Craig Topper [Mon, 9 Apr 2018 19:17:54 +0000 (19:17 +0000)]
[X86] Emit native IR for pmuldq/pmuludq builtins.
I believe all the pieces are now in place in the backend to make this work correctly. We can either mask the input to 32 bits for pmuludg or shl/ashr for pmuldq and use a regular mul instruction. The backend should combine this to PMULUDQ/PMULDQ and then SimplifyDemandedBits will remove the and/shifts.
Differential Revision: https://reviews.llvm.org/D45421
llvm-svn: 329605
Craig Topper [Mon, 9 Apr 2018 19:17:38 +0000 (19:17 +0000)]
[X86] Remove GCCBuiltin name from pmuldq/pmuludq intrinsics so clang can custom lower to native IR. Update fast-isel intrinsic tests for clang's new codegen.
In somes cases fast-isel fails to remove the and/shifts and uses blends or conditional moves.
But once masking gets involved, fast-isel aborts on the mask portion and we DAG combine more thorougly.
llvm-svn: 329604
Alexey Bataev [Mon, 9 Apr 2018 19:02:34 +0000 (19:02 +0000)]
[SLP] Additional tests for reorder reuse vectorization, NFC.
llvm-svn: 329603
Daniel Sanders [Mon, 9 Apr 2018 18:42:19 +0000 (18:42 +0000)]
Fix type mismatch between MachineMemOperand constructor and accessors. NFC
This allows MachineMemOperand::getSize()'s result to be fed directly into
MachineMemOperand::MachineMemOperand() without a narrowing type conversion
warning.
llvm-svn: 329602
Erik Pilkington [Mon, 9 Apr 2018 18:33:01 +0000 (18:33 +0000)]
[demangler] Support for fold expressions.
llvm-svn: 329601
Erik Pilkington [Mon, 9 Apr 2018 18:32:25 +0000 (18:32 +0000)]
[demangler] Support for <data-member-prefix>.
llvm-svn: 329600
Erik Pilkington [Mon, 9 Apr 2018 18:31:50 +0000 (18:31 +0000)]
[demangler] Support for partially substituted sizeof....
llvm-svn: 329599
Rumeet Dhindsa [Mon, 9 Apr 2018 17:56:07 +0000 (17:56 +0000)]
Added support for LTO options: sample_profile, new_pass_manager and debug_pass_manager
Differential Revision: https://reviews.llvm.org/D45275
llvm-svn: 329598
Tatyana Krasnukha [Mon, 9 Apr 2018 17:32:56 +0000 (17:32 +0000)]
Fix compilation error caused by tgmath.h.
On CentOS calling functions from <tgmath.h> produces multiple errors "'void*' is not a pointer-to-object type".
llvm-svn: 329597
Aditya Nandakumar [Mon, 9 Apr 2018 17:30:56 +0000 (17:30 +0000)]
[GISel] Refactor MachineIRBuilder to allow transformations while
building.
https://reviews.llvm.org/D45067
This change attempts to do two things:
1) It separates out the state that is stored in the
MachineIRBuilder(InsertionPt, MF, MRI, InsertFunction etc) into a
separate object called MachineIRBuilderState.
2) Add the ability to constant fold operations while building instructions
(optionally). MachineIRBuilder is now refactored into a MachineIRBuilderBase
which contains lots of non foldable build methods and their implementation.
Instructions which can be constant folded/transformed are now in a class
called FoldableInstructionBuilder which uses CRTP to use the implementation
of the derived class for buildBinaryOps. Additionally buildInstr in the derived
class can be used to implement other kinds of transformations.
Also because of separation of state, given a MachineIRBuilder in an API,
if one wishes to use another MachineIRBuilder, a new one can be
constructed from the state locally. For eg,
void doFoo(MachineIRBuilder &B) {
MyCustomBuilder CustomB(B.getState());
// Use CustomB for building.
}
reviewed by : aemerson
llvm-svn: 329596
Kostya Kortchinsky [Mon, 9 Apr 2018 17:25:57 +0000 (17:25 +0000)]
[sanitizer] Allow BackgroundThread to not depend on StackDepot
Summary:
Still pursuing the ultimate goal of splitting the Symbolizer code from
RTSanitizerCommon core, allow `BackgroundThread` to work even when not linked
with `sanitizer_stackdepot.cc`. There is no reason this function should pull in
the whole `StackDepot` if symbolization is not supported.
Currently this has no functional change as the depot is always linked anyway.
Reviewers: alekseyshl
Reviewed By: alekseyshl
Subscribers: kubamracek, delcypher, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D45296
llvm-svn: 329595
Michal Gorny [Mon, 9 Apr 2018 17:08:14 +0000 (17:08 +0000)]
[cmake] Include LLVMTestingSupport when doing stand-alone build
Explicitly include and build lib/Testing/Support from LLVM sources when
doing a stand-alone build. This is necessary since clangd tests started
to depend on LLVMTestingSupport library which is neither installed
by LLVM, nor built by clang itself.
Since completely separate build of clang-tools-extra is not supported,
this relies on variables set by clang CMakeLists.
Differential Revision: https://reviews.llvm.org/D45409
llvm-svn: 329594
Craig Topper [Mon, 9 Apr 2018 17:07:40 +0000 (17:07 +0000)]
[X86] Revert the SLM part of r328914.
