Dmitri Gribenko [Mon, 23 Mar 2015 19:23:50 +0000 (19:23 +0000)]
Refactor: Simplify boolean expressions in libclang
Simplify boolean expresions using true and false with clang-tidy.
Patch by Richard.
llvm-svn: 232996
Benjamin Kramer [Mon, 23 Mar 2015 19:06:17 +0000 (19:06 +0000)]
[ctorutils] Update and sort includes. NFC.
llvm-svn: 232995
Benjamin Kramer [Mon, 23 Mar 2015 18:57:17 +0000 (18:57 +0000)]
[winehprepare] Update and sort includes. NFC.
llvm-svn: 232994
Benjamin Kramer [Mon, 23 Mar 2015 18:45:56 +0000 (18:45 +0000)]
Another set of missing raw_ostream.h. Still no functional change.
llvm-svn: 232993
Matt Arsenault [Mon, 23 Mar 2015 18:45:41 +0000 (18:45 +0000)]
R600/SI: Merge tables for commuting
Don't use a separate table for compares anymore,
and use the same VOP2_REV class.
llvm-svn: 232992
Matt Arsenault [Mon, 23 Mar 2015 18:45:38 +0000 (18:45 +0000)]
R600/SI: Only use one range of isCommutable for compares
Also don't count the class instructions as isCompare anymore.
llvm-svn: 232991
Matt Arsenault [Mon, 23 Mar 2015 18:45:36 +0000 (18:45 +0000)]
R600/SI: Remove redundant unsetting of hasSideEffects
These are already set in the base class for the instruction.
llvm-svn: 232990
Matt Arsenault [Mon, 23 Mar 2015 18:45:35 +0000 (18:45 +0000)]
R600/SI: Move hasSideEffects setting into VOPCX classes
llvm-svn: 232989
Matt Arsenault [Mon, 23 Mar 2015 18:45:30 +0000 (18:45 +0000)]
R600/SI: Allow commuting compares
This enables very common cases to switch to the
smaller encoding.
All of the standard LLVM canonicalizations of comparisons
are the opposite of what we want. Compares with constants
are moved to the RHS, but the first operand can be an inline
immediate, literal constant, or SGPR using the 32-bit VOPC
encoding.
There are additional bad canonicalizations that should
also be fixed, such as canonicalizing ge x, k to gt x, (k + 1)
if this makes k no longer an inline immediate value.
llvm-svn: 232988
Matt Arsenault [Mon, 23 Mar 2015 18:45:23 +0000 (18:45 +0000)]
R600/SI: Use right class for cmpsx f64 instructions
Use VOPCX_F64 to not need the let Defs = [EXEC]
llvm-svn: 232987
Matt Arsenault [Mon, 23 Mar 2015 18:45:20 +0000 (18:45 +0000)]
R600/SI: Remove cond operand to VOPCX classes
It isn't used, and these will probably never be directly selected.
llvm-svn: 232986
Rafael Auler [Mon, 23 Mar 2015 18:39:37 +0000 (18:39 +0000)]
[ELF] Remove extraneous semicolon in DefaultLayout
llvm-svn: 232985
David Blaikie [Mon, 23 Mar 2015 18:39:02 +0000 (18:39 +0000)]
Refactor: simplify boolean expressions in llvm-objdump
Simplify boolean expressions involving `true` and `false` with `clang-tidy`.
Actually upon inspection a bunch of these boolean variables could be
factored away entirely anyway - using find_if and then testing the
result before using it. This also helps reduce indentation in the code
anyway - and a bunch of other related simplification fell out nearby so
I just committed all of that.
Patch by Richard Thomson (legalize@xmission.com)
Differential Revision: http://reviews.llvm.org/D8517
llvm-svn: 232984
Stephane Sezer [Mon, 23 Mar 2015 18:36:54 +0000 (18:36 +0000)]
Initialize ObjC runtime at the right location.
Summary:
Saw this while reading some code in DynamicLoader classes. Looks like this has
been a FIXME since 2011 at least.
Test Plan: Run unit tests.
Reviewers: clayborg
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D8558
llvm-svn: 232983
Yaron Keren [Mon, 23 Mar 2015 18:35:01 +0000 (18:35 +0000)]
Add missing ELFObjectWriter::reset() override, like other MC classes.
