 0;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If this is a forward reference for the value, see if we already created a
// forward ref record.
if (Val == 0) {
std::map<unsigned, std::pair<Value*, LocTy> >::iterator
I = ForwardRefValIDs.find(ID);
if (I != ForwardRefValIDs.end())
Val = I->second.first;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If we have the value in the symbol table or fwd-ref table, return it.
if (Val) {
if (Val->getType() == Ty) return Val;
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (Ty->isLabelTy())
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
P.Error(Loc, "'%" + Twine(ID) + "' is not a basic block");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
else
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
P.Error(Loc, "'%" + Twine(ID) + "' defined with type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Val->getType()) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!Ty->isFirstClassType() && !Ty->isLabelTy()) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
P.Error(Loc, "invalid use of a non-first-class type");
return 0;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Otherwise, create a new forward reference for this value and remember it.
Value *FwdVal;
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (Ty->isLabelTy())
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
FwdVal = BasicBlock::Create(F.getContext(), "", &F);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
else
FwdVal = new Argument(Ty);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefValIDs[ID] = std::make_pair(FwdVal, Loc);
return FwdVal;
}
/// SetInstName - After an instruction is parsed and inserted into its
/// basic block, this installs its name.
bool LLParser::PerFunctionState::SetInstName(int NameID,
const std::string &NameStr,
LocTy NameLoc, Instruction *Inst) {
// If this instruction has void type, it cannot have a name or ID specified.
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (Inst->getType()->isVoidTy()) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (NameID != -1 || !NameStr.empty())
return P.Error(NameLoc, "instructions returning void cannot have a name");
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If this was a numbered instruction, verify that the instruction is the
// expected value and resolve any forward references.
if (NameStr.empty()) {
// If neither a name nor an ID was specified, just use the next ID.
if (NameID == -1)
NameID = NumberedVals.size();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (unsigned(NameID) != NumberedVals.size())
return P.Error(NameLoc, "instruction expected to be numbered '%" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(NumberedVals.size()) + "'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
std::map<unsigned, std::pair<Value*, LocTy> >::iterator FI =
ForwardRefValIDs.find(NameID);
if (FI != ForwardRefValIDs.end()) {
if (FI->second.first->getType() != Inst->getType())
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
return P.Error(NameLoc, "instruction forward referenced with type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(FI->second.first->getType()) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
FI->second.first->replaceAllUsesWith(Inst);
plug another leak in LLParser::PerFunctionState::SetInstName() llvm-svn: 80792
2009-09-02 23:02:57 +08:00
delete FI->second.first;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefValIDs.erase(FI);
}
NumberedVals.push_back(Inst);
return false;
}
// Otherwise, the instruction had a name. Resolve forward refs and set it.
std::map<std::string, std::pair<Value*, LocTy> >::iterator
FI = ForwardRefVals.find(NameStr);
if (FI != ForwardRefVals.end()) {
if (FI->second.first->getType() != Inst->getType())
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
return P.Error(NameLoc, "instruction forward referenced with type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(FI->second.first->getType()) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
FI->second.first->replaceAllUsesWith(Inst);
plug memory leak in LLParser::PerFunctionState::SetInstName() by deleting a value after replacing it llvm-svn: 80790
2009-09-02 22:22:03 +08:00
delete FI->second.first;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefVals.erase(FI);
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Set the name on the instruction.
Inst->setName(NameStr);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Eliminate some calls to Value::getNameStr. llvm-svn: 116670
2010-10-16 19:28:23 +08:00
if (Inst->getName() != NameStr)
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
return P.Error(NameLoc, "multiple definition of local value named '" +
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
NameStr + "'");
return false;
}
/// GetBB - Get a basic block with the specified name or ID, creating a
/// forward reference record if needed.
BasicBlock *LLParser::PerFunctionState::GetBB(const std::string &Name,
LocTy Loc) {
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
return cast_or_null<BasicBlock>(GetVal(Name,
Type::getLabelTy(F.getContext()), Loc));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
BasicBlock *LLParser::PerFunctionState::GetBB(unsigned ID, LocTy Loc) {
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
return cast_or_null<BasicBlock>(GetVal(ID,
Type::getLabelTy(F.getContext()), Loc));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// DefineBB - Define the specified basic block, which is either named or
/// unnamed. If there is an error, this returns null otherwise it returns
/// the block being defined.
BasicBlock *LLParser::PerFunctionState::DefineBB(const std::string &Name,
LocTy Loc) {
BasicBlock *BB;
if (Name.empty())
BB = GetBB(NumberedVals.size(), Loc);
else
BB = GetBB(Name, Loc);
if (BB == 0) return 0; // Already diagnosed error.
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Move the block to the end of the function. Forward ref'd blocks are
// inserted wherever they happen to be referenced.
F.getBasicBlockList().splice(F.end(), F.getBasicBlockList(), BB);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Remove the block from forward ref sets.
if (Name.empty()) {
ForwardRefValIDs.erase(NumberedVals.size());
NumberedVals.push_back(BB);
} else {
// BB forward references are already in the function symbol table.
ForwardRefVals.erase(Name);
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return BB;
}
//===----------------------------------------------------------------------===//
// Constants.
//===----------------------------------------------------------------------===//
/// ParseValID - Parse an abstract value that doesn't necessarily have a
/// type implied. For example, if we parse "4" we don't know what integer type
/// it has. The value will later be combined with its type and checked for
Compute isFunctionLocal in MDNode ctor or via argument in new function getWhenValsUnresolved(). Document PFS argument to ParseValID() and ConvertGlobalOrMetadataValIDToValue(). llvm-svn: 93108
2010-01-10 15:14:18 +08:00
/// sanity. PFS is used to convert function-local operands of metadata (since
/// metadata operands are not just parsed here but also converted to values).
/// PFS can be null when we are not parsing metadata values inside a function.
Re-add parsing of function-local metadata; this time with testcase. llvm-svn: 92793
2010-01-06 06:22:14 +08:00
bool LLParser::ParseValID(ValID &ID, PerFunctionState *PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Loc = Lex.getLoc();
switch (Lex.getKind()) {
default: return TokError("expected value token");
case lltok::GlobalID: // @42
ID.UIntVal = Lex.getUIntVal();
ID.Kind = ValID::t_GlobalID;
break;
case lltok::GlobalVar: // @foo
ID.StrVal = Lex.getStrVal();
ID.Kind = ValID::t_GlobalName;
break;
case lltok::LocalVarID: // %42
ID.UIntVal = Lex.getUIntVal();
ID.Kind = ValID::t_LocalID;
break;
case lltok::LocalVar: // %foo
ID.StrVal = Lex.getStrVal();
ID.Kind = ValID::t_LocalName;
break;
Factor out metadata parsing into a separate function. llvm-svn: 108343
2010-07-15 02:26:50 +08:00
case lltok::exclaim: // !42, !{...}, or !"foo"
return ParseMetadataValue(ID, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::APSInt:
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
ID.APSIntVal = Lex.getAPSIntVal();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_APSInt;
break;
case lltok::APFloat:
ID.APFloatVal = Lex.getAPFloatVal();
ID.Kind = ValID::t_APFloat;
break;
case lltok::kw_true:
Move getTrue() and getFalse() to 2.5-like APIs. llvm-svn: 77685
2009-08-01 01:39:07 +08:00
ID.ConstantVal = ConstantInt::getTrue(Context);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
break;
case lltok::kw_false:
Move getTrue() and getFalse() to 2.5-like APIs. llvm-svn: 77685
2009-08-01 01:39:07 +08:00
ID.ConstantVal = ConstantInt::getFalse(Context);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
break;
case lltok::kw_null: ID.Kind = ValID::t_Null; break;
case lltok::kw_undef: ID.Kind = ValID::t_Undef; break;
case lltok::kw_zeroinitializer: ID.Kind = ValID::t_Zero; break;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::lbrace: {
// ValID ::= '{' ConstVector '}'
Lex.Lex();
SmallVector<Constant*, 16> Elts;
if (ParseGlobalValueVector(Elts) ||
ParseToken(lltok::rbrace, "expected end of struct constant"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
ID.ConstantStructElts = new Constant*[Elts.size()];
ID.UIntVal = Elts.size();
memcpy(ID.ConstantStructElts, Elts.data(), Elts.size()*sizeof(Elts[0]));
ID.Kind = ValID::t_ConstantStruct;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
case lltok::less: {
// ValID ::= '<' ConstVector '>' --> Vector.
// ValID ::= '<' '{' ConstVector '}' '>' --> Packed Struct.
Lex.Lex();
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
bool isPackedStruct = EatIfPresent(lltok::lbrace);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
SmallVector<Constant*, 16> Elts;
LocTy FirstEltLoc = Lex.getLoc();
if (ParseGlobalValueVector(Elts) ||
(isPackedStruct &&
ParseToken(lltok::rbrace, "expected end of packed struct")) ||
ParseToken(lltok::greater, "expected end of constant"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (isPackedStruct) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
ID.ConstantStructElts = new Constant*[Elts.size()];
memcpy(ID.ConstantStructElts, Elts.data(), Elts.size()*sizeof(Elts[0]));
ID.UIntVal = Elts.size();
ID.Kind = ValID::t_PackedConstantStruct;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Elts.empty())
return Error(ID.Loc, "constant vector must not be empty");
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!Elts[0]->getType()->isIntegerTy() &&
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
!Elts[0]->getType()->isFloatingPointTy() &&
!Elts[0]->getType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(FirstEltLoc,
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
"vector elements must have integer, pointer or floating point type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Verify that all the vector elements have the same type.
for (unsigned i = 1, e = Elts.size(); i != e; ++i)
if (Elts[i]->getType() != Elts[0]->getType())
return Error(FirstEltLoc,
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
"vector element #" + Twine(i) +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
" is not of type '" + getTypeString(Elts[0]->getType()));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
convert ConstantVector::get to use ArrayRef. llvm-svn: 125537
2011-02-15 08:14:00 +08:00
ID.ConstantVal = ConstantVector::get(Elts);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
case lltok::lsquare: { // Array Constant
Lex.Lex();
SmallVector<Constant*, 16> Elts;
LocTy FirstEltLoc = Lex.getLoc();
if (ParseGlobalValueVector(Elts) ||
ParseToken(lltok::rsquare, "expected end of array constant"))
return true;
// Handle empty element.
if (Elts.empty()) {
// Use undef instead of an array because it's inconvenient to determine
// the element type at this point, there being no elements to examine.
fix PR3281:accepted0[02].ll: represent empty arrays distinctly, and diagnose attempts to initialize non-empty arrays with them. This produces: llvm-as: accepted02.ll:1:28: invalid empty array initializer @"o" = global [5 x double] [] ^ llvm-as: accepted00.ll:1:32: invalid empty array initializer @"za" = thread_local global {} [] ^ [ llvm-svn: 61676
2009-01-05 15:52:51 +08:00
ID.Kind = ValID::t_EmptyArray;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!Elts[0]->getType()->isFirstClassType())
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
return Error(FirstEltLoc, "invalid array element type: " +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Elts[0]->getType()));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Move types back to the 2.5 API. llvm-svn: 77516
2009-07-30 06:17:13 +08:00
ArrayType *ATy = ArrayType::get(Elts[0]->getType(), Elts.size());
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Verify all elements are correct type!
Fix a really horrible typo, which caused undefined behavior. llvm-svn: 61566
2009-01-02 16:49:06 +08:00
for (unsigned i = 0, e = Elts.size(); i != e; ++i) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Elts[i]->getType() != Elts[0]->getType())
return Error(FirstEltLoc,
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
"array element #" + Twine(i) +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
" is not of type '" + getTypeString(Elts[0]->getType()));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Replace the existing forms of ConstantArray::get() with a single form that takes an ArrayRef. llvm-svn: 133615
2011-06-22 17:24:39 +08:00
ID.ConstantVal = ConstantArray::get(ATy, Elts);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
case lltok::kw_c: // c "foo"
Lex.Lex();
reapply the patches reverted in r149470 that reenable ConstantDataArray, but with a critical fix to the SelectionDAG code that optimizes copies from strings into immediate stores: the previous code was stopping reading string data at the first nul. Address this by adding a new argument to llvm::getConstantStringInfo, preserving the behavior before the patch. llvm-svn: 149800
2012-02-05 10:29:43 +08:00
ID.ConstantVal = ConstantDataArray::getString(Context, Lex.getStrVal(),
false);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseToken(lltok::StringConstant, "expected string")) return true;
ID.Kind = ValID::t_Constant;
return false;
case lltok::kw_asm: {
Rename msasm to alignstack per review. llvm-svn: 84795
2009-10-22 07:28:00 +08:00
// ValID ::= 'asm' SideEffect? AlignStack? STRINGCONSTANT ',' STRINGCONSTANT
[ms-inline asm] Enumerate the InlineAsm dialects and rename the nsdialect to inteldialect. llvm-svn: 163231
2012-09-06 03:00:49 +08:00
bool HasSideEffect, AlignStack, AsmDialect;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
if (ParseOptionalToken(lltok::kw_sideeffect, HasSideEffect) ||
Rename msasm to alignstack per review. llvm-svn: 84795
2009-10-22 07:28:00 +08:00
ParseOptionalToken(lltok::kw_alignstack, AlignStack) ||
[ms-inline asm] Enumerate the InlineAsm dialects and rename the nsdialect to inteldialect. llvm-svn: 163231
2012-09-06 03:00:49 +08:00
ParseOptionalToken(lltok::kw_inteldialect, AsmDialect) ||
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
ParseStringConstant(ID.StrVal) ||
ParseToken(lltok::comma, "expected comma in inline asm expression") ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::StringConstant, "expected constraint string"))
return true;
ID.StrVal2 = Lex.getStrVal();
[ms-inline asm] Emit the (new) inline asm Non-Standard Dialect attribute. llvm-svn: 163181
2012-09-05 08:08:17 +08:00
ID.UIntVal = unsigned(HasSideEffect) | (unsigned(AlignStack)<<1) |
[ms-inline asm] Enumerate the InlineAsm dialects and rename the nsdialect to inteldialect. llvm-svn: 163231
2012-09-06 03:00:49 +08:00
(unsigned(AsmDialect)<<2);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_InlineAsm;
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
case lltok::kw_blockaddress: {
// ValID ::= 'blockaddress' '(' @foo ',' %bar ')'
Lex.Lex();
ValID Fn, Label;
LocTy FnLoc, LabelLoc;
if (ParseToken(lltok::lparen, "expected '(' in block address expression") ||
ParseValID(Fn) ||
ParseToken(lltok::comma, "expected comma in block address expression")||
ParseValID(Label) ||
ParseToken(lltok::rparen, "expected ')' in block address expression"))
return true;
Revert 85678/85680. The decision is to stay with the current form of indirectbr, thus we don't need "blockaddr(@func, null)". Eliminate it for simplicity. llvm-svn: 85699
2009-11-01 09:27:45 +08:00
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
if (Fn.Kind != ValID::t_GlobalID && Fn.Kind != ValID::t_GlobalName)
return Error(Fn.Loc, "expected function name in blockaddress");
Revert 85678/85680. The decision is to stay with the current form of indirectbr, thus we don't need "blockaddr(@func, null)". Eliminate it for simplicity. llvm-svn: 85699
2009-11-01 09:27:45 +08:00
if (Label.Kind != ValID::t_LocalID && Label.Kind != ValID::t_LocalName)
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
return Error(Label.Loc, "expected basic block name in blockaddress");
// Make a global variable as a placeholder for this reference.
