Chandler Carruth [Sun, 26 Mar 2017 14:24:06 +0000 (14:24 +0000)]
[IR] Switch to more normal template parameter names ending in `T`
instead of `Ty`.
The `Ty` suffix is much more commonly used for LLVM `Type` variable
names, so this seemed like a particularly confusing collision.
llvm-svn: 298808
Simon Pilgrim [Sun, 26 Mar 2017 13:59:17 +0000 (13:59 +0000)]
Fix signed/unsigned comparison warnings.
llvm-svn: 298807
Simon Pilgrim [Sun, 26 Mar 2017 13:17:55 +0000 (13:17 +0000)]
[X86][SSE] Add computeKnownBitsForTargetNode support for (V)PSLL/(V)PSRL instructions
llvm-svn: 298806
Simon Pilgrim [Sun, 26 Mar 2017 12:52:28 +0000 (12:52 +0000)]
[X86][AVX512F] Fix reg class for VMOVSSZrr/VMOVSSZrrk and VMOVSDZrr/VMOVSDZrrk
Fixed -verify-machineinstrs errors in fast-isel-select-sse.ll (one of many in PR27481)
The VMOVSSZrr/VMOVSSZrrk and VMOVSDZrr/VMOVSDZrrk instructions were assuming both source registers were V128X when the second is actually supposed to be FR32X/FR64X
Differential Revision: https://reviews.llvm.org/D31200
llvm-svn: 298805
Simon Pilgrim [Sun, 26 Mar 2017 11:17:05 +0000 (11:17 +0000)]
Fix MSVC signed/unsigned comparison warnings.
llvm-svn: 298804
Simon Pilgrim [Sun, 26 Mar 2017 10:33:03 +0000 (10:33 +0000)]
Regenerate test
llvm-svn: 298803
Simon Pilgrim [Sun, 26 Mar 2017 10:31:37 +0000 (10:31 +0000)]
Regenerate test
The CHECK-DAG aren't necessary and get in the way of automated checks
llvm-svn: 298802
Simon Pilgrim [Sun, 26 Mar 2017 10:28:39 +0000 (10:28 +0000)]
Regenerate tests to remove duplicated checks
llvm-svn: 298801
Igor Breger [Sun, 26 Mar 2017 08:11:12 +0000 (08:11 +0000)]
[GlobalISel][X86] support G_FRAME_INDEX instruction selection.
Summary:
Support G_FRAME_INDEX instruction selection.
Reviewers: zvi, rovka, ab, qcolombet
Reviewed By: ab
Subscribers: llvm-commits, dberris, kristof.beyls, eladcohen, guyblank
Differential Revision: https://reviews.llvm.org/D30980
llvm-svn: 298800
Joerg Sonnenberger [Sun, 26 Mar 2017 06:44:08 +0000 (06:44 +0000)]
Split the SimplifyCFG pass into two variants.
The first variant contains all current transformations except
transforming switches into lookup tables. The second variant
contains all current transformations.
The switch-to-lookup-table conversion results in code that is more
difficult to analyze and optimize by other passes. Most importantly,
it can inhibit Dead Code Elimination. As such it is often beneficial to
only apply this transformation very late. A common example is inlining,
which can often result in range restrictions for the switch expression.
Changes in execution time according to LNT:
SingleSource/Benchmarks/Misc/fp-convert +3.03%
MultiSource/Benchmarks/ASC_Sequoia/CrystalMk/CrystalMk -11.20%
MultiSource/Benchmarks/Olden/perimeter/perimeter -10.43%
and a couple of smaller changes. For perimeter it also results 2.6%
a smaller binary.
Differential Revision: https://reviews.llvm.org/D30333
llvm-svn: 298799
Andrew Wilkins [Sun, 26 Mar 2017 05:58:48 +0000 (05:58 +0000)]
Add check for BSD when setting LIB_NAMES for GNU ld
Patch by Koop Mast and Alex Arslan!
