Commit Graph

632 Commits

Author SHA1 Message Date
Fangrui Song 6907ce2f8f Remove trailing space
sed -Ei 's/[[:space:]]+$//' include/**/*.{def,h,td} lib/**/*.{cpp,h}

llvm-svn: 338291
2018-07-30 19:24:48 +00:00
Erich Keane 3efe00206f Implement cpu_dispatch/cpu_specific Multiversioning
As documented here: https://software.intel.com/en-us/node/682969 and
https://software.intel.com/en-us/node/523346. cpu_dispatch multiversioning
is an ICC feature that provides for function multiversioning.

This feature is implemented with two attributes: First, cpu_specific,
which specifies the individual function versions. Second, cpu_dispatch,
which specifies the location of the resolver function and the list of
resolvable functions.

This is valuable since it provides a mechanism where the resolver's TU
can be specified in one location, and the individual implementions
each in their own translation units.

The goal of this patch is to be source-compatible with ICC, so this
implementation diverges from the ICC implementation in a few ways:
1- Linux x86/64 only: This implementation uses ifuncs in order to
properly dispatch functions. This is is a valuable performance benefit
over the ICC implementation. A future patch will be provided to enable
this feature on Windows, but it will obviously more closely fit ICC's
implementation.
2- CPU Identification functions: ICC uses a set of custom functions to identify
the feature list of the host processor. This patch uses the cpu_supports
functionality in order to better align with 'target' multiversioning.
1- cpu_dispatch function def/decl: ICC's cpu_dispatch requires that the function
marked cpu_dispatch be an empty definition. This patch supports that as well,
however declarations are also permitted, since the linker will solve the
issue of multiple emissions.

Differential Revision: https://reviews.llvm.org/D47474

llvm-svn: 337552
2018-07-20 14:13:28 +00:00
Peter Collingbourne e44acadf6a Implement CFI for indirect calls via a member function pointer.
Similarly to CFI on virtual and indirect calls, this implementation
tries to use program type information to make the checks as precise
as possible.  The basic way that it works is as follows, where `C`
is the name of the class being defined or the target of a call and
the function type is assumed to be `void()`.

For virtual calls:
- Attach type metadata to the addresses of function pointers in vtables
  (not the functions themselves) of type `void (B::*)()` for each `B`
  that is a recursive dynamic base class of `C`, including `C` itself.
  This type metadata has an annotation that the type is for virtual
  calls (to distinguish it from the non-virtual case).
- At the call site, check that the computed address of the function
  pointer in the vtable has type `void (C::*)()`.

For non-virtual calls:
- Attach type metadata to each non-virtual member function whose address
  can be taken with a member function pointer. The type of a function
  in class `C` of type `void()` is each of the types `void (B::*)()`
  where `B` is a most-base class of `C`. A most-base class of `C`
  is defined as a recursive base class of `C`, including `C` itself,
  that does not have any bases.
- At the call site, check that the function pointer has one of the types
  `void (B::*)()` where `B` is a most-base class of `C`.

Differential Revision: https://reviews.llvm.org/D47567

llvm-svn: 335569
2018-06-26 02:15:47 +00:00
Peter Collingbourne d914fd2163 IRgen: Mark aliases of ctors and dtors as unnamed_addr.
This is not only semantically correct but ensures that they will not
be marked as address-significant once D48155 lands.

Differential Revision: https://reviews.llvm.org/D48206

llvm-svn: 334982
2018-06-18 20:58:54 +00:00
Gabor Buella 1a83d06768 [CodeGen] Improve diagnostics related to target attributes
Summary:
When requirement imposed by __target__ attributes on functions
are not satisfied, prefer printing those requirements, which
are explicitly mentioned in the attributes.

