Tim Shen [Mon, 22 Aug 2016 22:30:27 +0000 (22:30 +0000)]
Migrate from NodeType * to NodeRef.
llvm-svn: 279488
Pete Cooper [Mon, 22 Aug 2016 22:27:05 +0000 (22:27 +0000)]
Fix crash from assert in r279466.
The assert in r279466 checks that we call the correct version of
Intrinsic::getName. The version which accepts only an ID should not
be used for intrinsics with overloaded types. The global-isel
code was calling the wrong version. The test CodeGen/AArch64/GlobalISel/arm64-irtranslator.ll
will ensure that we call the correct version from now on.
llvm-svn: 279487
Richard Smith [Mon, 22 Aug 2016 22:25:03 +0000 (22:25 +0000)]
Fix regression introduced by r279164: only pass definitions as the PatternDef
to DiagnoseUninstantiableTemplate, teach hasVisibleDefinition to correctly
determine whether a function definition is visible, and mark both the function
and the template as visible when merging function template definitions to
provide hasVisibleDefinition with the relevant information.
The change to always pass the right declaration as the PatternDef to
DiagnoseUninstantiableTemplate also caused those checks to happen before other
diagnostics in InstantiateFunctionDefinition, giving worse diagnostics for the
same situations, so I sunk the relevant diagnostics into
DiagnoseUninstantiableTemplate. Those parts of this patch are based on changes
in reviews.llvm.org/D23492 by Vassil Vassilev.
llvm-svn: 279486
Adrian Prantl [Mon, 22 Aug 2016 22:23:58 +0000 (22:23 +0000)]
Module debug info: Don't assert when encountering an incomplete definition
in isDefinedInClangModule() and assume that the incomplete definition
is not defined in the module.
This broke the -gmodules self host recently.
rdar://problem/
27894367
llvm-svn: 279485
Duncan P. N. Exon Smith [Mon, 22 Aug 2016 22:21:07 +0000 (22:21 +0000)]
ADT: Separate some list manipulation API into ilist_base, NFC
Separate algorithms in iplist<T> that don't depend on T into ilist_base,
and unit test them.
While I was adding unit tests for these algorithms anyway, I also added
unit tests for ilist_node_base and ilist_sentinel<T>.
To make the algorithms and unit tests easier to write, I also did the
following minor changes as a drive-by:
- encapsulate Prev/Next in ilist_node_base to so that algorithms are
easier to read, and
- update ilist_node_access API to take nodes by reference.
There should be no real functionality change here.
llvm-svn: 279484
Duncan P. N. Exon Smith [Mon, 22 Aug 2016 22:04:16 +0000 (22:04 +0000)]
Fix header comment for unittests/ADT/ilistTest.cpp
llvm-svn: 279483
Tim Shen [Mon, 22 Aug 2016 21:59:26 +0000 (21:59 +0000)]
[ADT] Actually mutate the iterator VisitStack.back().second, not its copy.
Summary: Before the change, *Opt never actually gets updated by the end
of toNext(), so for every next time the loop has to start over from
child_begin(). This bug doesn't affect the correctness, since Visited prevents
it from re-entering the same node again; but it's slow.
Reviewers: dberris, dblaikie, dannyb
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D23649
llvm-svn: 279482
Bruno Cardoso Lopes [Mon, 22 Aug 2016 21:50:22 +0000 (21:50 +0000)]
[SemaObjC] Do not RebuildObjCMessageExpr without valid method decl
Fix crash-on-invalid in ObjC Sema by avoiding to rebuild a message
expression to a 'super' class in case the method to call does not exist
(i.e. comes from another missing identifier).
In this case, the typo transform is invoked upon the message expression
in an attempt to solve a typo in a 'super' call parameters, but it
crashes since it assumes the method to call has a valid declaration.
rdar://problem/
27305403
llvm-svn: 279481
Rafael Espindola [Mon, 22 Aug 2016 21:33:35 +0000 (21:33 +0000)]
Add support for '|' in expressions.
llvm-svn: 279480
Davide Italiano [Mon, 22 Aug 2016 21:33:12 +0000 (21:33 +0000)]
[AST] Remove unused function, to silence a GCC7 warning.
llvm-svn: 279479
Sanjay Patel [Mon, 22 Aug 2016 21:24:29 +0000 (21:24 +0000)]
[InstCombine] change param type from Instruction to BinaryOperator for icmp helpers; NFCI
This saves some casting in the helper functions and eases some further refactoring.
llvm-svn: 279478
Tim Shen [Mon, 22 Aug 2016 21:09:30 +0000 (21:09 +0000)]
[GraphTraits] Replace all NodeType usage with NodeRef
This should finish the GraphTraits migration.
