Commits · 4b25c231cee6925725e268d9fb61841b6340dcf6 · HPC2SE-Project / pacxx-clang

Apr 27, 2015

[MS ABI] Use 'continue' instead of 'return false' where intended · 4b25c231

Reid Kleckner authored 9 years ago

This was a bug in r218285 that prevented us from seeing subsequent
virtual bases in the class hierarchy, leading to crashes later.

Also add some comments to this function, now that we better understand
what it's trying to do.

Fixes PR21062 and PR21064.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235899 91177308-0d34-0410-b5e6-96231b3b80d8

4b25c231

[OPENMP] Simplified iteration over clauses, NFC. · e9fda642

Alexey Bataev authored 9 years ago

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235838 91177308-0d34-0410-b5e6-96231b3b80d8

e9fda642

[OPENMP] Codegen for 'taskwait' directive. · 4c83c827

Alexey Bataev authored 9 years ago

Emit the following code for 'taskwait' directive within tied task:
call i32 @__kmpc_omp_taskwait(<loc>, i32 <thread_id>);
Differential Revision: http://reviews.llvm.org/D9245

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235836 91177308-0d34-0410-b5e6-96231b3b80d8

4c83c827

[OPENMP] Codegen for 'reduction' clause in 'sections' directive. · 05856ab8

Alexey Bataev authored 9 years ago

Emit a code for reduction clause. Next code should be emitted for reductions:

static kmp_critical_name lock = { 0 };

void reduce_func(void *lhs[<n>], void *rhs[<n>]) {
    *(Type0*)lhs[0] = ReductionOperation0(*(Type0*)lhs[0], *(Type0*)rhs[0]);
      ...
        *(Type<n>-1*)lhs[<n>-1] =
          ReductionOperation<n>-1(*(Type<n>-1*)lhs[<n>-1],
            *(Type<n>-1*)rhs[<n>-1]);
}

...
void *RedList[<n>] = {&<RHSExprs>[0], ..., &<RHSExprs>[<n>-1]};
switch (__kmpc_reduce{_nowait}(<loc>, <gtid>, <n>, sizeof(RedList), RedList, reduce_func, &<lock>)) {
case 1:
  <LHSExprs>[0] = ReductionOperation0(*<LHSExprs>[0], *<RHSExprs>[0]);
  ...
  <LHSExprs>[<n>-1] = ReductionOperation<n>-1(*<LHSExprs>[<n>-1], *<RHSExprs>[<n>-1]);
  __kmpc_end_reduce{_nowait}(<loc>, <gtid>, &<lock>);
  break;
case 2:
  Atomic(<LHSExprs>[0] = ReductionOperation0(*<LHSExprs>[0], *<RHSExprs>[0]));
  ...
  Atomic(<LHSExprs>[<n>-1] = ReductionOperation<n>-1(*<LHSExprs>[<n>-1], *<RHSExprs>[<n>-1]));
  break;
default:;
}
Reduction variables are a kind of a private variables, they have private copies, but initial values are chosen in accordance with the reduction operation.
If sections directive has only single section, then original shared variables are used instead with barrier at the end of the directive.
Differential Revision: http://reviews.llvm.org/D9242



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235835 91177308-0d34-0410-b5e6-96231b3b80d8

05856ab8

[OPENMP] Codegen for 'lastprivate' clause in 'sections' directive. · 7b00e6cf

Alexey Bataev authored 9 years ago

#pragma omp sections lastprivate(<var>)
<BODY>;
This construct is translated into something like:

<last_iter> = alloca i32
<init for lastprivates>;
<last_iter> = 0
; No initializer for simple variables or a default constructor is called for objects.
; For arrays perform element by element initialization by the call of the default constructor.
...
OMP_FOR_START(...,<last_iter>, ..); sets <last_iter> to 1 if this is the last iteration.
<BODY>
...
OMP_FOR_END
if (<last_iter> != 0) {
  <final copy for lastprivate>; Update original variable with the lastprivate value.
}
call __kmpc_cancel_barrier() ; an implicit barrier to avoid possible data race.
If there is only one section, there is no special code generation, original shared variables are used + barrier is emitted at the end of the directive.
Differential Revision: http://reviews.llvm.org/D9240


