llvm-project/llvm/lib/IR/DebugInfoMetadata.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1287 lines
50 KiB
C++
Raw Normal View History

//===- DebugInfoMetadata.cpp - Implement debug info metadata --------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file implements the debug info Metadata classes.
//
//===----------------------------------------------------------------------===//
#include "llvm/IR/DebugInfoMetadata.h"
#include "LLVMContextImpl.h"
#include "MetadataImpl.h"
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/IR/DIBuilder.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Instructions.h"
#include <numeric>
using namespace llvm;
const DIExpression::FragmentInfo DebugVariable::DefaultFragment = {
std::numeric_limits<uint64_t>::max(), std::numeric_limits<uint64_t>::min()};
DILocation::DILocation(LLVMContext &C, StorageType Storage, unsigned Line,
unsigned Column, ArrayRef<Metadata *> MDs,
bool ImplicitCode)
: MDNode(C, DILocationKind, Storage, MDs) {
assert((MDs.size() == 1 || MDs.size() == 2) &&
"Expected a scope and optional inlined-at");
// Set line and column.
assert(Column < (1u << 16) && "Expected 16-bit column");
SubclassData32 = Line;
SubclassData16 = Column;
setImplicitCode(ImplicitCode);
}
static void adjustColumn(unsigned &Column) {
// Set to unknown on overflow. We only have 16 bits to play with here.
if (Column >= (1u << 16))
Column = 0;
}
DILocation *DILocation::getImpl(LLVMContext &Context, unsigned Line,
unsigned Column, Metadata *Scope,
Metadata *InlinedAt, bool ImplicitCode,
StorageType Storage, bool ShouldCreate) {
// Fixup column.
adjustColumn(Column);
if (Storage == Uniqued) {
if (auto *N = getUniqued(Context.pImpl->DILocations,
DILocationInfo::KeyTy(Line, Column, Scope,
InlinedAt, ImplicitCode)))
return N;
if (!ShouldCreate)
return nullptr;
} else {
assert(ShouldCreate && "Expected non-uniqued nodes to always be created");
}
SmallVector<Metadata *, 2> Ops;
Ops.push_back(Scope);
if (InlinedAt)
Ops.push_back(InlinedAt);
return storeImpl(new (Ops.size()) DILocation(Context, Storage, Line, Column,
Ops, ImplicitCode),
Storage, Context.pImpl->DILocations);
}
const
DILocation *DILocation::getMergedLocations(ArrayRef<const DILocation *> Locs) {
if (Locs.empty())
return nullptr;
if (Locs.size() == 1)
return Locs[0];
auto *Merged = Locs[0];
for (auto I = std::next(Locs.begin()), E = Locs.end(); I != E; ++I) {
Merged = getMergedLocation(Merged, *I);
if (Merged == nullptr)
break;
}
return Merged;
}
const DILocation *DILocation::getMergedLocation(const DILocation *LocA,
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
const DILocation *LocB) {
if (!LocA || !LocB)
return nullptr;
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
if (LocA == LocB)
return LocA;
SmallPtrSet<DILocation *, 5> InlinedLocationsA;
for (DILocation *L = LocA->getInlinedAt(); L; L = L->getInlinedAt())
InlinedLocationsA.insert(L);
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
SmallSet<std::pair<DIScope *, DILocation *>, 5> Locations;
DIScope *S = LocA->getScope();
DILocation *L = LocA->getInlinedAt();
while (S) {
Locations.insert(std::make_pair(S, L));
S = S->getScope();
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
if (!S && L) {
S = L->getScope();
L = L->getInlinedAt();
}
}
const DILocation *Result = LocB;
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
S = LocB->getScope();
L = LocB->getInlinedAt();
while (S) {
if (Locations.count(std::make_pair(S, L)))
break;
S = S->getScope();
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
if (!S && L) {
S = L->getScope();
L = L->getInlinedAt();
}
}
// If the two locations are irreconsilable, just pick one. This is misleading,
// but on the other hand, it's a "line 0" location.
if (!S || !isa<DILocalScope>(S))
S = LocA->getScope();
DebugInfo: Improve debug location merging Fix a set of related bugs: * Considering two locations as equivalent when their lines are the same but their scopes are different causes erroneous debug info that attributes a commoned call to be attributed to one of the two calls it was commoned from. * The previous code to compute a new location's scope was inaccurate and would use the inlinedAt that was the /parent/ of the inlinedAt that is the nearest common one, and also used that parent scope instead of the nearest common scope. * Not generating new locations generally seemed like a lower quality choice There was some risk that generating more new locations could hurt object size by making more fine grained line table entries, but it looks like that was offset by the decrease in line table (& address & ranges) size caused by more accurately computing the scope - which likely lead to fewer range entries (more contiguous ranges) & reduced size that way. All up with these changes I saw minor reductions (-1.21%, -1.77%) in .rela.debug_ranges and .rela.debug_addr (in a fission, compressed debug info build) as well as other minor size changes (generally reductinos) across the board (-1.32% debug_info.dwo, -1.28% debug_loc.dwo). Measured in an optimized (-O2) build of the clang binary. If you are investigating a size regression in an optimized debug builds, this is certainly a patch to look into - and I'd be happy to look into any major regressions found & see what we can do to address them. llvm-svn: 340583
2018-08-24 06:35:58 +08:00
return DILocation::get(Result->getContext(), 0, 0, S, L);
}
Optional<unsigned> DILocation::encodeDiscriminator(unsigned BD, unsigned DF, unsigned CI) {
SmallVector<unsigned, 3> Components = {BD, DF, CI};
uint64_t RemainingWork = 0U;
// We use RemainingWork to figure out if we have no remaining components to
// encode. For example: if BD != 0 but DF == 0 && CI == 0, we don't need to
// encode anything for the latter 2.
// Since any of the input components is at most 32 bits, their sum will be
// less than 34 bits, and thus RemainingWork won't overflow.
RemainingWork = std::accumulate(Components.begin(), Components.end(), RemainingWork);
int I = 0;
unsigned Ret = 0;
unsigned NextBitInsertionIndex = 0;
while (RemainingWork > 0) {
unsigned C = Components[I++];
RemainingWork -= C;
unsigned EC = encodeComponent(C);
Ret |= (EC << NextBitInsertionIndex);
NextBitInsertionIndex += encodingBits(C);
}
// Encoding may be unsuccessful because of overflow. We determine success by
// checking equivalence of components before & after encoding. Alternatively,
// we could determine Success during encoding, but the current alternative is
// simpler.
unsigned TBD, TDF, TCI = 0;
decodeDiscriminator(Ret, TBD, TDF, TCI);
if (TBD == BD && TDF == DF && TCI == CI)
return Ret;
return None;
}
void DILocation::decodeDiscriminator(unsigned D, unsigned &BD, unsigned &DF,
unsigned &CI) {
BD = getUnsignedFromPrefixEncoding(D);
DF = getUnsignedFromPrefixEncoding(getNextComponentInDiscriminator(D));
CI = getUnsignedFromPrefixEncoding(
getNextComponentInDiscriminator(getNextComponentInDiscriminator(D)));
}
DINode::DIFlags DINode::getFlag(StringRef Flag) {
return StringSwitch<DIFlags>(Flag)
#define HANDLE_DI_FLAG(ID, NAME) .Case("DIFlag" #NAME, Flag##NAME)
#include "llvm/IR/DebugInfoFlags.def"
.Default(DINode::FlagZero);
}
StringRef DINode::getFlagString(DIFlags Flag) {
switch (Flag) {
#define HANDLE_DI_FLAG(ID, NAME) \
case Flag##NAME: \
return "DIFlag" #NAME;
#include "llvm/IR/DebugInfoFlags.def"
}
return "";
}
DINode::DIFlags DINode::splitFlags(DIFlags Flags,
SmallVectorImpl<DIFlags> &SplitFlags) {
// Flags that are packed together need to be specially handled, so
// that, for example, we emit "DIFlagPublic" and not
// "DIFlagPrivate | DIFlagProtected".