Differential Revision: http://reviews.llvm.org/D23730
llvm-svn: 279475
Duncan P. N. Exon Smith [Mon, 22 Aug 2016 20:51:00 +0000 (20:51 +0000)]
ADT: Remove ilist_*sentinel_traits, NFC
Remove all the dead code around ilist_*sentinel_traits. This is a
follow-up to gutting them as part of r279314 (originally r278974),
staged to prevent broken builds in sub-projects.
Uses were removed from clang in r279457 and lld in r279458.
llvm-svn: 279473
Sanjay Patel [Mon, 22 Aug 2016 20:45:06 +0000 (20:45 +0000)]
[InstCombine] use m_APInt to allow icmp (shr exact X, Y), 0 folds for splat constant vectors
llvm-svn: 279472
Pete Cooper [Mon, 22 Aug 2016 20:38:53 +0000 (20:38 +0000)]
Add ADT headers to the cmake headers directory for LLVMSupport. NFC.
Xcode and MSVC list the headers and source files for each library.
LLVMSupport lists included the source files for ADT but not the headers. This
add the ADT headers so that they are browsable by the UI.
llvm-svn: 279470
Francis Ricci [Mon, 22 Aug 2016 20:33:47 +0000 (20:33 +0000)]
[compiler-rt] Use flags found when configuring builtins during compilation
Summary: This fixes the omission of -fPIC when building the builtins.
Reviewers: compnerd, beanz
Subscribers: dberris, llvm-commits
Differential Revision: https://reviews.llvm.org/D23729
llvm-svn: 279469
Francis Ricci [Mon, 22 Aug 2016 20:27:21 +0000 (20:27 +0000)]
[compiler-rt] Don't build ubsan cxxabi sources when unused
Summary:
On apple targets, when SANITIZER_CAN_USE_CXXABI is false,
the ubsan cxxabi sources aren't built, since they're unused.
Do this on non-apple targets as well.
This fixes errors when linking sanitizers if c++ abi is
unavailable.
Reviewers: pcc, kubabrecka, beanz
Subscribers: rnk, llvm-commits, kubabrecka, compnerd, dberris
Differential Revision: https://reviews.llvm.org/D23638
llvm-svn: 279467
Pete Cooper [Mon, 22 Aug 2016 20:18:28 +0000 (20:18 +0000)]
Add comments and an assert to follow-up on r279113. NFC.
Philip commented on r279113 to ask for better comments as to
when to use the different versions of getName. Its also possible
to assert in the simple case that we aren't an overloaded intrinsic
as those have to use the more capable version of getName.
Thanks for the comments Philip.
llvm-svn: 279466
Daniel Berlin [Mon, 22 Aug 2016 19:52:23 +0000 (19:52 +0000)]
IDFCalculator: Remove unused field.
llvm-svn: 279465
Matt Arsenault [Mon, 22 Aug 2016 19:33:16 +0000 (19:33 +0000)]
AMDGPU: Split SILowerControlFlow into two pieces
Do most of the lowering in a pre-RA pass. Keep the skip jump
insertion late, plus a few other things that require more
work to move out.
One concern I have is now there may be COPY instructions
which do not have the necessary implicit exec uses
if they will be lowered to v_mov_b32.
This has a positive effect on SGPR usage in shader-db.
llvm-svn: 279464
Matt Arsenault [Mon, 22 Aug 2016 19:25:59 +0000 (19:25 +0000)]
AMDGPU: Handle structs directly in AMDGPUABIInfo
Structs are currently handled as pointer + byval, which makes AMDGPU
LLVM backend generate incorrect code when structs are used. This patch
changes struct argument to be handled directly and without flattening,
which Clover (Mesa 3D Gallium OpenCL state tracker) will be able to
handle. Flattening would expand the struct to individual elements and
pass each as a separate argument, which Clover can not
handle. Furthermore, such expansion does not fit the OpenCL
programming model which requires to explicitely specify each argument
index, size and memory location.