This makes such messages more useful, e.g. printing avx512f instead of avx2
in the following scenario:

```
$ cat foo.c
static inline void __attribute__((__always_inline__, __target__("avx512f")))
x(void)
{
}

int main(void)
{
            x();
}
$ clang foo.c
foo.c:7:2: error: always_inline function 'x' requires target feature 'avx2', but would be inlined into function 'main' that is compiled without support for 'avx2'
        x();
    ^
1 error generated.
```

bugzilla: https://bugs.llvm.org/show_bug.cgi?id=37338

Reviewers: craig.topper, echristo, dblaikie

Reviewed By: craig.topper, echristo

Differential Revision: https://reviews.llvm.org/D46541

llvm-svn: 334174
2018-06-07 08:48:36 +00:00
Yaxun Liu daceb1ea0f CodeGen: Emit string literal in constant address space
Some targets have constant address space (e.g. amdgcn). For them string literal should be
emitted in constant address space then casted to default address space.

Differential Revision: https://reviews.llvm.org/D46643

llvm-svn: 332279
2018-05-14 19:20:12 +00:00
Adrian Prantl 9fc8faf9e6 Remove \brief commands from doxygen comments.
This is similar to the LLVM change https://reviews.llvm.org/D46290.

We've been running doxygen with the autobrief option for a couple of
years now. This makes the \brief markers into our comments
redundant. Since they are a visual distraction and we don't want to
encourage more \brief markers in new code either, this patch removes
them all.

Patch produced by

for i in $(git grep -l '\@brief'); do perl -pi -e 's/\@brief //g' $i & done
for i in $(git grep -l '\\brief'); do perl -pi -e 's/\\brief //g' $i & done

Differential Revision: https://reviews.llvm.org/D46320

llvm-svn: 331834
2018-05-09 01:00:01 +00:00
Akira Hatanaka 617e26152d Add a command line option 'fregister_global_dtors_with_atexit' to
register destructor functions annotated with __attribute__((destructor))
using __cxa_atexit or atexit.

Register destructor functions annotated with __attribute__((destructor))
calling __cxa_atexit in a synthesized constructor function instead of
emitting references to the functions in a special section.

The primary reason for adding this option is that we are planning to
deprecate the __mod_term_funcs section on Darwin in the future. This
feature is enabled by default only on Darwin. Users who do not want this
can use command line option 'fno_register_global_dtors_with_atexit' to
disable it.

rdar://problem/33887655

Differential Revision: https://reviews.llvm.org/D45578

llvm-svn: 330199
2018-04-17 18:41:52 +00:00
Alexander Kornienko 2a8c18d991 Fix typos in clang
Found via codespell -q 3 -I ../clang-whitelist.txt
Where whitelist consists of:

  archtype
  cas
  classs
  checkk
  compres
  definit
  frome
  iff
  inteval
  ith
  lod
  methode
  nd
  optin
  ot
  pres
  statics
  te
  thru

Patch by luzpaz! (This is a subset of D44188 that applies cleanly with a few
files that have dubious fixes reverted.)

Differential revision: https://reviews.llvm.org/D44188

llvm-svn: 329399
2018-04-06 15:14:32 +00:00
Reid Kleckner 399d96e39c [MS] Emit vftable thunks for functions with incomplete prototypes
Summary:
The following class hierarchy requires that we be able to emit a
this-adjusting thunk for B::foo in C's vftable:

  struct Incomplete;
  struct A {
    virtual A* foo(Incomplete p) = 0;
  };
  struct B : virtual A {
    void foo(Incomplete p) override;
  };
  struct C : B { int c; };

This TU is valid, but lacks a definition of 'Incomplete', which makes it
hard to build a thunk for the final overrider, B::foo.

Before this change, Clang gives up attempting to emit the thunk, because
it assumes that if the parameter types are incomplete, it must be
emitting the thunk for optimization purposes. This is untrue for the MS
ABI, where the implementation of B::foo has no idea what thunks C's
vftable may require. Clang needs to emit the thunk without necessarily
having access to the complete prototype of foo.

This change makes Clang emit a musttail variadic call when it needs such
a thunk. I call these "unprototyped" thunks, because they only prototype
the "this" parameter, which must always come first in the MS C++ ABI.