GlobalVariable *FwdRef = new GlobalVariable(*M, Type::getInt8Ty(Context),
false, GlobalValue::InternalLinkage,
0, "");
ForwardRefBlockAddresses[Fn].push_back(std::make_pair(Label, FwdRef));
ID.ConstantVal = FwdRef;
ID.Kind = ValID::t_Constant;
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_trunc:
case lltok::kw_zext:
case lltok::kw_sext:
case lltok::kw_fptrunc:
case lltok::kw_fpext:
case lltok::kw_bitcast:
case lltok::kw_uitofp:
case lltok::kw_sitofp:
case lltok::kw_fptoui:
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
case lltok::kw_fptosi:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_inttoptr:
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
case lltok::kw_ptrtoint: {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
unsigned Opc = Lex.getUIntVal();
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *DestTy = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Constant *SrcVal;
Lex.Lex();
if (ParseToken(lltok::lparen, "expected '(' after constantexpr cast") ||
ParseGlobalTypeAndValue(SrcVal) ||
Fix a typo in a diagnostic. llvm-svn: 73429
2009-06-16 05:52:11 +08:00
ParseToken(lltok::kw_to, "expected 'to' in constantexpr cast") ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseType(DestTy) ||
ParseToken(lltok::rparen, "expected ')' at end of constantexpr cast"))
return true;
if (!CastInst::castIsValid((Instruction::CastOps)Opc, SrcVal, DestTy))
return Error(ID.Loc, "invalid cast opcode for cast from '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(SrcVal->getType()) + "' to '" +
getTypeString(DestTy) + "'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
ID.ConstantVal = ConstantExpr::getCast((Instruction::CastOps)Opc,
Try again at converting the LLParser to use LLVMContext, without massive breakage this time. llvm-svn: 74671
2009-07-02 07:57:11 +08:00
SrcVal, DestTy);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
case lltok::kw_extractvalue: {
Lex.Lex();
Constant *Val;
SmallVector<unsigned, 4> Indices;
if (ParseToken(lltok::lparen, "expected '(' in extractvalue constantexpr")||
ParseGlobalTypeAndValue(Val) ||
ParseIndexList(Indices) ||
ParseToken(lltok::rparen, "expected ')' in extractvalue constantexpr"))
return true;
Parse debug info attached with insertvalue and extractvalue instructions. llvm-svn: 85921
2009-11-04 03:06:07 +08:00
Add support for a union type in LLVM IR. Patch by Talin! llvm-svn: 96011
2010-02-13 04:49:41 +08:00
if (!Val->getType()->isAggregateType())
return Error(ID.Loc, "extractvalue operand must be aggregate type");
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
if (!ExtractValueInst::getIndexedType(Val->getType(), Indices))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "invalid indices for extractvalue");
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
ID.ConstantVal = ConstantExpr::getExtractValue(Val, Indices);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
case lltok::kw_insertvalue: {
Lex.Lex();
Constant *Val0, *Val1;
SmallVector<unsigned, 4> Indices;
if (ParseToken(lltok::lparen, "expected '(' in insertvalue constantexpr")||
ParseGlobalTypeAndValue(Val0) ||
ParseToken(lltok::comma, "expected comma in insertvalue constantexpr")||
ParseGlobalTypeAndValue(Val1) ||
ParseIndexList(Indices) ||
ParseToken(lltok::rparen, "expected ')' in insertvalue constantexpr"))
return true;
Add support for a union type in LLVM IR. Patch by Talin! llvm-svn: 96011
2010-02-13 04:49:41 +08:00
if (!Val0->getType()->isAggregateType())
return Error(ID.Loc, "insertvalue operand must be aggregate type");
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
if (!ExtractValueInst::getIndexedType(Val0->getType(), Indices))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "invalid indices for insertvalue");
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
ID.ConstantVal = ConstantExpr::getInsertValue(Val0, Val1, Indices);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
case lltok::kw_icmp:
Remove the vicmp and vfcmp instructions. Because we never had a release with these instructions, no autoupgrade or backwards compatibility support is provided. llvm-svn: 74991
2009-07-08 11:04:38 +08:00
case lltok::kw_fcmp: {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
unsigned PredVal, Opc = Lex.getUIntVal();
Constant *Val0, *Val1;
Lex.Lex();
if (ParseCmpPredicate(PredVal, Opc) ||
ParseToken(lltok::lparen, "expected '(' in compare constantexpr") ||
ParseGlobalTypeAndValue(Val0) ||
ParseToken(lltok::comma, "expected comma in compare constantexpr") ||
ParseGlobalTypeAndValue(Val1) ||
ParseToken(lltok::rparen, "expected ')' in compare constantexpr"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Val0->getType() != Val1->getType())
return Error(ID.Loc, "compare operands must have the same type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
CmpInst::Predicate Pred = (CmpInst::Predicate)PredVal;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Opc == Instruction::FCmp) {
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!Val0->getType()->isFPOrFPVectorTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "fcmp requires floating point operands");
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ID.ConstantVal = ConstantExpr::getFCmp(Pred, Val0, Val1);
Remove the vicmp and vfcmp instructions. Because we never had a release with these instructions, no autoupgrade or backwards compatibility support is provided. llvm-svn: 74991
2009-07-08 11:04:38 +08:00
} else {
assert(Opc == Instruction::ICmp && "Unexpected opcode for CmpInst!");
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!Val0->getType()->isIntOrIntVectorTy() &&
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
!Val0->getType()->getScalarType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "icmp requires pointer or integer operands");
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ID.ConstantVal = ConstantExpr::getICmp(Pred, Val0, Val1);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
ID.Kind = ValID::t_Constant;
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Binary Operators.
case lltok::kw_add:
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 06:49:04 +08:00
case lltok::kw_fadd:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_sub:
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 06:49:04 +08:00
case lltok::kw_fsub:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_mul:
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 06:49:04 +08:00
case lltok::kw_fmul:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_udiv:
case lltok::kw_sdiv:
case lltok::kw_fdiv:
case lltok::kw_urem:
case lltok::kw_srem:
implement .ll and .bc support for nsw/nuw on shl and exact on lshr/ashr. Factor some code better. llvm-svn: 125006
2011-02-08 00:40:21 +08:00
case lltok::kw_frem:
case lltok::kw_shl:
case lltok::kw_lshr:
case lltok::kw_ashr: {
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
bool NUW = false;
bool NSW = false;
bool Exact = false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
unsigned Opc = Lex.getUIntVal();
Constant *Val0, *Val1;
Lex.Lex();
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
LocTy ModifierLoc = Lex.getLoc();
implement .ll and .bc support for nsw/nuw on shl and exact on lshr/ashr. Factor some code better. llvm-svn: 125006
2011-02-08 00:40:21 +08:00
if (Opc == Instruction::Add || Opc == Instruction::Sub ||
Opc == Instruction::Mul || Opc == Instruction::Shl) {
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
if (EatIfPresent(lltok::kw_nuw))
NUW = true;
if (EatIfPresent(lltok::kw_nsw)) {
NSW = true;
if (EatIfPresent(lltok::kw_nuw))
NUW = true;
}
implement .ll and .bc support for nsw/nuw on shl and exact on lshr/ashr. Factor some code better. llvm-svn: 125006
2011-02-08 00:40:21 +08:00
} else if (Opc == Instruction::SDiv || Opc == Instruction::UDiv ||
Opc == Instruction::LShr || Opc == Instruction::AShr) {
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
if (EatIfPresent(lltok::kw_exact))
Exact = true;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseToken(lltok::lparen, "expected '(' in binary constantexpr") ||
ParseGlobalTypeAndValue(Val0) ||
ParseToken(lltok::comma, "expected comma in binary constantexpr") ||
ParseGlobalTypeAndValue(Val1) ||
ParseToken(lltok::rparen, "expected ')' in binary constantexpr"))
return true;
if (Val0->getType() != Val1->getType())
return Error(ID.Loc, "operands of constexpr must have same type");
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!Val0->getType()->isIntOrIntVectorTy()) {
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
if (NUW)
return Error(ModifierLoc, "nuw only applies to integer operations");
if (NSW)
return Error(ModifierLoc, "nsw only applies to integer operations");
}
Remove the API compatibility layer which converted add, sub, and mul to fadd, fsub, and fmul, when used with a floating-point type. LLVM has supported the new instructions since 2.6, so it's time to get on board. llvm-svn: 102971
2010-05-04 06:44:19 +08:00
// Check that the type is valid for the operator.
switch (Opc) {
case Instruction::Add:
case Instruction::Sub:
case Instruction::Mul:
case Instruction::UDiv:
case Instruction::SDiv:
case Instruction::URem:
case Instruction::SRem:
implement .ll and .bc support for nsw/nuw on shl and exact on lshr/ashr. Factor some code better. llvm-svn: 125006
2011-02-08 00:40:21 +08:00
case Instruction::Shl:
case Instruction::AShr:
case Instruction::LShr:
Remove the API compatibility layer which converted add, sub, and mul to fadd, fsub, and fmul, when used with a floating-point type. LLVM has supported the new instructions since 2.6, so it's time to get on board. llvm-svn: 102971
2010-05-04 06:44:19 +08:00
if (!Val0->getType()->isIntOrIntVectorTy())
return Error(ID.Loc, "constexpr requires integer operands");
break;
case Instruction::FAdd:
case Instruction::FSub:
case Instruction::FMul:
case Instruction::FDiv:
case Instruction::FRem:
if (!Val0->getType()->isFPOrFPVectorTy())
return Error(ID.Loc, "constexpr requires fp operands");
break;
default: llvm_unreachable("Unknown binary operator!");
}
Reappy r80998, now that the GlobalOpt bug that it exposed on MiniSAT is fixed. llvm-svn: 81172
2009-09-08 07:54:19 +08:00
unsigned Flags = 0;
if (NUW) Flags |= OverflowingBinaryOperator::NoUnsignedWrap;
if (NSW) Flags |= OverflowingBinaryOperator::NoSignedWrap;
enhance vmcore to know that udiv's can be exact, and add a trivial instcombine xform to exercise this. Nothing forms exact udivs yet though. This is progress on PR8862 llvm-svn: 124992
2011-02-07 05:44:57 +08:00
if (Exact) Flags |= PossiblyExactOperator::IsExact;
Reappy r80998, now that the GlobalOpt bug that it exposed on MiniSAT is fixed. llvm-svn: 81172
2009-09-08 07:54:19 +08:00
Constant *C = ConstantExpr::get(Opc, Val0, Val1, Flags);
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
ID.ConstantVal = C;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Logical Operations
case lltok::kw_and:
case lltok::kw_or:
case lltok::kw_xor: {
unsigned Opc = Lex.getUIntVal();
Constant *Val0, *Val1;
Lex.Lex();
if (ParseToken(lltok::lparen, "expected '(' in logical constantexpr") ||
ParseGlobalTypeAndValue(Val0) ||
ParseToken(lltok::comma, "expected comma in logical constantexpr") ||
ParseGlobalTypeAndValue(Val1) ||
ParseToken(lltok::rparen, "expected ')' in logical constantexpr"))
return true;
if (Val0->getType() != Val1->getType())
return Error(ID.Loc, "operands of constexpr must have same type");
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!Val0->getType()->isIntOrIntVectorTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc,
"constexpr requires integer or integer vector operands");
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ID.ConstantVal = ConstantExpr::get(Opc, Val0, Val1);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_getelementptr:
case lltok::kw_shufflevector:
case lltok::kw_insertelement:
case lltok::kw_extractelement:
case lltok::kw_select: {
unsigned Opc = Lex.getUIntVal();
SmallVector<Constant*, 16> Elts;
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
bool InBounds = false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
if (Opc == Instruction::GetElementPtr)
Minor code simplification. llvm-svn: 77459
2009-07-29 23:58:36 +08:00
InBounds = EatIfPresent(lltok::kw_inbounds);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseToken(lltok::lparen, "expected '(' in constantexpr") ||
ParseGlobalValueVector(Elts) ||
ParseToken(lltok::rparen, "expected ')' in constantexpr"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Opc == Instruction::GetElementPtr) {
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
if (Elts.size() == 0 ||
!Elts[0]->getType()->getScalarType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "getelementptr requires pointer operand");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Convert ConstantExpr::getGetElementPtr and ConstantExpr::getInBoundsGetElementPtr to use ArrayRef. llvm-svn: 135673
2011-07-21 22:31:17 +08:00
ArrayRef<Constant *> Indices(Elts.begin() + 1, Elts.end());
Convert GetElementPtrInst to use ArrayRef. llvm-svn: 135904
2011-07-25 17:48:08 +08:00
if (!GetElementPtrInst::getIndexedType(Elts[0]->getType(), Indices))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "invalid indices for getelementptr");
Make better use of ConstantExpr::getGetElementPtr's InBounds parameter. llvm-svn: 135676
2011-07-21 23:15:37 +08:00
ID.ConstantVal = ConstantExpr::getGetElementPtr(Elts[0], Indices,
InBounds);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else if (Opc == Instruction::Select) {
if (Elts.size() != 3)
return Error(ID.Loc, "expected three operands to select");
if (const char *Reason = SelectInst::areInvalidOperands(Elts[0], Elts[1],
Elts[2]))
return Error(ID.Loc, Reason);
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ID.ConstantVal = ConstantExpr::getSelect(Elts[0], Elts[1], Elts[2]);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else if (Opc == Instruction::ShuffleVector) {
if (Elts.size() != 3)
return Error(ID.Loc, "expected three operands to shufflevector");
if (!ShuffleVectorInst::isValidOperands(Elts[0], Elts[1], Elts[2]))
return Error(ID.Loc, "invalid operands to shufflevector");
Try again at converting the LLParser to use LLVMContext, without massive breakage this time. llvm-svn: 74671
2009-07-02 07:57:11 +08:00
ID.ConstantVal =
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ConstantExpr::getShuffleVector(Elts[0], Elts[1],Elts[2]);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else if (Opc == Instruction::ExtractElement) {
if (Elts.size() != 2)
return Error(ID.Loc, "expected two operands to extractelement");
if (!ExtractElementInst::isValidOperands(Elts[0], Elts[1]))
return Error(ID.Loc, "invalid extractelement operands");
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ID.ConstantVal = ConstantExpr::getExtractElement(Elts[0], Elts[1]);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
} else {
assert(Opc == Instruction::InsertElement && "Unknown opcode");
if (Elts.size() != 3)
return Error(ID.Loc, "expected three operands to insertelement");
if (!InsertElementInst::isValidOperands(Elts[0], Elts[1], Elts[2]))
return Error(ID.Loc, "invalid insertelement operands");
Try again at converting the LLParser to use LLVMContext, without massive breakage this time. llvm-svn: 74671
2009-07-02 07:57:11 +08:00
ID.ConstantVal =
Move ConstantExpr to 2.5 API. llvm-svn: 77494
2009-07-30 02:55:55 +08:00
ConstantExpr::getInsertElement(Elts[0], Elts[1],Elts[2]);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ID.Kind = ValID::t_Constant;
return false;
}
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex();
return false;
}
/// ParseGlobalValue - Parse a global value with the specified type.
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
bool LLParser::ParseGlobalValue(Type *Ty, Constant *&C) {
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
C = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ValID ID;
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
Value *V = NULL;
bool Parsed = ParseValID(ID) ||
ConvertValIDToValue(Ty, ID, V, NULL);
if (V && !(C = dyn_cast<Constant>(V)))
return Error(ID.Loc, "global values must be constants");
return Parsed;
}
bool LLParser::ParseGlobalTypeAndValue(Constant *&V) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Ty = 0;
return ParseType(Ty) ||
ParseGlobalValue(Ty, V);
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
}
/// ParseGlobalValueVector
/// ::= /*empty*/
/// ::= TypeAndValue (',' TypeAndValue)*
bool LLParser::ParseGlobalValueVector(SmallVectorImpl<Constant*> &Elts) {
// Empty list.
if (Lex.getKind() == lltok::rbrace ||
Lex.getKind() == lltok::rsquare ||
Lex.getKind() == lltok::greater ||
Lex.getKind() == lltok::rparen)
return false;
Constant *C;
if (ParseGlobalTypeAndValue(C)) return true;
Elts.push_back(C);
while (EatIfPresent(lltok::comma)) {
if (ParseGlobalTypeAndValue(C)) return true;
Elts.push_back(C);
}
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Extend function-local metadata to be usable as attachments. llvm-svn: 111895
2010-08-24 10:24:03 +08:00
bool LLParser::ParseMetadataListValue(ValID &ID, PerFunctionState *PFS) {
assert(Lex.getKind() == lltok::lbrace);
Lex.Lex();
SmallVector<Value*, 16> Elts;
if (ParseMDNodeVector(Elts, PFS) ||
ParseToken(lltok::rbrace, "expected end of metadata node"))
return true;
PR9214: Convert Metadata API to use ArrayRef. llvm-svn: 129932
2011-04-22 03:59:31 +08:00
ID.MDNodeVal = MDNode::get(Context, Elts);
Extend function-local metadata to be usable as attachments. llvm-svn: 111895
2010-08-24 10:24:03 +08:00
ID.Kind = ValID::t_MDNode;
return false;
}
Factor out metadata parsing into a separate function. llvm-svn: 108343
2010-07-15 02:26:50 +08:00
/// ParseMetadataValue
/// ::= !42
/// ::= !{...}
/// ::= !"string"
bool LLParser::ParseMetadataValue(ValID &ID, PerFunctionState *PFS) {
assert(Lex.getKind() == lltok::exclaim);
Lex.Lex();
// MDNode:
// !{ ... }
Extend function-local metadata to be usable as attachments. llvm-svn: 111895
2010-08-24 10:24:03 +08:00
if (Lex.getKind() == lltok::lbrace)
return ParseMetadataListValue(ID, PFS);
Factor out metadata parsing into a separate function. llvm-svn: 108343
2010-07-15 02:26:50 +08:00
// Standalone metadata reference
// !42
if (Lex.getKind() == lltok::APSInt) {
if (ParseMDNodeID(ID.MDNodeVal)) return true;
ID.Kind = ValID::t_MDNode;
return false;
}
// MDString:
// ::= '!' STRINGCONSTANT
if (ParseMDString(ID.MDStringVal)) return true;
ID.Kind = ValID::t_MDString;
return false;
}
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
//===----------------------------------------------------------------------===//
// Function Parsing.