See detailed discussion at
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-
20140915/235418.html
and r217907, r217948:
http://llvm.org/viewvc/llvm-project?view=revision&revision=217907
http://llvm.org/viewvc/llvm-project?view=revision&revision=217948
llvm-svn: 232982
Benjamin Kramer [Mon, 23 Mar 2015 18:23:08 +0000 (18:23 +0000)]
More missing includes only visible to MSVC.
NFC.
llvm-svn: 232981
Benjamin Kramer [Mon, 23 Mar 2015 18:19:41 +0000 (18:19 +0000)]
Add missing include that MSVC complains about.
Also reorder includes a bit, NFC.
llvm-svn: 232980
Stephane Sezer [Mon, 23 Mar 2015 18:15:07 +0000 (18:15 +0000)]
Add a missing null pointer check in CommandObjectThread.cpp.
Reviewers: clayborg
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D8554
llvm-svn: 232979
Tom Stellard [Mon, 23 Mar 2015 18:10:04 +0000 (18:10 +0000)]
Implement fast_distance builtin
This implementation was ported from the AMD builtin library
and has been tested with piglit, OpenCV, and the ocl conformance tests.
llvm-svn: 232978
Tom Stellard [Mon, 23 Mar 2015 18:10:02 +0000 (18:10 +0000)]
Implement fast_length builtin
This implementation was ported from the AMD builtin library
and has been tested with piglit, OpenCV, and the ocl conformance tests.
llvm-svn: 232977
Benjamin Kramer [Mon, 23 Mar 2015 18:07:13 +0000 (18:07 +0000)]
Purge unused includes throughout libSupport.
NFC.
llvm-svn: 232976
Benjamin Kramer [Mon, 23 Mar 2015 18:05:43 +0000 (18:05 +0000)]
Explicitly include raw_ostream.h instead of relying on transitive inclusion.
NFC.
llvm-svn: 232975
Zachary Turner [Mon, 23 Mar 2015 18:00:07 +0000 (18:00 +0000)]
Rely on Makefile.rules in lldb/test/driver/batch_mode.
This was hardcoding some make rules instead of relying on
Makefile.rules, which was causing some of the logic to be
incorrect for Windows.
Patch by: Adrian McCarthy
Differential Revision: http://reviews.llvm.org/D8363
llvm-svn: 232974
Ed Maste [Mon, 23 Mar 2015 17:57:42 +0000 (17:57 +0000)]
Remove non-standard Environment header section
llvm-svn: 232973
Alexander Potapenko [Mon, 23 Mar 2015 17:55:19 +0000 (17:55 +0000)]
[sanitizer] Fix file access modes in SanitizerCommon.InternalMmapWithOffset and SanitizerCommon.FileOps
llvm-svn: 232972
Ahmed Bougacha [Mon, 23 Mar 2015 17:54:16 +0000 (17:54 +0000)]
[CodeGen] Properly support the half FP type with non-native operations.
On AArch64, the -fallow-half-args-and-returns option is the default.
With it, the half type is considered legal (rather than the i16 used
normally for __fp16), but no operation is, except conversions and
load/stores and such.
The previous behavior was tantamount to saying LangOpts.NativeHalfType
was implied by LangOpts.HalfArgsAndReturns, which isn't true.
Instead, teach the various parts of CodeGen that already know about
half (using the intrinsics or not) about this weird in-between case,
where the "half" type is legal, but operations on it aren't.
This is a smaller intermediate step to the end-goal of removing the
intrinsic, always using "half", and letting the backend legalize.
Builds on r232968.
rdar://
20045970, rdar://
17468714
Differential Revision: http://reviews.llvm.org/D8367
llvm-svn: 232971
Ed Maste [Mon, 23 Mar 2015 17:52:38 +0000 (17:52 +0000)]
Use .so library extension by default if platform is not Windows or Darwin
llvm-svn: 232970
Ed Maste [Mon, 23 Mar 2015 17:52:10 +0000 (17:52 +0000)]
Fix whitespace in finishSwigWrapperClasses.py
- replace hard tabs with 4-space indents
- delete EOL whitespace
llvm-svn: 232969
Ahmed Bougacha [Mon, 23 Mar 2015 17:48:07 +0000 (17:48 +0000)]
[CodeGen] Convert double -> __fp16 in one step.
Fix the CodeGen so that for types bigger than float, instead of
converting to fp16 via the sequence "InTy -> float -> fp16", we
perform conversions in just one step. This avoids the double
rounding which potentially changes results from a natural
IEEE-754 operation.
rdar://
17594379, rdar://
17468714
Differential Revision: http://reviews.llvm.org/D4602
Part of: http://reviews.llvm.org/D8367
llvm-svn: 232968
Chad Rosier [Mon, 23 Mar 2015 17:19:34 +0000 (17:19 +0000)]
[AArch64] Enable rematerialization of float 0 values.
