Commit Graph

245562 Commits

Author SHA1 Message Date
Simon Pilgrim 6ac1e98b09 [X86][SSE] Add SSE41/AVX1 costs for vector shifts.
We were defaulting to SSE2 costs which weren't taking into account the availability of PBLENDW/PBLENDVB to improve merging of per-element shift results.

llvm-svn: 284939
2016-10-23 16:49:04 +00:00
Simon Pilgrim e16b1e2271 [CostModel][X86] Added tests for current integer trunc costs
llvm-svn: 284938
2016-10-23 15:17:52 +00:00
Simon Pilgrim 96ef0c1103 Use APInt::isAllOnesValue instead of popcnt. NFCI.
More obvious implementation and faster too.

llvm-svn: 284937
2016-10-23 15:09:44 +00:00
Craig Topper 0c5da26572 [AVX-512] Replace 512-bit pmovzx/sx builtins with native IR.
llvm-svn: 284936
2016-10-23 07:35:47 +00:00
Craig Topper 4ef879ac2c [AVX-512] Remove masked 128/256-bit packss/packus builtins and replace with selects and the older unmasked builtins.
llvm-svn: 284935
2016-10-23 07:35:39 +00:00
Ekaterina Romanova 06477bf035 Add more doxygen comments to emmintrin.h's intrinsics.
With this patch, all intrinsics in this file (with an exception of a handful of a recently added ones) will be documented. I will send out a patch for 4 missining intrisics later.

The doxygen comments are automatically generated based on Sony's intrinsics document.

I got an OK from Eric Christopher to commit doxygen comments without prior code
review upstream. This patch was internally reviewed by Yunzhong Gao.

llvm-svn: 284934
2016-10-23 07:30:50 +00:00
Richard Smith c7f576fc91 Fix mangling numbers for varargs lambdas; varargs and non-varargs lambdas get
different lambda-sigs, so they should have different counters.

llvm-svn: 284933
2016-10-23 04:53:03 +00:00
Mandeep Singh Grang 249237cb2e [compiler-rt] Remove redundant --check-prefix=CHECK from test
Reviewers: eugenis, rengolin

Subscribers: dberris

Differential Revision: https://reviews.llvm.org/D25892

llvm-svn: 284932
2016-10-23 00:55:12 +00:00
Mandeep Singh Grang 66215e97e3 [clang] Limit clang test to ARM and AArch64 only
Summary: Limit clang/test/Frontend/gnu-mcount.c to ARM and AArch64 only.

Reviewers: abdulras, honggyu.kim, rengolin

Subscribers: aemerson, rengolin, cfe-commits

Tags: #clang-c

Differential Revision: https://reviews.llvm.org/D25842

llvm-svn: 284931
2016-10-23 00:53:03 +00:00
Dylan McKay 479a13c0aa [AVR] Add the machine code disassembler
This adds a super basic implementation of a machine code disassembler.

It doesn't support any operands with custom encoding.

llvm-svn: 284930
2016-10-22 23:57:59 +00:00
Craig Topper 4d63dfc286 [AVX-512] Replace masked 128/256-bit pavg builtins and replace with select and older unmasked builtins.
llvm-svn: 284929
2016-10-22 21:24:56 +00:00
Craig Topper 622c63614d [AVX-512] Replace masked 128/256-bit saturating add/sub builtins with select and older unmasked builtins.
llvm-svn: 284928
2016-10-22 21:24:52 +00:00
Craig Topper 11dda92405 [AVX-512] Replace masked 128/256-bit vpmovzx/vpmovsx builtins with native IR.
llvm-svn: 284927
2016-10-22 21:24:48 +00:00
Craig Topper f742445eb4 [AVX-512] Remove duplicate test cases from the avx512vlbw intrinsic test. These tests already exist in the avx512vl test and represent avx512vl instructions.
llvm-svn: 284926
2016-10-22 21:24:44 +00:00
Craig Topper eb1c0afa90 [AVX-512] Remove masked 128/256-bit pshufb builtins. Replace with a select and the older unmaksed builtins.
llvm-svn: 284925
2016-10-22 21:24:42 +00:00
Craig Topper 78a9c40326 [AVX-512] Remove builtins for 128/256-bit pabsb/pabsw. We can use a select and the older non-masked versions instead.
llvm-svn: 284924
2016-10-22 21:24:38 +00:00
Craig Topper c2c7e42bfe [AVX-512] Add typecasts to alignr intrinsics that were modified in r284920.
llvm-svn: 284923
2016-10-22 21:24:34 +00:00
Simon Pilgrim d3829c89bc [X86][AVX512VL] Added support for combining target 256-bit shuffles to AVX512VL VPERMV3
llvm-svn: 284922
2016-10-22 20:15:39 +00:00
Simon Pilgrim 56c0524f0f [X86][AVX512] Added support for combining target shuffles to AVX512 VPERMV3
llvm-svn: 284921
2016-10-22 19:53:59 +00:00
Craig Topper f6373bc6fd [AVX-512] Remove masked 128/256-bit palignr builtins. We can just use a select in the header file with the older unmasked versions instead.
llvm-svn: 284920
2016-10-22 18:32:33 +00:00
Brian Gesiak ae8c2517b4 [lit] Add more testing instructions to README
Summary:
r283710 introduced two regressions, one to llvm-lit, and the other to
lit executables that were installed via setuptools. Add instructions on
how to test for these regressions in the future.