Differential Revision: https://reviews.llvm.org/D28215
llvm-svn: 298798
Rui Ueyama [Sun, 26 Mar 2017 04:10:43 +0000 (04:10 +0000)]
Sort. NFC.
llvm-svn: 298797
Rui Ueyama [Sun, 26 Mar 2017 03:42:00 +0000 (03:42 +0000)]
Remove a redundant local variable.
llvm-svn: 298796
Rui Ueyama [Sun, 26 Mar 2017 03:41:41 +0000 (03:41 +0000)]
Add a comment.
llvm-svn: 298795
Rui Ueyama [Sun, 26 Mar 2017 03:21:08 +0000 (03:21 +0000)]
Inline a function that is used only once.
llvm-svn: 298794
Rui Ueyama [Sun, 26 Mar 2017 03:20:49 +0000 (03:20 +0000)]
De-template a few functions in which ELFT is not needed.
llvm-svn: 298793
Rui Ueyama [Sun, 26 Mar 2017 03:20:30 +0000 (03:20 +0000)]
Add comments and return early.
llvm-svn: 298792
Chandler Carruth [Sun, 26 Mar 2017 02:49:23 +0000 (02:49 +0000)]
[IR] Make SwitchInst::CaseIt almost a normal iterator.
This moves it to the iterator facade utilities giving it full random
access semantics, etc. It can also now be used with standard algorithms
like std::all_of and std::any_of and range adaptors like llvm::reverse.
Also make the semantics of iterating match what every other iterator
uses and forbid decrementing past the begin iterator. This was used as
a hacky way to work around iterator invalidation. However, every
instance trying to do this failed to actually avoid touching invalid
iterators despite the clear documentation that the removed and all
subsequent iterators become invalid including the end iterator. So I've
added a return of the next iterator to removeCase and rewritten the
loops that were doing this to correctly follow the iterator pattern of
either incremneting or removing and assigning fresh values to the
iterator and the end.
In one case we were trying to go backwards to make this cleaner but it
doesn't actually work. I've made that code match the code we use
everywhere else to remove cases as we iterate. This changes the order of
cases in one test output and I moved that test to CHECK-DAG so it
wouldn't care -- the order isn't semantically meaningful anyways.
llvm-svn: 298791
Rui Ueyama [Sun, 26 Mar 2017 02:28:09 +0000 (02:28 +0000)]
Remove unused parameter.
llvm-svn: 298790
Rui Ueyama [Sun, 26 Mar 2017 02:27:50 +0000 (02:27 +0000)]
Remove redundant local variables.
llvm-svn: 298789
Rui Ueyama [Sun, 26 Mar 2017 02:27:30 +0000 (02:27 +0000)]
Remove a parameter from adjustExpr. NFC.
llvm-svn: 298788
Rui Ueyama [Sun, 26 Mar 2017 02:27:11 +0000 (02:27 +0000)]
Inline a small lambda.
llvm-svn: 298787
Rui Ueyama [Sun, 26 Mar 2017 02:26:52 +0000 (02:26 +0000)]
Rename C -> Sec and RI -> Rel.
C is short for Chunk, but we are no longer using that term.
RI is probably short for relocation iterator, but this is not an interator.
llvm-svn: 298786
Rui Ueyama [Sun, 26 Mar 2017 02:26:33 +0000 (02:26 +0000)]
Simplify relocation offset adjustment.
Previously, relocation offsets are recalculated for .eh_frame sections
inside the main loop, and that messed up the main loop. This patch
separates that logic into a dedicated class.
llvm-svn: 298785
Gor Nishanov [Sun, 26 Mar 2017 02:18:05 +0000 (02:18 +0000)]
[coroutines] Add codegen for await and yield expressions
Details:
Emit suspend expression which roughly looks like:
auto && x = CommonExpr();
if (!x.await_ready()) {
llvm_coro_save();
x.await_suspend(...); (*)
llvm_coro_suspend(); (**)
}
x.await_resume();
where the result of the entire expression is the result of x.await_resume()
(*) If x.await_suspend return type is bool, it allows to veto a suspend:
if (x.await_suspend(...))
llvm_coro_suspend();
(**) llvm_coro_suspend() encodes three possible continuations as a switch instruction:
%where-to = call i8 @llvm.coro.suspend(...)
switch i8 %where-to, label %coro.ret [ ; jump to epilogue to suspend
i8 0, label %yield.ready ; go here when resumed
i8 1, label %yield.cleanup ; go here when destroyed
]
llvm-svn: 298784
Simon Pilgrim [Sat, 25 Mar 2017 21:22:12 +0000 (21:22 +0000)]
[X86] Pull out repeated ScalarValueSizeInBits code. NFCI.
llvm-svn: 298783
Simon Pilgrim [Sat, 25 Mar 2017 20:43:01 +0000 (20:43 +0000)]
[X86][SSE] Combine (VSRLI (VSRAI X, Y), (NumSignBits-1)) -> (VSRLI X, (NumSignBits-1))
Part 3 of 3.