Patch by Geoff Berry<gberry@codeaurora.org>.
llvm-svn: 232967
Stephane Sezer [Mon, 23 Mar 2015 17:05:41 +0000 (17:05 +0000)]
Do not assert on POSIXDYLD double-eAdd.
Summary:
This has been discovered while experimenting with the gecko linker on android.
In general, assert()'ing on "user input" is a bad idea.
Test Plan: Run unit tests.
Reviewers: clayborg, tfiala
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D8495
llvm-svn: 232966
Tom Stellard [Mon, 23 Mar 2015 17:01:37 +0000 (17:01 +0000)]
Implement half_sqrt builtin v2
This is a generic implementation which just calls sqrt. Targets should
override this if they want a faster implementation.
v2:
- Alphabetize SOURCES
llvm-svn: 232965
Tom Stellard [Mon, 23 Mar 2015 17:01:35 +0000 (17:01 +0000)]
Implement distance builtin v2
This implementation was ported from the AMD builtin library
and has been tested with piglit, OpenCV, and the ocl conformance tests.
v2:
- Remove unnecessary copyright.
llvm-svn: 232964
Tom Stellard [Mon, 23 Mar 2015 17:01:33 +0000 (17:01 +0000)]
Fix implementation of length builtin v2
v2:
- Move common code into a macro
- Use the same constant for all vector types.
llvm-svn: 232963
Bradley Smith [Mon, 23 Mar 2015 16:52:52 +0000 (16:52 +0000)]
Revert "[ARM] Add more pattern matching for f16 <-> f64 conversions"
This change is incorrect since it converts double rounding into single rounding,
which can produce different results. Instead this optimization will be done by
modifying Clang's codegen to not produce double rounding in the first place.
This reverts commit r232954.
llvm-svn: 232962
Eli Bendersky [Mon, 23 Mar 2015 16:26:23 +0000 (16:26 +0000)]
Simplify boolean expressions with true and false using clang-tidy
Patch by Richard (legalize@xmission.com)
Differential Revision: http://reviews.llvm.org/D8521
llvm-svn: 232961
Tom Stellard [Mon, 23 Mar 2015 16:20:24 +0000 (16:20 +0000)]
Add __clc_ prefix to functions in sincos_helpers.cl
This will help avoid naming conflicts with functions defined in
kernels linking with libclc.
llvm-svn: 232960
Tamas Berghammer [Mon, 23 Mar 2015 16:17:47 +0000 (16:17 +0000)]
Shorten executable name in WatchpointLLDBCommandTestCase
It is required because the name of the executable exceeded the maximum
allowed file name on android.
llvm-svn: 232959
James Molloy [Mon, 23 Mar 2015 16:15:16 +0000 (16:15 +0000)]
[ARM] Remove target-specific ITOFP/FPTOI nodes
Anton tried this 5 years ago but it was reverted due to extra VMOVs
being emitted. This can be easily fixed with a liberal application
of patterns - matching loads/stores and extractelts.
llvm-svn: 232958
Tom Stellard [Mon, 23 Mar 2015 16:06:01 +0000 (16:06 +0000)]
R600/SI: Fix crash in SIInstrInfo::areLoadsFromSameBasePtr()
This function assumed that SMRD instructions always have immediate
offsets, which is not always the case.
llvm-svn: 232957
NAKAMURA Takumi [Mon, 23 Mar 2015 16:05:50 +0000 (16:05 +0000)]
OpenMPClause.h: Fix typo in \param. [-Wdocumentation]
llvm-svn: 232956
Colin LeMahieu [Mon, 23 Mar 2015 16:01:03 +0000 (16:01 +0000)]
[Hexagon] Simplify boolean expression
Patch by Richard
http://reviews.llvm.org/D8523
llvm-svn: 232955
Bradley Smith [Mon, 23 Mar 2015 15:59:54 +0000 (15:59 +0000)]
[ARM] Add more pattern matching for f16 <-> f64 conversions
Specifically when the conversion is done in two steps, f16 -> f32 -> f64.
For example:
%1 = tail call float @llvm.convert.from.fp16.f32(i16 %0)
%conv = fpext float %1 to double
to:
vcvtb.f64.f16
llvm-svn: 232954
Benjamin Kramer [Mon, 23 Mar 2015 15:55:00 +0000 (15:55 +0000)]
Fix the CMake shared build.
