Commit Graph

292372 Commits

Author SHA1 Message Date
Jan Vesely 8382e5bc48 atom: Use volatile pointers for cl_khr_{global,local}_int32_{base,extended}_atomics
int64 versions were switched to volatile pointers in cl1.1
cl1.1 also renamed atom_ functions to atomic_ that use volatile pointers.
CTS and applications use volatile pointers.

Passes CTS on carrizo
no return piglit tests still pass on turks.

Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335280
2018-06-21 19:27:39 +00:00
Jan Vesely 65e3541b78 atom: Consolidate cl_khr_{local,global}_int32_{base,extended}_atomics implementation
These are just atomic_* wrappers.
Switch inc, dec to use atomic_* wrappers as well.

Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335279
2018-06-21 19:27:33 +00:00
Jan Vesely f965b46c8e atomic: Provide function implementation of atomic_{dec,inc}
Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335278
2018-06-21 19:27:26 +00:00
Jan Vesely b9cbe0bf51 atom: Consolidate cl_khr_int64_{base,extended}_atomics declarations
Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335277
2018-06-21 19:27:23 +00:00
Jan Vesely d1c3811ff7 atom: Consolidate cl_khr_{local,global}_int32_{base,extended}_atomics declarations
Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335276
2018-06-21 19:27:18 +00:00
Jan Vesely fe08de0c89 atomic: Cleanup atomic_cmpxchg header
It's easier to just list the four function declarations

Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335275
2018-06-21 19:27:12 +00:00
Jan Vesely eabc110372 atomic: Move define cleanup to shared include
Reviewed-By: Aaron Watry <awatry@gmail.com>
Tested-By: Aaron Watry <awatry@gmail.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
llvm-svn: 335274
2018-06-21 19:27:07 +00:00
Tatyana Krasnukha be73fd6b5d Remove duplicated check and shared_ptr copying.
llvm-svn: 335273
2018-06-21 19:19:57 +00:00
Francis Visoiu Mistrih ac599b6951 Revert r335206 "Recommit r333268: [IPSCCP] Use PredicateInfo to propagate facts from cmp instructions."
This reverts commit r335206.

As discussed here: https://reviews.llvm.org/rL333740, a fix will come
tomorrow. In the meanwhile, revert this to fix some bots.

llvm-svn: 335272
2018-06-21 19:18:36 +00:00
Jonas Devlieghere 702e140d68 [DataFormatter] Add CFDictionary data formatter
Add data formatter for NSCFDictionary/CFDictionaryRef.

Differential revision: https://reviews.llvm.org/D48450

llvm-svn: 335271
2018-06-21 19:13:47 +00:00
Craig Topper 1763dbb278 [X86] Correct the inline assembly implementations of __movsb/w/d/q and __stosw/d/q to mark registers/memory as modified
The inline assembly for these didn't mark that edi, esi, ecx are modified by movs/stos instruction. It also didn't mark that memory is modified.

This issue was reported to llvm-dev last year http://lists.llvm.org/pipermail/cfe-dev/2017-November/055863.html but no bug was ever filed.

Differential Revision: https://reviews.llvm.org/D48448

llvm-svn: 335270
2018-06-21 18:56:30 +00:00
Simon Dardis 3505045b42 [mips] Modify comment to test new email address (NFC).
llvm-svn: 335269
2018-06-21 18:52:32 +00:00
Scott Linder 5792dd0f39 [AMDGPU] Fix bug with tracking processed blocks in SIInsertWaitcnts
BlockWaitcntProcessedSet was not being cleared between calls, so it was
producing incorrect counts in cases where MBB addresses happened to coincide
across multiple calls.

Differential Revision: https://reviews.llvm.org/D48391

llvm-svn: 335268
2018-06-21 18:48:48 +00:00
Konstantin Zhuravlyov 766c77efd7 AMDGPU/AMDHSA: Remove GridWorkGroupCountX/Y/Z
and everything that comes with it from implementation
and v3 header files.

Leave definition in v2 header files for backwards
compatibility.