These thunks work, but they create ugly LLVM IR. If the call to the
thunk is devirtualized, it will be a call to a bitcast of a function
pointer. Today, LLVM cannot inline through such a call, but I want to
address that soon, because we also use this pattern for virtual member
pointer thunks.

This change also implements an old FIXME in the code about reusing the
thunk's computed CGFunctionInfo as much as possible. Now we don't end up
computing the thunk's mangled name and arranging it's prototype up to
around three times.

Fixes PR25641

Reviewers: rjmccall, rsmith, hans

Subscribers: Prazek, cfe-commits

Differential Revision: https://reviews.llvm.org/D45112

llvm-svn: 329009
2018-04-02 20:20:33 +00:00
Rafael Espindola b735004615 Start setting dllimport/dllexport in setGVProperties.
This is the next step in setting dso_local for COFF.

The patches changes setGVProperties to first set dllimport/dllexport
and changes a few cases that were setting dllimport/dllexport
manually. With this a few more GVs are marked dso_local.

llvm-svn: 326397
2018-03-01 00:35:47 +00:00
Rafael Espindola 75c649c960 Pass a GlobalDecl to SetCommonAttributes. NFC.
Part of D43900.

llvm-svn: 326392
2018-03-01 00:06:55 +00:00
Rafael Espindola e4e78135ac Inline a trivial function. NFC.
llvm-svn: 326391
2018-03-01 00:00:02 +00:00
Rafael Espindola 285271cb78 Pass a GlobalDecl to setNonAliasAttributes. NFC.
Also part of D43900.

llvm-svn: 326388
2018-02-28 23:54:23 +00:00
Rafael Espindola 51ec5a9ce3 Pass a GlobalDecl to SetInternalFunctionAttributes. NFC.
This just reduces the noise in a followup patch.

Part of D43900.

llvm-svn: 326385
2018-02-28 23:46:35 +00:00
Rafael Espindola fa2fc906d2 Pass a GlobalDecl to setAliasAttributes. NFC.
This just makes a followup change easier to read.

llvm-svn: 326270
2018-02-28 00:06:01 +00:00
Rafael Espindola 64393cfca0 Pass a GlobalDecl to setFunctionDefinitionAttributes. NFC.
This just makes a followup patch easier to read.

llvm-svn: 326265
2018-02-27 23:44:36 +00:00
Rafael Espindola 922f2aa9b2 Bring r325915 back.
The tests that failed on a windows host have been fixed.

Original message:

Start setting dso_local for COFF.

With this there are still some GVs where we don't set dso_local
because setGVProperties is never called. I intend to fix that in
followup commits. This is just the bare minimum to teach
shouldAssumeDSOLocal what it should do for COFF.

llvm-svn: 325940
2018-02-23 19:30:48 +00:00
Rafael Espindola 43ce3a3a4d Revert "Start setting dso_local for COFF."
This reverts commit r325915.

It will take some time to fix the failures on a windows host.

llvm-svn: 325929
2018-02-23 18:09:29 +00:00
Rafael Espindola 004d240b6a Start setting dso_local for COFF.
With this there are still some GVs where we don't set dso_local
because setGVProperties is never called. I intend to fix that in
followup commits. This is just the bare minimum to teach
shouldAssumeDSOLocal what it should do for COFF.

llvm-svn: 325915
2018-02-23 15:32:32 +00:00
Rafael Espindola 3dd4981298 Simplify setting dso_local. NFC.
The value of dso_local can be computed from just IR properties and
global information (object file type, command line options, etc).

With this patch we no longer pass in the Decl. It was almost unused
and making it fully unused guarantees that dso_local is consistent
with the rest of the IR.

llvm-svn: 325846
2018-02-23 00:22:15 +00:00
Erich Keane 93e58667ee Make attribute-target on a Definition-after-use update the LLVM attributes
As reported here: https://bugs.llvm.org/show_bug.cgi?id=36301
The issue is that the 'use' causes the plain declaration to emit
the attributes to LLVM-IR. However, if the definition added it
later, these would silently disappear.