Reviewers: ddunbar, delcypher, beanz, chapuni, cmatthews, echristo

Subscribers: llvm-commits, mehdi_amini

Differential Revision: https://reviews.llvm.org/D25459

llvm-svn: 284919
2016-10-22 17:27:31 +00:00
James Molloy 2bae8640d7 [ARM] Fix crash in ConstantIslands
tPCRelJT may not be the first instruction in a block. Check that instead of dereferencing a broken iterator.

llvm-svn: 284917
2016-10-22 09:58:37 +00:00
Zvi Rackover 8eb99d88d4 [X86] Apply the Update LLC Test Checks tool on the mmx-bitcast test
llvm-svn: 284916
2016-10-22 07:15:00 +00:00
Craig Topper b084c90a18 [X86] Add support for printing shuffle comments for VALIGN instructions.
llvm-svn: 284915
2016-10-22 06:51:56 +00:00
Craig Topper 7b2b8db438 [X86] Add support for lowering v4i64 and v8i64 shuffles directly to PALIGNR. I think shuffle combine can figure it out later, but we should try to get it right up front.
llvm-svn: 284914
2016-10-22 06:51:52 +00:00
Craig Topper 9f374533e3 [X86] Remove unnecessary AVX2 check that was already covered by an assertion earlier in the function. NFC
llvm-svn: 284913
2016-10-22 06:51:49 +00:00
Craig Topper bea5cb5491 [X86] Remove 128-bit lane handling from the main loop of matchVectorShuffleAsByteRotate. Instead check for is128LaneRepeatedSuffleMask before the loop and just loop over the repeated mask.
I plan to use the loop to support VALIGND/Q shuffles so this makes it easier to reuse.

llvm-svn: 284912
2016-10-22 06:51:44 +00:00
Simon Pilgrim 0d376bcbf0 [X86][SSE] Use getConstVector helper for VPERMV mask generation. NFCI.
llvm-svn: 284911
2016-10-22 06:18:36 +00:00
Daniel Berlin f5361139bb Now that VS2013 is gone, make a memoryssa structure an anonymous union again
llvm-svn: 284910
2016-10-22 04:15:41 +00:00
Kostya Serebryany 65f102d4d2 [libFuzzer] mutation: insert the size of the input in bytes as one of the ways to mutate a binary integer
llvm-svn: 284909
2016-10-22 03:48:53 +00:00
Gerolf Hoflehner 9e2afa8bd7 [BasicAA] Fix - missed alias in GEP expressions
In BasicAA GEP operand values get adjusted ("wrap-around") based on the
pointersize. Otherwise, in non-64b modes, AA could report false negatives.
However, a wrap-around is valid only for a fully evaluated expression.
It had been introduced to fix an alias problem in
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20160118/326163.html.
This commit restricts the wrap-around to constant gep operands only where the
value is known at compile-time.

llvm-svn: 284908
2016-10-22 02:41:39 +00:00
Richard Smith 6294a768e5 Mark P0012R1 as "Partial" since it's now substantially complete.
llvm-svn: 284907
2016-10-22 01:35:34 +00:00
Richard Smith 58b4ab845a Add another testcase missed from r284905.
llvm-svn: 284906
2016-10-22 01:32:45 +00:00
Richard Smith 1be59c5106 [c++1z] P0012R1: Implement a few remaining pieces: downgrade diagnostic for
mismatched dynamic exception specifications in expressions from an error to a
warning, since this is no longer ill-formed in C++1z.