if (DIFlags A = Flags & FlagAccessibility) {
if (A == FlagPrivate)
SplitFlags.push_back(FlagPrivate);
else if (A == FlagProtected)
SplitFlags.push_back(FlagProtected);
else
SplitFlags.push_back(FlagPublic);
Flags &= ~A;
}
if (DIFlags R = Flags & FlagPtrToMemberRep) {
if (R == FlagSingleInheritance)
SplitFlags.push_back(FlagSingleInheritance);
else if (R == FlagMultipleInheritance)
SplitFlags.push_back(FlagMultipleInheritance);
else
SplitFlags.push_back(FlagVirtualInheritance);
Flags &= ~R;
}
if ((Flags & FlagIndirectVirtualBase) == FlagIndirectVirtualBase) {
Flags &= ~FlagIndirectVirtualBase;
SplitFlags.push_back(FlagIndirectVirtualBase);
}
#define HANDLE_DI_FLAG(ID, NAME) \
if (DIFlags Bit = Flags & Flag##NAME) { \
SplitFlags.push_back(Bit); \
Flags &= ~Bit; \
}
#include "llvm/IR/DebugInfoFlags.def"
return Flags;
}
DIScope *DIScope::getScope() const {
if (auto *T = dyn_cast<DIType>(this))
return T->getScope();
if (auto *SP = dyn_cast<DISubprogram>(this))
return SP->getScope();
if (auto *LB = dyn_cast<DILexicalBlockBase>(this))
return LB->getScope();
if (auto *NS = dyn_cast<DINamespace>(this))
return NS->getScope();
Add LLVM IR debug info support for Fortran COMMON blocks COMMON blocks are a feature of Fortran that has no direct analog in C languages, but they are similar to data sections in assembly language programming. A COMMON block is a named area of memory that holds a collection of variables. Fortran subprograms may map the COMMON block memory area to their own, possibly distinct, non-empty list of variables. A Fortran COMMON block might look like the following example. COMMON /ALPHA/ I, J For this construct, the compiler generates a new scope-like DI construct (!DICommonBlock) into which variables (see I, J above) can be placed. As the common block implies a range of storage with global lifetime, the !DICommonBlock refers to a !DIGlobalVariable. The Fortran variable that comprise the COMMON block are also linked via metadata to offsets within the global variable that stands for the entire common block. @alpha_ = common global %alphabytes_ zeroinitializer, align 64, !dbg !27, !dbg !30, !dbg !33 !14 = distinct !DISubprogram(…) !20 = distinct !DICommonBlock(scope: !14, declaration: !25, name: "alpha") !25 = distinct !DIGlobalVariable(scope: !20, name: "common alpha", type: !24) !27 = !DIGlobalVariableExpression(var: !25, expr: !DIExpression()) !29 = distinct !DIGlobalVariable(scope: !20, name: "i", file: !3, type: !28) !30 = !DIGlobalVariableExpression(var: !29, expr: !DIExpression()) !31 = distinct !DIGlobalVariable(scope: !20, name: "j", file: !3, type: !28) !32 = !DIExpression(DW_OP_plus_uconst, 4) !33 = !DIGlobalVariableExpression(var: !31, expr: !32) The DWARF generated for this is as follows. DW_TAG_common_block: DW_AT_name: alpha DW_AT_location: @alpha_+0 DW_TAG_variable: DW_AT_name: common alpha DW_AT_type: array of 8 bytes DW_AT_location: @alpha_+0 DW_TAG_variable: DW_AT_name: i DW_AT_type: integer*4 DW_AT_location: @Alpha+0 DW_TAG_variable: DW_AT_name: j DW_AT_type: integer*4 DW_AT_location: @Alpha+4 Patch by Eric Schweitz! Differential Revision: https://reviews.llvm.org/D54327 llvm-svn: 357934
2019-04-09 03:13:55 +08:00
if (auto *CB = dyn_cast<DICommonBlock>(this))
return CB->getScope();
if (auto *M = dyn_cast<DIModule>(this))
return M->getScope();
assert((isa<DIFile>(this) || isa<DICompileUnit>(this)) &&
"Unhandled type of scope.");
return nullptr;
}
StringRef DIScope::getName() const {
if (auto *T = dyn_cast<DIType>(this))
return T->getName();
if (auto *SP = dyn_cast<DISubprogram>(this))
return SP->getName();
if (auto *NS = dyn_cast<DINamespace>(this))
return NS->getName();
Add LLVM IR debug info support for Fortran COMMON blocks COMMON blocks are a feature of Fortran that has no direct analog in C languages, but they are similar to data sections in assembly language programming. A COMMON block is a named area of memory that holds a collection of variables. Fortran subprograms may map the COMMON block memory area to their own, possibly distinct, non-empty list of variables. A Fortran COMMON block might look like the following example. COMMON /ALPHA/ I, J For this construct, the compiler generates a new scope-like DI construct (!DICommonBlock) into which variables (see I, J above) can be placed. As the common block implies a range of storage with global lifetime, the !DICommonBlock refers to a !DIGlobalVariable. The Fortran variable that comprise the COMMON block are also linked via metadata to offsets within the global variable that stands for the entire common block. @alpha_ = common global %alphabytes_ zeroinitializer, align 64, !dbg !27, !dbg !30, !dbg !33 !14 = distinct !DISubprogram(…) !20 = distinct !DICommonBlock(scope: !14, declaration: !25, name: "alpha") !25 = distinct !DIGlobalVariable(scope: !20, name: "common alpha", type: !24) !27 = !DIGlobalVariableExpression(var: !25, expr: !DIExpression()) !29 = distinct !DIGlobalVariable(scope: !20, name: "i", file: !3, type: !28) !30 = !DIGlobalVariableExpression(var: !29, expr: !DIExpression()) !31 = distinct !DIGlobalVariable(scope: !20, name: "j", file: !3, type: !28) !32 = !DIExpression(DW_OP_plus_uconst, 4) !33 = !DIGlobalVariableExpression(var: !31, expr: !32) The DWARF generated for this is as follows. DW_TAG_common_block: DW_AT_name: alpha DW_AT_location: @alpha_+0 DW_TAG_variable: DW_AT_name: common alpha DW_AT_type: array of 8 bytes DW_AT_location: @alpha_+0 DW_TAG_variable: DW_AT_name: i DW_AT_type: integer*4 DW_AT_location: @Alpha+0 DW_TAG_variable: DW_AT_name: j DW_AT_type: integer*4 DW_AT_location: @Alpha+4 Patch by Eric Schweitz! Differential Revision: https://reviews.llvm.org/D54327 llvm-svn: 357934
2019-04-09 03:13:55 +08:00
if (auto *CB = dyn_cast<DICommonBlock>(this))
return CB->getName();
if (auto *M = dyn_cast<DIModule>(this))
return M->getName();
assert((isa<DILexicalBlockBase>(this) || isa<DIFile>(this) ||
isa<DICompileUnit>(this)) &&
"Unhandled type of scope.");
return "";
}
#ifndef NDEBUG
static bool isCanonical(const MDString *S) {
return !S || !S->getString().empty();
}
#endif
GenericDINode *GenericDINode::getImpl(LLVMContext &Context, unsigned Tag,
MDString *Header,
ArrayRef<Metadata *> DwarfOps,
StorageType Storage, bool ShouldCreate) {
unsigned Hash = 0;
if (Storage == Uniqued) {
GenericDINodeInfo::KeyTy Key(Tag, Header, DwarfOps);
if (auto *N = getUniqued(Context.pImpl->GenericDINodes, Key))
return N;
if (!ShouldCreate)
return nullptr;
Hash = Key.getHash();
} else {
assert(ShouldCreate && "Expected non-uniqued nodes to always be created");
}
// Use a nullptr for empty headers.