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235834 91177308-0d34-0410-b5e6-96231b3b80d8

7b00e6cf

[OPENMP] Codegen for 'private' clause in 'sections' directive. · 47709fb9

Alexey Bataev authored 9 years ago

If there are 2 or more sections in a 'section' directive the following code is generated:

<default init for privates>
@__kmpc_for_static_init_4();
<BODY for sections directive>
@__kmpc_for_static_fini()
If there is only one section, the following code is generated:

if (@__kmpc_single()) {
  <default init for privates>
  @__kmpc_end_single();
}
Differential Revision: http://reviews.llvm.org/D9239


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235833 91177308-0d34-0410-b5e6-96231b3b80d8

47709fb9

[OPENMP] Codegen for 'private' clause in 'single' directive. · f22d311e

Alexey Bataev authored 9 years ago

Emit the following code for 'single' directive with 'private' clause:

if (@__kmpc_single()) {
  <default init for privates>
  @__kmpc_end_single();
}
Differential Revision: http://reviews.llvm.org/D9238


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235832 91177308-0d34-0410-b5e6-96231b3b80d8

f22d311e

[MS ABI] Rephrase the mangling of array types in parameters · 8e9cfe28

David Majnemer authored 9 years ago

Make the canonicalization of array types more consistent.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235831 91177308-0d34-0410-b5e6-96231b3b80d8

8e9cfe28

Apr 26, 2015

[Sema] Do not permit binding a reference to a compound literal · 1dc7490a

David Majnemer authored 9 years ago

We could probably make this work if we cared enough.  However, we are
far outside any language rules at this point.

This fixes PR21834.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235818 91177308-0d34-0410-b5e6-96231b3b80d8

1dc7490a

[Sema] Don't allow unverified bitfields in FieldDecls · 3ca1dc1f

David Majnemer authored 9 years ago

VerifyBitField must be called if we are to form a bitfield FieldDecl.
We will not verify the bitfield if the decl is known to be malformed in
other ways; pretend that we don't have a bitfield if this happens.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235816 91177308-0d34-0410-b5e6-96231b3b80d8

3ca1dc1f

Correctly handle zero-sized but non-empty base classes in IRGen. · 1f7af499

John McCall authored 9 years ago

Fixes rdar://20621065.

A more elegant fix would preclude this case by defining the
rules such that zero-size classes are always formally empty.
I believe the only extensions which create zero-size classes
right now are flexible arrays and zero-length arrays; it's
not abstractly unreasonable to say that those don't count
as members for the purposes of emptiness, just as zero-width
bitfields don't count.  But that's an ABI-affecting change
and requires further discussion; in the meantime, let's not
assert / miscompile.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235815 91177308-0d34-0410-b5e6-96231b3b80d8

1f7af499

Apr 25, 2015

[Sema] Check if a builtin is FunctionPrototype(). · ca305f5b

Davide Italiano authored 9 years ago

Don't assume it's always is. This prevents a crash in Sema while
trying to merge return type for a builtin w/out function prototype.
 
PR:		23086
Differential Revision:	http://reviews.llvm.org/D9235
Reviewed by:	rsmith


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235806 91177308-0d34-0410-b5e6-96231b3b80d8

ca305f5b

InstrProf: Fix coverage maps for conditional operators · ec83e369

Justin Bogner authored 9 years ago

This fixes a crash when we're emitting coverage and a macro appears
between two binary conditional operators, ie, "foo ?: MACRO ?: bar",
and fixes the interaction of macros and conditional operators in
general.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235793 91177308-0d34-0410-b5e6-96231b3b80d8

ec83e369

clang-cl: Don't look up absolute paths in %LIB%. · 00b4e295

Nico Weber authored 9 years ago

Before this patch, passing a non-existent absolute path to clang-cl would cause
stat'ing of impossible paths. For example, `clang-cl -c d:\adsfasdf.txt` would
cause a stat of
C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\LIBd:\asdfadsf.cc

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235787 91177308-0d34-0410-b5e6-96231b3b80d8

00b4e295

Apr 24, 2015

[opaque pointer type] Update test cases now that the type for an invoke is... · c40864e6