Differential Revision: https://reviews.llvm.org/D31347
llvm-svn: 298782
Eric Christopher [Sat, 25 Mar 2017 20:20:23 +0000 (20:20 +0000)]
Change the default attributes for llvm.prefetch to inaccessiblemem_or_argmemonly
so that we can perform some optimizations across it.
Fixes PR32365
llvm-svn: 298781
Simon Pilgrim [Sat, 25 Mar 2017 19:58:36 +0000 (19:58 +0000)]
[X86][SSE] Added ComputeNumSignBitsForTargetNode support for (V)PSRAI
Part 2 of 3.
Differential Revision: https://reviews.llvm.org/D31347
llvm-svn: 298780
Simon Pilgrim [Sat, 25 Mar 2017 19:50:14 +0000 (19:50 +0000)]
[X86][SSE] Generalised CMP+AND1 combine to ZERO/ALLBITS+MASK
Patch to generalize combinePCMPAnd1 (for handling SETCC + ZEXT cases) to work for any input that has zero/all bits set masked with an 'all low bits' mask.
Replaced the implicit assumption of shift availability with a call to SupportedVectorShiftWithImm.
Part 1 of 3.
Differential Revision: https://reviews.llvm.org/D31347
llvm-svn: 298779
Eric Christopher [Sat, 25 Mar 2017 19:26:04 +0000 (19:26 +0000)]
_CALL_LINUX is only defined on 64-bit ppc linux platforms, not 32-bit.
Adjust and add a test for the 32-bit side.
llvm-svn: 298778
Michal Gorny [Sat, 25 Mar 2017 18:51:37 +0000 (18:51 +0000)]
PluginUnwindAssemblyX86: add missing linkage to MCDisasm
Add missing linkage of the lldbPluginUnwindAssemblyX86 to LLVMMCDisasm
library. This fixes the following build failure when linking against
shared libraries:
lib64/liblldbPluginUnwindAssemblyX86.a(x86AssemblyInspectionEngine.cpp.o):x86AssemblyInspectionEngine.cpp:function lldb_private::x86AssemblyInspectionEngine::instruction_length(unsigned char*, int&): error: undefined reference to 'LLVMDisasmInstruction'
lib64/liblldbPluginUnwindAssemblyX86.a(x86AssemblyInspectionEngine.cpp.o):x86AssemblyInspectionEngine.cpp:function lldb_private::x86AssemblyInspectionEngine::~x86AssemblyInspectionEngine(): error: undefined reference to 'LLVMDisasmDispose'
lib64/liblldbPluginUnwindAssemblyX86.a(x86AssemblyInspectionEngine.cpp.o):x86AssemblyInspectionEngine.cpp:function lldb_private::x86AssemblyInspectionEngine::x86AssemblyInspectionEngine(lldb_private::ArchSpec const&): error: undefined reference to 'LLVMCreateDisasm'
Differential Revision: https://reviews.llvm.org/D31369
llvm-svn: 298777
Michal Gorny [Sat, 25 Mar 2017 18:51:29 +0000 (18:51 +0000)]
Expression: add missing linkage to RuntimeDyld component
Add missing linkage from lldbExpression library to LLVMRuntimeDyld.
Otherwise the build against shared LLVM libraries fails with:
lib64/liblldbExpression.a(IRExecutionUnit.cpp.o):IRExecutionUnit.cpp:function llvm::RTDyldMemoryManager::deregisterEHFrames(unsigned char*, unsigned long, unsigned long): error: undefined reference to 'llvm::RTDyldMemoryManager::deregisterEHFramesInProcess(unsigned char*, unsigned long)'
Differential Revision: https://reviews.llvm.org/D31367
llvm-svn: 298776
Sanjay Patel [Sat, 25 Mar 2017 16:05:33 +0000 (16:05 +0000)]
[x86] use PMOVMSK to replace memcmp libcalls for 16-byte equality
This is the payoff for D31156 - if a target has efficient comparison instructions for vector-sized equality,
we can replace memcmp calls with inline code that is both smaller and faster.