Patch by Vedran Miletić
llvm-svn: 279463
Daniel Berlin [Mon, 22 Aug 2016 19:14:30 +0000 (19:14 +0000)]
MSSA: Factor out phi node placement
llvm-svn: 279462
Daniel Berlin [Mon, 22 Aug 2016 19:14:16 +0000 (19:14 +0000)]
MSSA: Only rename accesses whose defining access is nullptr
llvm-svn: 279461
James Molloy [Mon, 22 Aug 2016 19:07:15 +0000 (19:07 +0000)]
[SimplifyCFG] Rewrite SinkThenElseCodeToEnd
[Recommitting now an unrelated assertion in SROA is sorted out]
The new version has several advantages:
1) IMSHO it's more readable and neater
2) It handles loads and stores properly
3) It can handle any number of incoming blocks rather than just two. I'll be taking advantage of this in a followup patch.
With this change we can now finally sink load-modify-store idioms such as:
if (a)
return *b += 3;
else
return *b += 4;
=>
%z = load i32, i32* %y
%.sink = select i1 %a, i32 5, i32 7
%b = add i32 %z, %.sink
store i32 %b, i32* %y
ret i32 %b
When this works for switches it'll be even more powerful.
Round 4. This time we should handle all instructions correctly, and not replace any operands that need to be constant with variables.
This was really hard to determine safely, so the helper function should be put into the Instruction API. I'll do that as a followup.
llvm-svn: 279460
Petr Hosek [Mon, 22 Aug 2016 19:01:53 +0000 (19:01 +0000)]
[ELF] Only print symbol name when it is available
Not only symbols (like sections) have names, in case where we
fail to create relocation against such symbol, we should not
print out an empty string, instead we should print a generic
message.
Differential Revision: https://reviews.llvm.org/D23731
llvm-svn: 279459
Duncan P. N. Exon Smith [Mon, 22 Aug 2016 18:59:33 +0000 (18:59 +0000)]
ADT: Remove uses of ilist_*sentinel_traits, NFC
llvm-svn: 279458
Duncan P. N. Exon Smith [Mon, 22 Aug 2016 18:57:44 +0000 (18:57 +0000)]
ADT: Remove uses of ilist_*sentinel_traits, NFC
llvm-svn: 279457
Petr Hosek [Mon, 22 Aug 2016 18:53:09 +0000 (18:53 +0000)]
[ELF] ICF should respect section alignment
When performing ICF, we have to respect the alignment requirement
of each section within each group.
Differential Revision: https://reviews.llvm.org/D23732
llvm-svn: 279456
Artem Belevich [Mon, 22 Aug 2016 18:50:34 +0000 (18:50 +0000)]
[CUDA] Collapsed offload actions should not be top-level jobs.
If they are, we end up with the last intermediary output preserved
in the current directory after compilation.
Added a test case to verify that we're using appropriate filenames
for outputs of different phases.
Differential Revision: https://reviews.llvm.org/D23526
llvm-svn: 279455
James Molloy [Mon, 22 Aug 2016 18:49:42 +0000 (18:49 +0000)]
[SROA] Remove incorrect assertion
Confirmed with aprantl, this assertion is incorrect - code can get here (for example 80-bit FP types) and if it does it's benign. This is exposed by a completely unrelated patch of mine, so stop the compiler falling over.
Original differential: http://reviews.llvm.org/D16187
aprantl's advice to remove assertion: http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-
20160815/382129.html
llvm-svn: 279454
Marshall Clow [Mon, 22 Aug 2016 18:45:31 +0000 (18:45 +0000)]
Add missing include that caused a test failure on Windows. Thanks to STL for the patch. No functional change.
llvm-svn: 279453
Petr Hosek [Mon, 22 Aug 2016 18:44:04 +0000 (18:44 +0000)]
[ELF] Define __ehdr_start symbol
This symbol can be used by the program to examine its own headers.
Differential Revision: https://reviews.llvm.org/D23750
llvm-svn: 279452
Francis Ricci [Mon, 22 Aug 2016 18:31:37 +0000 (18:31 +0000)]
[compiler-rt] Make cxxabi availability configurable in cmake
Summary:
Allow for use of cxxabi to be disabled from cmake.
This will make sanitizers usable when c++ abi is unavailable.