David Blaikie authored 9 years ago

[opaque pointer type] Update test cases now that the type for an invoke is just a function type, not a pointer-to-function type

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235756 91177308-0d34-0410-b5e6-96231b3b80d8

c40864e6

Revert r235749 - Accidentally commited cruft from the wrong path. · 67196e47
Lang Hames authored 9 years ago
```
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235750 91177308-0d34-0410-b5e6-96231b3b80d8
```
67196e47

[CodeGen] Make AsmPrinter's OutStreamer member a unique_ptr. · c7338374

Lang Hames authored 9 years ago

AsmPrinter owns the OutStreamer, so an owning pointer makes sense here. Using a
reference for this is crufty.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235749 91177308-0d34-0410-b5e6-96231b3b80d8

c7338374

clang-format: Don't wrap after short first segments of builder calls. · 521c1564

Daniel Jasper authored 9 years ago

Before:
  a()
      .aaaaa()
      .aaaaa()
      .aaaaa();

After:
  a().aaaaa()
      .aaaaa()
      .aaaaa();

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235707 91177308-0d34-0410-b5e6-96231b3b80d8

521c1564

clang-format: More selectively detect QT's "signals". · c425488a
Daniel Jasper authored 9 years ago
```
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235702 91177308-0d34-0410-b5e6-96231b3b80d8
```
c425488a
[TableGen] Clang changes for r235697 to stop leaking Expanders and Operators in SetTheory. · 5e92fff2
Craig Topper authored 9 years ago
```
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235698 91177308-0d34-0410-b5e6-96231b3b80d8
```
5e92fff2

[OPENMP] Codegen for 'firstprivate' clause in 'single' directive. · e1bb9207

Alexey Bataev authored 9 years ago

Emit the following code for 'single' directive with 'firtstprivate' clause:

if (@__kmpc_single()) {
  <init for firstprivates>
  @__kmpc_end_single();
}
@__kmpc_cancel_barrier(); // To avoid data race in firstprivate init
Differential Revision: http://reviews.llvm.org/D9223


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235694 91177308-0d34-0410-b5e6-96231b3b80d8

e1bb9207

Remove unused variable to silence GCC warning · 4e55e7d4

David Majnemer authored 9 years ago

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235693 91177308-0d34-0410-b5e6-96231b3b80d8

4e55e7d4

[OPENMP] Do not emit implicit barrier for single directive with 'copyprivate' clause(s). · b44793c0

Alexey Bataev authored 9 years ago

Runtime function for 'copyprivate' directive generates implicit barriers, so no need to emit it.
Differential Revision: http://reviews.llvm.org/D9215

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235692 91177308-0d34-0410-b5e6-96231b3b80d8

b44793c0

[OPENMP] Codegen for 'firstprivate' clause in 'sections' directive. · 8cd55277

Alexey Bataev authored 9 years ago

If there are 2 or more sections in a 'section' directive the following code is generated:

<init for firstprivates>
@__kmpc_cancel_barrier();// To avoid data race in firstprivate init
@__kmpc_for_static_init_4();
<BODY for sections directive>
@__kmpc_for_static_fini()
If there is only one section, the following code is generated:

if (@__kmpc_single()) {
  <init for firstprivates>
  @__kmpc_end_single();
}
@__kmpc_cancel_barrier(); // To avoid data race in firstprivate init
Differential Revision: http://reviews.llvm.org/D9214


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235691 91177308-0d34-0410-b5e6-96231b3b80d8

8cd55277

Replace getPointeeType()->isFunctionType with isMemberDataPointerType · 3fb626d1
David Majnemer authored 9 years ago
```
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235682 91177308-0d34-0410-b5e6-96231b3b80d8
```
3fb626d1

[MS ABI] Fix the preferred alignment of member pointers · a0f93234

David Majnemer authored 9 years ago

Member pointers in the MS ABI have different alignment depending on
whether they were created on the stack or live in a record.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235681 91177308-0d34-0410-b5e6-96231b3b80d8

a0f93234

Cleanup some MS-ABI specific code · f3cb3aa2

David Majnemer authored 9 years ago

No functional change intended.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235680 91177308-0d34-0410-b5e6-96231b3b80d8

f3cb3aa2

[modules] Partial revert of r235669: don't create ModuleMacros for imported local macros. · a9f636ee