assert(isCanonical(Header) && "Expected canonical MDString");
Metadata *PreOps[] = {Header};
return storeImpl(new (DwarfOps.size() + 1) GenericDINode(
Context, Storage, Hash, Tag, PreOps, DwarfOps),
Storage, Context.pImpl->GenericDINodes);
}
void GenericDINode::recalculateHash() {
setHash(GenericDINodeInfo::KeyTy::calculateHash(this));
}
#define UNWRAP_ARGS_IMPL(...) __VA_ARGS__
#define UNWRAP_ARGS(ARGS) UNWRAP_ARGS_IMPL ARGS
#define DEFINE_GETIMPL_LOOKUP(CLASS, ARGS) \
do { \
if (Storage == Uniqued) { \
if (auto *N = getUniqued(Context.pImpl->CLASS##s, \
CLASS##Info::KeyTy(UNWRAP_ARGS(ARGS)))) \
return N; \
if (!ShouldCreate) \
return nullptr; \
} else { \
assert(ShouldCreate && \
"Expected non-uniqued nodes to always be created"); \
} \
} while (false)
#define DEFINE_GETIMPL_STORE(CLASS, ARGS, OPS) \
return storeImpl(new (array_lengthof(OPS)) \
CLASS(Context, Storage, UNWRAP_ARGS(ARGS), OPS), \
Storage, Context.pImpl->CLASS##s)
#define DEFINE_GETIMPL_STORE_NO_OPS(CLASS, ARGS) \
return storeImpl(new (0u) CLASS(Context, Storage, UNWRAP_ARGS(ARGS)), \
Storage, Context.pImpl->CLASS##s)
#define DEFINE_GETIMPL_STORE_NO_CONSTRUCTOR_ARGS(CLASS, OPS) \
return storeImpl(new (array_lengthof(OPS)) CLASS(Context, Storage, OPS), \
Storage, Context.pImpl->CLASS##s)
#define DEFINE_GETIMPL_STORE_N(CLASS, ARGS, OPS, NUM_OPS) \
return storeImpl(new (NUM_OPS) \
CLASS(Context, Storage, UNWRAP_ARGS(ARGS), OPS), \
Storage, Context.pImpl->CLASS##s)
DISubrange *DISubrange::getImpl(LLVMContext &Context, int64_t Count, int64_t Lo,
StorageType Storage, bool ShouldCreate) {
auto *CountNode = ConstantAsMetadata::get(
ConstantInt::getSigned(Type::getInt64Ty(Context), Count));
return getImpl(Context, CountNode, Lo, Storage, ShouldCreate);
}
DISubrange *DISubrange::getImpl(LLVMContext &Context, Metadata *CountNode,
int64_t Lo, StorageType Storage,
bool ShouldCreate) {
DEFINE_GETIMPL_LOOKUP(DISubrange, (CountNode, Lo));
Metadata *Ops[] = { CountNode };
DEFINE_GETIMPL_STORE(DISubrange, (CountNode, Lo), Ops);
}
DIEnumerator *DIEnumerator::getImpl(LLVMContext &Context, int64_t Value,
bool IsUnsigned, MDString *Name,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIEnumerator, (Value, IsUnsigned, Name));
Metadata *Ops[] = {Name};
DEFINE_GETIMPL_STORE(DIEnumerator, (Value, IsUnsigned), Ops);
}
DIBasicType *DIBasicType::getImpl(LLVMContext &Context, unsigned Tag,
MDString *Name, uint64_t SizeInBits,
uint32_t AlignInBits, unsigned Encoding,
DIFlags Flags, StorageType Storage,
bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIBasicType,
(Tag, Name, SizeInBits, AlignInBits, Encoding, Flags));
Metadata *Ops[] = {nullptr, nullptr, Name};
DEFINE_GETIMPL_STORE(DIBasicType, (Tag, SizeInBits, AlignInBits, Encoding,
Flags), Ops);
}
Optional<DIBasicType::Signedness> DIBasicType::getSignedness() const {
switch (getEncoding()) {
case dwarf::DW_ATE_signed:
case dwarf::DW_ATE_signed_char:
return Signedness::Signed;
case dwarf::DW_ATE_unsigned:
case dwarf::DW_ATE_unsigned_char:
return Signedness::Unsigned;
default:
return None;
}
}
DIDerivedType *DIDerivedType::getImpl(
LLVMContext &Context, unsigned Tag, MDString *Name, Metadata *File,
unsigned Line, Metadata *Scope, Metadata *BaseType, uint64_t SizeInBits,
uint32_t AlignInBits, uint64_t OffsetInBits,
Optional<unsigned> DWARFAddressSpace, DIFlags Flags, Metadata *ExtraData,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIDerivedType,
(Tag, Name, File, Line, Scope, BaseType, SizeInBits,
AlignInBits, OffsetInBits, DWARFAddressSpace, Flags,
ExtraData));
Metadata *Ops[] = {File, Scope, Name, BaseType, ExtraData};
DEFINE_GETIMPL_STORE(
DIDerivedType, (Tag, Line, SizeInBits, AlignInBits, OffsetInBits,
DWARFAddressSpace, Flags), Ops);
}
DICompositeType *DICompositeType::getImpl(
LLVMContext &Context, unsigned Tag, MDString *Name, Metadata *File,
unsigned Line, Metadata *Scope, Metadata *BaseType, uint64_t SizeInBits,
uint32_t AlignInBits, uint64_t OffsetInBits, DIFlags Flags,
Metadata *Elements, unsigned RuntimeLang, Metadata *VTableHolder,
Metadata *TemplateParams, MDString *Identifier, Metadata *Discriminator,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
// Keep this in sync with buildODRType.
DEFINE_GETIMPL_LOOKUP(
DICompositeType, (Tag, Name, File, Line, Scope, BaseType, SizeInBits,
AlignInBits, OffsetInBits, Flags, Elements, RuntimeLang,
VTableHolder, TemplateParams, Identifier, Discriminator));
Metadata *Ops[] = {File, Scope, Name, BaseType,
Elements, VTableHolder, TemplateParams, Identifier,
Discriminator};
DEFINE_GETIMPL_STORE(DICompositeType, (Tag, Line, RuntimeLang, SizeInBits,
AlignInBits, OffsetInBits, Flags),
Ops);
}
DICompositeType *DICompositeType::buildODRType(
LLVMContext &Context, MDString &Identifier, unsigned Tag, MDString *Name,
Metadata *File, unsigned Line, Metadata *Scope, Metadata *BaseType,
uint64_t SizeInBits, uint32_t AlignInBits, uint64_t OffsetInBits,
DIFlags Flags, Metadata *Elements, unsigned RuntimeLang,
Metadata *VTableHolder, Metadata *TemplateParams, Metadata *Discriminator) {
assert(!Identifier.getString().empty() && "Expected valid identifier");
if (!Context.isODRUniquingDebugTypes())
return nullptr;
auto *&CT = (*Context.pImpl->DITypeMap)[&Identifier];
if (!CT)
return CT = DICompositeType::getDistinct(
Context, Tag, Name, File, Line, Scope, BaseType, SizeInBits,
AlignInBits, OffsetInBits, Flags, Elements, RuntimeLang,
VTableHolder, TemplateParams, &Identifier, Discriminator);
// Only mutate CT if it's a forward declaration and the new operands aren't.
assert(CT->getRawIdentifier() == &Identifier && "Wrong ODR identifier?");
if (!CT->isForwardDecl() || (Flags & DINode::FlagFwdDecl))
return CT;
// Mutate CT in place. Keep this in sync with getImpl.
CT->mutate(Tag, Line, RuntimeLang, SizeInBits, AlignInBits, OffsetInBits,
Flags);
Metadata *Ops[] = {File, Scope, Name, BaseType,
Elements, VTableHolder, TemplateParams, &Identifier,
Discriminator};
assert((std::end(Ops) - std::begin(Ops)) == (int)CT->getNumOperands() &&
"Mismatched number of operands");
for (unsigned I = 0, E = CT->getNumOperands(); I != E; ++I)
if (Ops[I] != CT->getOperand(I))
CT->setOperand(I, Ops[I]);
return CT;
}
DICompositeType *DICompositeType::getODRType(
LLVMContext &Context, MDString &Identifier, unsigned Tag, MDString *Name,
Metadata *File, unsigned Line, Metadata *Scope, Metadata *BaseType,
uint64_t SizeInBits, uint32_t AlignInBits, uint64_t OffsetInBits,
DIFlags Flags, Metadata *Elements, unsigned RuntimeLang,
Metadata *VTableHolder, Metadata *TemplateParams, Metadata *Discriminator) {
assert(!Identifier.getString().empty() && "Expected valid identifier");
if (!Context.isODRUniquingDebugTypes())
return nullptr;
auto *&CT = (*Context.pImpl->DITypeMap)[&Identifier];
if (!CT)
CT = DICompositeType::getDistinct(
Context, Tag, Name, File, Line, Scope, BaseType, SizeInBits,
AlignInBits, OffsetInBits, Flags, Elements, RuntimeLang, VTableHolder,
TemplateParams, &Identifier, Discriminator);
return CT;
}
DICompositeType *DICompositeType::getODRTypeIfExists(LLVMContext &Context,
MDString &Identifier) {
assert(!Identifier.getString().empty() && "Expected valid identifier");
if (!Context.isODRUniquingDebugTypes())
return nullptr;
return Context.pImpl->DITypeMap->lookup(&Identifier);
}
DISubroutineType *DISubroutineType::getImpl(LLVMContext &Context, DIFlags Flags,
uint8_t CC, Metadata *TypeArray,
StorageType Storage,
bool ShouldCreate) {
DEFINE_GETIMPL_LOOKUP(DISubroutineType, (Flags, CC, TypeArray));
Metadata *Ops[] = {nullptr, nullptr, nullptr, TypeArray};
DEFINE_GETIMPL_STORE(DISubroutineType, (Flags, CC), Ops);
}
// FIXME: Implement this string-enum correspondence with a .def file and macros,
// so that the association is explicit rather than implied.
static const char *ChecksumKindName[DIFile::CSK_Last] = {
"CSK_MD5",
"CSK_SHA1",
"CSK_SHA256",
};
StringRef DIFile::getChecksumKindAsString(ChecksumKind CSKind) {
assert(CSKind <= DIFile::CSK_Last && "Invalid checksum kind");
// The first space was originally the CSK_None variant, which is now
// obsolete, but the space is still reserved in ChecksumKind, so we account
// for it here.