The ELF backends now depend on lld::script::Sema, which is in libReaderWriter.
Link it explicitly.
llvm-svn: 232953
Tamas Berghammer [Mon, 23 Mar 2015 15:50:03 +0000 (15:50 +0000)]
Remove virtual and add override to all virtual functions in Process/gdb-remote.
llvm-svn: 232952
Ilia K [Mon, 23 Mar 2015 14:22:27 +0000 (14:22 +0000)]
Fix ObjCDataFormatterTestCase.test_nsdate_* tests (ver. 2)
This commit reverts r232946 because it caused an another error with absolute time.
llvm-svn: 232951
Ilia K [Mon, 23 Mar 2015 14:00:54 +0000 (14:00 +0000)]
Convert CRLF to LF in CODE_OWNERS.txt
llvm-svn: 232950
Benjamin Kramer [Mon, 23 Mar 2015 13:59:13 +0000 (13:59 +0000)]
[gcov] Move formatBranchInfo into an anonymous namespace.
NFC.
llvm-svn: 232949
Benjamin Kramer [Mon, 23 Mar 2015 12:49:15 +0000 (12:49 +0000)]
Make helpers static. clang-tools edition.
Also purge dead code found by it. NFC.
llvm-svn: 232948
Benjamin Kramer [Mon, 23 Mar 2015 12:47:15 +0000 (12:47 +0000)]
[tooling] Move ArgumentsAdjustingCompilations into an anonymous namespace.
NFC.
llvm-svn: 232947
Ilia K [Mon, 23 Mar 2015 12:44:54 +0000 (12:44 +0000)]
Fix ObjCDataFormatterTestCase.test_nsdate_* tests
The time/date strings (like "6pm April 10, 1985") are interpreted as a local time but CFDateGetAbsoluteTime() returns time in UTC. It caused a problem when local time was UTC+0100 or more (0200, 0300 etc.):
```
======================================================================
FAIL: test_nsdate_with_dsym_and_run_command (TestDataFormatterObjC.ObjCDataFormatterTestCase)
Test formatters for NSDate.
----------------------------------------------------------------------
Traceback (most recent call last):
File "/Users/IliaK/p/llvm/tools/lldb/test/lldbtest.py", line 462, in wrapper
return func(self, *args, **kwargs)
File "/Users/IliaK/p/llvm/tools/lldb/test/functionalities/data-formatter/data-formatter-objc/TestDataFormatterObjC.py", line 157, in test_nsdate_with_dsym_and_run_command
self.appkit_tester_impl(self.buildDsym,self.nsdate_data_formatter_commands)
File "/Users/IliaK/p/llvm/tools/lldb/test/functionalities/data-formatter/data-formatter-objc/TestDataFormatterObjC.py", line 34, in appkit_tester_impl
commands()
File "/Users/IliaK/p/llvm/tools/lldb/test/functionalities/data-formatter/data-formatter-objc/TestDataFormatterObjC.py", line 475, in nsdate_data_formatter_commands
substrs = ['1985-04','2011-01'])
File "/Users/IliaK/p/llvm/tools/lldb/test/lldbtest.py", line 2146, in expect
msg if msg else EXP_MSG(str, exe))
AssertionError: False is not True : '2011-01' returns expected result
Config=x86_64-clang
======================================================================
FAIL: test_nsdate_with_dwarf_and_run_command (TestDataFormatterObjC.ObjCDataFormatterTestCase)
Test formatters for NSDate.
----------------------------------------------------------------------
Traceback (most recent call last):
File "/Users/IliaK/p/llvm/tools/lldb/test/lldbtest.py", line 479, in wrapper
return func(self, *args, **kwargs)
File "/Users/IliaK/p/llvm/tools/lldb/test/functionalities/data-formatter/data-formatter-objc/TestDataFormatterObjC.py", line 163, in test_nsdate_with_dwarf_and_run_command
self.appkit_tester_impl(self.buildDwarf,self.nsdate_data_formatter_commands)
File "/Users/IliaK/p/llvm/tools/lldb/test/functionalities/data-formatter/data-formatter-objc/TestDataFormatterObjC.py", line 34, in appkit_tester_impl
commands()
File "/Users/IliaK/p/llvm/tools/lldb/test/functionalities/data-formatter/data-formatter-objc/TestDataFormatterObjC.py", line 475, in nsdate_data_formatter_commands
substrs = ['1985-04','2011-01'])
File "/Users/IliaK/p/llvm/tools/lldb/test/lldbtest.py", line 2146, in expect
msg if msg else EXP_MSG(str, exe))
AssertionError: False is not True : '2011-01' returns expected result
Config=x86_64-clang
```
llvm-svn: 232946
Benjamin Kramer [Mon, 23 Mar 2015 12:31:05 +0000 (12:31 +0000)]
Move private classes into anonymous namespaces
Also merge anonymous namespaces in Targets.cpp a bit. NFC.