Reviewers: pcc, rnk, samsonov, beanz, compnerd
Subscribers: llvm-commits, compnerd, dberris
Differential Revision: https://reviews.llvm.org/D23639
llvm-svn: 279451
Filipe Cabecinhas [Mon, 22 Aug 2016 18:30:37 +0000 (18:30 +0000)]
[cmake] Fix append_rtti_flag macro and bad var name
llvm-svn: 279450
Tim Shen [Mon, 22 Aug 2016 18:26:27 +0000 (18:26 +0000)]
[SSP] Do not set __guard_local to hidden for OpenBSD SSP
__guard_local is defined as long on OpenBSD. If the source file contains
a definition of __guard_local, it mismatches with the int8 pointer type
used in LLVM. In that case, Module::getOrInsertGlobal() returns a
cast operation instead of a GlobalVariable. Trying to set the
visibility on the cast operation leads to random segfaults (seen when
compiling the OpenBSD kernel, which also runs with stack protection).
In the kernel, the hidden attribute does not matter. For userspace code,
__guard_local is defined as hidden in the startup code. If a program
re-defines __guard_local, the definition from the startup code will
either win or the linker complains about multiple definitions
(depending on whether the re-defined __guard_local is placed in the
common segment or not).
It also matches what gcc on OpenBSD does.
Thanks Stefan Kempf <sisnkemp@gmail.com> for the patch!
Differential Revision: http://reviews.llvm.org/D23674
llvm-svn: 279449
Jun Bum Lim [Mon, 22 Aug 2016 18:21:56 +0000 (18:21 +0000)]
[InstCombine] Allow sinking from unique predecessor with multiple edges
Summary: We can allow sinking if the single user block has only one unique predecessor, regardless of the number of edges. Note that a switch statement with multiple cases can have the same destination.
Reviewers: mcrosier, majnemer, spatel, reames
Subscribers: reames, mcrosier, llvm-commits
Differential Revision: https://reviews.llvm.org/D23722
llvm-svn: 279448
James Molloy [Mon, 22 Aug 2016 18:13:12 +0000 (18:13 +0000)]
Revert "[SimplifyCFG] Rewrite SinkThenElseCodeToEnd"
This reverts commit r279443. It caused buildbot failures.
llvm-svn: 279447
Enrico Granata [Mon, 22 Aug 2016 18:07:52 +0000 (18:07 +0000)]
Add a non-code-running data formatter for __NSCFBoolean
llvm-svn: 279446
David Blaikie [Mon, 22 Aug 2016 17:49:56 +0000 (17:49 +0000)]
PR29086: DebugInfo: Improve support for fixed array dimensions in variable length arrays
llvm-svn: 279445
David Blaikie [Mon, 22 Aug 2016 17:49:50 +0000 (17:49 +0000)]
Remove redundant test
test/CodeGenCXX/debug-info-zero-length-arrays.cpp tests this
functionality more comprehensively
llvm-svn: 279444
James Molloy [Mon, 22 Aug 2016 17:40:23 +0000 (17:40 +0000)]
[SimplifyCFG] Rewrite SinkThenElseCodeToEnd
The new version has several advantages:
1) IMSHO it's more readable and neater
2) It handles loads and stores properly
3) It can handle any number of incoming blocks rather than just two. I'll be taking advantage of this in a followup patch.
With this change we can now finally sink load-modify-store idioms such as:
if (a)
return *b += 3;
else
return *b += 4;
=>
%z = load i32, i32* %y
%.sink = select i1 %a, i32 5, i32 7
%b = add i32 %z, %.sink
store i32 %b, i32* %y
ret i32 %b
When this works for switches it'll be even more powerful.
Round 4. This time we should handle all instructions correctly, and not replace any operands that need to be constant with variables.
This was really hard to determine safely, so the helper function should be put into the Instruction API. I'll do that as a followup.
llvm-svn: 279443
Alexander Kornienko [Mon, 22 Aug 2016 17:19:23 +0000 (17:19 +0000)]
[clang-tidy docs] Further cleanup of options
llvm-svn: 279442
Simon Pilgrim [Mon, 22 Aug 2016 16:47:55 +0000 (16:47 +0000)]
[X86][AVX] Don't use SubVectorBroadcast if there are additional users of the chain (PR29088)
We could improve on this by making X86SubVBroadcast a full memory intrinsic similar to X86vzload
llvm-svn: 279441
Mehdi Amini [Mon, 22 Aug 2016 16:41:58 +0000 (16:41 +0000)]
Fix Gold Plugin after API change in the LTO API (constify callback type)
llvm-svn: 279440
Simon Atanasyan [Mon, 22 Aug 2016 16:18:42 +0000 (16:18 +0000)]
[mips][ias] Support .dtprel[d]word and .tprel[d]word directives
Assembler directives .dtprelword, .dtpreldword, .tprelword, and
.tpreldword generates relocations R_MIPS_TLS_DTPREL32, R_MIPS_TLS_DTPREL64,
R_MIPS_TLS_TPREL32, and R_MIPS_TLS_TPREL64 respectively.