return ChecksumKindName[CSKind - 1];
}
Optional<DIFile::ChecksumKind> DIFile::getChecksumKind(StringRef CSKindStr) {
return StringSwitch<Optional<DIFile::ChecksumKind>>(CSKindStr)
.Case("CSK_MD5", DIFile::CSK_MD5)
.Case("CSK_SHA1", DIFile::CSK_SHA1)
.Case("CSK_SHA256", DIFile::CSK_SHA256)
.Default(None);
}
DIFile *DIFile::getImpl(LLVMContext &Context, MDString *Filename,
MDString *Directory,
Optional<DIFile::ChecksumInfo<MDString *>> CS,
Optional<MDString *> Source, StorageType Storage,
bool ShouldCreate) {
assert(isCanonical(Filename) && "Expected canonical MDString");
assert(isCanonical(Directory) && "Expected canonical MDString");
assert((!CS || isCanonical(CS->Value)) && "Expected canonical MDString");
assert((!Source || isCanonical(*Source)) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIFile, (Filename, Directory, CS, Source));
Metadata *Ops[] = {Filename, Directory, CS ? CS->Value : nullptr,
Source.getValueOr(nullptr)};
DEFINE_GETIMPL_STORE(DIFile, (CS, Source), Ops);
}
DICompileUnit *DICompileUnit::getImpl(
LLVMContext &Context, unsigned SourceLanguage, Metadata *File,
MDString *Producer, bool IsOptimized, MDString *Flags,
unsigned RuntimeVersion, MDString *SplitDebugFilename,
unsigned EmissionKind, Metadata *EnumTypes, Metadata *RetainedTypes,
Metadata *GlobalVariables, Metadata *ImportedEntities, Metadata *Macros,
uint64_t DWOId, bool SplitDebugInlining, bool DebugInfoForProfiling,
unsigned NameTableKind, bool RangesBaseAddress, MDString *SysRoot,
MDString *SDK, StorageType Storage, bool ShouldCreate) {
assert(Storage != Uniqued && "Cannot unique DICompileUnit");
assert(isCanonical(Producer) && "Expected canonical MDString");
assert(isCanonical(Flags) && "Expected canonical MDString");
assert(isCanonical(SplitDebugFilename) && "Expected canonical MDString");
Metadata *Ops[] = {File,
Producer,
Flags,
SplitDebugFilename,
EnumTypes,
RetainedTypes,
GlobalVariables,
ImportedEntities,
Macros,
SysRoot,
SDK};
return storeImpl(new (array_lengthof(Ops)) DICompileUnit(
Context, Storage, SourceLanguage, IsOptimized,
RuntimeVersion, EmissionKind, DWOId, SplitDebugInlining,
DebugInfoForProfiling, NameTableKind, RangesBaseAddress,
Ops),
Storage);
}
Optional<DICompileUnit::DebugEmissionKind>
DICompileUnit::getEmissionKind(StringRef Str) {
return StringSwitch<Optional<DebugEmissionKind>>(Str)
.Case("NoDebug", NoDebug)
.Case("FullDebug", FullDebug)
.Case("LineTablesOnly", LineTablesOnly)
.Case("DebugDirectivesOnly", DebugDirectivesOnly)
.Default(None);
}
Optional<DICompileUnit::DebugNameTableKind>
DICompileUnit::getNameTableKind(StringRef Str) {
return StringSwitch<Optional<DebugNameTableKind>>(Str)
.Case("Default", DebugNameTableKind::Default)
.Case("GNU", DebugNameTableKind::GNU)
.Case("None", DebugNameTableKind::None)
.Default(None);
}
const char *DICompileUnit::emissionKindString(DebugEmissionKind EK) {
switch (EK) {
case NoDebug: return "NoDebug";
case FullDebug: return "FullDebug";
case LineTablesOnly: return "LineTablesOnly";
case DebugDirectivesOnly: return "DebugDirectivesOnly";
}
return nullptr;
}
const char *DICompileUnit::nameTableKindString(DebugNameTableKind NTK) {
switch (NTK) {
case DebugNameTableKind::Default:
return nullptr;
case DebugNameTableKind::GNU:
return "GNU";
case DebugNameTableKind::None:
return "None";
}
return nullptr;
}
DISubprogram *DILocalScope::getSubprogram() const {
if (auto *Block = dyn_cast<DILexicalBlockBase>(this))
return Block->getScope()->getSubprogram();
return const_cast<DISubprogram *>(cast<DISubprogram>(this));
}
DILocalScope *DILocalScope::getNonLexicalBlockFileScope() const {
if (auto *File = dyn_cast<DILexicalBlockFile>(this))
return File->getScope()->getNonLexicalBlockFileScope();
return const_cast<DILocalScope *>(this);
}
DISubprogram::DISPFlags DISubprogram::getFlag(StringRef Flag) {
return StringSwitch<DISPFlags>(Flag)
#define HANDLE_DISP_FLAG(ID, NAME) .Case("DISPFlag" #NAME, SPFlag##NAME)
#include "llvm/IR/DebugInfoFlags.def"
.Default(SPFlagZero);
}
StringRef DISubprogram::getFlagString(DISPFlags Flag) {
switch (Flag) {
// Appease a warning.
case SPFlagVirtuality:
return "";
#define HANDLE_DISP_FLAG(ID, NAME) \
case SPFlag##NAME: \
return "DISPFlag" #NAME;
#include "llvm/IR/DebugInfoFlags.def"
}
return "";
}
DISubprogram::DISPFlags
DISubprogram::splitFlags(DISPFlags Flags,
SmallVectorImpl<DISPFlags> &SplitFlags) {
// Multi-bit fields can require special handling. In our case, however, the
// only multi-bit field is virtuality, and all its values happen to be
// single-bit values, so the right behavior just falls out.
#define HANDLE_DISP_FLAG(ID, NAME) \
if (DISPFlags Bit = Flags & SPFlag##NAME) { \
SplitFlags.push_back(Bit); \
Flags &= ~Bit; \
}
#include "llvm/IR/DebugInfoFlags.def"
return Flags;
}
DISubprogram *DISubprogram::getImpl(
LLVMContext &Context, Metadata *Scope, MDString *Name,
MDString *LinkageName, Metadata *File, unsigned Line, Metadata *Type,
unsigned ScopeLine, Metadata *ContainingType, unsigned VirtualIndex,
int ThisAdjustment, DIFlags Flags, DISPFlags SPFlags, Metadata *Unit,
Metadata *TemplateParams, Metadata *Declaration, Metadata *RetainedNodes,
Metadata *ThrownTypes, StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
assert(isCanonical(LinkageName) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DISubprogram,
(Scope, Name, LinkageName, File, Line, Type, ScopeLine,
ContainingType, VirtualIndex, ThisAdjustment, Flags,
SPFlags, Unit, TemplateParams, Declaration,
RetainedNodes, ThrownTypes));
SmallVector<Metadata *, 11> Ops = {
File, Scope, Name, LinkageName, Type, Unit,
Declaration, RetainedNodes, ContainingType, TemplateParams, ThrownTypes};
if (!ThrownTypes) {
Ops.pop_back();
if (!TemplateParams) {
Ops.pop_back();
if (!ContainingType)
Ops.pop_back();
}
}
DEFINE_GETIMPL_STORE_N(
DISubprogram,
(Line, ScopeLine, VirtualIndex, ThisAdjustment, Flags, SPFlags), Ops,
Ops.size());
}
bool DISubprogram::describes(const Function *F) const {
assert(F && "Invalid function");
return F->getSubprogram() == this;
}
DILexicalBlock *DILexicalBlock::getImpl(LLVMContext &Context, Metadata *Scope,
Metadata *File, unsigned Line,
unsigned Column, StorageType Storage,
bool ShouldCreate) {
// Fixup column.
adjustColumn(Column);
assert(Scope && "Expected scope");
DEFINE_GETIMPL_LOOKUP(DILexicalBlock, (Scope, File, Line, Column));
Metadata *Ops[] = {File, Scope};
DEFINE_GETIMPL_STORE(DILexicalBlock, (Line, Column), Ops);
}
DILexicalBlockFile *DILexicalBlockFile::getImpl(LLVMContext &Context,
Metadata *Scope, Metadata *File,
unsigned Discriminator,
StorageType Storage,
bool ShouldCreate) {
assert(Scope && "Expected scope");
DEFINE_GETIMPL_LOOKUP(DILexicalBlockFile, (Scope, File, Discriminator));
Metadata *Ops[] = {File, Scope};
DEFINE_GETIMPL_STORE(DILexicalBlockFile, (Discriminator), Ops);
}
DINamespace *DINamespace::getImpl(LLVMContext &Context, Metadata *Scope,
MDString *Name, bool ExportSymbols,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DINamespace, (Scope, Name, ExportSymbols));
// The nullptr is for DIScope's File operand. This should be refactored.
Metadata *Ops[] = {nullptr, Scope, Name};
DEFINE_GETIMPL_STORE(DINamespace, (ExportSymbols), Ops);
}
Add LLVM IR debug info support for Fortran COMMON blocks COMMON blocks are a feature of Fortran that has no direct analog in C languages, but they are similar to data sections in assembly language programming. A COMMON block is a named area of memory that holds a collection of variables. Fortran subprograms may map the COMMON block memory area to their own, possibly distinct, non-empty list of variables. A Fortran COMMON block might look like the following example. COMMON /ALPHA/ I, J For this construct, the compiler generates a new scope-like DI construct (!DICommonBlock) into which variables (see I, J above) can be placed. As the common block implies a range of storage with global lifetime, the !DICommonBlock refers to a !DIGlobalVariable. The Fortran variable that comprise the COMMON block are also linked via metadata to offsets within the global variable that stands for the entire common block. @alpha_ = common global %alphabytes_ zeroinitializer, align 64, !dbg !27, !dbg !30, !dbg !33 !14 = distinct !DISubprogram(…) !20 = distinct !DICommonBlock(scope: !14, declaration: !25, name: "alpha") !25 = distinct !DIGlobalVariable(scope: !20, name: "common alpha", type: !24) !27 = !DIGlobalVariableExpression(var: !25, expr: !DIExpression()) !29 = distinct !DIGlobalVariable(scope: !20, name: "i", file: !3, type: !28) !30 = !DIGlobalVariableExpression(var: !29, expr: !DIExpression()) !31 = distinct !DIGlobalVariable(scope: !20, name: "j", file: !3, type: !28) !32 = !DIExpression(DW_OP_plus_uconst, 4) !33 = !DIGlobalVariableExpression(var: !31, expr: !32) The DWARF generated for this is as follows. DW_TAG_common_block: DW_AT_name: alpha DW_AT_location: @alpha_+0 DW_TAG_variable: DW_AT_name: common alpha DW_AT_type: array of 8 bytes DW_AT_location: @alpha_+0 DW_TAG_variable: DW_AT_name: i DW_AT_type: integer*4 DW_AT_location: @Alpha+0 DW_TAG_variable: DW_AT_name: j DW_AT_type: integer*4 DW_AT_location: @Alpha+4 Patch by Eric Schweitz! Differential Revision: https://reviews.llvm.org/D54327 llvm-svn: 357934
2019-04-09 03:13:55 +08:00
DICommonBlock *DICommonBlock::getImpl(LLVMContext &Context, Metadata *Scope,
Metadata *Decl, MDString *Name,
Metadata *File, unsigned LineNo,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DICommonBlock, (Scope, Decl, Name, File, LineNo));
// The nullptr is for DIScope's File operand. This should be refactored.