llvm-svn: 232945
Benjamin Kramer [Mon, 23 Mar 2015 12:30:58 +0000 (12:30 +0000)]
Move private classes into anonymous namespaces
NFC.
llvm-svn: 232944
Petar Jovanovic [Mon, 23 Mar 2015 12:28:13 +0000 (12:28 +0000)]
Fix sign extension for MIPS64 in makeLibCall function
Fixing sign extension in makeLibCall for MIPS64. In MIPS64 architecture all
32 bit arguments (int, unsigned int, float 32 (soft float)) must be sign
extended. This fixes test "MultiSource/Applications/oggenc/".
Patch by Strahinja Petrovic.
Differential Revision: http://reviews.llvm.org/D7791
llvm-svn: 232943
Denis Protivensky [Mon, 23 Mar 2015 11:38:24 +0000 (11:38 +0000)]
[ARM] Fix entry point for Thumb code
Test cases for both entry functions in ARM and Thumb
code are added.
llvm-svn: 232942
Daniel Sanders [Mon, 23 Mar 2015 11:33:15 +0000 (11:33 +0000)]
[aarch64] Distinguish the 'Q' and 'm' inline assembly memory constraints.
Summary:
But still handle them the same way since I don't know how they differ on
this target.
Clang also has code for 'Ump', 'Utf', 'Usa', and 'Ush' but calls
llvm_unreachable() on this code path so they are not converted to a
constraint id at the moment.
No functional change intended.
Reviewers: t.p.northover
Subscribers: aemerson, llvm-commits
Differential Revision: http://reviews.llvm.org/D8177
llvm-svn: 232941
Scott Douglass [Mon, 23 Mar 2015 10:54:24 +0000 (10:54 +0000)]
handle armeb/thumb/thumbeb consistently in gnutools::Assemble::ConstructJob
Differential Revision: http://reviews.llvm.org/D8196
llvm-svn: 232940
Scott Douglass [Mon, 23 Mar 2015 10:46:28 +0000 (10:46 +0000)]
Convert cascading if-else-if to switch. NFC
Differential Revision: http://reviews.llvm.org/D8485
llvm-svn: 232939
Simon Atanasyan [Mon, 23 Mar 2015 10:33:46 +0000 (10:33 +0000)]
[ELF] Add missing override keyword
No functional changes.
llvm-svn: 232938
Alexander Potapenko [Mon, 23 Mar 2015 10:32:28 +0000 (10:32 +0000)]
[libsanitizer] Fix OpenFile() usage in TSan and DFSan.
This is a follow-up for r232936.
llvm-svn: 232937
Alexander Potapenko [Mon, 23 Mar 2015 10:10:46 +0000 (10:10 +0000)]
[ASan] Distinguish between read, write and read-write file access modes in OpenFile.
This is to fix mapping coverage files into memory on OSX.
llvm-svn: 232936
Hal Finkel [Mon, 23 Mar 2015 08:22:43 +0000 (08:22 +0000)]
[SDAG] Don't widen VSETCC during type legalization for split operands
Because the operands of a vector SETCC node can be of a different type from the
result (and often are), it can happen that even if we'd prefer to widen the
result type of the SETCC, the operands have been split instead. In this case,
the SETCC result also must be split. This mirrors what is done in
WidenVecRes_SELECT, and should be NFC elsewhere because if the operands are not
widened the following calls to GetWidenedVector will assert (which is what was
happening in the test case).
llvm-svn: 232935
Tobias Grosser [Mon, 23 Mar 2015 07:00:36 +0000 (07:00 +0000)]
Drop -polly-vectorizer-unroll-only option
This options was earlier used for experiments with the vectorizer, but to my
knowledge is not really used anymore. If anybody needs this, we can always
reintroduce this feature.
llvm-svn: 232934
Alexey Bataev [Mon, 23 Mar 2015 06:40:37 +0000 (06:40 +0000)]
[OPENMP] Fixed test for 'single' directive codegen.
llvm-svn: 232933
Alexey Bataev [Mon, 23 Mar 2015 06:18:07 +0000 (06:18 +0000)]
[OPENMP] Codegen for 'copyprivate' clause ('single' directive).