The main motivation for this patch is to be able to write test cases
for checking correctness of the LLD linker's behaviour.
Differential Revision: https://reviews.llvm.org/D23669
llvm-svn: 279439
Mehdi Amini [Mon, 22 Aug 2016 16:17:40 +0000 (16:17 +0000)]
[LTO] Constify the Module Hook function (NFC)
It use to be non-const for the sole purpose of custom handling of
commons symbol. This is moved now in the regular LTO handling now
and such we can constify the callback.
llvm-svn: 279438
Krzysztof Parzyszek [Mon, 22 Aug 2016 14:50:12 +0000 (14:50 +0000)]
Reset isUndef when removing subreg from a def operand
llvm-svn: 279437
Martin Probst [Mon, 22 Aug 2016 14:23:30 +0000 (14:23 +0000)]
clang-format: [JS] supports casts to types starting with punctuation ("{[(").
Before:
x as{x: number}
After:
x as {x: number}
Reviewers: djasper
Subscribers: klimek, cfe-commits
Differential Revision: https://reviews.llvm.org/D23761
llvm-svn: 279436
Simon Pilgrim [Mon, 22 Aug 2016 13:18:56 +0000 (13:18 +0000)]
[X86] Only accept SM_SentinelUndef (-1) as an undefined shuffle mask in range
As discussed on D23027 we should be trying to be more strict on what is an undefined mask value.
llvm-svn: 279435
Artur Pilipenko [Mon, 22 Aug 2016 13:18:19 +0000 (13:18 +0000)]
Remove missing file from r279433 reversal
llvm-svn: 279434
Artur Pilipenko [Mon, 22 Aug 2016 13:14:07 +0000 (13:14 +0000)]
Revert -r278267 [ValueTracking] An improvement to IR ValueTracking on Non-negative Integers
This change cause performance regression on MultiSource/Benchmarks/TSVC/Symbolics-flt/Symbolics-flt from LNT and some other bechmarks.
See https://reviews.llvm.org/D18777 for details.
llvm-svn: 279433
Artur Pilipenko [Mon, 22 Aug 2016 13:12:07 +0000 (13:12 +0000)]
Revert -r278269 [IndVarSimplify] Eliminate zext of a signed IV when the IV is known to be non-negative
This change needs to be reverted in order to revert -r278267 which cause performance regression on MultiSource/Benchmarks/TSVC/Symbolics-flt/Symbolics-flt from LNT and some other bechmarks.
See comments on https://reviews.llvm.org/D18777 for details.
llvm-svn: 279432
Balaram Makam [Mon, 22 Aug 2016 12:59:58 +0000 (12:59 +0000)]
[PM] Port LoopDataPrefetch AArch64 tests to new pass manager
Reviewers: mcrosier, tejohnson
Subscribers: aemerson, rengolin, mcrosier, llvm-commits
Differential Revision: https://reviews.llvm.org/D23724
llvm-svn: 279431
Simon Pilgrim [Mon, 22 Aug 2016 12:56:54 +0000 (12:56 +0000)]
[X86][SSE] Avoid specifying unused arguments in SHUFPD lowering
As discussed on PR26491, we are missing the opportunity to make use of the smaller MOVHLPS instruction because we set both arguments of a SHUFPD when using it to lower a single input shuffle.
This patch sets the lowered argument to UNDEF if that shuffle element is undefined. This in turn makes it easier for target shuffle combining to decode UNDEF shuffle elements, allowing combines to MOVHLPS to occur.
A fix to match against MOVHPD stores was necessary as well.
This builds on the improved MOVLHPS/MOVHLPS lowering and memory folding support added in D16956
Adding similar support for SHUFPS will have to wait until have better support for target combining of binary shuffles.
Differential Revision: https://reviews.llvm.org/D23027
llvm-svn: 279430
Hrvoje Varga [Mon, 22 Aug 2016 12:17:59 +0000 (12:17 +0000)]
[mips][microMIPS] Implement BLTZC, BLEZC, BGEZC and BGTZC instructions, fix disassembly and add operand checking to existing B<cond>C implementations
Differential Revision: https://reviews.llvm.org/D22667
llvm-svn: 279429
Davide Italiano [Mon, 22 Aug 2016 11:55:22 +0000 (11:55 +0000)]
[MC] Remove guard(s). NFCI.