Differential Revision: https://reviews.llvm.org/D48191

llvm-svn: 335267
2018-06-21 18:36:04 +00:00
Sanjay Patel 3382dc644e [InstCombine] add tests for shuffled cmps; NFC
llvm-svn: 335266
2018-06-21 18:07:38 +00:00
Kuba Mracek 14c786a960 [tsan] Use DARWIN_osx_LINK_FLAGS when building unit tests to match ASan behavior.
llvm-svn: 335265
2018-06-21 18:00:51 +00:00
Matt Davis d041f21810 [DebugInfo] Ignore DBG_VALUE instructions in PostRA Machine Sink
Summary:
The logic for handling the sinking of COPY instructions was generating
different code when building with debug flags.

The original code did not take into consideration debug instructions.  This
resulted in the registers in the DBG_VALUE instructions being treated as used,
and prevented the COPY from being sunk.  This patch avoids analyzing debug
instructions when trying to sink COPY instructions.

This patch also creates a routine from the code in MachineSinking::SinkInstruction to
perform the logic of sinking an instruction along with its debug instructions.
This functionality is used in multiple places, including the code for sinking COPY instrs.


Reviewers: junbuml, javed.absar, MatzeB, bjope

Reviewed By: bjope

Subscribers: aprantl, probinson, thegameg, jonpa, bjope, vsk, kristof.beyls, JDevlieghere, llvm-commits

Tags: #debug-info

Differential Revision: https://reviews.llvm.org/D45637

llvm-svn: 335264
2018-06-21 17:59:52 +00:00
Greg Clayton 59a1be3acf Fix an issue where DW_OP_deref might be dereferencing a file address. Convert the file address to a load address so this works.
https://bugs.llvm.org/show_bug.cgi?id=36871

llvm-svn: 335263
2018-06-21 17:58:06 +00:00
Sanjay Patel 3244537a3c [InstCombine] use constant pattern matchers with icmp+sext
The previous code worked with vectors, but it failed when the
vector constants contained undef elements. 
The matchers handle those cases.

llvm-svn: 335262
2018-06-21 17:51:44 +00:00
Sanjay Patel 5522e968ad [InstCombine] add vector icmp tests with undefs; NFC
llvm-svn: 335261
2018-06-21 17:37:14 +00:00
Pavel Labath d68983e3d5 Partially revert r335236
Jim pointed out that XCode has build configurations that build without
python and removing the ifdefs around the python code breaks them.

This reverts the #ifdef part of the above patch, while keeping the cmake
parts.

llvm-svn: 335260
2018-06-21 17:36:32 +00:00
Craig Topper b2431c6c33 [Intrinsics] Add/move some builtin declarations in intrin.h to get ms-intrinsics.c to not issue warnings
ud2 and int2c were missing declarations entirely. And the bitscans were only under x86_64, but they seem to be in BuiltinsARM.def as well and are tested by ms_intrinsics.c

Differential Revision: https://reviews.llvm.org/D48187

llvm-svn: 335259
2018-06-21 17:07:04 +00:00
Sanjay Patel 7b0fc75f73 [InstCombine] simplify binops before trying other folds
This is outwardly NFC from what I can tell, but it should be more efficient 
to simplify first (despite the name, SimplifyAssociativeOrCommutative does
not actually simplify as InstSimplify does - it creates/morphs instructions).

This should make it easier to refactor duplicated code that runs for all binops.

llvm-svn: 335258
2018-06-21 17:06:36 +00:00
Sanjay Patel 447e8ece4d [LoopVectorize] regenerate full checks; NFC
llvm-svn: 335257
2018-06-21 16:54:32 +00:00
Craig Topper 8014053cbd [X86] Update fast-isel tests for clang r335253.
The new IR fixes a mismatch in the final extractelement for the i32 intrinsics. Previously we extracted a 64-bit element even though we only wanted 32 bits.

SimplifyDemandedElts isn't able to make FP elements undef now and the shuffle mask I used prevents the use of horizontal add we had before. Not sure we should have been using horizontal add anyway. It's implemented on Intel with two port 5 shuffles and an add. So we have on less shuffle now, but an additional instruction to decode.

Differential Revision: https://reviews.llvm.org/D48347

llvm-svn: 335256
2018-06-21 16:54:18 +00:00
Anastasis Grammenos dfe8fe503c [DebugInfo] Inline for without DebugLocation
Summary:
This test is a strip down version of a function inside the
amalgamated sqlite source. When converted to IR clang produces
a phi instruction without debug location.

This patch fixes the above issue.