//===----------------------------------------------------------------------===//
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
bool LLParser::ConvertValIDToValue(Type *Ty, ValID &ID, Value *&V,
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
PerFunctionState *PFS) {
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (Ty->isFunctionTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "functions are not values, refer to them as pointers");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
switch (ID.Kind) {
case ValID::t_LocalID:
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
if (!PFS) return Error(ID.Loc, "invalid use of function-local name");
V = PFS->GetVal(ID.UIntVal, Ty, ID.Loc);
return (V == 0);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case ValID::t_LocalName:
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
if (!PFS) return Error(ID.Loc, "invalid use of function-local name");
V = PFS->GetVal(ID.StrVal, Ty, ID.Loc);
return (V == 0);
case ValID::t_InlineAsm: {
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
PointerType *PTy = dyn_cast<PointerType>(Ty);
FunctionType *FTy =
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
PTy ? dyn_cast<FunctionType>(PTy->getElementType()) : 0;
if (!FTy || !InlineAsm::Verify(FTy, ID.StrVal2))
return Error(ID.Loc, "invalid type for inline asm constraint string");
[ms-inline asm] Emit the (new) inline asm Non-Standard Dialect attribute. llvm-svn: 163181
2012-09-05 08:08:17 +08:00
V = InlineAsm::get(FTy, ID.StrVal, ID.StrVal2, ID.UIntVal&1,
[ms-inline asm] Enumerate the InlineAsm dialects and rename the nsdialect to inteldialect. llvm-svn: 163231
2012-09-06 03:00:49 +08:00
(ID.UIntVal>>1)&1, (InlineAsm::AsmDialect(ID.UIntVal>>2)));
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
return false;
}
case ValID::t_MDNode:
if (!Ty->isMetadataTy())
return Error(ID.Loc, "metadata value must have metadata type");
V = ID.MDNodeVal;
return false;
case ValID::t_MDString:
if (!Ty->isMetadataTy())
return Error(ID.Loc, "metadata value must have metadata type");
V = ID.MDStringVal;
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case ValID::t_GlobalName:
V = GetGlobalVal(ID.StrVal, Ty, ID.Loc);
return V == 0;
case ValID::t_GlobalID:
V = GetGlobalVal(ID.UIntVal, Ty, ID.Loc);
return V == 0;
case ValID::t_APSInt:
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Ty->isIntegerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "integer constant must have integer type");
PR5207: Change APInt methods trunc(), sext(), zext(), sextOrTrunc() and zextOrTrunc(), and APSInt methods extend(), extOrTrunc() and new method trunc(), to be const and to return a new value instead of modifying the object in place. llvm-svn: 121120
2010-12-07 16:25:19 +08:00
ID.APSIntVal = ID.APSIntVal.extOrTrunc(Ty->getPrimitiveSizeInBits());
Revert the ConstantInt constructors back to their 2.5 forms where possible, thanks to contexts-on-types. More to come. llvm-svn: 77011
2009-07-25 07:12:02 +08:00
V = ConstantInt::get(Context, ID.APSIntVal);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
case ValID::t_APFloat:
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!Ty->isFloatingPointTy() ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
!ConstantFP::isValueValidForType(Ty, ID.APFloatVal))
return Error(ID.Loc, "floating point constant invalid for type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
The powers that be have decided that LLVM IR should now support 16-bit "half precision" floating-point with a first-class type. This patch adds basic IR support (but not codegen support). llvm-svn: 146786
2011-12-17 08:04:22 +08:00
// The lexer has no type info, so builds all half, float, and double FP
// constants as double. Fix this here. Long double does not need this.
if (&ID.APFloatVal.getSemantics() == &APFloat::IEEEdouble) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool Ignored;
The powers that be have decided that LLVM IR should now support 16-bit "half precision" floating-point with a first-class type. This patch adds basic IR support (but not codegen support). llvm-svn: 146786
2011-12-17 08:04:22 +08:00
if (Ty->isHalfTy())
ID.APFloatVal.convert(APFloat::IEEEhalf, APFloat::rmNearestTiesToEven,
&Ignored);
else if (Ty->isFloatTy())
ID.APFloatVal.convert(APFloat::IEEEsingle, APFloat::rmNearestTiesToEven,
&Ignored);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Move ConstantFP construction back to the 2.5-ish API. llvm-svn: 77247
2009-07-28 04:59:43 +08:00
V = ConstantFP::get(Context, ID.APFloatVal);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
reject PR3281:crash10.ll with: llvm-as: crash10.ll:3:35: floating point constant does not have type 'ppc_fp128' "dumy" = fcmp ult ppc_fp128 "j",9209.4 ^ llvm-svn: 61721
2009-01-06 02:24:23 +08:00
if (V->getType() != Ty)
return Error(ID.Loc, "floating point constant does not have type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Ty) + "'");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
case ValID::t_Null:
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Ty->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "null must be a pointer type");
Move more code back to 2.5 APIs. llvm-svn: 77635
2009-07-31 07:03:37 +08:00
V = ConstantPointerNull::get(cast<PointerType>(Ty));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
case ValID::t_Undef:
reject undef/zero labels. This fixes PR3281:crash0[56].ll with these diagnostics: llvm-as: crash05.ll:1:14: invalid type for null constant global label zeroinitializer addrspace (75), section "c" ^ llvm-as: crash06.ll:2:14: invalid type for null constant udiv label zeroinitializer, @0 ^ llvm-svn: 61681
2009-01-05 16:13:38 +08:00
// FIXME: LabelTy should not be a first-class type.
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!Ty->isFirstClassType() || Ty->isLabelTy())
reject undef/zero labels. This fixes PR3281:crash0[56].ll with these diagnostics: llvm-as: crash05.ll:1:14: invalid type for null constant global label zeroinitializer addrspace (75), section "c" ^ llvm-as: crash06.ll:2:14: invalid type for null constant udiv label zeroinitializer, @0 ^ llvm-svn: 61681
2009-01-05 16:13:38 +08:00
return Error(ID.Loc, "invalid type for undef constant");
Move more code back to 2.5 APIs. llvm-svn: 77635
2009-07-31 07:03:37 +08:00
V = UndefValue::get(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
fix PR3281:accepted0[02].ll: represent empty arrays distinctly, and diagnose attempts to initialize non-empty arrays with them. This produces: llvm-as: accepted02.ll:1:28: invalid empty array initializer @"o" = global [5 x double] [] ^ llvm-as: accepted00.ll:1:32: invalid empty array initializer @"za" = thread_local global {} [] ^ [ llvm-svn: 61676
2009-01-05 15:52:51 +08:00
case ValID::t_EmptyArray:
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Ty->isArrayTy() || cast<ArrayType>(Ty)->getNumElements() != 0)
fix PR3281:accepted0[02].ll: represent empty arrays distinctly, and diagnose attempts to initialize non-empty arrays with them. This produces: llvm-as: accepted02.ll:1:28: invalid empty array initializer @"o" = global [5 x double] [] ^ llvm-as: accepted00.ll:1:32: invalid empty array initializer @"za" = thread_local global {} [] ^ [ llvm-svn: 61676
2009-01-05 15:52:51 +08:00
return Error(ID.Loc, "invalid empty array initializer");
Move more code back to 2.5 APIs. llvm-svn: 77635
2009-07-31 07:03:37 +08:00
V = UndefValue::get(Ty);
fix PR3281:accepted0[02].ll: represent empty arrays distinctly, and diagnose attempts to initialize non-empty arrays with them. This produces: llvm-as: accepted02.ll:1:28: invalid empty array initializer @"o" = global [5 x double] [] ^ llvm-as: accepted00.ll:1:32: invalid empty array initializer @"za" = thread_local global {} [] ^ [ llvm-svn: 61676
2009-01-05 15:52:51 +08:00
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case ValID::t_Zero:
reject undef/zero labels. This fixes PR3281:crash0[56].ll with these diagnostics: llvm-as: crash05.ll:1:14: invalid type for null constant global label zeroinitializer addrspace (75), section "c" ^ llvm-as: crash06.ll:2:14: invalid type for null constant udiv label zeroinitializer, @0 ^ llvm-svn: 61681
2009-01-05 16:13:38 +08:00
// FIXME: LabelTy should not be a first-class type.
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (!Ty->isFirstClassType() || Ty->isLabelTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "invalid type for null constant");
Move a few more APIs back to 2.5 forms. The only remaining ones left to change back are metadata related, which I'm waiting on to avoid conflicting with Devang. llvm-svn: 77721
2009-08-01 04:28:14 +08:00
V = Constant::getNullValue(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
case ValID::t_Constant:
remove unions from LLVM IR. They are severely buggy and not being actively maintained, improved, or extended. llvm-svn: 112356
2010-08-28 12:09:24 +08:00
if (ID.ConstantVal->getType() != Ty)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ID.Loc, "constant expression type mismatch");
Add support for a union type in LLVM IR. Patch by Talin! llvm-svn: 96011
2010-02-13 04:49:41 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
V = ID.ConstantVal;
return false;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
case ValID::t_ConstantStruct:
case ValID::t_PackedConstantStruct:
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
if (StructType *ST = dyn_cast<StructType>(Ty)) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ST->getNumElements() != ID.UIntVal)
return Error(ID.Loc,
"initializer with struct type has wrong # elements");
if (ST->isPacked() != (ID.Kind == ValID::t_PackedConstantStruct))
return Error(ID.Loc, "packed'ness of initializer and type don't match");
// Verify that the elements are compatible with the structtype.
for (unsigned i = 0, e = ID.UIntVal; i != e; ++i)
if (ID.ConstantStructElts[i]->getType() != ST->getElementType(i))
return Error(ID.Loc, "element " + Twine(i) +
" of struct initializer doesn't match struct element type");
Migrate LLVM and Clang to use the new makeArrayRef(...) functions where previously explicit non-default constructors were used. Mostly mechanical with some manual reformatting. llvm-svn: 135390
2011-07-18 20:00:32 +08:00
V = ConstantStruct::get(ST, makeArrayRef(ID.ConstantStructElts,
ID.UIntVal));
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
} else
return Error(ID.Loc, "constant expression type mismatch");
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Add 'llvm_unreachable' to passify GCC's understanding of the constraints of several newly un-defaulted switches. This also helps optimizers (including LLVM's) recognize that every case is covered, and we should assume as much. llvm-svn: 147861
2012-01-11 02:08:01 +08:00
llvm_unreachable("Invalid ValID");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
bool LLParser::ParseValue(Type *Ty, Value *&V, PerFunctionState *PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
V = 0;
ValID ID;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
return ParseValID(ID, PFS) ||
ConvertValIDToValue(Ty, ID, V, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
bool LLParser::ParseTypeAndValue(Value *&V, PerFunctionState *PFS) {
Type *Ty = 0;
return ParseType(Ty) ||
ParseValue(Ty, V, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
bool LLParser::ParseTypeAndBasicBlock(BasicBlock *&BB, LocTy &Loc,
PerFunctionState &PFS) {
Value *V;
Loc = Lex.getLoc();
if (ParseTypeAndValue(V, PFS)) return true;
if (!isa<BasicBlock>(V))
return Error(Loc, "expected a basic block");
BB = cast<BasicBlock>(V);
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// FunctionHeader
/// ::= OptionalLinkage OptionalVisibility OptionalCallingConv OptRetAttrs
First step in fixing PR8927: Add a unnamed_addr bit to global variables and functions. This will be used to indicate that the address is not significant and therefore the constant or function can be merged with others. If an optimization pass can show that an address is not used, it can set this. Examples of things that can have this set by the FE are globals created to hold string literals and C++ constructors. Adding unnamed_addr to a non-const global should have no effect unless an optimization can transform that global into a constant. Aliases are not allowed to have unnamed_addr since I couldn't figure out any use for it. llvm-svn: 123063
2011-01-09 00:42:36 +08:00
/// OptUnnamedAddr Type GlobalName '(' ArgList ')' OptFuncAttrs OptSection
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// OptionalAlign OptGC
bool LLParser::ParseFunctionHeader(Function *&Fn, bool isDefine) {
// Parse the linkage.
LocTy LinkageLoc = Lex.getLoc();
unsigned Linkage;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
unsigned Visibility;
Attributes RetAttrs;
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
CallingConv::ID CC;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *RetType = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy RetTypeLoc = Lex.getLoc();
if (ParseOptionalLinkage(Linkage) ||
ParseOptionalVisibility(Visibility) ||
ParseOptionalCallingConv(CC) ||
ParseOptionalAttrs(RetAttrs, 1) ||
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
ParseType(RetType, RetTypeLoc, true /*void allowed*/))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
// Verify that the linkage is ok.
switch ((GlobalValue::LinkageTypes)Linkage) {
case GlobalValue::ExternalLinkage:
break; // always ok.
case GlobalValue::DLLImportLinkage:
Remove the one-definition-rule version of extern_weak linkage: this linkage type only applies to declarations, but ODR is only relevant to globals with definitions. llvm-svn: 66650
2009-03-11 16:08:06 +08:00
case GlobalValue::ExternalWeakLinkage:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (isDefine)
return Error(LinkageLoc, "invalid linkage for function definition");
break;
Add the private linkage. llvm-svn: 62279
2009-01-16 04:18:42 +08:00
case GlobalValue::PrivateLinkage:
Add plumbing for the `linker_private' linkage type. This type is meant for "private" symbols which the assember shouldn't strip, but which the linker may remove after evaluation. This is mostly useful for Objective-C metadata. This is plumbing, so we don't have a use of it yet. More to come, etc. llvm-svn: 76385
2009-07-20 09:03:30 +08:00
case GlobalValue::LinkerPrivateLinkage:
Implement the "linker_private_weak" linkage type. This will be used for Objective-C metadata types which should be marked as "weak", but which the linker will remove upon final linkage. However, this linkage isn't specific to Objective-C. For example, the "objc_msgSend_fixup_alloc" symbol is defined like this: .globl l_objc_msgSend_fixup_alloc .weak_definition l_objc_msgSend_fixup_alloc .section __DATA, __objc_msgrefs, coalesced .align 3 l_objc_msgSend_fixup_alloc: .quad _objc_msgSend_fixup .quad L_OBJC_METH_VAR_NAME_1 This is different from the "linker_private" linkage type, because it can't have the metadata defined with ".weak_definition". Currently only supported on Darwin platforms. llvm-svn: 107433
2010-07-02 05:55:59 +08:00
case GlobalValue::LinkerPrivateWeakLinkage:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case GlobalValue::InternalLinkage:
Fix warning in .ll parser, detect and reject available_externally on function declarations. llvm-svn: 68944
2009-04-13 15:02:02 +08:00
case GlobalValue::AvailableExternallyLinkage:
Introduce new linkage types linkonce_odr, weak_odr, common_odr and extern_weak_odr. These are the same as the non-odr versions, except that they indicate that the global will only be overridden by an *equivalent* global. In C, a function with weak linkage can be overridden by a function which behaves completely differently. This means that IP passes have to skip weak functions, since any deductions made from the function definition might be wrong, since the definition could be replaced by something completely different at link time. This is not allowed in C++, thanks to the ODR (One-Definition-Rule): if a function is replaced by another at link-time, then the new function must be the same as the original function. If a language knows that a function or other global can only be overridden by an equivalent global, it can give it the weak_odr linkage type, and the optimizers will understand that it is alright to make deductions based on the function body. The code generators on the other hand map weak and weak_odr linkage to the same thing. llvm-svn: 66339
2009-03-07 23:45:40 +08:00
case GlobalValue::LinkOnceAnyLinkage:
case GlobalValue::LinkOnceODRLinkage:
Change the `linker_private_weak_def_auto' linkage to `linkonce_odr_auto_hide' to make it more consistent with its intended semantics. The `linker_private_weak_def_auto' linkage type was meant to automatically hide globals which never had their addresses taken. It has nothing to do with the `linker_private' linkage type, which outputs the symbols with a `l' (ell) prefix among other things. The intended semantic is more like the `linkonce_odr' linkage type. Change the name of the linkage type to `linkonce_odr_auto_hide'. And therefore changing the semantics so that it produces the correct output for the linker. Note: The old linkage name `linker_private_weak_def_auto' will still parse but is not a synonym for `linkonce_odr_auto_hide'. This should be removed in 4.0. <rdar://problem/11754934> llvm-svn: 162114
2012-08-18 02:33:14 +08:00
case GlobalValue::LinkOnceODRAutoHideLinkage:
Introduce new linkage types linkonce_odr, weak_odr, common_odr and extern_weak_odr. These are the same as the non-odr versions, except that they indicate that the global will only be overridden by an *equivalent* global. In C, a function with weak linkage can be overridden by a function which behaves completely differently. This means that IP passes have to skip weak functions, since any deductions made from the function definition might be wrong, since the definition could be replaced by something completely different at link time. This is not allowed in C++, thanks to the ODR (One-Definition-Rule): if a function is replaced by another at link-time, then the new function must be the same as the original function. If a language knows that a function or other global can only be overridden by an equivalent global, it can give it the weak_odr linkage type, and the optimizers will understand that it is alright to make deductions based on the function body. The code generators on the other hand map weak and weak_odr linkage to the same thing. llvm-svn: 66339
2009-03-07 23:45:40 +08:00
case GlobalValue::WeakAnyLinkage:
case GlobalValue::WeakODRLinkage:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case GlobalValue::DLLExportLinkage:
if (!isDefine)
return Error(LinkageLoc, "invalid linkage for function declaration");
break;
case GlobalValue::AppendingLinkage:
It makes no sense to have a ODR version of common linkage, so remove it. llvm-svn: 66690
2009-03-12 04:14:15 +08:00
case GlobalValue::CommonLinkage:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(LinkageLoc, "invalid function linkage type");
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!FunctionType::isValidReturnType(RetType))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(RetTypeLoc, "invalid function return type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
add proper asmwriter and asmparser support for anonymous functions. llvm-svn: 64953
2009-02-19 05:48:13 +08:00
LocTy NameLoc = Lex.getLoc();
std::string FunctionName;
if (Lex.getKind() == lltok::GlobalVar) {
FunctionName = Lex.getStrVal();
} else if (Lex.getKind() == lltok::GlobalID) { // @42 is ok.