If there is at least one 'copyprivate' clause is associated with the single directive, the following code is generated:
```
i32 did_it = 0; \\ for 'copyprivate' clause
if(__kmpc_single(ident_t *, gtid)) {
SingleOpGen();
__kmpc_end_single(ident_t *, gtid);
did_it = 1; \\ for 'copyprivate' clause
}
<copyprivate_list>[0] = &var0;
...
<copyprivate_list>[n] = &varn;
call __kmpc_copyprivate(ident_t *, gtid, <copyprivate_list_size>,
<copyprivate_list>, <copy_func>, did_it);
...
void<copy_func>(void *LHSArg, void *RHSArg) {
Dst = (void * [n])(LHSArg);
Src = (void * [n])(RHSArg);
Dst[0] = Src[0];
... Dst[n] = Src[n];
}
```
All list items from all 'copyprivate' clauses are gathered into single <copyprivate list> (<copyprivate_list_size> is a size in bytes of this list) and <copy_func> is used to propagate values of private or threadprivate variables from the 'single' region to other implicit threads from outer 'parallel' region.
Differential Revision: http://reviews.llvm.org/D8410
llvm-svn: 232932
Lang Hames [Mon, 23 Mar 2015 06:02:49 +0000 (06:02 +0000)]
[Orc] Add missing -use-orcmcjit flag to a number of Orc regression tests.
llvm-svn: 232931
Daniel Jasper [Mon, 23 Mar 2015 05:52:28 +0000 (05:52 +0000)]
Revert "Add CodeGen support for adding cpu attributes on functions based on"
This breaks CodeGen for an internal target. I'll get repro instructions
to you.
llvm-svn: 232930
Craig Topper [Mon, 23 Mar 2015 04:17:11 +0000 (04:17 +0000)]
Fix typo 'AVX too' instead of 'AVX2'
llvm-svn: 232929
Richard Smith [Mon, 23 Mar 2015 03:25:59 +0000 (03:25 +0000)]
[modules] Remove redundant import of lexical decls when building a lookup table
for a DeclContext, and fix propagation of exception specifications along
redeclaration chains.
This reverts r232905, r232907, and r232907, which reverted r232793, r232853,
and r232853.
One additional change is present here to resolve issues with LLDB: distinguish
between whether lexical decls missing from the lookup table are local or are
provided by the external AST source, and still look in the external source if
that's where they came from.
llvm-svn: 232928
Craig Topper [Mon, 23 Mar 2015 00:15:06 +0000 (00:15 +0000)]
[X86] Add one stepping of Broadwell to the CPU name autodetection for march=native.
llvm-svn: 232927
Davide Italiano [Sun, 22 Mar 2015 23:43:58 +0000 (23:43 +0000)]
Use POSIX pid_t and not lldb::pid_t.
The latter is uint64_t beacuse lldb supports arbitrary pid/platforms
but in this case we're using it as return value for fork() which might
return -1 to the parent in case the syscall fails.
Differential Revision: http://reviews.llvm.org/D8491
llvm-svn: 232926
Richard Smith [Sun, 22 Mar 2015 23:18:46 +0000 (23:18 +0000)]
Add missing #include to fix libc++ Linux build.
llvm-svn: 232925
Benjamin Kramer [Sun, 22 Mar 2015 21:57:53 +0000 (21:57 +0000)]
[ASTMatchers] Factor wrapping matcher classes into a common base class.
The deduplication here is negligible, but it allows the compiler to
skip emission of many templated base class destructors. Shrinks
clang-query by 53k. No functionality change intended.
llvm-svn: 232924
David Majnemer [Sun, 22 Mar 2015 21:27:10 +0000 (21:27 +0000)]
Silence a GCC warning
llvm-svn: 232923
Benjamin Kramer [Sun, 22 Mar 2015 18:22:33 +0000 (18:22 +0000)]
FoldingSet: Make FoldingSetImpl's dtor protected and non-virtual
It's not intended to be polymorphically deleted. Make FoldingSet
and ContextualFoldingSet final to avoid noise from -Wnon-virtual-dtor.
No functional change intended.
llvm-svn: 232922
Benjamin Kramer [Sun, 22 Mar 2015 18:16:22 +0000 (18:16 +0000)]
[Analyzer] Don't inherit from FoldingSet.