Differential Revision: https://reviews.llvm.org/D47720

llvm-svn: 335255
2018-06-21 16:53:48 +00:00
Paul Robinson 547adcaaac [DWARF] Warn on and ignore ".file 0" for DWARF v4 and earlier.
This had been messing with the directory table for prior versions, and
also could induce a crash when generating asm output.

llvm-svn: 335254
2018-06-21 16:42:03 +00:00
Craig Topper ddfe69cc99 [X86] Rewrite the add/mul/or/and reduction intrinsics to make better use of other intrinsics and remove undef shuffle indices.
Similar to what was done to max/min recently.

These already reduced the vector width to 256 and 128 bit as we go unlike the original max/min code.

Differential Revision: https://reviews.llvm.org/D48346

llvm-svn: 335253
2018-06-21 16:41:28 +00:00
Alexander Kornienko 437e026631 [clang-tidy] Remove the google-readability-redundant-smartptr-get alias
I don't remember why I added it, but it's definitely not needed, since the check
doesn't have any options and the check doesn't have any special relation to the
Google C++ style.

llvm-svn: 335252
2018-06-21 16:14:27 +00:00
Sirish Pande b60acb9e48 Revert "[AArch64] Coalesce Copy Zero during instruction selection"
This reverts commit d8f57105010cc7e78026e511d5def873fc91e0e7.

Original Commit:

Author: Haicheng Wu <haicheng@codeaurora.org>
Date:   Sun Feb 18 13:51:33 2018 +0000

    [AArch64] Coalesce Copy Zero during instruction selection

    Add special case for copy of zero to avoid a double copy.

    Differential Revision: https://reviews.llvm.org/D36104

Author's intention is to remove a BB that has one mov instruction. In
order to do that, d8f571050 pessmizes MachineSinking by introducing a
copy, such that mov instruction is NOT moved to the BB. Optimization
downstream gets rid of the BB with only mov instruction. This works well
if we have only one fall through branch as there is only one "extra"
mov instruction.

If we have multiple fall throughs, we will have a lot of redundant movs.
In such a case, it's better to have this BB which has one mov instruction.

This is causing degradation in jpeg, fft and other codebases. I believe
if we want to remove a BB with only one branch instruction, we should not
pessimize Machine Sinking at all, and find some other solution.

llvm-svn: 335251
2018-06-21 16:05:24 +00:00
Stanislav Mekhanoshin 22ee191c3e DAG combine "and|or (select c, -1, 0), x" -> "select c, x, 0|-1"
Allowed folding for "and/or" binops with non-constant operand if
arguments of select are 0/-1 values.

Normally this code with "and" opcode does not get to a DAG combiner
and simplified yet in the InstCombine. However AMDGPU produces it
during lowering and InstCombine has no chance to optimize it out.

In turn the same pattern with "or" opcode can reach DAG.

Differential Revision: https://reviews.llvm.org/D48301

llvm-svn: 335250
2018-06-21 16:02:05 +00:00
David Green 21a2973cc4 [ARM] Enable useAA() for the in-order Cortex-R52
This option allows codegen (such as DAGCombine or MI scheduling) to use alias
analysis information, which can help with the codegen on in-order cpu's,
especially machine scheduling. Here I have done things the same way as AArch64,
adding a subtarget feature to enable this for specific cores, and enabled it for
the R52 where we have a schedule to make use of it.

Differential Revision: https://reviews.llvm.org/D48074

llvm-svn: 335249
2018-06-21 15:48:29 +00:00
Pavel Labath 9092cc96d4 Fix macos build for r335244
I've made the code accept only 16 byte UUIDs, which is technically not
NFC (previously it would also accept 20 byte ones, but use only the
first 16 bytes), but this should be more correct as mac UUIDs are always
16 byte long.

llvm-svn: 335247
2018-06-21 15:40:33 +00:00
Pavel Labath a174bcbf03 Remove UUID::SetFromCString
Replace uses with SetFromStringRef. NFC.

llvm-svn: 335246
2018-06-21 15:24:39 +00:00
Kuba Mracek bb9dedfa8a [sanitizer] Stop running tests against 32-bit iOS simulator
llvm-svn: 335245
2018-06-21 15:21:24 +00:00
Pavel Labath 470b286ee5 Modernize UUID class
Instead of a separate GetBytes + GetByteSize methods I introduce a
single GetBytes method returning an ArrayRef.