All the methods are already marked with
LLVM_DUMP_METHOD.
llvm-svn: 279428
Gabor Horvath [Mon, 22 Aug 2016 11:21:30 +0000 (11:21 +0000)]
Reapply "[analyzer] Added valist related checkers."
Differential Revision: https://reviews.llvm.org/D15227
llvm-svn: 279427
Simon Pilgrim [Mon, 22 Aug 2016 10:49:37 +0000 (10:49 +0000)]
[ThinLTO][X86] Fix windows build
Windows 'rm' complains about non-existent files if a wildcard is used. Be more explicit about the files deleted to avoid this.
llvm-svn: 279426
Gabor Horvath [Mon, 22 Aug 2016 10:07:32 +0000 (10:07 +0000)]
[analyzer] Correctly add assumptions based on array bounds.
Also simplify the constraints generated by the checker.
Differential Revision: https://reviews.llvm.org/D23112
llvm-svn: 279425
Craig Topper [Mon, 22 Aug 2016 07:38:50 +0000 (07:38 +0000)]
[X86] Create a new instruction format to handle 4VOp3 encoding. This saves one bit in TSFlags and simplifies MRMSrcMem/MRMSrcReg format handling.
llvm-svn: 279424
Craig Topper [Mon, 22 Aug 2016 07:38:45 +0000 (07:38 +0000)]
[X86] Create a new instruction format to handle MemOp4 encoding. This saves one bit in TSFlags and simplifies MRMSrcMem/MRMSrcReg format handling.
llvm-svn: 279423
Craig Topper [Mon, 22 Aug 2016 07:38:41 +0000 (07:38 +0000)]
[X86] Space out the encodings of X86 instruction formats. I plan to add some new encodings in future commits and this will reduce the size of those commits. NFC
This tries to keep all the ModRM memory and register forms in their own regions of the encodings. Hoping to make it simple on some of the switch statements that operate on these encodings.
llvm-svn: 279422
Craig Topper [Mon, 22 Aug 2016 07:38:36 +0000 (07:38 +0000)]
[X86] Merge small helper function into the switch that calls it since they both operate on the same variable. NFC
llvm-svn: 279421
Craig Topper [Mon, 22 Aug 2016 07:38:30 +0000 (07:38 +0000)]
[X86] Explicitly list all X86 instruction forms in switch statement so its easier to detect when one is missing. NFC
llvm-svn: 279420
Elad Cohen [Mon, 22 Aug 2016 07:34:21 +0000 (07:34 +0000)]
Test commit (Removing trailing whitespace).
llvm-svn: 279419
Mehdi Amini [Mon, 22 Aug 2016 06:37:41 +0000 (06:37 +0000)]
Add REQUIRES:X86 to test/tools/llvm-lto2/common.ll
llvm-svn: 279418
Mehdi Amini [Mon, 22 Aug 2016 06:25:46 +0000 (06:25 +0000)]
[LTO] Handles commons in monolithic LTO
The gold-plugin was doing this internally, now the API is handling
commons correctly based on the given resolution.
Differential Revision: https://reviews.llvm.org/D23739
llvm-svn: 279417
Mehdi Amini [Mon, 22 Aug 2016 06:25:41 +0000 (06:25 +0000)]
[LTO] Add a "CodeGenOnly" option. Allows the client to skip the optimizer.
Summary: Slowly getting on par with libLTO
Reviewers: tejohnson
Subscribers: llvm-commits, mehdi_amini
Differential Revision: https://reviews.llvm.org/D23615
llvm-svn: 279416
Rui Ueyama [Mon, 22 Aug 2016 04:55:41 +0000 (04:55 +0000)]
Format.
llvm-svn: 279415
Rui Ueyama [Mon, 22 Aug 2016 04:55:20 +0000 (04:55 +0000)]
Do not add .interp, .dynamic nor .eh_frame_hdr to segments just by type.
Summary:
We previously added these output sections to segments just by type.
Therefore, if there's a PHDRS command like this
PHDRS {
headers PT_PHDR PHDRS;
interp PT_INTERP;
}
SECTIONS {
. = SIZEOF_HEADERS;
.interp : { *(.interp) } :text
}
then .interp was added to "interp" segment even though the linker
is not instructed to do so by SECTIONS command. This patch removes
the default behavior to simplify.