That's not really necessary here. NFCI.
llvm-svn: 232921
Kuba Brecka [Sun, 22 Mar 2015 18:00:58 +0000 (18:00 +0000)]
Fix a flaky heap-overflow-large.cc test
Reviewed at http://reviews.llvm.org/D8515
llvm-svn: 232920
Simon Atanasyan [Sun, 22 Mar 2015 16:35:08 +0000 (16:35 +0000)]
[Mips] Pass type of relocation and addendum to the `calculateRelocation`
separately
This change reduce difference between the trunk and upcoming patch and
simplify the future code review.
No functional changes.
llvm-svn: 232919
Benjamin Kramer [Sun, 22 Mar 2015 16:05:46 +0000 (16:05 +0000)]
Update unit test for r232916.
llvm-svn: 232918
Duncan P. N. Exon Smith [Sun, 22 Mar 2015 15:58:21 +0000 (15:58 +0000)]
Prevent CHECK-NOTs from matching file paths
A build directory with a name like `build-Werror` would hit a false
positive on these `CHECK-NOT`s before, since the actual error line looks
like:
.../build-Werror/bin/llvm-as <stdin>:1:2: error: ...
Switch to using:
CHECK-NOT: error:
(note the trailing semi-colon) to avoid matching almost any file path.
llvm-svn: 232917
Benjamin Kramer [Sun, 22 Mar 2015 15:56:12 +0000 (15:56 +0000)]
[multilib] Turn virtual functor into functin_ref
And update code to use lambdas where possible, plus random cleanup. NFCI.
llvm-svn: 232916
Simon Atanasyan [Sun, 22 Mar 2015 15:41:16 +0000 (15:41 +0000)]
[Mips] Rename template argument ELFType => ELFT
No functional changes.
llvm-svn: 232915
Simon Atanasyan [Sun, 22 Mar 2015 15:41:10 +0000 (15:41 +0000)]
[Mips] Use allocator from the base class
No functional changes.
llvm-svn: 232914
Simon Pilgrim [Sun, 22 Mar 2015 13:38:36 +0000 (13:38 +0000)]
Fixed MSVC compile warning issue introduced in r232837
- was reporting 'warning C4715: 'getType32' : not all control paths return a value'
llvm-svn: 232913
Faisal Vali [Sun, 22 Mar 2015 13:35:56 +0000 (13:35 +0000)]
Add a dump function to Stmt that takes only an output stream. No functionality change.
This allows dumping to any given output stream but without requiring a SourceManager, similar to the interface provided by Decl.
It's useful when writing certain generic debug functions, external to the clang code base (for e.g.).
llvm-svn: 232912
Tobias Grosser [Sun, 22 Mar 2015 12:06:39 +0000 (12:06 +0000)]
Use schedule trees to perform post-scheduling transformations
Replacing the old band_tree based code with code that is based on the new
schedule tree [1] interface makes applying complex schedule transformations a lot
more straightforward. We now do not need to reason about the meaning of flat
schedules, but can use a more straightforward tree structure. We do not yet
exploit this a lot in the current code, but hopefully we will be able to do so
soon.
This change also allows us to drop some code, as isl now provides some higher
level interfaces to apply loop transformations such as tiling.
This change causes some small test case changes as isl uses a slightly different
way to perform loop tiling, but no significant functional changes are intended.
[1] http://impact.gforge.inria.fr/impact2014/papers/impact2014-verdoolaege.pdf
llvm-svn: 232911
Kuba Brecka [Sun, 22 Mar 2015 11:38:55 +0000 (11:38 +0000)]
Demangling for DlAddrSymbolizer
On OS X, dladdr() provides mangled names only, so we need need to demangle in
DlAddrSymbolizer::SymbolizePC.
Reviewed at http://reviews.llvm.org/D8291
llvm-svn: 232910
Benjamin Kramer [Sun, 22 Mar 2015 10:08:31 +0000 (10:08 +0000)]
Remove empty files left behind by r232907.
llvm-svn: 232909
Kuba Brecka [Sun, 22 Mar 2015 10:02:50 +0000 (10:02 +0000)]
Add AtosSymbolizer and DlAddrSymbolizer as fallbacks for OS X
This patch changes the symbolizer chain on OS X (which currently only uses 1
symbolizer at most) to use this behavior:
* By default, use LLVMSymbolizer -> DlAddrSymbolizer.