This is NFC cleanup now, but it should make handling arbitrarily-sized
UUIDs cleaner, should we choose to go that way. I also took the
opportunity to add some unit tests for this class.

llvm-svn: 335244
2018-06-21 15:07:43 +00:00
Sam Clegg 084d360f69 [WebAssembly] Only mark non-hidden symbols as live if they are also defined
Previously we were also marking undefined symbols (i.e. imports)
as live.

Differential Revision: https://reviews.llvm.org/D48299

llvm-svn: 335243
2018-06-21 15:00:00 +00:00
Sanjay Patel 3e5c051a06 [InstCombine] make div/rem vector constant utility function; NFCI
This was originally in D48401 and will be used there.

llvm-svn: 335242
2018-06-21 14:59:35 +00:00
Sam Parker ad71333b92 [NFC][ARM] ldrd/strd negative tests
Add negative tests for load and stores of alignment 2.

llvm-svn: 335241
2018-06-21 14:53:06 +00:00
Clement Courbet 1ef6aa814d [llvm-exegesis][NFC] Simplify BenchmarkRunner.
Get rid of createExecutableFunction().

llvm-svn: 335240
2018-06-21 14:49:04 +00:00
Sameer AbuAsal e01e711c64 [RISCV] Tail calls don't need to save return address
Summary:
 When expanding the PseudoTail in expandFunctionCall() we were using X6
 to save the return address. Since this is a tail call the return
 address is not needed, this patch replaces it with X0 to be ignored.

 This matches the behaviour listed in the ISA V2.2 document page 110.
 tail offset -----> jalr x0, x6, offset

 GCC exhibits the same behavior.

Reviewers: apazos, asb, mgrang

Reviewed By: asb

Subscribers: rbar, johnrusso, simoncook, niosHD, kito-cheng, shiva0217, zzheng, edward-jones, rogfer01

Differential Revision: https://reviews.llvm.org/D48343

llvm-svn: 335239
2018-06-21 14:37:09 +00:00
Mikhail Dvoretckii 22c82af5c8 [x86] Lower some trunc + shuffle patterns to vpmov[q|d][b|w]
This should help in lowering the following four intrinsics:
 _mm256_cvtepi32_epi8
 _mm256_cvtepi64_epi16
 _mm256_cvtepi64_epi8
 _mm512_cvtepi64_epi8

Differential Revision: https://reviews.llvm.org/D46957

llvm-svn: 335238
2018-06-21 14:16:45 +00:00
Clement Courbet 760d1d5741 [llvm-exegesis][NFC] Simplify LLVMState.
Summary: Pretty much everything we need is in llvm::TargetMachine.

Reviewers: gchatelet

Subscribers: llvm-commits, tschuett

Differential Revision: https://reviews.llvm.org/D48428

llvm-svn: 335237
2018-06-21 14:11:09 +00:00
Pavel Labath bcadb5a3d1 ScriptInterpreterPython cleanup
Instead of #ifdef-ing the contents of all files in the plugin for all
non-python builds, just disable the plugin at the cmake level. Also,
remove spurious extra linking of the Python plugin in liblldb. This
plugin is already included as a part of LLDB_ALL_PLUGINS variable.

llvm-svn: 335236
2018-06-21 14:09:15 +00:00
Pavel Labath 30f2e61b97 Disable gmodules tests on linux
These tests are extremely environment-dependent. if the environment is
not module-enabled (which is the likely scenario), they won't test
anything. If one happens to have a module-enabled libc++, then the he
will start running into problems.

The first one is that the debug info in pcm file contains relocations
that ObjectFileELF doesn't handle (particularly on non-x86
architectures), but even after that is resolved, it seems we still are
unable to pull debug info out of the pcm file. I've filed pr37893 to
track that, and I am disabling gmodules tests on linux until these
issues are resolved.