This commit extracts that logic to its own function in CodeGenModule,
and has the attribute-applications done during 'definition' update
the attributes properly.

Differential Revision: https://reviews.llvm.org/D43095

llvm-svn: 324907
2018-02-12 17:01:41 +00:00
Rafael Espindola 699f5d6bbc Recommit r324107 again.
The difference from the previous try is that we no longer directly
access function declarations from position independent executables. It
should work, but currently doesn't with some linkers.

It now includes a fix to not mark available_externally definitions as
dso_local.

Original message:

Start setting dso_local in clang.

This starts adding dso_local to clang.

The hope is to eventually have TargetMachine::shouldAssumeDsoLocal go
away. My objective for now is to move enough of it to clang to remove
the need for the TargetMachine one to handle PIE copy relocations and
-fno-plt. With that it should then be easy to implement a
-fno-copy-reloc in clang.

This patch just adds the cases where we assume a symbol to be local
based on the file being compiled for an executable or a shared
library.

llvm-svn: 324535
2018-02-07 22:15:33 +00:00
Rafael Espindola 880c3b24c5 Revert "Recommit r324107."
This reverts commit r324500.

The bots found two failures:

    ThreadSanitizer-x86_64 :: Linux/pie_no_aslr.cc
    ThreadSanitizer-x86_64 :: pie_test.cc

when using gold. The issue is a limitation in gold when building pie
binaries. I will investigate how to work around it.

llvm-svn: 324505
2018-02-07 19:44:15 +00:00
Rafael Espindola fa9874c33b Recommit r324107.
It now includes a fix to not mark available_externally definitions as
dso_local.

Original message:

Start setting dso_local in clang.

This starts adding dso_local to clang.

The hope is to eventually have TargetMachine::shouldAssumeDsoLocal go
away. My objective for now is to move enough of it to clang to remove
the need for the TargetMachine one to handle PIE copy relocations and
-fno-plt. With that it should then be easy to implement a
-fno-copy-reloc in clang.

This patch just adds the cases where we assume a symbol to be local
based on the file being compiled for an executable or a shared
library.

llvm-svn: 324500
2018-02-07 19:16:49 +00:00
Rafael Espindola deb10becf3 Don't pass ForDefinition_t in places it is redundant.
I found this while looking at the ppc failures caused by the dso_local
change.

The issue was that the patch would produce the wrong answer for
available_externally. Having ForDefinition_t available in places where
the code can just check the linkage is a bit of a foot gun.

This patch removes the ForDefiniton_t argument in places where the
linkage is already know.

llvm-svn: 324499
2018-02-07 19:04:41 +00:00
Saleem Abdulrasool fd4db5331e Support `#pragma comment(lib, "name")` in the frontend for ELF
This adds the frontend support required to support the use of the
comment pragma to enable auto linking on ELFish targets. This is a
generic ELF extension supported by LLVM. We need to change the handling
for the "dependentlib" in order to accommodate the previously discussed
encoding for the dependent library descriptor. Without the custom
handling of the PCK_Lib directive, the -l prefixed option would be
encoded into the resulting object (which is treated as a frontend
error).

llvm-svn: 324438
2018-02-07 01:46:46 +00:00
Rafael Espindola 9f34b7b93b Revert "Start setting dso_local in clang."
This reverts commit r324107.

I will have to test it on OS X.

llvm-svn: 324108
2018-02-02 17:29:22 +00:00
Rafael Espindola 7e34a308ff Start setting dso_local in clang.
This starts adding dso_local to clang.

The hope is to eventually have TargetMachine::shouldAssumeDsoLocal go
away. My objective for now is to move enough of it to clang to remove
the need for the TargetMachine one to handle PIE copy relocations and
-fno-plt. With that it should then be easy to implement a
-fno-copy-reloc in clang.