unsigned NameID = Lex.getUIntVal();
if (NameID != NumberedVals.size())
return TokError("function expected to be numbered '%" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(NumberedVals.size()) + "'");
add proper asmwriter and asmparser support for anonymous functions. llvm-svn: 64953
2009-02-19 05:48:13 +08:00
} else {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return TokError("expected function name");
add proper asmwriter and asmparser support for anonymous functions. llvm-svn: 64953
2009-02-19 05:48:13 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (Lex.getKind() != lltok::lparen)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return TokError("expected '(' in function argument list");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
SmallVector<ArgInfo, 8> ArgList;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool isVarArg;
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
Attributes FuncAttrs;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
std::string Section;
unsigned Alignment;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
std::string GC;
Move unnamed_addr after the function arguments on Sabre's request. llvm-svn: 124209
2011-01-26 03:09:56 +08:00
bool UnnamedAddr;
LocTy UnnamedAddrLoc;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseArgumentList(ArgList, isVarArg) ||
Move unnamed_addr after the function arguments on Sabre's request. llvm-svn: 124209
2011-01-26 03:09:56 +08:00
ParseOptionalToken(lltok::kw_unnamed_addr, UnnamedAddr,
&UnnamedAddrLoc) ||
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
ParseOptionalAttrs(FuncAttrs, 2) ||
(EatIfPresent(lltok::kw_section) &&
ParseStringConstant(Section)) ||
ParseOptionalAlignment(Alignment) ||
(EatIfPresent(lltok::kw_gc) &&
ParseStringConstant(GC)))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If the alignment was parsed as an attribute, move to the alignment field.
if (FuncAttrs & Attribute::Alignment) {
Alignment = Attribute::getAlignmentFromAttrs(FuncAttrs);
FuncAttrs &= ~Attribute::Alignment;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Okay, if we got here, the function is syntactically valid. Convert types
// and do semantic checks.
Second attempt at de-constifying LLVM Types in FunctionType::get(), StructType::get() and TargetData::getIntPtrType(). llvm-svn: 134982
2011-07-12 22:06:48 +08:00
std::vector<Type*> ParamTypeList;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
SmallVector<AttributeWithIndex, 8> Attrs;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (RetAttrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(0, RetAttrs));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
for (unsigned i = 0, e = ArgList.size(); i != e; ++i) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
ParamTypeList.push_back(ArgList[i].Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ArgList[i].Attrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(i+1, ArgList[i].Attrs));
}
if (FuncAttrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(~0, FuncAttrs));
switch AttrListPtr::get to take an ArrayRef, simplifying a lot of clients. llvm-svn: 157556
2012-05-28 09:47:44 +08:00
AttrListPtr PAL = AttrListPtr::get(Attrs);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Avoid going through the LLVMContext for type equality where it's safe to dereference the type pointer. llvm-svn: 92726
2010-01-05 21:12:22 +08:00
if (PAL.paramHasAttr(1, Attribute::StructRet) && !RetType->isVoidTy())
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
return Error(RetTypeLoc, "functions with 'sret' argument must return void");
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
FunctionType *FT =
Move types back to the 2.5 API. llvm-svn: 77516
2009-07-30 06:17:13 +08:00
FunctionType::get(RetType, ParamTypeList, isVarArg);
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
PointerType *PFT = PointerType::getUnqual(FT);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Fn = 0;
if (!FunctionName.empty()) {
// If this was a definition of a forward reference, remove the definition
// from the forward reference table and fill in the forward ref.
std::map<std::string, std::pair<GlobalValue*, LocTy> >::iterator FRVI =
ForwardRefVals.find(FunctionName);
if (FRVI != ForwardRefVals.end()) {
Fn = M->getFunction(FunctionName);
reapply 'reject forward references to functions whose type don't match' now that the testsuite has been updated. llvm-svn: 101866
2010-04-20 12:49:11 +08:00
if (Fn->getType() != PFT)
return Error(FRVI->second.second, "invalid forward reference to "
"function '" + FunctionName + "' with wrong type!");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefVals.erase(FRVI);
} else if ((Fn = M->getFunction(FunctionName))) {
make the asmparser reject function and type redefinitions. 'Merging' hasn't been needed since llvm-gcc 3.4 days. llvm-svn: 133248
2011-06-17 15:06:44 +08:00
// Reject redefinitions.
return Error(NameLoc, "invalid redefinition of function '" +
FunctionName + "'");
fix PR5295 where the .ll parser didn't reject a function after a global or global after a function with conflicting names. Update some testcases that were accidentally depending on this behavior. llvm-svn: 85081
2009-10-26 07:22:50 +08:00
} else if (M->getNamedValue(FunctionName)) {
return Error(NameLoc, "redefinition of function '@" + FunctionName + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Minor logic simplification. llvm-svn: 80453
2009-08-30 07:37:49 +08:00
} else {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If this is a definition of a forward referenced function, make sure the
// types agree.
std::map<unsigned, std::pair<GlobalValue*, LocTy> >::iterator I
= ForwardRefValIDs.find(NumberedVals.size());
if (I != ForwardRefValIDs.end()) {
Fn = cast<Function>(I->second.first);
if (Fn->getType() != PFT)
return Error(NameLoc, "type of definition and forward reference of '@" +
Push twines deeper into SourceMgr's error handling methods. llvm-svn: 114847
2010-09-28 01:42:11 +08:00
Twine(NumberedVals.size()) + "' disagree");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ForwardRefValIDs.erase(I);
}
}
if (Fn == 0)
Fn = Function::Create(FT, GlobalValue::ExternalLinkage, FunctionName, M);
else // Move the forward-reference to the correct spot in the module.
M->getFunctionList().splice(M->end(), M->getFunctionList(), Fn);
if (FunctionName.empty())
NumberedVals.push_back(Fn);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Fn->setLinkage((GlobalValue::LinkageTypes)Linkage);
Fn->setVisibility((GlobalValue::VisibilityTypes)Visibility);
Fn->setCallingConv(CC);
Fn->setAttributes(PAL);
First step in fixing PR8927: Add a unnamed_addr bit to global variables and functions. This will be used to indicate that the address is not significant and therefore the constant or function can be merged with others. If an optimization pass can show that an address is not used, it can set this. Examples of things that can have this set by the FE are globals created to hold string literals and C++ constructors. Adding unnamed_addr to a non-const global should have no effect unless an optimization can transform that global into a constant. Aliases are not allowed to have unnamed_addr since I couldn't figure out any use for it. llvm-svn: 123063
2011-01-09 00:42:36 +08:00
Fn->setUnnamedAddr(UnnamedAddr);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Fn->setAlignment(Alignment);
Fn->setSection(Section);
if (!GC.empty()) Fn->setGC(GC.c_str());
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Add all of the arguments we parsed to the function.
Function::arg_iterator ArgIt = Fn->arg_begin();
for (unsigned i = 0, e = ArgList.size(); i != e; ++i, ++ArgIt) {
// If the argument has a name, insert it into the argument symbol table.
if (ArgList[i].Name.empty()) continue;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Set the name, if it conflicted, it will be auto-renamed.
ArgIt->setName(ArgList[i].Name);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Eliminate some calls to Value::getNameStr. llvm-svn: 116670
2010-10-16 19:28:23 +08:00
if (ArgIt->getName() != ArgList[i].Name)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(ArgList[i].Loc, "redefinition of argument '%" +
ArgList[i].Name + "'");
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
/// ParseFunctionBody
/// ::= '{' BasicBlock+ '}'
///
bool LLParser::ParseFunctionBody(Function &Fn) {
stop accepting begin/end around function bodies in the .ll parser, this isn't pascal anymore. llvm-svn: 133244
2011-06-17 14:42:57 +08:00
if (Lex.getKind() != lltok::lbrace)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return TokError("expected '{' in function body");
Lex.Lex(); // eat the {.
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
int FunctionNumber = -1;
if (!Fn.hasName()) FunctionNumber = NumberedVals.size()-1;
PerFunctionState PFS(*this, Fn, FunctionNumber);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
fix PR5983, reject function bodies with no blocks per the grammar. llvm-svn: 93067
2010-01-10 03:20:07 +08:00
// We need at least one basic block.
stop accepting begin/end around function bodies in the .ll parser, this isn't pascal anymore. llvm-svn: 133244
2011-06-17 14:42:57 +08:00
if (Lex.getKind() == lltok::rbrace)
fix PR5983, reject function bodies with no blocks per the grammar. llvm-svn: 93067
2010-01-10 03:20:07 +08:00
return TokError("function body requires at least one basic block");
stop accepting begin/end around function bodies in the .ll parser, this isn't pascal anymore. llvm-svn: 133244
2011-06-17 14:42:57 +08:00
while (Lex.getKind() != lltok::rbrace)
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseBasicBlock(PFS)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Eat the }.
Lex.Lex();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Verify function is ok.
full asmparser support for blockaddress. We can now do: $ llvm-as foo.ll -d -disable-output which reads and prints the .ll file. BC encoding is the next project. Testcase will go in once that works. llvm-svn: 85368
2009-10-28 11:39:23 +08:00
return PFS.FinishFunction();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseBasicBlock
/// ::= LabelStr? Instruction*
bool LLParser::ParseBasicBlock(PerFunctionState &PFS) {
// If this basic block starts out with a name, remember it.
std::string Name;
LocTy NameLoc = Lex.getLoc();
if (Lex.getKind() == lltok::LabelStr) {
Name = Lex.getStrVal();
Lex.Lex();
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
BasicBlock *BB = PFS.DefineBB(Name, NameLoc);
if (BB == 0) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
std::string NameStr;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Parse the instructions in this block until we get a terminator.
Instruction *Inst;
now that instruction metadata is only parsed in one place, eliminate the parser-global MDsOnInst vector and make ParseInstructionMetadata return its result by-ref through an argument like the entire rest of the parser. llvm-svn: 92302
2009-12-30 13:48:36 +08:00
SmallVector<std::pair<unsigned, MDNode *>, 4> MetadataOnInst;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
do {
// This instruction may have three possibilities for a name: a) none
// specified, b) name specified "%foo =", c) number specified: "%4 =".
LocTy NameLoc = Lex.getLoc();
int NameID = -1;
NameStr = "";
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Lex.getKind() == lltok::LocalVarID) {
NameID = Lex.getUIntVal();
Lex.Lex();
if (ParseToken(lltok::equal, "expected '=' after instruction id"))
return true;
Remove support for using "foo" as symbols instead of %"foo". This is ancient syntax and has been long obsolete. As usual, updating the tests is the nasty part of this. llvm-svn: 133242
2011-06-17 14:36:20 +08:00
} else if (Lex.getKind() == lltok::LocalVar) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
NameStr = Lex.getStrVal();
Lex.Lex();
if (ParseToken(lltok::equal, "expected '=' after instruction name"))
return true;
}
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
switch (ParseInstruction(Inst, BB, PFS)) {
Convert assert(0) to llvm_unreachable llvm-svn: 149967
2012-02-07 13:05:23 +08:00
default: llvm_unreachable("Unknown ParseInstruction result!");
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
case InstError: return true;
case InstNormal:
fix a crash on invalid metadata, e.g.: call i32 @foo(), XXXX We would return the error without inserting the new instruction into the program, so it wouldn't get deallocated, and an abort would trigger when the module was deleted. llvm-svn: 100602
2010-04-07 12:08:57 +08:00
BB->getInstList().push_back(Inst);
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
// With a normal result, we check to see if the instruction is followed by
// a comma and metadata.
if (EatIfPresent(lltok::comma))
Give ParseInstructionMetadata access to the PerFunctionState object. This is in preparation for generalizing its parsing of function-local values. llvm-svn: 111893
2010-08-24 10:05:17 +08:00
if (ParseInstructionMetadata(Inst, &PFS))
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
return true;
break;
case InstExtraComma:
fix a crash on invalid metadata, e.g.: call i32 @foo(), XXXX We would return the error without inserting the new instruction into the program, so it wouldn't get deallocated, and an abort would trigger when the module was deleted. llvm-svn: 100602
2010-04-07 12:08:57 +08:00
BB->getInstList().push_back(Inst);
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
// If the instruction parser ate an extra comma at the end of it, it
// *must* be followed by metadata.
Give ParseInstructionMetadata access to the PerFunctionState object. This is in preparation for generalizing its parsing of function-local values. llvm-svn: 111893
2010-08-24 10:05:17 +08:00
if (ParseInstructionMetadata(Inst, &PFS))
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
return true;
break;
}
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Set the name on the instruction.
if (PFS.SetInstName(NameID, NameStr, NameLoc, Inst)) return true;
} while (!isa<TerminatorInst>(Inst));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
//===----------------------------------------------------------------------===//
// Instruction Parsing.
//===----------------------------------------------------------------------===//
/// ParseInstruction - Parse one of the many different instructions.
///
add facilities to start factoring instruction metadata parsing out of each opcode's handler. Change ret over so far. llvm-svn: 92298
2009-12-30 13:23:43 +08:00
int LLParser::ParseInstruction(Instruction *&Inst, BasicBlock *BB,
PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
lltok::Kind Token = Lex.getKind();
if (Token == lltok::Eof)
return TokError("found end of file when expecting more instructions");
LocTy Loc = Lex.getLoc();
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
unsigned KeywordVal = Lex.getUIntVal();
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex(); // Eat the keyword.
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
switch (Token) {
default: return Error(Loc, "expected instruction opcode");
// Terminator Instructions.