Allow reference binding of a reference-to-non-noexcept function to a noexcept
function lvalue. As defect resolutions, also allow a conditional between
noexcept and non-noexcept function lvalues to produce a non-noexcept function
lvalue (rather than decaying to a function pointer), and allow function
template argument deduction to deduce a reference to non-noexcept function when
binding to a noexcept function type.

llvm-svn: 284905
2016-10-22 01:32:19 +00:00
Davide Italiano 738837eed9 [CtorUtils] Modernize. No functional changes intended.
llvm-svn: 284904
2016-10-22 01:21:24 +00:00
Kostya Serebryany 10ae9e23a3 [libFuzzer] typo in a test
llvm-svn: 284903
2016-10-22 01:07:38 +00:00
Kostya Serebryany 2bfff021ad [libFuzzer] add a test for asan's strict_string_checks=1
llvm-svn: 284902
2016-10-22 00:05:44 +00:00
Kostya Serebryany 110e2e52b7 [sanitizers] support strict_string_checks for strncmp
llvm-svn: 284901
2016-10-21 23:52:26 +00:00
Jason Molenda bc22c8d8d8 Add some additional logging to
DynamicLoaderDarwinKernel::CheckForKernelImageAtAddress to debug
corefiles that may not be correctly formed.

<rdar://problem/28884846> 

llvm-svn: 284900
2016-10-21 23:45:07 +00:00
Manman Ren c8c9415644 Module: correctly set the module file kind when emitting file_modified.
rdar://28503343

Differential Revision: http://reviews.llvm.org/D25806

llvm-svn: 284899
2016-10-21 23:35:03 +00:00
David L. Jones 0db4549a1e Fix map insertion that is elided in release build.
The assert() macro doesn't actually execute its body in Release builds, so using
it to check cache invariants requires that the insertion be outside of the
assert() statement. This change does that, and also makes sure to return the
actual map contents.

llvm-svn: 284898
2016-10-21 23:30:39 +00:00
Manman Ren 70a7738f84 Module: improve the diagnostic message for include of non-modular header.
Emit the actual path to the non-modular include.

rdar://28897010

llvm-svn: 284897
2016-10-21 23:27:37 +00:00
Sanjay Patel 5b7767f646 [x86] add test for missing vector SRA combine via computeKnownBits
llvm-svn: 284896
2016-10-21 23:02:31 +00:00
Richard Smith ce76629905 Remove unnecessary distinction between Ref_Compatible and
Ref_Compatible_With_Added_Qualification. We always treated these two values the
same way.

llvm-svn: 284895
2016-10-21 23:01:55 +00:00
Eugene Zelenko 566b34d9b7 [Release notes] Mention removed Clang-tidy misc-pointer-and-integral-operation check
llvm-svn: 284894
2016-10-21 22:35:58 +00:00
Chris Bieneman 7ba5581b84 [Test Suite] Allow overriding codesign identity
Summary: Not everyone names their code sign identity "lldb_codesign", so it is nice to allow this to be overridden.

Reviewers: zturner, tfiala

Subscribers: labath, mgorny, lldb-commits

Differential Revision: https://reviews.llvm.org/D25714

llvm-svn: 284893
2016-10-21 22:13:55 +00:00
Justin Lebar 7b15e2946f [ADT] Don't rely on string literals not being convertible to non-const char* in CachedHashString.
The build was breaking on some platforms because we assumed that
CachedHashString("foo") would match the CachedHashString(StringRef)
constructor rather than the CachedHashString(char*) constructor.

To fix this, provide a CachedHashString(const char*) constructor, and
add a dummy argument to the old CachedHashString(char*) constructor.

llvm-svn: 284892
2016-10-21 22:10:23 +00:00
Konstantin Zhuravlyov fda33eaf0c [AMDGPU] Perform uchar to float combine for ISD::SINT_TO_FP
This will prevent following regression when enabling i16 support (D18049):
  test/CodeGen/AMDGPU/cvt_f32_ubyte.ll

Differential Revision: https://reviews.llvm.org/D25805

llvm-svn: 284891
2016-10-21 22:10:03 +00:00
Richard Smith 5e9746f520 DR583, DR1512: Implement a rewrite to C++'s 'composite pointer type' rules.
This has two significant effects:

1) Direct relational comparisons between null pointer constants (0 and nullopt)
   and pointers are now ill-formed. This was always the case for C, and it
   appears that C++ only ever permitted by accident. For instance, cases like
     nullptr < &a
   are now rejected.

2) Comparisons and conditional operators between differently-cv-qualified
   pointer types now work, and produce a composite type that both source
   pointer types can convert to (when possible). For instance, comparison
   between 'int **' and 'const int **' is now valid, and uses an intermediate
   type of 'const int *const *'.

Clang previously supported #2 as an extension.

We do not accept the cases in #1 as an extension. I've tested a fair amount of
code to check that this doesn't break it, but if it turns out that someone is
relying on this, we can easily add it back as an extension.

This is a re-commit of r284800.

llvm-svn: 284890
2016-10-21 22:00:42 +00:00
Richard Smith 00c03c5de7 Remove 'misc-pointer-and-integral-operation' clang-tidy check. The only cases
it detects are ill-formed (some per C++ core issue 1512, others always have
been).

llvm-svn: 284888
2016-10-21 21:50:28 +00:00