Metadata *Ops[] = {Scope, Decl, Name, File};
DEFINE_GETIMPL_STORE(DICommonBlock, (LineNo), Ops);
}
DIModule *DIModule::getImpl(LLVMContext &Context, Metadata *Scope,
MDString *Name, MDString *ConfigurationMacros,
MDString *IncludePath, MDString *APINotesFile,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(
DIModule, (Scope, Name, ConfigurationMacros, IncludePath, APINotesFile));
Metadata *Ops[] = {Scope, Name, ConfigurationMacros, IncludePath,
APINotesFile};
DEFINE_GETIMPL_STORE_NO_CONSTRUCTOR_ARGS(DIModule, Ops);
}
DITemplateTypeParameter *
DITemplateTypeParameter::getImpl(LLVMContext &Context, MDString *Name,
Metadata *Type, bool isDefault,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DITemplateTypeParameter, (Name, Type, isDefault));
Metadata *Ops[] = {Name, Type};
DEFINE_GETIMPL_STORE(DITemplateTypeParameter, (isDefault), Ops);
}
DITemplateValueParameter *DITemplateValueParameter::getImpl(
LLVMContext &Context, unsigned Tag, MDString *Name, Metadata *Type,
bool isDefault, Metadata *Value, StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DITemplateValueParameter,
(Tag, Name, Type, isDefault, Value));
Metadata *Ops[] = {Name, Type, Value};
DEFINE_GETIMPL_STORE(DITemplateValueParameter, (Tag, isDefault), Ops);
}
DIGlobalVariable *
DIGlobalVariable::getImpl(LLVMContext &Context, Metadata *Scope, MDString *Name,
MDString *LinkageName, Metadata *File, unsigned Line,
Metadata *Type, bool IsLocalToUnit, bool IsDefinition,
Metadata *StaticDataMemberDeclaration,
Metadata *TemplateParams, uint32_t AlignInBits,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
assert(isCanonical(LinkageName) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIGlobalVariable, (Scope, Name, LinkageName, File, Line,
Type, IsLocalToUnit, IsDefinition,
StaticDataMemberDeclaration,
TemplateParams, AlignInBits));
Metadata *Ops[] = {Scope,
Name,
File,
Type,
Name,
LinkageName,
StaticDataMemberDeclaration,
TemplateParams};
DEFINE_GETIMPL_STORE(DIGlobalVariable,
(Line, IsLocalToUnit, IsDefinition, AlignInBits), Ops);
}
DILocalVariable *DILocalVariable::getImpl(LLVMContext &Context, Metadata *Scope,
MDString *Name, Metadata *File,
unsigned Line, Metadata *Type,
unsigned Arg, DIFlags Flags,
uint32_t AlignInBits,
StorageType Storage,
bool ShouldCreate) {
// 64K ought to be enough for any frontend.
assert(Arg <= UINT16_MAX && "Expected argument number to fit in 16-bits");
assert(Scope && "Expected scope");
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DILocalVariable,
(Scope, Name, File, Line, Type, Arg, Flags,
AlignInBits));
Metadata *Ops[] = {Scope, Name, File, Type};
DEFINE_GETIMPL_STORE(DILocalVariable, (Line, Arg, Flags, AlignInBits), Ops);
}
Optional<uint64_t> DIVariable::getSizeInBits() const {
// This is used by the Verifier so be mindful of broken types.
const Metadata *RawType = getRawType();
while (RawType) {
// Try to get the size directly.
if (auto *T = dyn_cast<DIType>(RawType))
if (uint64_t Size = T->getSizeInBits())
return Size;
if (auto *DT = dyn_cast<DIDerivedType>(RawType)) {
// Look at the base type.
RawType = DT->getRawBaseType();
continue;
}
// Missing type or size.
break;
}
// Fail gracefully.
return None;
}
DILabel *DILabel::getImpl(LLVMContext &Context, Metadata *Scope,
MDString *Name, Metadata *File, unsigned Line,
StorageType Storage,
bool ShouldCreate) {
assert(Scope && "Expected scope");
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DILabel,
(Scope, Name, File, Line));
Metadata *Ops[] = {Scope, Name, File};
DEFINE_GETIMPL_STORE(DILabel, (Line), Ops);
}
DIExpression *DIExpression::getImpl(LLVMContext &Context,
ArrayRef<uint64_t> Elements,
StorageType Storage, bool ShouldCreate) {
DEFINE_GETIMPL_LOOKUP(DIExpression, (Elements));
DEFINE_GETIMPL_STORE_NO_OPS(DIExpression, (Elements));
}
unsigned DIExpression::ExprOperand::getSize() const {
uint64_t Op = getOp();
if (Op >= dwarf::DW_OP_breg0 && Op <= dwarf::DW_OP_breg31)
return 2;
switch (Op) {
case dwarf::DW_OP_LLVM_convert:
case dwarf::DW_OP_LLVM_fragment:
case dwarf::DW_OP_bregx:
return 3;
case dwarf::DW_OP_constu:
case dwarf::DW_OP_consts:
case dwarf::DW_OP_deref_size:
case dwarf::DW_OP_plus_uconst:
case dwarf::DW_OP_LLVM_tag_offset:
[DebugInfo] Add a DW_OP_LLVM_entry_value operation Summary: Internally in LLVM's metadata we use DW_OP_entry_value operations with the same semantics as DWARF; that is, its operand specifies the number of bytes that the entry value covers. At the time of emitting entry values we don't know the emitted size of the DWARF expression that the entry value will cover. Currently the size is hardcoded to 1 in DIExpression, and other values causes the verifier to fail. As the size is 1, that effectively means that we can only have valid entry values for registers that can be encoded in one byte, which are the registers with DWARF numbers 0 to 31 (as they can be encoded as single-byte DW_OP_reg0..DW_OP_reg31 rather than a multi-byte DW_OP_regx). It is a bit confusing, but it seems like llvm-dwarfdump will print an operation "correctly", even if the byte size is less than that, which may make it seem that we emit correct DWARF for registers with DWARF numbers > 31. If you instead use readelf for such cases, it will interpret the number of specified bytes as a DWARF expression. This seems like a limitation in llvm-dwarfdump. As suggested in D66746, a way forward would be to add an internal variant of DW_OP_entry_value, DW_OP_LLVM_entry_value, whose operand instead specifies the number of operations that the entry value covers, and we then translate that into the byte size at the time of emission. In this patch that internal operation is added. This patch keeps the limitation that a entry value can only be applied to simple register locations, but it will fix the issue with the size operand being incorrect for DWARF numbers > 31. Reviewers: aprantl, vsk, djtodoro, NikolaPrica Reviewed By: aprantl Subscribers: jyknight, fedor.sergeev, hiraditya, llvm-commits Tags: #debug-info, #llvm Differential Revision: https://reviews.llvm.org/D67492 llvm-svn: 374881
2019-10-15 19:31:21 +08:00
case dwarf::DW_OP_LLVM_entry_value:
case dwarf::DW_OP_regx:
return 2;
default:
return 1;
}
}
bool DIExpression::isValid() const {
for (auto I = expr_op_begin(), E = expr_op_end(); I != E; ++I) {
// Check that there's space for the operand.
if (I->get() + I->getSize() > E->get())
return false;
uint64_t Op = I->getOp();
if ((Op >= dwarf::DW_OP_reg0 && Op <= dwarf::DW_OP_reg31) ||
(Op >= dwarf::DW_OP_breg0 && Op <= dwarf::DW_OP_breg31))
return true;
// Check that the operand is valid.
switch (Op) {
default:
return false;
case dwarf::DW_OP_LLVM_fragment:
// A fragment operator must appear at the end.
return I->get() + I->getSize() == E->get();
case dwarf::DW_OP_stack_value: {
// Must be the last one or followed by a DW_OP_LLVM_fragment.
if (I->get() + I->getSize() == E->get())
break;
auto J = I;
if ((++J)->getOp() != dwarf::DW_OP_LLVM_fragment)
return false;
break;
}
case dwarf::DW_OP_swap: {
// Must be more than one implicit element on the stack.