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
case lltok::kw_unreachable: Inst = new UnreachableInst(Context); return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_ret: return ParseRet(Inst, BB, PFS);
case lltok::kw_br: return ParseBr(Inst, PFS);
case lltok::kw_switch: return ParseSwitch(Inst, PFS);
rename indbr -> indirectbr to appease the residents of #llvm. llvm-svn: 85351
2009-10-28 08:19:10 +08:00
case lltok::kw_indirectbr: return ParseIndirectBr(Inst, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_invoke: return ParseInvoke(Inst, PFS);
Add the 'resume' instruction for the new EH rewrite. This adds the 'resume' instruction class, IR parsing, and bitcode reading and writing. The 'resume' instruction resumes propagation of an existing (in-flight) exception whose unwinding was interrupted with a 'landingpad' instruction (to be added later). llvm-svn: 136589
2011-07-31 14:30:59 +08:00
case lltok::kw_resume: return ParseResume(Inst, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Binary Operators.
case lltok::kw_add:
case lltok::kw_sub:
implement .ll and .bc support for nsw/nuw on shl and exact on lshr/ashr. Factor some code better. llvm-svn: 125006
2011-02-08 00:40:21 +08:00
case lltok::kw_mul:
case lltok::kw_shl: {
bool NUW = EatIfPresent(lltok::kw_nuw);
bool NSW = EatIfPresent(lltok::kw_nsw);
if (!NUW) NUW = EatIfPresent(lltok::kw_nuw);
if (ParseArithmetic(Inst, PFS, KeywordVal, 1)) return true;
if (NUW) cast<BinaryOperator>(Inst)->setHasNoUnsignedWrap(true);
if (NSW) cast<BinaryOperator>(Inst)->setHasNoSignedWrap(true);
return false;
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
}
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 06:49:04 +08:00
case lltok::kw_fadd:
case lltok::kw_fsub:
case lltok::kw_fmul: return ParseArithmetic(Inst, PFS, KeywordVal, 2);
enhance vmcore to know that udiv's can be exact, and add a trivial instcombine xform to exercise this. Nothing forms exact udivs yet though. This is progress on PR8862 llvm-svn: 124992
2011-02-07 05:44:57 +08:00
case lltok::kw_sdiv:
implement .ll and .bc support for nsw/nuw on shl and exact on lshr/ashr. Factor some code better. llvm-svn: 125006
2011-02-08 00:40:21 +08:00
case lltok::kw_udiv:
case lltok::kw_lshr:
case lltok::kw_ashr: {
bool Exact = EatIfPresent(lltok::kw_exact);
if (ParseArithmetic(Inst, PFS, KeywordVal, 1)) return true;
if (Exact) cast<BinaryOperator>(Inst)->setIsExact(true);
return false;
Change the assembly syntax for nsw, nuw, and exact, putting them after their associated opcodes rather than before. This makes them a little easier to read. llvm-svn: 77194
2009-07-28 00:11:46 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_urem:
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
case lltok::kw_srem: return ParseArithmetic(Inst, PFS, KeywordVal, 1);
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
case lltok::kw_fdiv:
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
case lltok::kw_frem: return ParseArithmetic(Inst, PFS, KeywordVal, 2);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_and:
case lltok::kw_or:
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
case lltok::kw_xor: return ParseLogical(Inst, PFS, KeywordVal);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_icmp:
Remove the vicmp and vfcmp instructions. Because we never had a release with these instructions, no autoupgrade or backwards compatibility support is provided. llvm-svn: 74991
2009-07-08 11:04:38 +08:00
case lltok::kw_fcmp: return ParseCompare(Inst, PFS, KeywordVal);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Casts.
case lltok::kw_trunc:
case lltok::kw_zext:
case lltok::kw_sext:
case lltok::kw_fptrunc:
case lltok::kw_fpext:
case lltok::kw_bitcast:
case lltok::kw_uitofp:
case lltok::kw_sitofp:
case lltok::kw_fptoui:
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
case lltok::kw_fptosi:
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_inttoptr:
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
case lltok::kw_ptrtoint: return ParseCast(Inst, PFS, KeywordVal);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Other.
case lltok::kw_select: return ParseSelect(Inst, PFS);
reject PR3281:crash07.ll with: llvm-as: crash07.ll:2:32: va_arg requires operand with first class type %y = va_arg [52 x <{}>] %43, double (...) sspreq ^ llvm-svn: 61683
2009-01-05 16:18:44 +08:00
case lltok::kw_va_arg: return ParseVA_Arg(Inst, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_extractelement: return ParseExtractElement(Inst, PFS);
case lltok::kw_insertelement: return ParseInsertElement(Inst, PFS);
case lltok::kw_shufflevector: return ParseShuffleVector(Inst, PFS);
case lltok::kw_phi: return ParsePHI(Inst, PFS);
Initial commit of the 'landingpad' instruction. This implements the 'landingpad' instruction. It's used to indicate that a basic block is a landing pad. There are several restrictions on its use (see LangRef.html for more detail). These restrictions allow the exception handling code to gather the information it needs in a much more sane way. This patch has the definition, implementation, C interface, parsing, and bitcode support in it. llvm-svn: 137501
2011-08-13 04:24:12 +08:00
case lltok::kw_landingpad: return ParseLandingPad(Inst, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_call: return ParseCall(Inst, PFS, false);
case lltok::kw_tail: return ParseCall(Inst, PFS, true);
// Memory.
Autoupgrade malloc insts to malloc calls. Update testcases that rely on malloc insts being present. Also prematurely remove MallocInst handling from IndMemRemoval and RaiseAllocations to help pass tests in this incremental step. llvm-svn: 84292
2009-10-17 08:00:19 +08:00
case lltok::kw_alloca: return ParseAlloc(Inst, PFS);
remove asmparsing and documentation support for "volatile load", which was only produced by LLVM 2.9 and earlier. LLVM 3.0 and later prefers "load volatile". llvm-svn: 145172
2011-11-27 14:56:53 +08:00
case lltok::kw_load: return ParseLoad(Inst, PFS);
case lltok::kw_store: return ParseStore(Inst, PFS);
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
case lltok::kw_cmpxchg: return ParseCmpXchg(Inst, PFS);
case lltok::kw_atomicrmw: return ParseAtomicRMW(Inst, PFS);
Initial implementation of 'fence' instruction, the new C++0x-style replacement for llvm.memory.barrier. This is just a LangRef entry and reading/writing/memory representation; optimizer+codegen support coming soon. llvm-svn: 136009
2011-07-26 07:16:38 +08:00
case lltok::kw_fence: return ParseFence(Inst, PFS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
case lltok::kw_getelementptr: return ParseGetElementPtr(Inst, PFS);
case lltok::kw_extractvalue: return ParseExtractValue(Inst, PFS);
case lltok::kw_insertvalue: return ParseInsertValue(Inst, PFS);
}
}
/// ParseCmpPredicate - Parse an integer or fp predicate, based on Kind.
bool LLParser::ParseCmpPredicate(unsigned &P, unsigned Opc) {
Remove the vicmp and vfcmp instructions. Because we never had a release with these instructions, no autoupgrade or backwards compatibility support is provided. llvm-svn: 74991
2009-07-08 11:04:38 +08:00
if (Opc == Instruction::FCmp) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
switch (Lex.getKind()) {
default: TokError("expected fcmp predicate (e.g. 'oeq')");
case lltok::kw_oeq: P = CmpInst::FCMP_OEQ; break;
case lltok::kw_one: P = CmpInst::FCMP_ONE; break;
case lltok::kw_olt: P = CmpInst::FCMP_OLT; break;
case lltok::kw_ogt: P = CmpInst::FCMP_OGT; break;
case lltok::kw_ole: P = CmpInst::FCMP_OLE; break;
case lltok::kw_oge: P = CmpInst::FCMP_OGE; break;
case lltok::kw_ord: P = CmpInst::FCMP_ORD; break;
case lltok::kw_uno: P = CmpInst::FCMP_UNO; break;
case lltok::kw_ueq: P = CmpInst::FCMP_UEQ; break;
case lltok::kw_une: P = CmpInst::FCMP_UNE; break;
case lltok::kw_ult: P = CmpInst::FCMP_ULT; break;
case lltok::kw_ugt: P = CmpInst::FCMP_UGT; break;
case lltok::kw_ule: P = CmpInst::FCMP_ULE; break;
case lltok::kw_uge: P = CmpInst::FCMP_UGE; break;
case lltok::kw_true: P = CmpInst::FCMP_TRUE; break;
case lltok::kw_false: P = CmpInst::FCMP_FALSE; break;
}
} else {
switch (Lex.getKind()) {
default: TokError("expected icmp predicate (e.g. 'eq')");
case lltok::kw_eq: P = CmpInst::ICMP_EQ; break;
case lltok::kw_ne: P = CmpInst::ICMP_NE; break;
case lltok::kw_slt: P = CmpInst::ICMP_SLT; break;
case lltok::kw_sgt: P = CmpInst::ICMP_SGT; break;
case lltok::kw_sle: P = CmpInst::ICMP_SLE; break;
case lltok::kw_sge: P = CmpInst::ICMP_SGE; break;
case lltok::kw_ult: P = CmpInst::ICMP_ULT; break;
case lltok::kw_ugt: P = CmpInst::ICMP_UGT; break;
case lltok::kw_ule: P = CmpInst::ICMP_ULE; break;
case lltok::kw_uge: P = CmpInst::ICMP_UGE; break;
}
}
Lex.Lex();
return false;
}
//===----------------------------------------------------------------------===//
// Terminator Instructions.
//===----------------------------------------------------------------------===//
/// ParseRet - Parse a return instruction.
Each instruction is allowed to have *multiple* different metadata objects on them. Though the entire compiler supports this, the asmparser didn't. llvm-svn: 92270
2009-12-30 05:25:40 +08:00
/// ::= 'ret' void (',' !dbg, !1)*
/// ::= 'ret' TypeAndValue (',' !dbg, !1)*
remove parser support for the obsolete "multiple return values" syntax, which was replaced with return of a "first class aggregate". llvm-svn: 133245
2011-06-17 14:49:41 +08:00
bool LLParser::ParseRet(Instruction *&Inst, BasicBlock *BB,
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
PerFunctionState &PFS) {
SMLoc TypeLoc = Lex.getLoc();
Type *Ty = 0;
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
if (ParseType(Ty, true /*void allowed*/)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *ResType = PFS.getFunction().getReturnType();
strength reduce a ton of type equality tests to check the typeid (Through the new predicates I added) instead of going through a context and doing a pointer comparison. Besides being cheaper, this allows a smart compiler to turn the if sequence into a switch. llvm-svn: 83297
2009-10-05 13:54:46 +08:00
if (Ty->isVoidTy()) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (!ResType->isVoidTy())
return Error(TypeLoc, "value doesn't match function result type '" +
getTypeString(ResType) + "'");
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
Inst = ReturnInst::Create(Context);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *RV;
if (ParseValue(Ty, RV, PFS)) return true;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ResType != RV->getType())
return Error(TypeLoc, "value doesn't match function result type '" +
getTypeString(ResType) + "'");
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
Inst = ReturnInst::Create(Context, RV);
remove parser support for the obsolete "multiple return values" syntax, which was replaced with return of a "first class aggregate". llvm-svn: 133245
2011-06-17 14:49:41 +08:00
return false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseBr
/// ::= 'br' TypeAndValue
/// ::= 'br' TypeAndValue ',' TypeAndValue ',' TypeAndValue
bool LLParser::ParseBr(Instruction *&Inst, PerFunctionState &PFS) {
LocTy Loc, Loc2;
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
Value *Op0;
BasicBlock *Op1, *Op2;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Op0, Loc, PFS)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (BasicBlock *BB = dyn_cast<BasicBlock>(Op0)) {
Inst = BranchInst::Create(BB);
return false;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
if (Op0->getType() != Type::getInt1Ty(Context))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "branch condition must have 'i1' type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseToken(lltok::comma, "expected ',' after branch condition") ||
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
ParseTypeAndBasicBlock(Op1, Loc, PFS) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::comma, "expected ',' after true destination") ||
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
ParseTypeAndBasicBlock(Op2, Loc2, PFS))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
Inst = BranchInst::Create(Op1, Op2, Op0);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
/// ParseSwitch
/// Instruction
/// ::= 'switch' TypeAndValue ',' TypeAndValue '[' JumpTable ']'
/// JumpTable
/// ::= (TypeAndValue ',' TypeAndValue)*
bool LLParser::ParseSwitch(Instruction *&Inst, PerFunctionState &PFS) {
LocTy CondLoc, BBLoc;
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
Value *Cond;
BasicBlock *DefaultBB;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Cond, CondLoc, PFS) ||
ParseToken(lltok::comma, "expected ',' after switch condition") ||
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
ParseTypeAndBasicBlock(DefaultBB, BBLoc, PFS) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::lsquare, "expected '[' with switch table"))
return true;
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Cond->getType()->isIntegerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(CondLoc, "switch condition must have integer type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Parse the jump table pairs.
SmallPtrSet<Value*, 32> SeenCases;
SmallVector<std::pair<ConstantInt*, BasicBlock*>, 32> Table;
while (Lex.getKind() != lltok::rsquare) {
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
Value *Constant;
BasicBlock *DestBB;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Constant, CondLoc, PFS) ||
ParseToken(lltok::comma, "expected ',' after case value") ||
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
ParseTypeAndBasicBlock(DestBB, PFS))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!SeenCases.insert(Constant))
return Error(CondLoc, "duplicate case value in switch");
if (!isa<ConstantInt>(Constant))
return Error(CondLoc, "case value is not a constant integer");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
Table.push_back(std::make_pair(cast<ConstantInt>(Constant), DestBB));
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Lex.Lex(); // Eat the ']'.
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
SwitchInst *SI = SwitchInst::Create(Cond, DefaultBB, Table.size());
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
for (unsigned i = 0, e = Table.size(); i != e; ++i)
SI->addCase(Table[i].first, Table[i].second);
Inst = SI;
return false;
}
rename indbr -> indirectbr to appease the residents of #llvm. llvm-svn: 85351
2009-10-28 08:19:10 +08:00
/// ParseIndirectBr
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
/// Instruction
rename indbr -> indirectbr to appease the residents of #llvm. llvm-svn: 85351
2009-10-28 08:19:10 +08:00
/// ::= 'indirectbr' TypeAndValue ',' '[' LabelList ']'
bool LLParser::ParseIndirectBr(Instruction *&Inst, PerFunctionState &PFS) {
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
LocTy AddrLoc;
Value *Address;
if (ParseTypeAndValue(Address, AddrLoc, PFS) ||
rename indbr -> indirectbr to appease the residents of #llvm. llvm-svn: 85351
2009-10-28 08:19:10 +08:00
ParseToken(lltok::comma, "expected ',' after indirectbr address") ||
ParseToken(lltok::lsquare, "expected '[' with indirectbr"))
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
return true;
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Address->getType()->isPointerTy())
rename indbr -> indirectbr to appease the residents of #llvm. llvm-svn: 85351
2009-10-28 08:19:10 +08:00
return Error(AddrLoc, "indirectbr address must have pointer type");
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
// Parse the destination list.
SmallVector<BasicBlock*, 16> DestList;
if (Lex.getKind() != lltok::rsquare) {
BasicBlock *DestBB;
if (ParseTypeAndBasicBlock(DestBB, PFS))
return true;
DestList.push_back(DestBB);
while (EatIfPresent(lltok::comma)) {
if (ParseTypeAndBasicBlock(DestBB, PFS))
return true;
DestList.push_back(DestBB);
}
}
if (ParseToken(lltok::rsquare, "expected ']' at end of block list"))
return true;
rename indbr -> indirectbr to appease the residents of #llvm. llvm-svn: 85351
2009-10-28 08:19:10 +08:00
IndirectBrInst *IBI = IndirectBrInst::Create(Address, DestList.size());
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
for (unsigned i = 0, e = DestList.size(); i != e; ++i)
IBI->addDestination(DestList[i]);
Inst = IBI;
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseInvoke
/// ::= 'invoke' OptionalCallingConv OptionalAttrs Type Value ParamList
/// OptionalAttrs 'to' TypeAndValue 'unwind' TypeAndValue
bool LLParser::ParseInvoke(Instruction *&Inst, PerFunctionState &PFS) {
LocTy CallLoc = Lex.getLoc();
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
Attributes RetAttrs, FnAttrs;
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
CallingConv::ID CC;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *RetType = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy RetTypeLoc;
ValID CalleeID;
SmallVector<ParamInfo, 16> ArgList;
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
BasicBlock *NormalBB, *UnwindBB;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseOptionalCallingConv(CC) ||
ParseOptionalAttrs(RetAttrs, 1) ||
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
ParseType(RetType, RetTypeLoc, true /*void allowed*/) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseValID(CalleeID) ||
ParseParameterList(ArgList, PFS) ||
ParseOptionalAttrs(FnAttrs, 2) ||
ParseToken(lltok::kw_to, "expected 'to' in invoke") ||
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
ParseTypeAndBasicBlock(NormalBB, PFS) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::kw_unwind, "expected 'unwind' in invoke") ||
add enough support for indirect branch for the feature test to pass (assembler,asmprinter, bc reader+writer) and document it. Codegen currently aborts on it. llvm-svn: 85274
2009-10-28 03:13:16 +08:00
ParseTypeAndBasicBlock(UnwindBB, PFS))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If RetType is a non-function pointer type, then this is the short syntax
// for the call, which means that RetType is just the return type. Infer the
// rest of the function argument types from the arguments that are present.
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
PointerType *PFTy = 0;
FunctionType *Ty = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!(PFTy = dyn_cast<PointerType>(RetType)) ||
!(Ty = dyn_cast<FunctionType>(PFTy->getElementType()))) {
// Pull out the types of all of the arguments...