Differential Revision: https://reviews.llvm.org/D31290
llvm-svn: 298775
Simon Pilgrim [Sat, 25 Mar 2017 15:43:36 +0000 (15:43 +0000)]
[X86][SSE] Add extra computeNumSignBits test case for D31311.
llvm-svn: 298774
Yaxun Liu [Sat, 25 Mar 2017 11:34:41 +0000 (11:34 +0000)]
[AMDGPU] Make AMDGPUTargetInfo::AS private
llvm-svn: 298773
Craig Topper [Sat, 25 Mar 2017 06:52:52 +0000 (06:52 +0000)]
[InstCombine] Change the interface of SimplifyDemandedBits so that it takes the instruction and operand instead of the Use.
The first thing it did was get the User for the Use to get the instruction back. This requires looking through the Uses for the User using the waymarking walk. That's pretty fast, but its probably still better to just pass the Instruction we already had.
llvm-svn: 298772
Eric Christopher [Sat, 25 Mar 2017 06:38:57 +0000 (06:38 +0000)]
Update the comment on not yet generated preprocessor defines to remove __LONGDOUBLE128.
llvm-svn: 298771
Eric Christopher [Sat, 25 Mar 2017 06:37:23 +0000 (06:37 +0000)]
Add the __LONGDOUBLE128 define for ppc targets that have 128 bit long doubles.
llvm-svn: 298770
Eric Christopher [Sat, 25 Mar 2017 05:40:13 +0000 (05:40 +0000)]
Define __HAVE_BSWAP__ on ppc to match gcc since we support both builtins as well.
llvm-svn: 298769
Balaram Makam [Sat, 25 Mar 2017 04:02:39 +0000 (04:02 +0000)]
[AArch64] Refine Falkor Machine Model - Part1
llvm-svn: 298768
Yaxun Liu [Sat, 25 Mar 2017 03:46:25 +0000 (03:46 +0000)]
[AMDGPU] Switch address space mapping by triple environment amdgiz
For target environment amdgiz and amdgizcl (giz means Generic Is Zero), AMDGPU will use new address space mapping where generic address space is 0 and private address space is 5. The data layout is also changed correspondingly.
Differential Revision: https://reviews.llvm.org/D31210
llvm-svn: 298767
Shoaib Meenai [Sat, 25 Mar 2017 03:42:20 +0000 (03:42 +0000)]
[libc++] Fix some comment typos
Remove a stray letter, add a missing letter. No functional change.
llvm-svn: 298766
Eric Christopher [Sat, 25 Mar 2017 03:33:59 +0000 (03:33 +0000)]
Add the _CALL_LINUX preprocessor define for ppc linux platforms.
This typically is only for a new enough linker (bfd >= 2.16.2 or gold), but
our ppc suppport post-dates this and it should work on all linux platforms. It
is guaranteed to work on all elfv2 platforms.
llvm-svn: 298765
Shoaib Meenai [Sat, 25 Mar 2017 03:29:51 +0000 (03:29 +0000)]
[libc++] Fix word transposition in comment
"to due" -> "due to". No functional change.
llvm-svn: 298764
Shoaib Meenai [Sat, 25 Mar 2017 03:22:35 +0000 (03:22 +0000)]
[libc++] Fix capitalization in comment
Fix a stray capital letter in the middle of a sentence. No functional
change.
llvm-svn: 298763
Shoaib Meenai [Sat, 25 Mar 2017 03:12:37 +0000 (03:12 +0000)]
[libc++] Update package version
Make it consistent with the rest of LLVM.
llvm-svn: 298762
Eric Christopher [Sat, 25 Mar 2017 02:55:21 +0000 (02:55 +0000)]
__BIGGEST_ALIGNMENT__ has always been 16 on all power platforms rather
than the default of 8 in clang, fix and update tests accordingly.
llvm-svn: 298761
Davide Italiano [Sat, 25 Mar 2017 02:40:02 +0000 (02:40 +0000)]
[NewGVN] Adjust NDEBUG markers.