// FIXME: A better way to implement this would be to add a local variable
// that keeps track of the stack depth and introduce something like a
// DW_LLVM_OP_implicit_location as a placeholder for the location this
// DIExpression is attached to, or else pass the number of implicit stack
// elements into isValid.
if (getNumElements() == 1)
return false;
break;
}
[DebugInfo] Add a DW_OP_LLVM_entry_value operation Summary: Internally in LLVM's metadata we use DW_OP_entry_value operations with the same semantics as DWARF; that is, its operand specifies the number of bytes that the entry value covers. At the time of emitting entry values we don't know the emitted size of the DWARF expression that the entry value will cover. Currently the size is hardcoded to 1 in DIExpression, and other values causes the verifier to fail. As the size is 1, that effectively means that we can only have valid entry values for registers that can be encoded in one byte, which are the registers with DWARF numbers 0 to 31 (as they can be encoded as single-byte DW_OP_reg0..DW_OP_reg31 rather than a multi-byte DW_OP_regx). It is a bit confusing, but it seems like llvm-dwarfdump will print an operation "correctly", even if the byte size is less than that, which may make it seem that we emit correct DWARF for registers with DWARF numbers > 31. If you instead use readelf for such cases, it will interpret the number of specified bytes as a DWARF expression. This seems like a limitation in llvm-dwarfdump. As suggested in D66746, a way forward would be to add an internal variant of DW_OP_entry_value, DW_OP_LLVM_entry_value, whose operand instead specifies the number of operations that the entry value covers, and we then translate that into the byte size at the time of emission. In this patch that internal operation is added. This patch keeps the limitation that a entry value can only be applied to simple register locations, but it will fix the issue with the size operand being incorrect for DWARF numbers > 31. Reviewers: aprantl, vsk, djtodoro, NikolaPrica Reviewed By: aprantl Subscribers: jyknight, fedor.sergeev, hiraditya, llvm-commits Tags: #debug-info, #llvm Differential Revision: https://reviews.llvm.org/D67492 llvm-svn: 374881
2019-10-15 19:31:21 +08:00
case dwarf::DW_OP_LLVM_entry_value: {
// An entry value operator must appear at the beginning and the number of
// operations it cover can currently only be 1, because we support only
// entry values of a simple register location. One reason for this is that
// we currently can't calculate the size of the resulting DWARF block for
// other expressions.
return I->get() == expr_op_begin()->get() && I->getArg(0) == 1 &&
getNumElements() == 2;
}
case dwarf::DW_OP_LLVM_convert:
case dwarf::DW_OP_LLVM_tag_offset:
case dwarf::DW_OP_constu:
case dwarf::DW_OP_plus_uconst:
case dwarf::DW_OP_plus:
case dwarf::DW_OP_minus:
case dwarf::DW_OP_mul:
case dwarf::DW_OP_div:
case dwarf::DW_OP_mod:
case dwarf::DW_OP_or:
case dwarf::DW_OP_and:
case dwarf::DW_OP_xor:
case dwarf::DW_OP_shl:
case dwarf::DW_OP_shr:
case dwarf::DW_OP_shra:
case dwarf::DW_OP_deref:
case dwarf::DW_OP_deref_size:
case dwarf::DW_OP_xderef:
case dwarf::DW_OP_lit0:
case dwarf::DW_OP_not:
case dwarf::DW_OP_dup:
case dwarf::DW_OP_regx:
case dwarf::DW_OP_bregx:
break;
}
}
return true;
}
bool DIExpression::isImplicit() const {
if (!isValid())
return false;
if (getNumElements() == 0)
return false;
for (const auto &It : expr_ops()) {
switch (It.getOp()) {
default:
break;
case dwarf::DW_OP_stack_value:
case dwarf::DW_OP_LLVM_tag_offset:
return true;
}
}
return false;
}
bool DIExpression::isComplex() const {
if (!isValid())
return false;
if (getNumElements() == 0)
return false;
// If there are any elements other than fragment or tag_offset, then some
// kind of complex computation occurs.
for (const auto &It : expr_ops()) {
switch (It.getOp()) {
case dwarf::DW_OP_LLVM_tag_offset:
case dwarf::DW_OP_LLVM_fragment:
continue;
default: return true;
}
}
return false;
}
Optional<DIExpression::FragmentInfo>
DIExpression::getFragmentInfo(expr_op_iterator Start, expr_op_iterator End) {
for (auto I = Start; I != End; ++I)
if (I->getOp() == dwarf::DW_OP_LLVM_fragment) {
DIExpression::FragmentInfo Info = {I->getArg(1), I->getArg(0)};
return Info;
}
return None;
}
void DIExpression::appendOffset(SmallVectorImpl<uint64_t> &Ops,
int64_t Offset) {
if (Offset > 0) {
Ops.push_back(dwarf::DW_OP_plus_uconst);
Ops.push_back(Offset);
} else if (Offset < 0) {
Ops.push_back(dwarf::DW_OP_constu);
Ops.push_back(-Offset);
Ops.push_back(dwarf::DW_OP_minus);
}
}
bool DIExpression::extractIfOffset(int64_t &Offset) const {
if (getNumElements() == 0) {
Offset = 0;
return true;
}
if (getNumElements() == 2 && Elements[0] == dwarf::DW_OP_plus_uconst) {
Offset = Elements[1];
return true;
}
if (getNumElements() == 3 && Elements[0] == dwarf::DW_OP_constu) {
if (Elements[2] == dwarf::DW_OP_plus) {
Offset = Elements[1];
return true;
}
if (Elements[2] == dwarf::DW_OP_minus) {
Offset = -Elements[1];
return true;
}
}
return false;
}
const DIExpression *DIExpression::extractAddressClass(const DIExpression *Expr,
unsigned &AddrClass) {
// FIXME: This seems fragile. Nothing that verifies that these elements
// actually map to ops and not operands.
const unsigned PatternSize = 4;
if (Expr->Elements.size() >= PatternSize &&
Expr->Elements[PatternSize - 4] == dwarf::DW_OP_constu &&
Expr->Elements[PatternSize - 2] == dwarf::DW_OP_swap &&
Expr->Elements[PatternSize - 1] == dwarf::DW_OP_xderef) {
AddrClass = Expr->Elements[PatternSize - 3];
if (Expr->Elements.size() == PatternSize)
return nullptr;
return DIExpression::get(Expr->getContext(),
makeArrayRef(&*Expr->Elements.begin(),
Expr->Elements.size() - PatternSize));
}
return Expr;
}
DIExpression *DIExpression::prepend(const DIExpression *Expr, uint8_t Flags,
int64_t Offset) {
SmallVector<uint64_t, 8> Ops;
if (Flags & DIExpression::DerefBefore)
Ops.push_back(dwarf::DW_OP_deref);
appendOffset(Ops, Offset);
if (Flags & DIExpression::DerefAfter)
Ops.push_back(dwarf::DW_OP_deref);
bool StackValue = Flags & DIExpression::StackValue;
bool EntryValue = Flags & DIExpression::EntryValue;
return prependOpcodes(Expr, Ops, StackValue, EntryValue);
}
DIExpression *DIExpression::prependOpcodes(const DIExpression *Expr,
SmallVectorImpl<uint64_t> &Ops,
bool StackValue,
bool EntryValue) {
assert(Expr && "Can't prepend ops to this expression");
if (EntryValue) {
[DebugInfo] Add a DW_OP_LLVM_entry_value operation Summary: Internally in LLVM's metadata we use DW_OP_entry_value operations with the same semantics as DWARF; that is, its operand specifies the number of bytes that the entry value covers. At the time of emitting entry values we don't know the emitted size of the DWARF expression that the entry value will cover. Currently the size is hardcoded to 1 in DIExpression, and other values causes the verifier to fail. As the size is 1, that effectively means that we can only have valid entry values for registers that can be encoded in one byte, which are the registers with DWARF numbers 0 to 31 (as they can be encoded as single-byte DW_OP_reg0..DW_OP_reg31 rather than a multi-byte DW_OP_regx). It is a bit confusing, but it seems like llvm-dwarfdump will print an operation "correctly", even if the byte size is less than that, which may make it seem that we emit correct DWARF for registers with DWARF numbers > 31. If you instead use readelf for such cases, it will interpret the number of specified bytes as a DWARF expression. This seems like a limitation in llvm-dwarfdump. As suggested in D66746, a way forward would be to add an internal variant of DW_OP_entry_value, DW_OP_LLVM_entry_value, whose operand instead specifies the number of operations that the entry value covers, and we then translate that into the byte size at the time of emission. In this patch that internal operation is added. This patch keeps the limitation that a entry value can only be applied to simple register locations, but it will fix the issue with the size operand being incorrect for DWARF numbers > 31. Reviewers: aprantl, vsk, djtodoro, NikolaPrica Reviewed By: aprantl Subscribers: jyknight, fedor.sergeev, hiraditya, llvm-commits Tags: #debug-info, #llvm Differential Revision: https://reviews.llvm.org/D67492 llvm-svn: 374881
2019-10-15 19:31:21 +08:00
Ops.push_back(dwarf::DW_OP_LLVM_entry_value);
// Add size info needed for entry value expression.
// Add plus one for target register operand.