This patch just adds the cases where we assume a symbol to be local
based on the file being compiled for an executable or a shared
library.

llvm-svn: 324107
2018-02-02 17:17:39 +00:00
Ivan A. Kosarev 1860b520a2 [CodeGen] Decorate aggregate accesses with TBAA tags
Differential Revision: https://reviews.llvm.org/D41539

llvm-svn: 323421
2018-01-25 14:21:55 +00:00
Erich Keane 281d20b601 Implement Attribute Target MultiVersioning
GCC's attribute 'target', in addition to being an optimization hint,
also allows function multiversioning. We currently have the former
implemented, this is the latter's implementation.

This works by enabling functions with the same name/signature to coexist,
so that they can all be emitted. Multiversion state is stored in the
FunctionDecl itself, and SemaDecl manages the definitions.
Note that it ends up having to permit redefinition of functions so
that they can all be emitted. Additionally, all versions of the function
must be emitted, so this also manages that.

Note that this includes some additional rules that GCC does not, since
defining something as a MultiVersion function after a usage has been made illegal.

The only 'history rewriting' that happens is if a function is emitted before
it has been converted to a multiversion'ed function, at which point its name
needs to be changed.

Function templates and virtual functions are NOT yet supported (not supported
in GCC either).

Additionally, constructors/destructors are disallowed, but the former is 
planned.

llvm-svn: 322028
2018-01-08 21:34:17 +00:00
Eli Friedman dbcfadedf9 [Coverage] Always emit unused coverage mappings in the same order.
Non-determinism is confusing at best.

Differential Revision: https://reviews.llvm.org/D41140

llvm-svn: 320533
2017-12-13 00:14:17 +00:00
Hal Finkel a5986b9e91 Revert "[CodeGen] Add initial support for union members in TBAA"
This reverts commit r319413. See PR35503.

We can't use "union member" as the access type here like this.

llvm-svn: 319629
2017-12-03 03:10:13 +00:00
Ivan A. Kosarev da34247a6a [CodeGen] Add initial support for union members in TBAA
The basic idea behind this patch is that since in strict aliasing
mode all accesses to union members require their outermost
enclosing union objects to be specified explicitly, then for a
couple given accesses to union members of the form

p->a.b.c...
q->x.y.z...

it is known they can only alias if both p and q point to the same
union type and offset ranges of members a.b.c... and x.y.z...
overlap. Note that the actual types of the members do not matter.

Specifically, in this patch we do the following:

* Make unions to be valid TBAA base access types. This enables
  generation of TBAA type descriptors for unions.

* Encode union types as structures with a single member of a
  special "union member" type. Currently we do not encode
  information about sizes of types, but conceptually such union
  members are considered to be of the size of the whole union.

* Encode accesses to direct and indirect union members, including
  member arrays, as accesses to these special members. All
  accesses to members of a union thus get the same offset, which
  is the offset of the union they are part of. This means the
  existing LLVM TBAA machinery is able to handle such accesses
  with no changes.

While this is already an improvement comparing to the current
situation, that is, representing all union accesses as may-alias
ones, there are further changes planned to complete the support
for unions. One of them is storing information about access sizes
so we can distinct accesses to non-overlapping union members,
including accesses to different elements of member arrays.
Another change is encoding type sizes in order to make it
possible to compute offsets within constant-indexed array
elements. These enhancements will be addressed with separate
patches.

Differential Revision: https://reviews.llvm.org/D39455

llvm-svn: 319413
2017-11-30 09:26:39 +00:00
Jake Ehrlich ba874ad83e Reland "Fix vtable not receiving hidden visibility when using push(visibility)"
I had to reland this change in order to make the test work on windows

This change should resolve https://bugs.llvm.org/show_bug.cgi?id=35022

https://reviews.llvm.org/D39627

llvm-svn: 319269
2017-11-29 00:54:20 +00:00
Ivan A. Kosarev 4e50e707bf [CodeGen] Collect information about sizes of accesses and access types for TBAA
The information about access and type sizes is necessary for
producing TBAA metadata in the new size-aware format. With this
patch, D39955 and D39956 in place we should be able to change
CodeGenTBAA::createScalarTypeNode() and
CodeGenTBAA::getBaseTypeInfo() to generate metadata in the new
format under the -new-struct-path-tbaa command-line option. For
now, this new information remains unused.