Differential Revision: https://reviews.llvm.org/D23702
llvm-svn: 279414
Vitaly Buka [Mon, 22 Aug 2016 04:16:14 +0000 (04:16 +0000)]
[asan] Use 1 byte aligned stores to poison shadow memory
Summary: r279379 introduced crash on arm 32bit bot. I suspect this is alignment issue.
Reviewers: eugenis
Subscribers: llvm-commits, aemerson
Differential Revision: https://reviews.llvm.org/D23762
llvm-svn: 279413
Craig Topper [Mon, 22 Aug 2016 01:37:19 +0000 (01:37 +0000)]
[X86] Merge hasVEX_i8ImmReg into the ImmFormat type which had extra unused encodings. This saves one bit in TSFlags. NFC
llvm-svn: 279412
Craig Topper [Mon, 22 Aug 2016 01:37:16 +0000 (01:37 +0000)]
[X86] Remove ignoreVEX_L from TSFlags. Only the disassembler needs it and the disassembler doesn't use TSFlags. NFC
llvm-svn: 279411
NAKAMURA Takumi [Mon, 22 Aug 2016 00:58:47 +0000 (00:58 +0000)]
Reformat.
llvm-svn: 279409
NAKAMURA Takumi [Mon, 22 Aug 2016 00:58:04 +0000 (00:58 +0000)]
Untabify.
llvm-svn: 279408
Saleem Abdulrasool [Sun, 21 Aug 2016 23:05:43 +0000 (23:05 +0000)]
COFF: hoist a local variable
Create a local variable for the rdata section. NFC.
llvm-svn: 279407
Simon Pilgrim [Sun, 21 Aug 2016 19:56:02 +0000 (19:56 +0000)]
[CostModel][X86] Removed shift tests
There are more thorough tests found in vshift-*-cost.ll
llvm-svn: 279406
Simon Pilgrim [Sun, 21 Aug 2016 19:44:44 +0000 (19:44 +0000)]
[CostModel][X86] Added costs for vXi16 and vXi8 vectors for add/sub/mul/and/or/xor tests
llvm-svn: 279405
Simon Pilgrim [Sun, 21 Aug 2016 19:14:48 +0000 (19:14 +0000)]
[CostModel][X86] Replaced SSSE3 with SSE2 costs to create a better baseline
llvm-svn: 279404
Simon Pilgrim [Sun, 21 Aug 2016 19:06:25 +0000 (19:06 +0000)]
[CostModel][X86] Added fsqrt and fma costs
llvm-svn: 279403
Simon Pilgrim [Sun, 21 Aug 2016 18:34:47 +0000 (18:34 +0000)]
[CostModel][X86] Split off float arithmetic cost tests
llvm-svn: 279402
Sanjay Patel [Sun, 21 Aug 2016 17:10:07 +0000 (17:10 +0000)]
[InstCombine] use m_APInt to allow icmp (shl X, Y), C folds for splat constant vectors, part 4
This concludes the fixes for icmp+shl in this series:
https://reviews.llvm.org/rL279339
https://reviews.llvm.org/rL279398
https://reviews.llvm.org/rL279399
llvm-svn: 279401
Sanjay Patel [Sun, 21 Aug 2016 16:40:42 +0000 (16:40 +0000)]
remove FIXME comment; fixed by previous commit
llvm-svn: 279400
Sanjay Patel [Sun, 21 Aug 2016 16:35:34 +0000 (16:35 +0000)]
[InstCombine] use m_APInt to allow icmp (shl X, Y), C folds for splat constant vectors, part 3
This is a partial enablement (move the ConstantInt guard down).
llvm-svn: 279399
Sanjay Patel [Sun, 21 Aug 2016 16:28:22 +0000 (16:28 +0000)]
[InstCombine] use m_APInt to allow icmp (shl X, Y), C folds for splat constant vectors, part 2
This is a partial enablement (move the ConstantInt guard down).
llvm-svn: 279398
Simon Pilgrim [Sun, 21 Aug 2016 15:39:45 +0000 (15:39 +0000)]
[X86][AVX] Dropped combineShuffle256 - this can now be performed by EltsFromConsecutiveLoads
llvm-svn: 279397
Sanjay Patel [Sun, 21 Aug 2016 15:07:45 +0000 (15:07 +0000)]
[InstCombine] use APInt instead of ConstantInt in isSignBitCheck(); NFCI
The callers still have ConstantInt guards, so there is no functional change
intended from this change. But relaxing the callers will allow more folds
for vector types.
llvm-svn: 279396
Roman Gareev [Sun, 21 Aug 2016 11:20:39 +0000 (11:20 +0000)]
Add a flag to dump SCoP optimized with the IslScheduleOptimizer pass
Dump polyhedral descriptions of Scops optimized with the isl scheduling
optimizer and the set of post-scheduling transformations applied
on the schedule tree to be able to check the work of the IslScheduleOptimizer
pass at the polyhedral level.