Ops.push_back(Expr->getNumElements() + 1);
}
[DebugInfo] Corrections for salvageDebugInfo Summary: When salvaging a dbg.declare/dbg.addr we should not add DW_OP_stack_value to the DIExpression (see test/Transforms/InstCombine/salvage-dbg-declare.ll). Consider this example %vla = alloca i32, i64 2 call void @llvm.dbg.declare(metadata i32* %vla, metadata !1, metadata !DIExpression()) Instcombine will turn it into %vla1 = alloca [2 x i32] %vla1.sub = getelementptr inbounds [2 x i32], [2 x i32]* %vla, i64 0, i64 0 call void @llvm.dbg.declare(metadata [2 x i32]* %vla1.sub, metadata !19, metadata !DIExpression()) If the GEP can be eliminated, then the dbg.declare will be salvaged and we should get %vla1 = alloca [2 x i32] call void @llvm.dbg.declare(metadata [2 x i32]* %vla1, metadata !19, metadata !DIExpression()) The problem was that salvageDebugInfo did not recognize dbg.declare as being indirect (%vla1 points to the value, it does not hold the value), so we incorrectly got call void @llvm.dbg.declare(metadata [2 x i32]* %vla1, metadata !19, metadata !DIExpression(DW_OP_stack_value)) I also made sure that llvm::salvageDebugInfo and DIExpression::prependOpcodes do not add DW_OP_stack_value to the DIExpression in case no new operands are added to the DIExpression. That way we avoid to, unneccessarily, turn a register location expression into an implicit location expression in some situations (see test11 in test/Transforms/LICM/sinking.ll). Reviewers: aprantl, vsk Reviewed By: aprantl, vsk Subscribers: JDevlieghere, llvm-commits Differential Revision: https://reviews.llvm.org/D48837 llvm-svn: 336191
2018-07-03 19:29:00 +08:00
// If there are no ops to prepend, do not even add the DW_OP_stack_value.
if (Ops.empty())
StackValue = false;
for (auto Op : Expr->expr_ops()) {
// A DW_OP_stack_value comes at the end, but before a DW_OP_LLVM_fragment.
if (StackValue) {
if (Op.getOp() == dwarf::DW_OP_stack_value)
StackValue = false;
else if (Op.getOp() == dwarf::DW_OP_LLVM_fragment) {
Ops.push_back(dwarf::DW_OP_stack_value);
StackValue = false;
}
}
Op.appendToVector(Ops);
}
if (StackValue)
Ops.push_back(dwarf::DW_OP_stack_value);
return DIExpression::get(Expr->getContext(), Ops);
}
[DebugInfo] LowerDbgDeclare: Add derefs when handling CallInst users LowerDbgDeclare inserts a dbg.value before each use of an address described by a dbg.declare. When inserting a dbg.value before a CallInst use, however, it fails to append DW_OP_deref to the DIExpression. The DW_OP_deref is needed to reflect the fact that a dbg.value describes a source variable directly (as opposed to a dbg.declare, which relies on pointer indirection). This patch adds in the DW_OP_deref where needed. This results in the correct values being shown during a debug session for a program compiled with ASan and optimizations (see https://reviews.llvm.org/D49520). Note that ConvertDebugDeclareToDebugValue is already correct -- no changes there were needed. One complication is that SelectionDAG is unable to distinguish between direct and indirect frame-index (FRAMEIX) SDDbgValues. This patch also fixes this long-standing issue in order to not regress integration tests relying on the incorrect assumption that all frame-index SDDbgValues are indirect. This is a necessary fix: the newly-added DW_OP_derefs cannot be lowered properly otherwise. Basically the fix prevents a direct SDDbgValue with DIExpression(DW_OP_deref) from being dereferenced twice by a debugger. There were a handful of tests relying on this incorrect "FRAMEIX => indirect" assumption which actually had incorrect DW_AT_locations: these are all fixed up in this patch. Testing: - check-llvm, and an end-to-end test using lldb to debug an optimized program. - Existing unit tests for DIExpression::appendToStack fully cover the new DIExpression::append utility. - check-debuginfo (the debug info integration tests) Differential Revision: https://reviews.llvm.org/D49454 llvm-svn: 338069
2018-07-27 04:56:53 +08:00
DIExpression *DIExpression::append(const DIExpression *Expr,
ArrayRef<uint64_t> Ops) {
assert(Expr && !Ops.empty() && "Can't append ops to this expression");
// Copy Expr's current op list.
SmallVector<uint64_t, 16> NewOps;
for (auto Op : Expr->expr_ops()) {
// Append new opcodes before DW_OP_{stack_value, LLVM_fragment}.
if (Op.getOp() == dwarf::DW_OP_stack_value ||
Op.getOp() == dwarf::DW_OP_LLVM_fragment) {
NewOps.append(Ops.begin(), Ops.end());
// Ensure that the new opcodes are only appended once.
Ops = None;
}
Op.appendToVector(NewOps);
}
NewOps.append(Ops.begin(), Ops.end());
auto *result = DIExpression::get(Expr->getContext(), NewOps);
assert(result->isValid() && "concatenated expression is not valid");
return result;
[DebugInfo] LowerDbgDeclare: Add derefs when handling CallInst users LowerDbgDeclare inserts a dbg.value before each use of an address described by a dbg.declare. When inserting a dbg.value before a CallInst use, however, it fails to append DW_OP_deref to the DIExpression. The DW_OP_deref is needed to reflect the fact that a dbg.value describes a source variable directly (as opposed to a dbg.declare, which relies on pointer indirection). This patch adds in the DW_OP_deref where needed. This results in the correct values being shown during a debug session for a program compiled with ASan and optimizations (see https://reviews.llvm.org/D49520). Note that ConvertDebugDeclareToDebugValue is already correct -- no changes there were needed. One complication is that SelectionDAG is unable to distinguish between direct and indirect frame-index (FRAMEIX) SDDbgValues. This patch also fixes this long-standing issue in order to not regress integration tests relying on the incorrect assumption that all frame-index SDDbgValues are indirect. This is a necessary fix: the newly-added DW_OP_derefs cannot be lowered properly otherwise. Basically the fix prevents a direct SDDbgValue with DIExpression(DW_OP_deref) from being dereferenced twice by a debugger. There were a handful of tests relying on this incorrect "FRAMEIX => indirect" assumption which actually had incorrect DW_AT_locations: these are all fixed up in this patch. Testing: - check-llvm, and an end-to-end test using lldb to debug an optimized program. - Existing unit tests for DIExpression::appendToStack fully cover the new DIExpression::append utility. - check-debuginfo (the debug info integration tests) Differential Revision: https://reviews.llvm.org/D49454 llvm-svn: 338069
2018-07-27 04:56:53 +08:00
}
DIExpression *DIExpression::appendToStack(const DIExpression *Expr,
ArrayRef<uint64_t> Ops) {
assert(Expr && !Ops.empty() && "Can't append ops to this expression");
[DebugInfo] LowerDbgDeclare: Add derefs when handling CallInst users LowerDbgDeclare inserts a dbg.value before each use of an address described by a dbg.declare. When inserting a dbg.value before a CallInst use, however, it fails to append DW_OP_deref to the DIExpression. The DW_OP_deref is needed to reflect the fact that a dbg.value describes a source variable directly (as opposed to a dbg.declare, which relies on pointer indirection). This patch adds in the DW_OP_deref where needed. This results in the correct values being shown during a debug session for a program compiled with ASan and optimizations (see https://reviews.llvm.org/D49520). Note that ConvertDebugDeclareToDebugValue is already correct -- no changes there were needed. One complication is that SelectionDAG is unable to distinguish between direct and indirect frame-index (FRAMEIX) SDDbgValues. This patch also fixes this long-standing issue in order to not regress integration tests relying on the incorrect assumption that all frame-index SDDbgValues are indirect. This is a necessary fix: the newly-added DW_OP_derefs cannot be lowered properly otherwise. Basically the fix prevents a direct SDDbgValue with DIExpression(DW_OP_deref) from being dereferenced twice by a debugger. There were a handful of tests relying on this incorrect "FRAMEIX => indirect" assumption which actually had incorrect DW_AT_locations: these are all fixed up in this patch. Testing: - check-llvm, and an end-to-end test using lldb to debug an optimized program. - Existing unit tests for DIExpression::appendToStack fully cover the new DIExpression::append utility. - check-debuginfo (the debug info integration tests) Differential Revision: https://reviews.llvm.org/D49454 llvm-svn: 338069
2018-07-27 04:56:53 +08:00
assert(none_of(Ops,
[](uint64_t Op) {
return Op == dwarf::DW_OP_stack_value ||
Op == dwarf::DW_OP_LLVM_fragment;
}) &&
"Can't append this op");
// Append a DW_OP_deref after Expr's current op list if it's non-empty and
// has no DW_OP_stack_value.
//
// Match .* DW_OP_stack_value (DW_OP_LLVM_fragment A B)?.