This avoids 'used but not defined' warnings in Release builds
with GCC.
llvm-svn: 298760
Eric Christopher [Sat, 25 Mar 2017 02:29:18 +0000 (02:29 +0000)]
Add preprocessor defines for a bare powerpc64le triple/cpu.
The le triple didn't exist until power8, so use that as a default (this
also matches what gcc does).
llvm-svn: 298759
Yaxun Liu [Sat, 25 Mar 2017 02:05:44 +0000 (02:05 +0000)]
[AMDGPU] Switch data layout by triple environment amdgiz
Switch data layout by target triple environment amdgiz and amdgizcl indicating using of an address space mapping in which generic address space is 0.
amdgiz is for non-OpenCL environment where generic address space is 0.
amdgizcl is for OpenCL environment where generic address space is 0.
Differential Revision: https://reviews.llvm.org/D31211
llvm-svn: 298758
Vitaly Buka [Sat, 25 Mar 2017 01:19:45 +0000 (01:19 +0000)]
[libFuzzer] Fix test on Windows.
llvm-svn: 298757
Evgeniy Stepanov [Sat, 25 Mar 2017 01:01:11 +0000 (01:01 +0000)]
[asan] Put ctor/dtor in comdat.
When possible, put ASan ctor/dtor in comdat.
The only reason not to is global registration, which can be
TU-specific. This is not the case when there are no instrumented
globals. This is also limited to ELF targets, because MachO does
not have comdat, and COFF linkers may GC comdat constructors.
The benefit of this is a lot less __asan_init() calls: one per DSO
instead of one per TU. It's also necessary for the upcoming
gc-sections-for-globals change on Linux, where multiple references to
section start symbols trigger quadratic behaviour in gold linker.
llvm-svn: 298756
Kostya Serebryany [Sat, 25 Mar 2017 00:56:08 +0000 (00:56 +0000)]
[libFuzzer] read asan's dedup_token while minimizing a crash and stop minimization if another bug was found during minimization (https://github.com/google/oss-fuzz/issues/452)
llvm-svn: 298755
Richard Trieu [Sat, 25 Mar 2017 00:48:52 +0000 (00:48 +0000)]
[ODRHash] Add support for array and decayed types.
llvm-svn: 298754
Evgeniy Stepanov [Sat, 25 Mar 2017 00:42:25 +0000 (00:42 +0000)]
Fix sanitizer tests with LLVM_TOOL_LLD_BUILD=OFF.
Only depend on LLD if it is going to be built.
Re-land of r298174 which got reverted in r298287.
llvm-svn: 298753
Eli Friedman [Sat, 25 Mar 2017 00:13:24 +0000 (00:13 +0000)]
[ARM] Fix mixup between Lo and Hi in SMLALBB formation.
llvm-svn: 298752
Simon Pilgrim [Fri, 24 Mar 2017 23:47:41 +0000 (23:47 +0000)]
Apply clang-format as commented in D31311. NFCI.
llvm-svn: 298751
Reid Kleckner [Fri, 24 Mar 2017 23:28:42 +0000 (23:28 +0000)]
[codeview] Don't assert when the user violates the ODR
If we have an array of a user-defined aggregates for which there was an
ODR violation, then the array size will not necessarily match the number
of elements times the size of the element.
Fixes PR32383
llvm-svn: 298750
Rui Ueyama [Fri, 24 Mar 2017 23:21:34 +0000 (23:21 +0000)]
Use raw_fd_ostream::has_colors instead of StandardErrHasColors().
I just didn't know that raw_fd_ostream has has_colors() function.
llvm-svn: 298749
Rui Ueyama [Fri, 24 Mar 2017 23:21:14 +0000 (23:21 +0000)]
Update comment.
llvm-svn: 298748
Jessica Paquette [Fri, 24 Mar 2017 23:00:21 +0000 (23:00 +0000)]
[Outliner] Revert r298734.