Differential Revision: https://reviews.llvm.org/D40176

llvm-svn: 319012
2017-11-27 09:39:29 +00:00
Petr Hosek 32c9de009a Revert "[CodeGen] Fix vtable not receiving hidden visibility when using push(visibility)"
This reverts commit r318853: tests are failing on Windows bots

llvm-svn: 318866
2017-11-22 19:50:17 +00:00
Petr Hosek 9696dbb988 [CodeGen] Fix vtable not receiving hidden visibility when using push(visibility)
This change should resolve https://bugs.llvm.org/show_bug.cgi?id=35022

Patch by Jake Ehrlich

Differential Revision: https://reviews.llvm.org/D39627

llvm-svn: 318853
2017-11-22 17:59:30 +00:00
Vlad Tsyrklevich 634c601fe3 [CFI] Add CFI-icall pointer type generalization
Summary:
This change allows generalizing pointers in type signatures used for
cfi-icall by enabling the -fsanitize-cfi-icall-generalize-pointers flag.
This works by 1) emitting an additional generalized type signature
metadata node for functions and 2) llvm.type.test()ing for the
generalized type for translation units with the flag specified.

This flag is incompatible with -fsanitize-cfi-cross-dso because it would
require emitting twice as many type hashes which would increase artifact
size.

Reviewers: pcc, eugenis

Reviewed By: pcc

Subscribers: kcc

Differential Revision: https://reviews.llvm.org/D39358

llvm-svn: 317044
2017-10-31 22:39:44 +00:00
Ivan A. Kosarev b9c59f36fc [CodeGen] Propagate may-alias'ness of lvalues with TBAA info
This patch fixes various places in clang to propagate may-alias
TBAA access descriptors during construction of lvalues, thus
eliminating the need for the LValueBaseInfo::MayAlias flag.

This is part of D38126 reworked to be a separate patch to
simplify review.

Differential Revision: https://reviews.llvm.org/D39008

llvm-svn: 316988
2017-10-31 11:05:34 +00:00
Ivan A. Kosarev ed141bab63 [CodeGen] EmitPointerWithAlignment() to generate TBAA info along with LValue base info
Differential Revision: https://reviews.llvm.org/D38796

llvm-svn: 315984
2017-10-17 09:12:13 +00:00
Alexander Richardson 6d989436d0 Convert clang::LangAS to a strongly typed enum
Summary:
Convert clang::LangAS to a strongly typed enum

Currently both clang AST address spaces and target specific address spaces
are represented as unsigned which can lead to subtle errors if the wrong
type is passed. It is especially confusing in the CodeGen files as it is
not possible to see what kind of address space should be passed to a
function without looking at the implementation.
I originally made this change for our LLVM fork for the CHERI architecture
where we make extensive use of address spaces to differentiate between
capabilities and pointers. When merging the upstream changes I usually
run into some test failures or runtime crashes because the wrong kind of
address space is passed to a function. By converting the LangAS enum to a
C++11 we can catch these errors at compile time. Additionally, it is now
obvious from the function signature which kind of address space it expects.