Second attempt at de-constifying LLVM Types in FunctionType::get(), StructType::get() and TargetData::getIntPtrType(). llvm-svn: 134982
2011-07-12 22:06:48 +08:00
std::vector<Type*> ParamTypes;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
for (unsigned i = 0, e = ArgList.size(); i != e; ++i)
ParamTypes.push_back(ArgList[i].V->getType());
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!FunctionType::isValidReturnType(RetType))
return Error(RetTypeLoc, "Invalid result type for LLVM function");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Move types back to the 2.5 API. llvm-svn: 77516
2009-07-30 06:17:13 +08:00
Ty = FunctionType::get(RetType, ParamTypes, false);
PFTy = PointerType::getUnqual(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Look up the callee.
Value *Callee;
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
if (ConvertValIDToValue(PFTy, CalleeID, Callee, &PFS)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Set up the Attributes for the function.
SmallVector<AttributeWithIndex, 8> Attrs;
if (RetAttrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(0, RetAttrs));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
SmallVector<Value*, 8> Args;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Loop through FunctionType's arguments and ensure they are specified
// correctly. Also, gather any parameter attributes.
FunctionType::param_iterator I = Ty->param_begin();
FunctionType::param_iterator E = Ty->param_end();
for (unsigned i = 0, e = ArgList.size(); i != e; ++i) {
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
Type *ExpectedTy = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (I != E) {
ExpectedTy = *I++;
} else if (!Ty->isVarArg()) {
return Error(ArgList[i].Loc, "too many arguments specified");
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ExpectedTy && ExpectedTy != ArgList[i].V->getType())
return Error(ArgList[i].Loc, "argument is not of expected type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(ExpectedTy) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Args.push_back(ArgList[i].V);
if (ArgList[i].Attrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(i+1, ArgList[i].Attrs));
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (I != E)
return Error(CallLoc, "not enough parameters specified for call");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (FnAttrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(~0, FnAttrs));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Finish off the Attributes and check them
switch AttrListPtr::get to take an ArrayRef, simplifying a lot of clients. llvm-svn: 157556
2012-05-28 09:47:44 +08:00
AttrListPtr PAL = AttrListPtr::get(Attrs);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Convert CallInst and InvokeInst APIs to use ArrayRef. llvm-svn: 135265
2011-07-15 16:37:34 +08:00
InvokeInst *II = InvokeInst::Create(Callee, NormalBB, UnwindBB, Args);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
II->setCallingConv(CC);
II->setAttributes(PAL);
Inst = II;
return false;
}
Add the 'resume' instruction for the new EH rewrite. This adds the 'resume' instruction class, IR parsing, and bitcode reading and writing. The 'resume' instruction resumes propagation of an existing (in-flight) exception whose unwinding was interrupted with a 'landingpad' instruction (to be added later). llvm-svn: 136589
2011-07-31 14:30:59 +08:00
/// ParseResume
/// ::= 'resume' TypeAndValue
bool LLParser::ParseResume(Instruction *&Inst, PerFunctionState &PFS) {
Value *Exn; LocTy ExnLoc;
if (ParseTypeAndValue(Exn, ExnLoc, PFS))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Add the 'resume' instruction for the new EH rewrite. This adds the 'resume' instruction class, IR parsing, and bitcode reading and writing. The 'resume' instruction resumes propagation of an existing (in-flight) exception whose unwinding was interrupted with a 'landingpad' instruction (to be added later). llvm-svn: 136589
2011-07-31 14:30:59 +08:00
ResumeInst *RI = ResumeInst::Create(Exn);
Inst = RI;
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
//===----------------------------------------------------------------------===//
// Binary Operators.
//===----------------------------------------------------------------------===//
/// ParseArithmetic
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
/// ::= ArithmeticOps TypeAndValue ',' Value
///
/// If OperandType is 0, then any FP or integer operand is allowed. If it is 1,
/// then any integer operand is allowed, if it is 2, any fp operand is allowed.
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
bool LLParser::ParseArithmetic(Instruction *&Inst, PerFunctionState &PFS,
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
unsigned Opc, unsigned OperandType) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy Loc; Value *LHS, *RHS;
if (ParseTypeAndValue(LHS, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' in arithmetic operation") ||
ParseValue(LHS->getType(), RHS, PFS))
return true;
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
bool Valid;
switch (OperandType) {
llvm_unreachable->llvm_unreachable(0), LLVM_UNREACHABLE->llvm_unreachable. This adds location info for all llvm_unreachable calls (which is a macro now) in !NDEBUG builds. In NDEBUG builds location info and the message is off (it only prints "UREACHABLE executed"). llvm-svn: 75640
2009-07-15 00:55:14 +08:00
default: llvm_unreachable("Unknown operand type!");
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
case 0: // int or FP.
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
Valid = LHS->getType()->isIntOrIntVectorTy() ||
LHS->getType()->isFPOrFPVectorTy();
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
break;
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
case 1: Valid = LHS->getType()->isIntOrIntVectorTy(); break;
case 2: Valid = LHS->getType()->isFPOrFPVectorTy(); break;
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Fix PR3281:crash08.ll with this diagnostic: llvm-as: crash08.ll:3:15: invalid operand type for instruction "qp" = sdiv fp128 0x1, %30 ^ llvm-svn: 61684
2009-01-05 16:24:46 +08:00
if (!Valid)
return Error(Loc, "invalid operand type for instruction");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Inst = BinaryOperator::Create((Instruction::BinaryOps)Opc, LHS, RHS);
return false;
}
/// ParseLogical
/// ::= ArithmeticOps TypeAndValue ',' Value {
bool LLParser::ParseLogical(Instruction *&Inst, PerFunctionState &PFS,
unsigned Opc) {
LocTy Loc; Value *LHS, *RHS;
if (ParseTypeAndValue(LHS, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' in logical operation") ||
ParseValue(LHS->getType(), RHS, PFS))
return true;
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!LHS->getType()->isIntOrIntVectorTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc,"instruction requires integer or integer vector operands");
Inst = BinaryOperator::Create((Instruction::BinaryOps)Opc, LHS, RHS);
return false;
}
/// ParseCompare
/// ::= 'icmp' IPredicates TypeAndValue ',' Value
/// ::= 'fcmp' FPredicates TypeAndValue ',' Value
bool LLParser::ParseCompare(Instruction *&Inst, PerFunctionState &PFS,
unsigned Opc) {
// Parse the integer/fp comparison predicate.
LocTy Loc;
unsigned Pred;
Value *LHS, *RHS;
if (ParseCmpPredicate(Pred, Opc) ||
ParseTypeAndValue(LHS, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' after compare value") ||
ParseValue(LHS->getType(), RHS, PFS))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (Opc == Instruction::FCmp) {
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!LHS->getType()->isFPOrFPVectorTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "fcmp requires floating point operands");
Eliminate the unused Context argument on one of the ICmpInst and FCmpInst constructors. llvm-svn: 80049
2009-08-26 07:17:54 +08:00
Inst = new FCmpInst(CmpInst::Predicate(Pred), LHS, RHS);
Remove the vicmp and vfcmp instructions. Because we never had a release with these instructions, no autoupgrade or backwards compatibility support is provided. llvm-svn: 74991
2009-07-08 11:04:38 +08:00
} else {
assert(Opc == Instruction::ICmp && "Unknown opcode for CmpInst!");
Uniformize the names of type predicates: rather than having isFloatTy and isInteger, we now have isFloatTy and isIntegerTy. Requested by Chris! llvm-svn: 96223
2010-02-16 00:12:20 +08:00
if (!LHS->getType()->isIntOrIntVectorTy() &&
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
!LHS->getType()->getScalarType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "icmp requires integer operands");
Eliminate the unused Context argument on one of the ICmpInst and FCmpInst constructors. llvm-svn: 80049
2009-08-26 07:17:54 +08:00
Inst = new ICmpInst(CmpInst::Predicate(Pred), LHS, RHS);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
return false;
}
//===----------------------------------------------------------------------===//
// Other Instructions.
//===----------------------------------------------------------------------===//
/// ParseCast
/// ::= CastOpc TypeAndValue 'to' Type
bool LLParser::ParseCast(Instruction *&Inst, PerFunctionState &PFS,
unsigned Opc) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
LocTy Loc;
Value *Op;
Type *DestTy = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Op, Loc, PFS) ||
ParseToken(lltok::kw_to, "expected 'to' after cast value") ||
ParseType(DestTy))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
if (!CastInst::castIsValid((Instruction::CastOps)Opc, Op, DestTy)) {
CastInst::castIsValid((Instruction::CastOps)Opc, Op, DestTy);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "invalid cast opcode for cast from '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(Op->getType()) + "' to '" +
getTypeString(DestTy) + "'");
Fix a pretty awesome bug that only happened in a strange case with anonymous types. This was reading the uint for the keyword after the token was advanced. llvm-svn: 65743
2009-03-01 08:53:13 +08:00
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Inst = CastInst::Create((Instruction::CastOps)Opc, Op, DestTy);
return false;
}
/// ParseSelect
/// ::= 'select' TypeAndValue ',' TypeAndValue ',' TypeAndValue
bool LLParser::ParseSelect(Instruction *&Inst, PerFunctionState &PFS) {
LocTy Loc;
Value *Op0, *Op1, *Op2;
if (ParseTypeAndValue(Op0, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' after select condition") ||
ParseTypeAndValue(Op1, PFS) ||
ParseToken(lltok::comma, "expected ',' after select value") ||
ParseTypeAndValue(Op2, PFS))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (const char *Reason = SelectInst::areInvalidOperands(Op0, Op1, Op2))
return Error(Loc, Reason);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Inst = SelectInst::Create(Op0, Op1, Op2);
return false;
}
reject PR3281:crash07.ll with: llvm-as: crash07.ll:2:32: va_arg requires operand with first class type %y = va_arg [52 x <{}>] %43, double (...) sspreq ^ llvm-svn: 61683
2009-01-05 16:18:44 +08:00
/// ParseVA_Arg
/// ::= 'va_arg' TypeAndValue ',' Type
bool LLParser::ParseVA_Arg(Instruction *&Inst, PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Op;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *EltTy = 0;
reject PR3281:crash07.ll with: llvm-as: crash07.ll:2:32: va_arg requires operand with first class type %y = va_arg [52 x <{}>] %43, double (...) sspreq ^ llvm-svn: 61683
2009-01-05 16:18:44 +08:00
LocTy TypeLoc;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Op, PFS) ||
ParseToken(lltok::comma, "expected ',' after vaarg operand") ||
reject PR3281:crash07.ll with: llvm-as: crash07.ll:2:32: va_arg requires operand with first class type %y = va_arg [52 x <{}>] %43, double (...) sspreq ^ llvm-svn: 61683
2009-01-05 16:18:44 +08:00
ParseType(EltTy, TypeLoc))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
reject PR3281:crash07.ll with: llvm-as: crash07.ll:2:32: va_arg requires operand with first class type %y = va_arg [52 x <{}>] %43, double (...) sspreq ^ llvm-svn: 61683
2009-01-05 16:18:44 +08:00
if (!EltTy->isFirstClassType())
return Error(TypeLoc, "va_arg requires operand with first class type");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Inst = new VAArgInst(Op, EltTy);
return false;
}
/// ParseExtractElement
/// ::= 'extractelement' TypeAndValue ',' TypeAndValue
bool LLParser::ParseExtractElement(Instruction *&Inst, PerFunctionState &PFS) {
LocTy Loc;
Value *Op0, *Op1;
if (ParseTypeAndValue(Op0, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' after extract value") ||
ParseTypeAndValue(Op1, PFS))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!ExtractElementInst::isValidOperands(Op0, Op1))
return Error(Loc, "invalid extractelement operands");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Move ExtractElementInst to ::Create instead of new. Update all uses. llvm-svn: 77044
2009-07-25 10:28:41 +08:00
Inst = ExtractElementInst::Create(Op0, Op1);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return false;
}
/// ParseInsertElement
/// ::= 'insertelement' TypeAndValue ',' TypeAndValue ',' TypeAndValue
bool LLParser::ParseInsertElement(Instruction *&Inst, PerFunctionState &PFS) {
LocTy Loc;
Value *Op0, *Op1, *Op2;
if (ParseTypeAndValue(Op0, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' after insertelement value") ||
ParseTypeAndValue(Op1, PFS) ||
ParseToken(lltok::comma, "expected ',' after insertelement value") ||
ParseTypeAndValue(Op2, PFS))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!InsertElementInst::isValidOperands(Op0, Op1, Op2))
Fix error message for correct opcode. llvm-svn: 76829
2009-07-23 09:01:32 +08:00
return Error(Loc, "invalid insertelement operands");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Inst = InsertElementInst::Create(Op0, Op1, Op2);
return false;
}
/// ParseShuffleVector
/// ::= 'shufflevector' TypeAndValue ',' TypeAndValue ',' TypeAndValue
bool LLParser::ParseShuffleVector(Instruction *&Inst, PerFunctionState &PFS) {
LocTy Loc;
Value *Op0, *Op1, *Op2;
if (ParseTypeAndValue(Op0, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' after shuffle mask") ||
ParseTypeAndValue(Op1, PFS) ||
ParseToken(lltok::comma, "expected ',' after shuffle value") ||
ParseTypeAndValue(Op2, PFS))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!ShuffleVectorInst::isValidOperands(Op0, Op1, Op2))
Typo llvm-svn: 149562
2012-02-02 07:43:12 +08:00
return Error(Loc, "invalid shufflevector operands");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Inst = new ShuffleVectorInst(Op0, Op1, Op2);
return false;
}
/// ParsePHI
remove some nonascii weird stuff llvm-svn: 84410
2009-10-18 13:27:44 +08:00
/// ::= 'phi' Type '[' Value ',' Value ']' (',' '[' Value ',' Value ']')*
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
int LLParser::ParsePHI(Instruction *&Inst, PerFunctionState &PFS) {
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Ty = 0; LocTy TypeLoc;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Op0, *Op1;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseType(Ty, TypeLoc) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::lsquare, "expected '[' in phi value list") ||
ParseValue(Ty, Op0, PFS) ||
ParseToken(lltok::comma, "expected ',' after insertelement value") ||
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
ParseValue(Type::getLabelTy(Context), Op1, PFS) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::rsquare, "expected ']' in phi value list"))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
bool AteExtraComma = false;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
SmallVector<std::pair<Value*, BasicBlock*>, 16> PHIVals;
while (1) {
PHIVals.push_back(std::make_pair(Op0, cast<BasicBlock>(Op1)));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (!EatIfPresent(lltok::comma))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
break;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
if (Lex.getKind() == lltok::MetadataVar) {
AteExtraComma = true;
Parse PHI instruction with attached metadata. llvm-svn: 84264
2009-10-17 02:45:49 +08:00
break;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
}
Parse PHI instruction with attached metadata. llvm-svn: 84264
2009-10-17 02:45:49 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseToken(lltok::lsquare, "expected '[' in phi value list") ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseValue(Ty, Op0, PFS) ||
ParseToken(lltok::comma, "expected ',' after insertelement value") ||
Push LLVMContexts through the IntegerType APIs. llvm-svn: 78948
2009-08-14 05:58:54 +08:00
ParseValue(Type::getLabelTy(Context), Op1, PFS) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseToken(lltok::rsquare, "expected ']' in phi value list"))
return true;
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!Ty->isFirstClassType())
return Error(TypeLoc, "phi node must have first class type");
Remove PHINode::reserveOperandSpace(). Instead, add a parameter to PHINode::Create() giving the (known or expected) number of operands. llvm-svn: 128537
2011-03-30 19:28:46 +08:00
PHINode *PN = PHINode::Create(Ty, PHIVals.size());
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
for (unsigned i = 0, e = PHIVals.size(); i != e; ++i)
PN->addIncoming(PHIVals[i].first, PHIVals[i].second);
Inst = PN;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Initial commit of the 'landingpad' instruction. This implements the 'landingpad' instruction. It's used to indicate that a basic block is a landing pad. There are several restrictions on its use (see LangRef.html for more detail). These restrictions allow the exception handling code to gather the information it needs in a much more sane way. This patch has the definition, implementation, C interface, parsing, and bitcode support in it. llvm-svn: 137501
2011-08-13 04:24:12 +08:00
/// ParseLandingPad
/// ::= 'landingpad' Type 'personality' TypeAndValue 'cleanup'? Clause+
/// Clause
/// ::= 'catch' TypeAndValue
/// ::= 'filter'
/// ::= 'filter' TypeAndValue ( ',' TypeAndValue )*
bool LLParser::ParseLandingPad(Instruction *&Inst, PerFunctionState &PFS) {
Type *Ty = 0; LocTy TyLoc;
Value *PersFn; LocTy PersFnLoc;
if (ParseType(Ty, TyLoc) ||
ParseToken(lltok::kw_personality, "expected 'personality'") ||
ParseTypeAndValue(PersFn, PersFnLoc, PFS))
return true;
LandingPadInst *LP = LandingPadInst::Create(Ty, PersFn, 0);
LP->setCleanup(EatIfPresent(lltok::kw_cleanup));
while (Lex.getKind() == lltok::kw_catch || Lex.getKind() == lltok::kw_filter){
LandingPadInst::ClauseType CT;
if (EatIfPresent(lltok::kw_catch))
CT = LandingPadInst::Catch;
else if (EatIfPresent(lltok::kw_filter))
CT = LandingPadInst::Filter;
else
return TokError("expected 'catch' or 'filter' clause type");
Value *V; LocTy VLoc;
if (ParseTypeAndValue(V, VLoc, PFS)) {
delete LP;
return true;
}
Add checks for the landingpad instruction's clause values to make sure that they're the correct type. llvm-svn: 137511
2011-08-13 04:52:25 +08:00
// A 'catch' type expects a non-array constant. A filter clause expects an
// array constant.