When I tested r298734, I thought that red zones were enabled by default like in
X86. Since red zones are behind a flag on AArch64 the testing wasn't true.
llvm-svn: 298747
Kostya Serebryany [Fri, 24 Mar 2017 22:19:52 +0000 (22:19 +0000)]
[libFuzzer] be more careful when calling strlen of strcmp parameters, PR32357
llvm-svn: 298746
Craig Topper [Fri, 24 Mar 2017 22:12:10 +0000 (22:12 +0000)]
Revert r298711 "[InstCombine] Provide a way to calculate KnownZero/One for Add/Sub in SimplifyDemandedUseBits without recursing into ComputeKnownBits"
Tsan bot is failing.
llvm-svn: 298745
Sanjay Patel [Fri, 24 Mar 2017 22:09:48 +0000 (22:09 +0000)]
[x86] add 32-bit RUN for better memcmp coverage; NFC
llvm-svn: 298744
Alex Shlyapnikov [Fri, 24 Mar 2017 21:43:56 +0000 (21:43 +0000)]
Fix flaky strtok.c test.
Asserting the result of strtok when we expect delimiter overflow is
flaky, the result depends on the random state of memory right after the
delimiters.
llvm-svn: 298743
Richard Trieu [Fri, 24 Mar 2017 21:17:48 +0000 (21:17 +0000)]
[ODRHash] Add error messages for mismatched parameters in methods.
llvm-svn: 298742
Kuba Mracek [Fri, 24 Mar 2017 21:12:24 +0000 (21:12 +0000)]
Follow-up for r298738: Use "0" instead of "false" because the variable is uptr.
llvm-svn: 298741
Kostya Serebryany [Fri, 24 Mar 2017 21:09:16 +0000 (21:09 +0000)]
[libFuzzer] honor -exact_artifact_path for all intermediate files during crash minimization (https://github.com/google/oss-fuzz/issues/250)
llvm-svn: 298740
Krzysztof Parzyszek [Fri, 24 Mar 2017 21:01:16 +0000 (21:01 +0000)]
Move spill size and alignment info from MC to TargetRegisterInfo
This is another step towards implementing register classes with
parametrized register/spill sizes and value types.
This is an updated version of r298652. The difference is that MCRegister-
Class still contains register size, available as getPhysRegSize(). The
old function getSize was retained as a temporary measure to avoid build
breakage for out-of-tree targets.
llvm-svn: 298739
Kuba Mracek [Fri, 24 Mar 2017 20:57:33 +0000 (20:57 +0000)]
Fix an uninitialized field in tsan_block_context_t/AllocContext in tsan_libdispatch_mac.cc.
llvm-svn: 298738
Matt Arsenault [Fri, 24 Mar 2017 20:57:10 +0000 (20:57 +0000)]
AMDGPU: Fix annotating loops with nested loop conditions
If the branch condition for a loop was a phi which itself
was fed from a phi from a loop, it isn't safe to try
to delete the phi until after the loop is handled.
llvm-svn: 298737
Davide Italiano [Fri, 24 Mar 2017 20:52:56 +0000 (20:52 +0000)]
[MachineScheduler] Add missing machine pass dependency.
llvm-svn: 298736
Ivan Krasin [Fri, 24 Mar 2017 20:49:43 +0000 (20:49 +0000)]
Revert r298620: [LV] Vectorize GEPs
Reason: breaks linking Chromium with LLD + ThinLTO (a pass crashes)
LLVM bug: https://bugs.llvm.org//show_bug.cgi?id=32413
Original change description:
[LV] Vectorize GEPs
This patch adds support for vectorizing GEPs. Previously, we only generated
vector GEPs on-demand when creating gather or scatter operations. All GEPs from
the original loop were scalarized by default, and if a pointer was to be stored
to memory, we would have to build up the pointer vector with insertelement
instructions.
With this patch, we will vectorize all GEPs that haven't already been marked
for scalarization.
The patch refines collectLoopScalars to more exactly identify the scalar GEPs.
The function now more closely resembles collectLoopUniforms. And the patch
moves vector GEP creation out of vectorizeMemoryInstruction and into the main
vectorization loop. The vector GEPs needed for gather and scatter operations
will have already been generated before vectoring the memory accesses.