I found the following errors while writing this patch:

- ItaniumRecordLayoutBuilder::LayoutField was passing a clang AST address
  space to  TargetInfo::getPointer{Width,Align}()
- TypePrinter::printAttributedAfter() prints the numeric value of the
  clang AST address space instead of the target address space.
  However, this code is not used so I kept the current behaviour
- initializeForBlockHeader() in CGBlocks.cpp was passing
  LangAS::opencl_generic to TargetInfo::getPointer{Width,Align}()
- CodeGenFunction::EmitBlockLiteral() was passing a AST address space to
  TargetInfo::getPointerWidth()
- CGOpenMPRuntimeNVPTX::translateParameter() passed a target address space
  to Qualifiers::addAddressSpace()
- CGOpenMPRuntimeNVPTX::getParameterAddress() was using
  llvm::Type::getPointerTo() with a AST address space
- clang_getAddressSpace() returns either a LangAS or a target address
  space. As this is exposed to C I have kept the current behaviour and
  added a comment stating that it is probably not correct.

Other than this the patch should not cause any functional changes.

Reviewers: yaxunl, pcc, bader

Reviewed By: yaxunl, bader

Subscribers: jlebar, jholewinski, nhaehnle, Anastasia, cfe-commits

Differential Revision: https://reviews.llvm.org/D38816

llvm-svn: 315871
2017-10-15 18:48:14 +00:00
Ivan A. Kosarev ce601eedf6 Revert "[CodeGen] EmitPointerWithAlignment() to generate TBAA info along with LValue base info", r315731.
With this change we fail on the clang-x86_64-linux-selfhost-modules builder.

Differential Revision: https://reviews.llvm.org/D38796

llvm-svn: 315739
2017-10-13 19:55:01 +00:00
Ivan A. Kosarev 0e528202b8 [CodeGen] EmitPointerWithAlignment() to generate TBAA info along with LValue base info
Differential Revision: https://reviews.llvm.org/D38796

llvm-svn: 315731
2017-10-13 18:40:18 +00:00
Reid Kleckner 3da37e05f7 [MS] Don't bail on replacing dllimport vbase dtors with base dtors
Fix PR32990 by effectively reverting r283063 and solving it a different
way.

We want to limit the hack to not replace equivalent available_externally
dtors specifically to libc++, which uses always_inline. It seems certain
versions of libc++ do not provide all the symbols that an explicit
template instantiation is expected to provide.

If we get to the code that forms a real alias, only *then* check if this
is available_externally, and do that by asking a better question, which
is "is this a declaration for the linker?", because *that's* what means
we can't form an alias to it.

As a follow-on simplification, remove the InEveryTU parameter. Its last
use guarded this code for forming aliases, but we should never form
aliases to declarations, regardless of what we know about every TU.

llvm-svn: 315656
2017-10-13 00:53:02 +00:00
Ivan A. Kosarev 383890bad4 Refine generation of TBAA information in clang
This patch is an attempt to clarify and simplify generation and
propagation of TBAA information. The idea is to pack all values
that describe a memory access, namely, base type, access type and
offset, into a single structure. This is supposed to make further
changes, such as adding support for unions and array members,
easier to prepare and review.

DecorateInstructionWithTBAA() is no more responsible for
converting types to tags. These implicit conversions not only
complicate reading the code, but also suggest assigning scalar
access tags while we generally prefer full-size struct-path tags.

TBAAPathTag is replaced with TBAAAccessInfo; the latter is now
the type of the keys of the cache map that translates access
descriptors to metadata nodes.

Fixed a bug with writing to a wrong map in
getTBAABaseTypeMetadata() (former getTBAAStructTypeInfo()).

We now check for valid base access types every time we
dereference a field. The original code only checks the top-level
base type. See isValidBaseType() / isTBAAPathStruct() calls.

Some entities have been renamed to sound more adequate and less
confusing/misleading in presence of path-aware TBAA information.

Now we do not lookup twice for the same cache entry in
getAccessTagInfo().

Refined relevant comments and descriptions.

Differential Revision: https://reviews.llvm.org/D37826

llvm-svn: 315048
2017-10-06 08:17:48 +00:00
Ivan A. Kosarev 3d68ce90f2 [CodeGen] Unify generation of scalar and struct-path TBAA tags
This patch makes it possible to produce access tags in a uniform
manner regardless whether the resulting tag will be a scalar or a
struct-path one. getAccessTagInfo() now takes care of the actual
translation of access descriptors to tags and can handle all
kinds of accesses. Facilities that specific to scalar accesses
are eliminated.