Richard Smith authored 9 years ago

The surrounding infrastructure isn't quite ready for this yet.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235677 91177308-0d34-0410-b5e6-96231b3b80d8

a9f636ee

[modules] Refactor creation of ModuleMacros and create them when importing from local submodules. · cd318b3e
Richard Smith authored 9 years ago
```
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235669 91177308-0d34-0410-b5e6-96231b3b80d8
```
cd318b3e

InstrProf: Stop using RegionCounter outside of CodeGenPGO (NFC) · aa53a4f2

Justin Bogner authored 9 years ago

The RegionCounter type does a lot of legwork, but most of it is only
meaningful within the implementation of CodeGenPGO. The uses elsewhere
in CodeGen generally just want to increment or read counters, so do
that directly.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235664 91177308-0d34-0410-b5e6-96231b3b80d8

aa53a4f2

[modules] Properly attribute macros to modules if they're in a file textually... · b58f3164

Richard Smith authored 9 years ago

[modules] Properly attribute macros to modules if they're in a file textually included into a file in the module.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235661 91177308-0d34-0410-b5e6-96231b3b80d8

b58f3164

Apr 23, 2015

Fix clang docs build. · 9fad2a17

Sergey Matveev authored 9 years ago

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235650 91177308-0d34-0410-b5e6-96231b3b80d8

9fad2a17

[modules] Remove the now-redundant import of all pending macros at the end of building a module. · f081abcc

Richard Smith authored 9 years ago

Since we now track module macros separately from their visibility state, this
is no longer necessary.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235648 91177308-0d34-0410-b5e6-96231b3b80d8

f081abcc

[modules] Store a ModuleMacro* on an imported macro directive rather than... · a09eec8c

Richard Smith authored 9 years ago

[modules] Store a ModuleMacro* on an imported macro directive rather than duplicating the info within it.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235644 91177308-0d34-0410-b5e6-96231b3b80d8

a09eec8c

Add clang/docs/SanitizerCoverage.rst · abe1c4b5

Sergey Matveev authored 9 years ago

Moved from https://code.google.com/p/address-sanitizer/wiki/AsanCoverage

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235643 91177308-0d34-0410-b5e6-96231b3b80d8

abe1c4b5

[modules] Determine the set of macros exported by a submodule at the end of that submodule. · eacc7339

Richard Smith authored 9 years ago

Previously we'd defer this determination until writing the AST, which doesn't
allow us to use this information when building other submodules of the same
module. This change also allows us to use a uniform mechanism for writing
module macro records, independent of whether they are local or imported.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235614 91177308-0d34-0410-b5e6-96231b3b80d8

eacc7339

Don't emit lifetime markers when msan is enabled · 9e7eb7e6

Reid Kleckner authored 9 years ago

In r235553, Clang started emitting lifetime markers more often. This
caused false negative in MSan, because MSan only poisons all allocas
once at function entry. Eventually, MSan should poison allocas at
lifetime start and probably also lifetime end, but until then, let's not
emit markers that aren't going to be useful.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235613 91177308-0d34-0410-b5e6-96231b3b80d8

9e7eb7e6

Extend format specifier checking to include field function pointers in... · 24aae829

Aaron Ballman authored 9 years ago

Extend format specifier checking to include field function pointers in addition to variable function pointers. Addresses PR21082.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235606 91177308-0d34-0410-b5e6-96231b3b80d8

24aae829

Diagnose variadic main() as an extension; addresses PR17905. · f92f0762
Aaron Ballman authored 9 years ago
```
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235605 91177308-0d34-0410-b5e6-96231b3b80d8
```
f92f0762

clang-format: Properly detect variable declarations with ObjC. · 418836a4

Daniel Jasper authored 9 years ago

Before:
  LoooooooooooooooooooooooooooooooooooooooongType
  LoooooooooooooooooooooooooooooooooooooongVariable([A a]);

After:
  LoooooooooooooooooooooooooooooooooooooooongType
      LoooooooooooooooooooooooooooooooooooooongVariable([A a]);

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@235599 91177308-0d34-0410-b5e6-96231b3b80d8

418836a4