if (CT == LandingPadInst::Catch) {
if (isa<ArrayType>(V->getType()))
Error(VLoc, "'catch' clause has an invalid type");
} else {
if (!isa<ArrayType>(V->getType()))
Error(VLoc, "'filter' clause has an invalid type");
}
Initial commit of the 'landingpad' instruction. This implements the 'landingpad' instruction. It's used to indicate that a basic block is a landing pad. There are several restrictions on its use (see LangRef.html for more detail). These restrictions allow the exception handling code to gather the information it needs in a much more sane way. This patch has the definition, implementation, C interface, parsing, and bitcode support in it. llvm-svn: 137501
2011-08-13 04:24:12 +08:00
LP->addClause(V);
}
Inst = LP;
return false;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseCall
/// ::= 'tail'? 'call' OptionalCallingConv OptionalAttrs Type Value
/// ParameterList OptionalAttrs
bool LLParser::ParseCall(Instruction *&Inst, PerFunctionState &PFS,
bool isTail) {
Extend Attributes to 64 bits Problem: LLVM needs more function attributes than currently available (32 bits). One such proposed attribute is "address_safety", which shows that a function is being checked for address safety (by AddressSanitizer, SAFECode, etc). Solution: - extend the Attributes from 32 bits to 64-bits - wrap the object into a class so that unsigned is never erroneously used instead - change "unsigned" to "Attributes" throughout the code, including one place in clang. - the class has no "operator uint64 ()", but it has "uint64_t Raw() " to support packing/unpacking. - the class has "safe operator bool()" to support the common idiom: if (Attributes attr = getAttrs()) useAttrs(attr); - The CTOR from uint64_t is marked explicit, so I had to add a few explicit CTOR calls - Add the new attribute "address_safety". Doing it in the same commit to check that attributes beyond first 32 bits actually work. - Some of the functions from the Attribute namespace are worth moving inside the class, but I'd prefer to have it as a separate commit. Tested: "make check" on Linux (32-bit and 64-bit) and Mac (10.6) built/run spec CPU 2006 on Linux with clang -O2. This change will break clang build in lib/CodeGen/CGCall.cpp. The following patch will fix it. llvm-svn: 148553
2012-01-21 01:56:17 +08:00
Attributes RetAttrs, FnAttrs;
Retype from unsigned to CallingConv::ID accordingly. Approved by Bob Wilson. llvm-svn: 80773
2009-09-02 16:44:58 +08:00
CallingConv::ID CC;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *RetType = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
LocTy RetTypeLoc;
ValID CalleeID;
SmallVector<ParamInfo, 16> ArgList;
LocTy CallLoc = Lex.getLoc();
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if ((isTail && ParseToken(lltok::kw_call, "expected 'tail call'")) ||
ParseOptionalCallingConv(CC) ||
ParseOptionalAttrs(RetAttrs, 1) ||
Fix two classes of bugs. First: validate an invariant so that the asmparser rejects a bad construct instead of the verifier. Before: llvm-as: assembly parsed, but does not verify as correct! Invalid struct return type! i64 (%struct.Type*, %struct.Type*)* @foo after: llvm-as: t.ll:5:8: functions with 'sret' argument must return void define i64 @foo(%struct.Type* noalias nocapture sret %agg.result, %struct.Type* nocapture byval %t) nounwind { ^ Second, check that void is only used where allowed (in function return types) not in arbitrary places, fixing PR3747 - Crash in llvm-as with void field in struct. We now reject that example with: $ llvm-as t.ll llvm-as: t.ll:1:12: struct element can not have void type %x = type {void} ^ llvm-svn: 66394
2009-03-09 12:49:14 +08:00
ParseType(RetType, RetTypeLoc, true /*void allowed*/) ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
ParseValID(CalleeID) ||
ParseParameterList(ArgList, PFS) ||
ParseOptionalAttrs(FnAttrs, 2))
return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// If RetType is a non-function pointer type, then this is the short syntax
// for the call, which means that RetType is just the return type. Infer the
// rest of the function argument types from the arguments that are present.
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
PointerType *PFTy = 0;
FunctionType *Ty = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!(PFTy = dyn_cast<PointerType>(RetType)) ||
!(Ty = dyn_cast<FunctionType>(PFTy->getElementType()))) {
// Pull out the types of all of the arguments...
Second attempt at de-constifying LLVM Types in FunctionType::get(), StructType::get() and TargetData::getIntPtrType(). llvm-svn: 134982
2011-07-12 22:06:48 +08:00
std::vector<Type*> ParamTypes;
Revert r109361; it's impossible to write a call with an argument with an invalid type. llvm-svn: 109365
2010-07-25 07:06:59 +08:00
for (unsigned i = 0, e = ArgList.size(); i != e; ++i)
ParamTypes.push_back(ArgList[i].V->getType());
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (!FunctionType::isValidReturnType(RetType))
return Error(RetTypeLoc, "Invalid result type for LLVM function");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Move types back to the 2.5 API. llvm-svn: 77516
2009-07-30 06:17:13 +08:00
Ty = FunctionType::get(RetType, ParamTypes, false);
PFTy = PointerType::getUnqual(Ty);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Look up the callee.
Value *Callee;
Consolidate ConvertGlobalValIDToValue, ConvertGlobalOrMetadataValIDToValue, and ConvertValIDToValue into a more powerful ConvertValIDToValue() that does all three's work llvm-svn: 93197
2010-01-12 06:31:58 +08:00
if (ConvertValIDToValue(PFTy, CalleeID, Callee, &PFS)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Set up the Attributes for the function.
SmallVector<AttributeWithIndex, 8> Attrs;
if (RetAttrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(0, RetAttrs));
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
SmallVector<Value*, 8> Args;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
// Loop through FunctionType's arguments and ensure they are specified
// correctly. Also, gather any parameter attributes.
FunctionType::param_iterator I = Ty->param_begin();
FunctionType::param_iterator E = Ty->param_end();
for (unsigned i = 0, e = ArgList.size(); i != e; ++i) {
land David Blaikie's patch to de-constify Type, with a few tweaks. llvm-svn: 135375
2011-07-18 12:54:35 +08:00
Type *ExpectedTy = 0;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (I != E) {
ExpectedTy = *I++;
} else if (!Ty->isVarArg()) {
return Error(ArgList[i].Loc, "too many arguments specified");
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ExpectedTy && ExpectedTy != ArgList[i].V->getType())
return Error(ArgList[i].Loc, "argument is not of expected type '" +
eliminate the Type::getDescription() method, using "<<" instead. This removes some gunk from LLVMContext. llvm-svn: 133360
2011-06-19 05:18:23 +08:00
getTypeString(ExpectedTy) + "'");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Args.push_back(ArgList[i].V);
if (ArgList[i].Attrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(i+1, ArgList[i].Attrs));
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (I != E)
return Error(CallLoc, "not enough parameters specified for call");
if (FnAttrs != Attribute::None)
Attrs.push_back(AttributeWithIndex::get(~0, FnAttrs));
// Finish off the Attributes and check them
switch AttrListPtr::get to take an ArrayRef, simplifying a lot of clients. llvm-svn: 157556
2012-05-28 09:47:44 +08:00
AttrListPtr PAL = AttrListPtr::get(Attrs);
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Convert CallInst and InvokeInst APIs to use ArrayRef. llvm-svn: 135265
2011-07-15 16:37:34 +08:00
CallInst *CI = CallInst::Create(Callee, Args);
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
CI->setTailCall(isTail);
CI->setCallingConv(CC);
CI->setAttributes(PAL);
Inst = CI;
return false;
}
//===----------------------------------------------------------------------===//
// Memory Instructions.
//===----------------------------------------------------------------------===//
/// ParseAlloc
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
/// ::= 'alloca' Type (',' TypeAndValue)? (',' OptionalInfo)?
Remove old backwards compatibility support from the parser for autoupgrading the old malloc/free instructions, and for 'sext' and 'zext' as function attributes (they are spelled signext/zeroext now), and support for result value attributes being specified after a function. Additionally, diagnose invalid attributes on functions with an error message instead of an abort in the verifier. llvm-svn: 133229
2011-06-17 11:16:47 +08:00
int LLParser::ParseAlloc(Instruction *&Inst, PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Size = 0;
switch the .ll parser to use SourceMgr. llvm-svn: 74735
2009-07-03 07:08:13 +08:00
LocTy SizeLoc;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
unsigned Alignment = 0;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
Type *Ty = 0;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseType(Ty)) return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
bool AteExtraComma = false;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (EatIfPresent(lltok::comma)) {
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
if (Lex.getKind() == lltok::kw_align) {
if (ParseOptionalAlignment(Alignment)) return true;
} else if (Lex.getKind() == lltok::MetadataVar) {
AteExtraComma = true;
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
} else {
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
if (ParseTypeAndValue(Size, SizeLoc, PFS) ||
ParseOptionalCommaAlign(Alignment, AteExtraComma))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
}
Eliminate the restriction that the array size in an alloca must be i32. This will help reduce the amount of casting required on 64-bit targets. llvm-svn: 104911
2010-05-28 09:14:11 +08:00
if (Size && !Size->getType()->isIntegerTy())
return Error(SizeLoc, "element count must have integer type");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Remove old backwards compatibility support from the parser for autoupgrading the old malloc/free instructions, and for 'sext' and 'zext' as function attributes (they are spelled signext/zeroext now), and support for result value attributes being specified after a function. Additionally, diagnose invalid attributes on functions with an error message instead of an abort in the verifier. llvm-svn: 133229
2011-06-17 11:16:47 +08:00
Inst = new AllocaInst(Ty, Size, Alignment);
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseLoad
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
/// ::= 'load' 'volatile'? TypeAndValue (',' 'align' i32)?
/// ::= 'load' 'atomic' 'volatile'? TypeAndValue
/// 'singlethread'? AtomicOrdering (',' 'align' i32)?
remove asmparsing and documentation support for "volatile load", which was only produced by LLVM 2.9 and earlier. LLVM 3.0 and later prefers "load volatile". llvm-svn: 145172
2011-11-27 14:56:53 +08:00
int LLParser::ParseLoad(Instruction *&Inst, PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Val; LocTy Loc;
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
unsigned Alignment = 0;
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
bool AteExtraComma = false;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
bool isAtomic = false;
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
AtomicOrdering Ordering = NotAtomic;
SynchronizationScope Scope = CrossThread;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
if (Lex.getKind() == lltok::kw_atomic) {
isAtomic = true;
Lex.Lex();
}
remove asmparsing and documentation support for "volatile load", which was only produced by LLVM 2.9 and earlier. LLVM 3.0 and later prefers "load volatile". llvm-svn: 145172
2011-11-27 14:56:53 +08:00
bool isVolatile = false;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
if (Lex.getKind() == lltok::kw_volatile) {
isVolatile = true;
Lex.Lex();
}
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
if (ParseTypeAndValue(Val, Loc, PFS) ||
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
ParseScopeAndOrdering(isAtomic, Scope, Ordering) ||
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
ParseOptionalCommaAlign(Alignment, AteExtraComma))
return true;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Val->getType()->isPointerTy() ||
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
!cast<PointerType>(Val->getType())->getElementType()->isFirstClassType())
return Error(Loc, "load operand must be a pointer to a first class type");
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
if (isAtomic && !Alignment)
return Error(Loc, "atomic load must have explicit non-zero alignment");
if (Ordering == Release || Ordering == AcquireRelease)
return Error(Loc, "atomic load cannot use Release ordering");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
Inst = new LoadInst(Val, "", isVolatile, Alignment, Ordering, Scope);
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseStore
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
/// ::= 'store' 'volatile'? TypeAndValue ',' TypeAndValue (',' 'align' i32)?
/// ::= 'store' 'atomic' 'volatile'? TypeAndValue ',' TypeAndValue
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
/// 'singlethread'? AtomicOrdering (',' 'align' i32)?
remove asmparsing and documentation support for "volatile load", which was only produced by LLVM 2.9 and earlier. LLVM 3.0 and later prefers "load volatile". llvm-svn: 145172
2011-11-27 14:56:53 +08:00
int LLParser::ParseStore(Instruction *&Inst, PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Val, *Ptr; LocTy Loc, PtrLoc;
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
unsigned Alignment = 0;
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
bool AteExtraComma = false;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
bool isAtomic = false;
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
AtomicOrdering Ordering = NotAtomic;
SynchronizationScope Scope = CrossThread;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
if (Lex.getKind() == lltok::kw_atomic) {
isAtomic = true;
Lex.Lex();
}
remove asmparsing and documentation support for "volatile load", which was only produced by LLVM 2.9 and earlier. LLVM 3.0 and later prefers "load volatile". llvm-svn: 145172
2011-11-27 14:56:53 +08:00
bool isVolatile = false;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
if (Lex.getKind() == lltok::kw_volatile) {
isVolatile = true;
Lex.Lex();
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Val, Loc, PFS) ||
ParseToken(lltok::comma, "expected ',' after store operand") ||
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
ParseTypeAndValue(Ptr, PtrLoc, PFS) ||
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
ParseScopeAndOrdering(isAtomic, Scope, Ordering) ||
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
ParseOptionalCommaAlign(Alignment, AteExtraComma))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Fix parsing of optional metadata for 'load', 'store' and 'alloc' instructions. llvm-svn: 82175
2009-09-18 07:04:48 +08:00
There are two ways of checking for a given type, for example isa<PointerType>(T) and T->isPointerTy(). Convert most instances of the first form to the second form. Requested by Chris. llvm-svn: 96344
2010-02-16 19:11:14 +08:00
if (!Ptr->getType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(PtrLoc, "store operand must be a pointer");
if (!Val->getType()->isFirstClassType())
return Error(Loc, "store operand must be a first class value");
if (cast<PointerType>(Ptr->getType())->getElementType() != Val->getType())
return Error(Loc, "stored value and pointer type do not match");
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
if (isAtomic && !Alignment)
return Error(Loc, "atomic store must have explicit non-zero alignment");
if (Ordering == Acquire || Ordering == AcquireRelease)
return Error(Loc, "atomic store cannot use Acquire ordering");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Representation of 'atomic load' and 'atomic store' in IR. llvm-svn: 137170
2011-08-10 07:02:53 +08:00
Inst = new StoreInst(Val, Ptr, isVolatile, Alignment, Ordering, Scope);
reimplement ParseOptionalInfo as ParseOptionalCommaAlign, correctly handle the comma case for metadata. llvm-svn: 92301
2009-12-30 13:44:30 +08:00
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
LangRef and basic memory-representation/reading/writing for 'cmpxchg' and 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
2011-07-29 05:48:00 +08:00
/// ParseCmpXchg
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
/// ::= 'cmpxchg' 'volatile'? TypeAndValue ',' TypeAndValue ',' TypeAndValue
/// 'singlethread'? AtomicOrdering
int LLParser::ParseCmpXchg(Instruction *&Inst, PerFunctionState &PFS) {
LangRef and basic memory-representation/reading/writing for 'cmpxchg' and 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
2011-07-29 05:48:00 +08:00
Value *Ptr, *Cmp, *New; LocTy PtrLoc, CmpLoc, NewLoc;
bool AteExtraComma = false;
AtomicOrdering Ordering = NotAtomic;
SynchronizationScope Scope = CrossThread;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
bool isVolatile = false;
if (EatIfPresent(lltok::kw_volatile))
isVolatile = true;
LangRef and basic memory-representation/reading/writing for 'cmpxchg' and 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
2011-07-29 05:48:00 +08:00
if (ParseTypeAndValue(Ptr, PtrLoc, PFS) ||
ParseToken(lltok::comma, "expected ',' after cmpxchg address") ||
ParseTypeAndValue(Cmp, CmpLoc, PFS) ||
ParseToken(lltok::comma, "expected ',' after cmpxchg cmp operand") ||
ParseTypeAndValue(New, NewLoc, PFS) ||
ParseScopeAndOrdering(true /*Always atomic*/, Scope, Ordering))
return true;
if (Ordering == Unordered)
return TokError("cmpxchg cannot be unordered");
if (!Ptr->getType()->isPointerTy())
return Error(PtrLoc, "cmpxchg operand must be a pointer");
if (cast<PointerType>(Ptr->getType())->getElementType() != Cmp->getType())
return Error(CmpLoc, "compare value and pointer type do not match");
if (cast<PointerType>(Ptr->getType())->getElementType() != New->getType())
return Error(NewLoc, "new value and pointer type do not match");
if (!New->getType()->isIntegerTy())
return Error(NewLoc, "cmpxchg operand must be an integer");
unsigned Size = New->getType()->getPrimitiveSizeInBits();
if (Size < 8 || (Size & (Size - 1)))
return Error(NewLoc, "cmpxchg operand must be power-of-two byte-sized"
" integer");
AtomicCmpXchgInst *CXI =
new AtomicCmpXchgInst(Ptr, Cmp, New, Ordering, Scope);
CXI->setVolatile(isVolatile);
Inst = CXI;
return AteExtraComma ? InstExtraComma : InstNormal;
}
/// ParseAtomicRMW
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
/// ::= 'atomicrmw' 'volatile'? BinOp TypeAndValue ',' TypeAndValue
/// 'singlethread'? AtomicOrdering
int LLParser::ParseAtomicRMW(Instruction *&Inst, PerFunctionState &PFS) {
LangRef and basic memory-representation/reading/writing for 'cmpxchg' and 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
2011-07-29 05:48:00 +08:00
Value *Ptr, *Val; LocTy PtrLoc, ValLoc;
bool AteExtraComma = false;
AtomicOrdering Ordering = NotAtomic;
SynchronizationScope Scope = CrossThread;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
bool isVolatile = false;
LangRef and basic memory-representation/reading/writing for 'cmpxchg' and 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
2011-07-29 05:48:00 +08:00
AtomicRMWInst::BinOp Operation;
Move "atomic" and "volatile" designations on instructions after the opcode of the instruction. Note that this change affects the existing non-atomic load and store instructions; the parser now accepts both forms, and the change is noted in the release notes. llvm-svn: 137527
2011-08-13 06:50:01 +08:00
if (EatIfPresent(lltok::kw_volatile))
isVolatile = true;
LangRef and basic memory-representation/reading/writing for 'cmpxchg' and 'atomicrmw' instructions, which allow representing all the current atomic rmw intrinsics. The allowed operands for these instructions are heavily restricted at the moment; we can probably loosen it a bit, but supporting general first-class types (where it makes sense) might get a bit complicated, given how SelectionDAG works. As an initial cut, these operations do not support specifying an alignment, but it would be possible to add if we think it's useful. Specifying an alignment lower than the natural alignment would be essentially impossible to support on anything other than x86, but specifying a greater alignment would be possible. I can't think of any useful optimizations which would use that information, but maybe someone else has ideas. Optimizer/codegen support coming soon. llvm-svn: 136404
2011-07-29 05:48:00 +08:00
switch (Lex.getKind()) {
default: return TokError("expected binary operation in atomicrmw");
case lltok::kw_xchg: Operation = AtomicRMWInst::Xchg; break;
case lltok::kw_add: Operation = AtomicRMWInst::Add; break;
case lltok::kw_sub: Operation = AtomicRMWInst::Sub; break;
case lltok::kw_and: Operation = AtomicRMWInst::And; break;
case lltok::kw_nand: Operation = AtomicRMWInst::Nand; break;
case lltok::kw_or: Operation = AtomicRMWInst::Or; break;
case lltok::kw_xor: Operation = AtomicRMWInst::Xor; break;
case lltok::kw_max: Operation = AtomicRMWInst::Max; break;
case lltok::kw_min: Operation = AtomicRMWInst::Min; break;
case lltok::kw_umax: Operation = AtomicRMWInst::UMax; break;
case lltok::kw_umin: Operation = AtomicRMWInst::UMin; break;
}
Lex.Lex(); // Eat the operation.