llvm-svn: 335235
2018-06-21 13:55:19 +00:00
Rainer Orth 4bd5c47601 [liblang] Remove DOS line endings in libclang.exports
Undefined                       first referenced
 symbol                             in file
clang_getCompletionFixIt           /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getTokenLocation             /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getToken                     /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getTemplateCursorKind        /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getTUResourceUsageName       /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionChunkKind       /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionChunkText       /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getSpellingLocation          /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionParent          /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionChunkCompletionString /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionPriority        /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionNumFixIts       /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getTokenExtent               /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getCompletionNumAnnotations  /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
clang_getTokenKind                 /var/gcc/llvm/obj/local/tools/clang/tools/libclang/libclang.exports
ld: fatal: symbol referencing errors
collect2: error: ld returned 1 exit status
make[2]: *** [tools/clang/tools/libclang/CMakeFiles/libclang.dir/build.make:651: lib/libclang.so.7] Error 1

It turns out that this is caused by https://reviews.llvm.org/D46862: it added a
couple of CRs (^M) to some lines.  Solaris ld takes them to be part of the symbol
names, which of course are missing from the input objects.  GNU ld handles this
just fine.  Fixed by removing the CRs.

Bootstrapped on i386-pc-solaris2.11.  I guess this is obvious.

Differential Revision: https://reviews.llvm.org/D48423

llvm-svn: 335234
2018-06-21 13:40:50 +00:00
Krzysztof Parzyszek 6a3229301f [CodeGen] Avoid handling DBG_VALUE in LiveRegUnits::stepBackward
Patch by Jesper Antonsson.

Differential Revision: https://reviews.llvm.org/D48420

llvm-svn: 335233
2018-06-21 13:38:43 +00:00
Nicolai Haehnle 15745ba5c1 AMDGPU: Remove redundant MIMG instruction variants
Summary:
For sample and gather ops, we can accurately determine the set of
vaddr-size instruction variants that are required. This reduces
the size of instruction tables by ~5%.

The number of machine instruction opcodes is reduced from 10002
to 9476.

Change-Id: Ie7fc65d3657b762c7816017fe70b2e9bec644a8a

Reviewers: arsenm, rampitec

Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, llvm-commits, t-tye

Differential Revision: https://reviews.llvm.org/D48168

llvm-svn: 335232
2018-06-21 13:37:55 +00:00
Nicolai Haehnle db6911a6f9 AMDGPU: Remove old-style image intrinsics
Summary:
This also removes the need for atomic pseudo instructions, since
we select the correct encoding directly in SITargetLowering::lowerImage
for dimension-aware image intrinsics.

Mesa uses dimension-aware image intrinsics since
commit a9a7993441.

Change-Id: I7473d20009476a4ed6d919cae4e6dca9ff42e77a

Reviewers: arsenm, rampitec, mareko, tpr, b-sumner

Subscribers: kzhuravl, wdng, yaxunl, dstuttard, t-tye, llvm-commits

Differential Revision: https://reviews.llvm.org/D48167

llvm-svn: 335231
2018-06-21 13:37:45 +00:00
Nicolai Haehnle b29ee70122 InstCombine/AMDGPU: Add dimension-aware image intrinsics to SimplifyDemanded
Summary:
Use the expanded features of the TableGen generic tables to avoid manually
adding the combinatorially exploded set of intrinsics. The
getAMDGPUImageDimIntrinsic lookup function is early-out,
i.e. non-AMDGPU intrinsics will never look at the underlying table.

Use a generic approach for getting the new intrinsic overload to keep the
code simple, and make the image dmask handling more generic:
- handle non-sampler image loads
- handle the case where the set of demanded elements is not a prefix

There is some overlap between this code and an optimization that happens
in the backend during code generation. They currently complement each other:

- only the codegen optimization can generate vec3 loads
- only the InstCombine optimization can handle D16

The InstCombine optimization also likely covers more cases since the
codegen optimization is fairly ad-hoc. Ideally, we'll remove the optimization
in codegen once the infrastructure for vec3 is in place (which will probably
take a long time).

Modify the test cases to use dimension-aware intrinsics. This makes it
easier to see that the test coverage for the new intrinsics is equivalent,
and the old style intrinsics will be removed in a follow-up commit anyway.

Change-Id: I4b91ea661413d13004956fe4ef7d13d41b8ce3ad

Reviewers: arsenm, rampitec, majnemer

Subscribers: kzhuravl, wdng, mgorny, yaxunl, dstuttard, tpr, t-tye, llvm-commits

Differential Revision: https://reviews.llvm.org/D48165

llvm-svn: 335230
2018-06-21 13:37:31 +00:00