Original Differential Revision: https://reviews.llvm.org/D30710
llvm-svn: 298735
Jessica Paquette [Fri, 24 Mar 2017 20:47:59 +0000 (20:47 +0000)]
[Outliner] Remove no red zone requirment for AArch64
AArch64 doesn't require -mno-red-zone; stack fixups are sufficient here. This was
unnecessarily copied over from the X86 target.
(You can now outline with red zones! Yay!)
Removing the requirement passes all Single/MultiSource tests.
llvm-svn: 298734
Reid Kleckner [Fri, 24 Mar 2017 20:47:41 +0000 (20:47 +0000)]
Document how to fetch monorepo SVN revision notes
llvm-svn: 298733
Evgeniy Stepanov [Fri, 24 Mar 2017 20:42:15 +0000 (20:42 +0000)]
[asan] Delay creation of asan ctor.
Create the constructor in the module pass.
This in needed for the GC-friendly globals change, where the constructor can be
put in a comdat in some cases, but we don't know about that in the function
pass.
llvm-svn: 298731
Matt Arsenault [Fri, 24 Mar 2017 20:04:18 +0000 (20:04 +0000)]
AMDGPU: Implement f16 fround
llvm-svn: 298730
Matt Arsenault [Fri, 24 Mar 2017 19:52:05 +0000 (19:52 +0000)]
AMDGPU: Unify divergent function exits.
StructurizeCFG can't handle cases with multiple
returns creating regions with multiple exits.
Create a copy of UnifyFunctionExitNodes that only
unifies exit nodes that skips exit nodes
with uniform branch sources.
llvm-svn: 298729
Peter Collingbourne [Fri, 24 Mar 2017 19:32:20 +0000 (19:32 +0000)]
Fix a test so that it actually checks the output.
llvm-svn: 298728
Krzysztof Parzyszek [Fri, 24 Mar 2017 19:18:29 +0000 (19:18 +0000)]
Revert r298652 on Quentin's request
llvm-svn: 298727
Adrian Prantl [Fri, 24 Mar 2017 19:11:31 +0000 (19:11 +0000)]
Make testcase less nonsensical while still exercising the same code paths.
llvm-svn: 298726
Matt Arsenault [Fri, 24 Mar 2017 19:04:57 +0000 (19:04 +0000)]
AMDGPU: Fold rcp/rsq of undef to undef
llvm-svn: 298725
Matt Arsenault [Fri, 24 Mar 2017 18:56:43 +0000 (18:56 +0000)]
TTI: Split IsSimple in MemIntrinsicInfo
All this did before was assert in EarlyCSE.
llvm-svn: 298724
Stanislav Mekhanoshin [Fri, 24 Mar 2017 18:55:20 +0000 (18:55 +0000)]
[AMDGPU] Fold V_CNDMASK with identical source operands
Such instructions sometimes appear after lowering and folding.
Differential Revision: https://reviews.llvm.org/D31318
llvm-svn: 298723
Konstantin Zhuravlyov [Fri, 24 Mar 2017 18:43:15 +0000 (18:43 +0000)]
[AMDGPU] Rename Kind to ValueKind in metadata to be consistent
llvm-svn: 298722
Stanislav Mekhanoshin [Fri, 24 Mar 2017 18:01:14 +0000 (18:01 +0000)]
[AMDGPU] Add AMDGPUAliasAnalysis to opt pipeline
Previously it was added only to the BE.
Differential Revision: https://reviews.llvm.org/D31323
llvm-svn: 298721
Teresa Johnson [Fri, 24 Mar 2017 17:59:06 +0000 (17:59 +0000)]
[ThinLTO] Correct counting of functions in inliner stats
Summary: Declarations need to be filtered out when counting functions.
Reviewers: eraman
Subscribers: Prazek, llvm-commits
Differential Revision: https://reviews.llvm.org/D31336
llvm-svn: 298720
Benjamin Kramer [Fri, 24 Mar 2017 17:53:06 +0000 (17:53 +0000)]
[AMDGPU] Don't enforce constexpr, there are still old standard libraries around that don't have a constexpr std::pair.
llvm-svn: 298719
Valery Pykhtin [Fri, 24 Mar 2017 17:49:05 +0000 (17:49 +0000)]
[AMDGPU] Remove double map lookups in SI scheduler
Patch by Axel Davy (axel.davy@normalesup.org)
Differential revision: https://reviews.llvm.org/D30382
llvm-svn: 298718
Reid Kleckner [Fri, 24 Mar 2017 17:26:38 +0000 (17:26 +0000)]
[PDB] Split item and type records when merging type streams
Summary: MSVC does this when producing a PDB.