Optional<FragmentInfo> FI = Expr->getFragmentInfo();
unsigned DropUntilStackValue = FI.hasValue() ? 3 : 0;
[DebugInfo] LowerDbgDeclare: Add derefs when handling CallInst users LowerDbgDeclare inserts a dbg.value before each use of an address described by a dbg.declare. When inserting a dbg.value before a CallInst use, however, it fails to append DW_OP_deref to the DIExpression. The DW_OP_deref is needed to reflect the fact that a dbg.value describes a source variable directly (as opposed to a dbg.declare, which relies on pointer indirection). This patch adds in the DW_OP_deref where needed. This results in the correct values being shown during a debug session for a program compiled with ASan and optimizations (see https://reviews.llvm.org/D49520). Note that ConvertDebugDeclareToDebugValue is already correct -- no changes there were needed. One complication is that SelectionDAG is unable to distinguish between direct and indirect frame-index (FRAMEIX) SDDbgValues. This patch also fixes this long-standing issue in order to not regress integration tests relying on the incorrect assumption that all frame-index SDDbgValues are indirect. This is a necessary fix: the newly-added DW_OP_derefs cannot be lowered properly otherwise. Basically the fix prevents a direct SDDbgValue with DIExpression(DW_OP_deref) from being dereferenced twice by a debugger. There were a handful of tests relying on this incorrect "FRAMEIX => indirect" assumption which actually had incorrect DW_AT_locations: these are all fixed up in this patch. Testing: - check-llvm, and an end-to-end test using lldb to debug an optimized program. - Existing unit tests for DIExpression::appendToStack fully cover the new DIExpression::append utility. - check-debuginfo (the debug info integration tests) Differential Revision: https://reviews.llvm.org/D49454 llvm-svn: 338069
2018-07-27 04:56:53 +08:00
ArrayRef<uint64_t> ExprOpsBeforeFragment =
Expr->getElements().drop_back(DropUntilStackValue);
bool NeedsDeref = (Expr->getNumElements() > DropUntilStackValue) &&
(ExprOpsBeforeFragment.back() != dwarf::DW_OP_stack_value);
bool NeedsStackValue = NeedsDeref || ExprOpsBeforeFragment.empty();
// Append a DW_OP_deref after Expr's current op list if needed, then append
// the new ops, and finally ensure that a single DW_OP_stack_value is present.
SmallVector<uint64_t, 16> NewOps;
if (NeedsDeref)
NewOps.push_back(dwarf::DW_OP_deref);
NewOps.append(Ops.begin(), Ops.end());
[DebugInfo] LowerDbgDeclare: Add derefs when handling CallInst users LowerDbgDeclare inserts a dbg.value before each use of an address described by a dbg.declare. When inserting a dbg.value before a CallInst use, however, it fails to append DW_OP_deref to the DIExpression. The DW_OP_deref is needed to reflect the fact that a dbg.value describes a source variable directly (as opposed to a dbg.declare, which relies on pointer indirection). This patch adds in the DW_OP_deref where needed. This results in the correct values being shown during a debug session for a program compiled with ASan and optimizations (see https://reviews.llvm.org/D49520). Note that ConvertDebugDeclareToDebugValue is already correct -- no changes there were needed. One complication is that SelectionDAG is unable to distinguish between direct and indirect frame-index (FRAMEIX) SDDbgValues. This patch also fixes this long-standing issue in order to not regress integration tests relying on the incorrect assumption that all frame-index SDDbgValues are indirect. This is a necessary fix: the newly-added DW_OP_derefs cannot be lowered properly otherwise. Basically the fix prevents a direct SDDbgValue with DIExpression(DW_OP_deref) from being dereferenced twice by a debugger. There were a handful of tests relying on this incorrect "FRAMEIX => indirect" assumption which actually had incorrect DW_AT_locations: these are all fixed up in this patch. Testing: - check-llvm, and an end-to-end test using lldb to debug an optimized program. - Existing unit tests for DIExpression::appendToStack fully cover the new DIExpression::append utility. - check-debuginfo (the debug info integration tests) Differential Revision: https://reviews.llvm.org/D49454 llvm-svn: 338069
2018-07-27 04:56:53 +08:00
if (NeedsStackValue)
NewOps.push_back(dwarf::DW_OP_stack_value);
return DIExpression::append(Expr, NewOps);
}
Optional<DIExpression *> DIExpression::createFragmentExpression(
const DIExpression *Expr, unsigned OffsetInBits, unsigned SizeInBits) {
SmallVector<uint64_t, 8> Ops;
// Copy over the expression, but leave off any trailing DW_OP_LLVM_fragment.
if (Expr) {
for (auto Op : Expr->expr_ops()) {
switch (Op.getOp()) {
default: break;
case dwarf::DW_OP_shr:
case dwarf::DW_OP_shra:
case dwarf::DW_OP_shl:
case dwarf::DW_OP_plus:
case dwarf::DW_OP_plus_uconst:
case dwarf::DW_OP_minus:
// We can't safely split arithmetic or shift operations into multiple
// fragments because we can't express carry-over between fragments.
//
// FIXME: We *could* preserve the lowest fragment of a constant offset
// operation if the offset fits into SizeInBits.
return None;
case dwarf::DW_OP_LLVM_fragment: {
// Make the new offset point into the existing fragment.
uint64_t FragmentOffsetInBits = Op.getArg(0);
uint64_t FragmentSizeInBits = Op.getArg(1);
(void)FragmentSizeInBits;
assert((OffsetInBits + SizeInBits <= FragmentSizeInBits) &&
"new fragment outside of original fragment");
OffsetInBits += FragmentOffsetInBits;
continue;
}
}
Op.appendToVector(Ops);
}
}
assert(Expr && "Unknown DIExpression");
Ops.push_back(dwarf::DW_OP_LLVM_fragment);
Ops.push_back(OffsetInBits);
Ops.push_back(SizeInBits);
return DIExpression::get(Expr->getContext(), Ops);
}
bool DIExpression::isConstant() const {
// Recognize DW_OP_constu C DW_OP_stack_value (DW_OP_LLVM_fragment Len Ofs)?.
if (getNumElements() != 3 && getNumElements() != 6)
return false;
if (getElement(0) != dwarf::DW_OP_constu ||
getElement(2) != dwarf::DW_OP_stack_value)
return false;
if (getNumElements() == 6 && getElement(3) != dwarf::DW_OP_LLVM_fragment)
return false;
return true;
}
DIExpression::ExtOps DIExpression::getExtOps(unsigned FromSize, unsigned ToSize,
bool Signed) {
dwarf::TypeKind TK = Signed ? dwarf::DW_ATE_signed : dwarf::DW_ATE_unsigned;
DIExpression::ExtOps Ops{{dwarf::DW_OP_LLVM_convert, FromSize, TK,
dwarf::DW_OP_LLVM_convert, ToSize, TK}};
return Ops;
}
DIExpression *DIExpression::appendExt(const DIExpression *Expr,
unsigned FromSize, unsigned ToSize,
bool Signed) {
return appendToStack(Expr, getExtOps(FromSize, ToSize, Signed));
}
DIGlobalVariableExpression *
DIGlobalVariableExpression::getImpl(LLVMContext &Context, Metadata *Variable,
Metadata *Expression, StorageType Storage,
bool ShouldCreate) {
DEFINE_GETIMPL_LOOKUP(DIGlobalVariableExpression, (Variable, Expression));
Metadata *Ops[] = {Variable, Expression};
DEFINE_GETIMPL_STORE_NO_CONSTRUCTOR_ARGS(DIGlobalVariableExpression, Ops);
}
DIObjCProperty *DIObjCProperty::getImpl(
LLVMContext &Context, MDString *Name, Metadata *File, unsigned Line,
MDString *GetterName, MDString *SetterName, unsigned Attributes,
Metadata *Type, StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
assert(isCanonical(GetterName) && "Expected canonical MDString");
assert(isCanonical(SetterName) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIObjCProperty, (Name, File, Line, GetterName,
SetterName, Attributes, Type));
Metadata *Ops[] = {Name, File, GetterName, SetterName, Type};
DEFINE_GETIMPL_STORE(DIObjCProperty, (Line, Attributes), Ops);
}
DIImportedEntity *DIImportedEntity::getImpl(LLVMContext &Context, unsigned Tag,
Metadata *Scope, Metadata *Entity,
Metadata *File, unsigned Line,
MDString *Name, StorageType Storage,
bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIImportedEntity,
(Tag, Scope, Entity, File, Line, Name));
Metadata *Ops[] = {Scope, Entity, Name, File};
DEFINE_GETIMPL_STORE(DIImportedEntity, (Tag, Line), Ops);
}
DIMacro *DIMacro::getImpl(LLVMContext &Context, unsigned MIType,
unsigned Line, MDString *Name, MDString *Value,
StorageType Storage, bool ShouldCreate) {
assert(isCanonical(Name) && "Expected canonical MDString");
DEFINE_GETIMPL_LOOKUP(DIMacro, (MIType, Line, Name, Value));
Metadata *Ops[] = { Name, Value };
DEFINE_GETIMPL_STORE(DIMacro, (MIType, Line), Ops);
}
DIMacroFile *DIMacroFile::getImpl(LLVMContext &Context, unsigned MIType,
unsigned Line, Metadata *File,
Metadata *Elements, StorageType Storage,
bool ShouldCreate) {
DEFINE_GETIMPL_LOOKUP(DIMacroFile,
(MIType, Line, File, Elements));
Metadata *Ops[] = { File, Elements };
DEFINE_GETIMPL_STORE(DIMacroFile, (MIType, Line), Ops);
}