* If the llvm-symbolizer binary is not found, use AtosSymbolizer
-> DlAddrSymbolizer.
* If the user specifies ASAN_SYMBOLIZER_PATH=.../atos, then use AtosSymbolizer
-> DlAddrSymbolizer.
* If neither llvm-symbolizer or atos is found, or external symbolication is
disabled with ASAN_SYMBOLIZER_PATH="", use DlAddrSymbolizer.
Reviewed at http://reviews.llvm.org/D8285
llvm-svn: 232908
Vince Harron [Sun, 22 Mar 2015 08:47:07 +0000 (08:47 +0000)]
Reverting 232853 and 232870 because they depend on 232793,
which was reverted because it was causing LLDB test failures
llvm-svn: 232907
David Majnemer [Sun, 22 Mar 2015 08:39:22 +0000 (08:39 +0000)]
MS ABI: Implement driver-level support for thread-safe statics
Decide whether or not to use thread-safe statics depending on whether or
not we have an explicit request from the driver. If we don't have an
explicit request, infer which behavior to use depending on the
compatibility version we are targeting.
N.B. CodeGen support is still ongoing.
llvm-svn: 232906
Vince Harron [Sun, 22 Mar 2015 05:59:59 +0000 (05:59 +0000)]
Reverting r232793 until some new LLDB failures are discussed
llvm-svn: 232905
Vince Harron [Sun, 22 Mar 2015 05:47:00 +0000 (05:47 +0000)]
Reverted r232883 due to failing tests.
llvm-svn: 232904
Benjamin Kramer [Sat, 21 Mar 2015 22:04:26 +0000 (22:04 +0000)]
[SimplifyLibCalls] Fix negative shifts being produced by the memchr -> bitfield transform.
llvm-svn: 232903
Benjamin Kramer [Sat, 21 Mar 2015 21:09:33 +0000 (21:09 +0000)]
[SimplifyLibCalls] Turn memchr(const, C, const) into a bitfield check.
strchr("123!", C) != nullptr is a common pattern to check if C is one
of 1, 2, 3 or !. If the largest element of the string is smaller than
the target's register size we can easily create a bitfield and just
do a simple test for set membership.
int foo(char C) { return strchr("123!", C) != nullptr; } now becomes
cmpl $64, %edi ## range check
sbbb %al, %al
movabsq $0xE000200000001, %rcx
btq %rdi, %rcx ## bit test
sbbb %cl, %cl
andb %al, %cl ## and the two conditions
andb $1, %cl
movzbl %cl, %eax ## returning an int
ret
(imho the backend should expand this into a series of branches, but
that's a different story)
The code is currently limited to bit fields that fit in a register, so
usually 64 or 32 bits. Sadly, this misses anything using alpha chars
or {}. This could be fixed by just emitting a i128 bit field, but that
can generate really ugly code so we have to find a better way. To some
degree this is also recreating switch lowering logic, but we can't
simply emit a switch instruction and thus change the CFG within
instcombine.
llvm-svn: 232902
Matt Arsenault [Sat, 21 Mar 2015 19:15:46 +0000 (19:15 +0000)]
R600: Cleanup test with multiple check prefixes
llvm-svn: 232901
Nico Weber [Sat, 21 Mar 2015 17:56:44 +0000 (17:56 +0000)]
-Wdynamic-class-memaccess: Also warn about array types.
It looks like not warning on this was an oversight in the original
implementation of this warning.
llvm-svn: 232900
Nico Weber [Sat, 21 Mar 2015 17:37:46 +0000 (17:37 +0000)]
Dedent code for -Wdynamic-class-memaccess warning. No behavior change.
The diff looks intimidating, but this just moves the -Wdynamic-class-memaccess
code out a scope, protected by a
if (PointeeTy == QualType())
continue;
check so that it still only runs when it should.
llvm-svn: 232899
Benjamin Kramer [Sat, 21 Mar 2015 16:42:35 +0000 (16:42 +0000)]
StringRef: Just forward StringRef::find to libc's memchr.
Modern libc's have an SSE version of memchr which is a lot faster than our
hand-rolled version. In the past I was reluctant to use it because Darwin's
memchr used a naive ridiculously slow implementation, but that has been fixed
some versions ago.
Should have zero functional impact.
llvm-svn: 232898
Benjamin Kramer [Sat, 21 Mar 2015 15:37:32 +0000 (15:37 +0000)]
Revert accidental commit.
While this is a fun change, I didn't really test it :)
llvm-svn: 232897