Some more details:
* DecorateInstructionWithTBAA() is not responsible for conversion
  of types to access tags anymore. Instead, it takes an access
  descriptor (TBAAAccessInfo) and generates corresponding access
  tag from it.
* getTBAAInfoForVTablePtr() reworked to
  getTBAAVTablePtrAccessInfo() that now returns the
  virtual-pointer access descriptor and not the virtual-point
  type metadata.
* Added function getTBAAMayAliasAccessInfo() that returns the
  descriptor for may-alias accesses.
* getTBAAStructTagInfo() renamed to getTBAAAccessTagInfo() as now
  it is the only way to generate access tag by a given access
  descriptor. It is capable of producing both scalar and
  struct-path tags, depending on options and availability of the
  base access type. getTBAAScalarTagInfo() and its cache
  ScalarTagMetadataCache are eliminated.
* Now that we do not need to care about whether the resulting
  access tag should be a scalar or struct-path one,
  getTBAAStructTypeInfo() is renamed to getBaseTypeInfo().
* Added function getTBAAAccessInfo() that constructs access
  descriptor by a given QualType access type.

This is part of D37826 reworked to be a separate patch to
simplify review.

Differential Revision: https://reviews.llvm.org/D38503

llvm-svn: 314979
2017-10-05 11:08:17 +00:00
Ivan A. Kosarev afc074cc41 Revert r314977 "[CodeGen] Unify generation of scalar and struct-path TBAA tags"
D37826 has been mistakenly committed where it should be the patch from D38503.

Differential Revision: https://reviews.llvm.org/D38503

llvm-svn: 314978
2017-10-05 11:05:43 +00:00
Ivan A. Kosarev 6fa20cfea3 [CodeGen] Unify generation of scalar and struct-path TBAA tags
This patch makes it possible to produce access tags in a uniform
manner regardless whether the resulting tag will be a scalar or a
struct-path one. getAccessTagInfo() now takes care of the actual
translation of access descriptors to tags and can handle all
kinds of accesses. Facilities that specific to scalar accesses
are eliminated.

Some more details:
* DecorateInstructionWithTBAA() is not responsible for conversion
  of types to access tags anymore. Instead, it takes an access
  descriptor (TBAAAccessInfo) and generates corresponding access
  tag from it.
* getTBAAInfoForVTablePtr() reworked to
  getTBAAVTablePtrAccessInfo() that now returns the
  virtual-pointer access descriptor and not the virtual-point
  type metadata.
* Added function getTBAAMayAliasAccessInfo() that returns the
  descriptor for may-alias accesses.
* getTBAAStructTagInfo() renamed to getTBAAAccessTagInfo() as now
  it is the only way to generate access tag by a given access
  descriptor. It is capable of producing both scalar and
  struct-path tags, depending on options and availability of the
  base access type. getTBAAScalarTagInfo() and its cache
  ScalarTagMetadataCache are eliminated.
* Now that we do not need to care about whether the resulting
  access tag should be a scalar or struct-path one,
  getTBAAStructTypeInfo() is renamed to getBaseTypeInfo().
* Added function getTBAAAccessInfo() that constructs access
  descriptor by a given QualType access type.

This is part of D37826 reworked to be a separate patch to
simplify review.

Differential Revision: https://reviews.llvm.org/D38503

llvm-svn: 314977
2017-10-05 10:47:51 +00:00
Ivan A. Kosarev a511ed7501 [CodeGen] Introduce generic TBAA access descriptors
With this patch we implement a concept of TBAA access descriptors
that are capable of representing both scalar and struct-path
accesses in a generic way.

This is part of D37826 reworked to be a separate patch to
simplify review.

Differential Revision: https://reviews.llvm.org/D38456

llvm-svn: 314780
2017-10-03 10:52:39 +00:00