Reviewers: ruiu
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D31316
llvm-svn: 298717
Simon Pilgrim [Fri, 24 Mar 2017 17:25:47 +0000 (17:25 +0000)]
[X86][SSE] Add ashr + mask test cases.
Test cases showing cases where we're missing an opportunity to lshr a value with an extended sign to avoid loading a mask
llvm-svn: 298716
Weiming Zhao [Fri, 24 Mar 2017 17:08:35 +0000 (17:08 +0000)]
Revert "builtins: Select correct code fragments when compiling for Thumb1/Thum2/ARM ISA."
This reverts commit
c3709191b6d36c4c936173f4a9a29a734b12cb15.
(commit by mistake)
llvm-svn: 298715
Weiming Zhao [Fri, 24 Mar 2017 17:06:05 +0000 (17:06 +0000)]
[Compiler-rt][Builtins] Implement lit-test support (part 2 of 2)
Summary:
Original r297566 (https://reviews.llvm.org/D30802) is splitted into two parts.
This part adds CMakefile/lit.cfg support.
Reviewers: rengolin, compnerd, jroelofs, erik.pilkington
Subscribers: srhines, dberris, mgorny
Differential Revision: https://reviews.llvm.org/D31259
llvm-svn: 298714
Weiming Zhao [Fri, 24 Mar 2017 17:06:00 +0000 (17:06 +0000)]
builtins: Select correct code fragments when compiling for Thumb1/Thum2/ARM ISA.
Summary:
Value of __ARM_ARCH_ISA_THUMB isn't based on the actual compilation
mode (-mthumb, -marm), it reflect's capability of given CPU.
Due to this:
- use __tbumb__ and __thumb2__ insteand of __ARM_ARCH_ISA_THUMB
- use '.thumb' directive consistently in all affected files
- decorate all thumb functions using
DEFINE_COMPILERRT_THUMB_FUNCTION()
---------
Note: This patch doesn't fix broken Thumb1 variant of __udivsi3 !
Reviewers: weimingz, rengolin, compnerd
Subscribers: aemerson, dim
Differential Revision: https://reviews.llvm.org/D30938
llvm-svn: 298713
Simon Pilgrim [Fri, 24 Mar 2017 16:59:14 +0000 (16:59 +0000)]
Fix MSVC 'not all control paths return a value' warning.
llvm-svn: 298712
Craig Topper [Fri, 24 Mar 2017 16:56:51 +0000 (16:56 +0000)]
[InstCombine] Provide a way to calculate KnownZero/One for Add/Sub in SimplifyDemandedUseBits without recursing into ComputeKnownBits
SimplifyDemandedUseBits for Add/Sub already recursed down LHS and RHS for simplifying bits. If that didn't provide any simplifications we fall back to calling computeKnownBits which will recurse again. Instead just take the known bits for LHS and RHS we already have and call into a new function in ValueTracking that can calculate the known bits given the LHS/RHS bits.
llvm-svn: 298711
Valery Pykhtin [Fri, 24 Mar 2017 16:45:50 +0000 (16:45 +0000)]
[AMDGPU] Fix SGPR usage count in SI scheduler
Patch by Axel Davy (axel.davy@normalesup.org)
Differential revision: https://reviews.llvm.org/D30149
llvm-svn: 298710
Anastasia Stulova [Fri, 24 Mar 2017 16:43:51 +0000 (16:43 +0000)]
[OpenCL][Bug 10573] Don't set CXXOperatorNames flag
The flag CXXOperatorNames was overwritten unconditionally
after being set for OpenCL.
There seems to be no necessity to set it, so removing the line.
llvm-svn: 298709
Valery Pykhtin [Fri, 24 Mar 2017 16:37:48 +0000 (16:37 +0000)]
[AMDGPU] Add a new line after a debug message
Patch by Axel Davy (axel.davy@normalesup.org)
Differential revision: https://reviews.llvm.org/D30146
llvm-svn: 298708