Reviewed-by: Tobias Grosser <tobias@grosser.es>
Differential Revision: https://reviews.llvm.org/D23740
llvm-svn: 279395
Roman Gareev [Sun, 21 Aug 2016 11:09:19 +0000 (11:09 +0000)]
Simplify AccFuncMap to vector<> AccessFunctions
getAccessFunctions() is dead code and the 'BB' argument
of getOrCreateAccessFunctions() is not used. This patch deletes
getAccessFunctions and transforms AccFuncMap into
a std::vector<std::unique_ptr<MemoryAccess>> AccessFunctions.
Reviewed-by: Tobias Grosser <tobias@grosser.es>
Differential Revision: https://reviews.llvm.org/D23759
llvm-svn: 279394
Guy Blank [Sun, 21 Aug 2016 08:02:27 +0000 (08:02 +0000)]
[AVX512][FastISel] Do not use K registers in TEST instructions
In some cases, FastIsel was emitting TEST instruction with K reg input, which is illegal.
Changed to using KORTEST when dealing with K regs.
Differential Revision: https://reviews.llvm.org/D23163
llvm-svn: 279393
Asaf Badouh [Sun, 21 Aug 2016 07:56:47 +0000 (07:56 +0000)]
[X86][AVX512F] minor fix of the parameter names
add "__" prefix
Bug 28842 https://llvm.org/bugs/show_bug.cgi?id=29040
Differential Revision: https://reviews.llvm.org/D23753
llvm-svn: 279392
Duncan P. N. Exon Smith [Sun, 21 Aug 2016 00:08:10 +0000 (00:08 +0000)]
ARM: Avoid dereferencing end() in ARMFrameLowering::emitEpilogue
This fixes the crash from PR29072, where the MachineBasicBlock::iterator
wasn't being properly checked against MachineBasicBlock::end() before
iterating. This was another bug exposed by the new
ilist::iterator::operator*() assertion from r279314.
This testcase is poor quality. bugpoint couldn't reduce any further,
and I haven't had time to dig into what's going on so I can't invent a
better one. I didn't even get good CHECK lines in: this is just a
crasher.
I'm committing anyway since this is a real crash with an obvious fix,
but I'll leave PR29072 open and ask an ARM maintainer to help improve
the testcase.
llvm-svn: 279391
Todd Fiala [Sat, 20 Aug 2016 23:24:02 +0000 (23:24 +0000)]
Fix broken macOS LLDB Xcode build from r279314
llvm-svn: 279390
Simon Pilgrim [Sat, 20 Aug 2016 23:09:57 +0000 (23:09 +0000)]
[X86][SSE] Regenerate 32-bit buildvector test
llvm-svn: 279389
Simon Pilgrim [Sat, 20 Aug 2016 22:00:53 +0000 (22:00 +0000)]
[X86][SSE] Regenerate subvector extraction widening test
llvm-svn: 279388
Simon Pilgrim [Sat, 20 Aug 2016 21:56:33 +0000 (21:56 +0000)]
[X86] Regenerate fp truncate tests
llvm-svn: 279387
Simon Pilgrim [Sat, 20 Aug 2016 21:37:30 +0000 (21:37 +0000)]
Regenerate test
llvm-svn: 279386
Simon Pilgrim [Sat, 20 Aug 2016 21:35:45 +0000 (21:35 +0000)]
Regenerate test
llvm-svn: 279385
Simon Pilgrim [Sat, 20 Aug 2016 21:07:41 +0000 (21:07 +0000)]
[X86][XOP] Tweak vpermil2pd test to stop it being combined away
llvm-svn: 279384
Vitaly Buka [Sat, 20 Aug 2016 20:23:50 +0000 (20:23 +0000)]
[asan] Minimize code size by using __asan_set_shadow_* for large blocks
Summary:
We can insert function call instead of multiple store operation.
Current default is blocks larger than 64 bytes.
Changes are hidden behind -asan-experimental-poisoning flag.
PR27453
Differential Revision: https://reviews.llvm.org/D23711
llvm-svn: 279383