if (ParseTypeAndValue(Ptr, PtrLoc, PFS) ||
ParseToken(lltok::comma, "expected ',' after atomicrmw address") ||
ParseTypeAndValue(Val, ValLoc, PFS) ||
ParseScopeAndOrdering(true /*Always atomic*/, Scope, Ordering))
return true;
if (Ordering == Unordered)
return TokError("atomicrmw cannot be unordered");
if (!Ptr->getType()->isPointerTy())
return Error(PtrLoc, "atomicrmw operand must be a pointer");
if (cast<PointerType>(Ptr->getType())->getElementType() != Val->getType())
return Error(ValLoc, "atomicrmw value and pointer type do not match");
if (!Val->getType()->isIntegerTy())
return Error(ValLoc, "atomicrmw operand must be an integer");
unsigned Size = Val->getType()->getPrimitiveSizeInBits();
if (Size < 8 || (Size & (Size - 1)))
return Error(ValLoc, "atomicrmw operand must be power-of-two byte-sized"
" integer");
AtomicRMWInst *RMWI =
new AtomicRMWInst(Operation, Ptr, Val, Ordering, Scope);
RMWI->setVolatile(isVolatile);
Inst = RMWI;
return AteExtraComma ? InstExtraComma : InstNormal;
}
Initial implementation of 'fence' instruction, the new C++0x-style replacement for llvm.memory.barrier. This is just a LangRef entry and reading/writing/memory representation; optimizer+codegen support coming soon. llvm-svn: 136009
2011-07-26 07:16:38 +08:00
/// ParseFence
/// ::= 'fence' 'singlethread'? AtomicOrdering
int LLParser::ParseFence(Instruction *&Inst, PerFunctionState &PFS) {
AtomicOrdering Ordering = NotAtomic;
SynchronizationScope Scope = CrossThread;
if (ParseScopeAndOrdering(true /*Always atomic*/, Scope, Ordering))
return true;
if (Ordering == Unordered)
return TokError("fence cannot be unordered");
if (Ordering == Monotonic)
return TokError("fence cannot be monotonic");
Inst = new FenceInst(Context, Ordering, Scope);
return InstNormal;
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
/// ParseGetElementPtr
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
/// ::= 'getelementptr' 'inbounds'? TypeAndValue (',' TypeAndValue)*
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
int LLParser::ParseGetElementPtr(Instruction *&Inst, PerFunctionState &PFS) {
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
Value *Ptr = 0;
Value *Val = 0;
LocTy Loc, EltLoc;
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
Minor code simplification. llvm-svn: 77459
2009-07-29 23:58:36 +08:00
bool InBounds = EatIfPresent(lltok::kw_inbounds);
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseTypeAndValue(Ptr, Loc, PFS)) return true;
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
if (!Ptr->getType()->getScalarType()->isPointerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "base of getelementptr must be a pointer");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
SmallVector<Value*, 16> Indices;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
bool AteExtraComma = false;
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
while (EatIfPresent(lltok::comma)) {
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
if (Lex.getKind() == lltok::MetadataVar) {
AteExtraComma = true;
Parse GEP with attached custom metadata. This happens during libprofile_rt.bca build. llvm-svn: 84002
2009-10-14 02:49:55 +08:00
break;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
}
minor cleanups and comment improvements. llvm-svn: 61564
2009-01-02 16:05:26 +08:00
if (ParseTypeAndValue(Val, EltLoc, PFS)) return true;
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
if (!Val->getType()->getScalarType()->isIntegerTy())
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(EltLoc, "getelementptr index must be an integer");
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
if (Val->getType()->isVectorTy() != Ptr->getType()->isVectorTy())
return Error(EltLoc, "getelementptr index type missmatch");
if (Val->getType()->isVectorTy()) {
unsigned ValNumEl = cast<VectorType>(Val->getType())->getNumElements();
unsigned PtrNumEl = cast<VectorType>(Ptr->getType())->getNumElements();
if (ValNumEl != PtrNumEl)
return Error(EltLoc,
"getelementptr vector index has a wrong number of elements");
}
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Indices.push_back(Val);
}
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Add support for vectors of pointers. llvm-svn: 145801
2011-12-05 14:29:09 +08:00
if (Val && Val->getType()->isVectorTy() && Indices.size() != 1)
return Error(EltLoc, "vector getelementptrs must have a single index");
Convert GetElementPtrInst to use ArrayRef. llvm-svn: 135904
2011-07-25 17:48:08 +08:00
if (!GetElementPtrInst::getIndexedType(Ptr->getType(), Indices))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "invalid getelementptr indices");
Convert GetElementPtrInst to use ArrayRef. llvm-svn: 135904
2011-07-25 17:48:08 +08:00
Inst = GetElementPtrInst::Create(Ptr, Indices);
Add a new keyword 'inbounds' for use with getelementptr. See the LangRef.html changes for details. llvm-svn: 77259
2009-07-28 05:53:46 +08:00
if (InBounds)
Reappy r80998, now that the GlobalOpt bug that it exposed on MiniSAT is fixed. llvm-svn: 81172
2009-09-08 07:54:19 +08:00
cast<GetElementPtrInst>(Inst)->setIsInBounds(true);
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseExtractValue
/// ::= 'extractvalue' TypeAndValue (',' uint32)+
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
int LLParser::ParseExtractValue(Instruction *&Inst, PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Val; LocTy Loc;
SmallVector<unsigned, 4> Indices;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
bool AteExtraComma;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Val, Loc, PFS) ||
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
ParseIndexList(Indices, AteExtraComma))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
Add support for a union type in LLVM IR. Patch by Talin! llvm-svn: 96011
2010-02-13 04:49:41 +08:00
if (!Val->getType()->isAggregateType())
return Error(Loc, "extractvalue operand must be aggregate type");
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
if (!ExtractValueInst::getIndexedType(Val->getType(), Indices))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc, "invalid indices for extractvalue");
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
Inst = ExtractValueInst::Create(Val, Indices);
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
/// ParseInsertValue
/// ::= 'insertvalue' TypeAndValue ',' TypeAndValue (',' uint32)+
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
int LLParser::ParseInsertValue(Instruction *&Inst, PerFunctionState &PFS) {
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
Value *Val0, *Val1; LocTy Loc0, Loc1;
SmallVector<unsigned, 4> Indices;
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
bool AteExtraComma;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
if (ParseTypeAndValue(Val0, Loc0, PFS) ||
ParseToken(lltok::comma, "expected comma after insertvalue operand") ||
ParseTypeAndValue(Val1, Loc1, PFS) ||
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
ParseIndexList(Indices, AteExtraComma))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return true;
reimplement insertvalue/extractvalue metadata handling to not blindly accept invalid input. Actually add a testcase. llvm-svn: 92297
2009-12-30 13:14:00 +08:00
Add support for a union type in LLVM IR. Patch by Talin! llvm-svn: 96011
2010-02-13 04:49:41 +08:00
if (!Val0->getType()->isAggregateType())
return Error(Loc0, "insertvalue operand must be aggregate type");
Tabs -> spaces, and remove trailing whitespace. llvm-svn: 82355
2009-09-20 10:20:51 +08:00
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
if (!ExtractValueInst::getIndexedType(Val0->getType(), Indices))
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
return Error(Loc0, "invalid indices for insertvalue");
Convert InsertValueInst and ExtractValueInst APIs to use ArrayRef. llvm-svn: 135040
2011-07-13 18:26:04 +08:00
Inst = InsertValueInst::Create(Val0, Val1, Indices);
convert 4 more instructions over. llvm-svn: 92299
2009-12-30 13:27:33 +08:00
return AteExtraComma ? InstExtraComma : InstNormal;
Reimplement the old and horrible bison parser for .ll files with a nice and clean recursive descent parser. This change has a couple of ramifications: 1. The parser code is about 400 lines shorter (in what we maintain, not including what is autogenerated). 2. The code should be significantly faster than the old code because we don't have to work around bison's poor handling of datatypes with ctors/dtors. This also makes the code much more resistant to memory leaks. 3. We now get caret diagnostics from the .ll parser, woo. 4. The actual diagnostics emited from the parser are completely different so a bunch of testcases had to be updated. 5. I now disallow "%ty = type opaque %ty = type i32". There was no good reason to support this, it was just an accident of the old implementation. I have no reason to think that anyone is actually using this. 6. The syntax for sticking a global variable has changed to make it unambiguous. I don't think anyone is depending on this since only clang supports this and it is not solid yet, so I'm not worried about anything breaking. 7. This gets rid of the last use of bison, and along with it the .cvs files. I'll prune this from the makefiles as a subsequent commit. There are a few minor cleanups that can be done after this commit (suggestions welcome!) but this passes dejagnu testing and is ready for its time in the limelight. llvm-svn: 61558
2009-01-02 15:01:27 +08:00
}
Add support for embedded metadata to LLVM. This introduces two new types of Constant, MDString and MDNode which can only be used by globals with a name that starts with "llvm." or as arguments to a function with the same naming restriction. llvm-svn: 68420
2009-04-04 15:22:01 +08:00
//===----------------------------------------------------------------------===//
// Embedded metadata.
//===----------------------------------------------------------------------===//
/// ParseMDNodeVector
Make MDNode use CallbackVH. Also change MDNode to store Value* instead of Constant* in preperation of a future change to support holding non-Constants in an MDNode. llvm-svn: 71407
2009-05-11 04:57:05 +08:00
/// ::= Element (',' Element)*
/// Element
/// ::= 'null' | TypeAndValue
Re-add parsing of function-local metadata; this time with testcase. llvm-svn: 92793
2010-01-06 06:22:14 +08:00
bool LLParser::ParseMDNodeVector(SmallVectorImpl<Value*> &Elts,
Compute isFunctionLocal in MDNode ctor or via argument in new function getWhenValsUnresolved(). Document PFS argument to ParseValID() and ConvertGlobalOrMetadataValIDToValue(). llvm-svn: 93108
2010-01-10 15:14:18 +08:00
PerFunctionState *PFS) {
Add support for empty metadata nodes: !{}. llvm-svn: 108259
2010-07-14 03:33:27 +08:00
// Check for an empty list.
if (Lex.getKind() == lltok::rbrace)
return false;
Add support for embedded metadata to LLVM. This introduces two new types of Constant, MDString and MDNode which can only be used by globals with a name that starts with "llvm." or as arguments to a function with the same naming restriction. llvm-svn: 68420
2009-04-04 15:22:01 +08:00
do {
rewrite ParseMDNodeVector to follow the normal patter used in the .ll parser. llvm-svn: 92292
2009-12-30 12:42:57 +08:00
// Null is a special case since it is typeless.
if (EatIfPresent(lltok::kw_null)) {
Elts.push_back(0);
continue;
Make MDNode use CallbackVH. Also change MDNode to store Value* instead of Constant* in preperation of a future change to support holding non-Constants in an MDNode. llvm-svn: 71407
2009-05-11 04:57:05 +08:00
}
rewrite ParseMDNodeVector to follow the normal patter used in the .ll parser. llvm-svn: 92292
2009-12-30 12:42:57 +08:00
Value *V = 0;
Land the long talked about "type system rewrite" patch. This patch brings numerous advantages to LLVM. One way to look at it is through diffstat: 109 files changed, 3005 insertions(+), 5906 deletions(-) Removing almost 3K lines of code is a good thing. Other advantages include: 1. Value::getType() is a simple load that can be CSE'd, not a mutating union-find operation. 2. Types a uniqued and never move once created, defining away PATypeHolder. 3. Structs can be "named" now, and their name is part of the identity that uniques them. This means that the compiler doesn't merge them structurally which makes the IR much less confusing. 4. Now that there is no way to get a cycle in a type graph without a named struct type, "upreferences" go away. 5. Type refinement is completely gone, which should make LTO much MUCH faster in some common cases with C++ code. 6. Types are now generally immutable, so we can use "Type *" instead "const Type *" everywhere. Downsides of this patch are that it removes some functions from the C API, so people using those will have to upgrade to (not yet added) new API. "LLVM 3.0" is the right time to do this. There are still some cleanups pending after this, this patch is large enough as-is. llvm-svn: 134829
2011-07-10 01:41:24 +08:00
if (ParseTypeAndValue(V, PFS)) return true;
Make MDNode use CallbackVH. Also change MDNode to store Value* instead of Constant* in preperation of a future change to support holding non-Constants in an MDNode. llvm-svn: 71407
2009-05-11 04:57:05 +08:00
Elts.push_back(V);
Add support for embedded metadata to LLVM. This introduces two new types of Constant, MDString and MDNode which can only be used by globals with a name that starts with "llvm." or as arguments to a function with the same naming restriction. llvm-svn: 68420
2009-04-04 15:22:01 +08:00
} while (EatIfPresent(lltok::comma));
return false;
}