While it appears to be correct information based on Intel's optimization manual and Agner's data, it causes perf regressions on a couple of the benchmarks in our internal list.
llvm-svn: 329593
Fangrui Song [Mon, 9 Apr 2018 17:06:57 +0000 (17:06 +0000)]
[llvm-mca] Fix MCACommentConsumer
llvm-svn: 329592
Marek Olsak [Mon, 9 Apr 2018 16:56:32 +0000 (16:56 +0000)]
AMDGPU: enable 128-bit for local addr space under an option
Author: Samuel Pitoiset
ds_read_b128 and ds_write_b128 have been recently enabled
under the amdgpu-ds128 option because the performance benefit
is unclear.
Though, using 128-bit loads/stores for the local address space
appears to introduce regressions in tessellation shaders. Not
sure what is broken, but as ds_read_b128/ds_write_b128 are not
enabled by default, just introduce a global option and enable
128-bit only if requested (until it's fixed/used correctly).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105464
llvm-svn: 329591
Andrea Di Biagio [Mon, 9 Apr 2018 16:39:52 +0000 (16:39 +0000)]
[llvm-mca] Add the ability to mark regions of code for analysis (PR36875)
This patch teaches llvm-mca how to parse code comments in search for special
"markers" used to select regions of code.
Example:
# LLVM-MCA-BEGIN My Code Region
....
# LLVM-MCA-END
The MCAsmLexer now delegates to an object of class MCACommentParser (i.e. an
AsmCommentConsumer) the parsing of code comments to search for begin/end code
region markers.
A comment starting with substring "LLVM-MCA-BEGIN" marks the beginning of a new
region of code. A comment starting with substring "LLVM-MCA-END" marks the end
of the last region.
This implementation doesn't allow regions to overlap. Each region can have a
optional description; internally, each region is identified by a range of source
code locations (SMLoc).
MCInst objects are added to a region R only if the source location for the
MCInst is in the range of locations specified by R.
By default, the tool allocates an implicit "Default" code region which contains
every source location. See new tests llvm-mca-marker-*.s for a few examples.
A new Backend object is created for every region. So, the analysis is conducted
on every parsed code region. The final report is the union of the reports
generated for every code region. Note that empty regions are skipped.
Special "[#] Code Region - ..." strings are used in the report to mark the
portion which is specific to a code region only. For example, see
llvm-mca-markers-5.s.
Differential Revision: https://reviews.llvm.org/D45433
llvm-svn: 329590
Tom Stellard [Mon, 9 Apr 2018 16:09:13 +0000 (16:09 +0000)]
AMDGPU: Initialize GlobalISel passes
Summary:
This fixes AMDGPU GlobalISel test failures when enabling the AMDGPU
target without any other targets that use GlobalISel.
Reviewers: arsenm
Subscribers: kzhuravl, wdng, nhaehnle, yaxunl, rovka, kristof.beyls, dstuttard, tpr, llvm-commits, t-tye
Differential Revision: https://reviews.llvm.org/D45353
llvm-svn: 329588
Simon Pilgrim [Mon, 9 Apr 2018 16:01:44 +0000 (16:01 +0000)]
[X86][SSE] Add floating point add/mul strict (ordered) vector.reduce tests (PR36732)
llvm-svn: 329587
Alex Shlyapnikov [Mon, 9 Apr 2018 16:00:10 +0000 (16:00 +0000)]
[ASan] NFC: make use of a new ErrorBase ctor
Summary:
Minor style changes to complement D44404:
- make use of a new ErrorBase ctor
- de-duplicate a comment about VS2013 support
Reviewers: eugenis
Subscribers: kubamracek, delcypher, llvm-commits, #sanitizers
Differential Revision: https://reviews.llvm.org/D45390
llvm-svn: 329586
Simon Pilgrim [Mon, 9 Apr 2018 15:44:20 +0000 (15:44 +0000)]
Support generic expansion of ordered vector reduction (PR36732)
Without the fast math flags, the llvm.experimental.vector.reduce.fadd/fmul intrinsic expansions must be expanded in order.
This patch scalarizes the reduction, applying the accumulator at the start of the sequence: ((((Acc + Scl[0]) + Scl[1]) + Scl[2]) + ) ... + Scl[NumElts-1]
Differential Revision: https://reviews.llvm.org/D45366
llvm-svn: 329585
Yaxun Liu [Mon, 9 Apr 2018 15:43:01 +0000 (15:43 +0000)]
[CUDA] Revert defining __CUDA_ARCH__ for amdgcn targets
amdgcn targets only support HIP, which does not define __CUDA_ARCH__.
this is a partial unroll of r329232 / D45277.
Differential Revision: https://reviews.llvm.org/D45387
llvm-svn: 329584
Haojian Wu [Mon, 9 Apr 2018 15:37:09 +0000 (15:37 +0000)]
[clangd] Bump v0.0.6 for vscode-clangd.
llvm-svn: 329583
Sam McCall [Mon, 9 Apr 2018 15:22:08 +0000 (15:22 +0000)]
[clang] Use compile-command interpolation to provide commands for header files.
Summary: This uses the inferring wrapper introduced in D45006.
Subscribers: klimek, ilya-biryukov, jkorous-apple, ioeric, cfe-commits
Differential Revision: https://reviews.llvm.org/D45007
llvm-svn: 329582
Max Moroz [Mon, 9 Apr 2018 15:20:35 +0000 (15:20 +0000)]
[llvm-cov] Implement -ignore-filename-regex= option for excluding source files.
Summary:
The option is helpful for large projects where it's not feasible to specify sources which
user would like to see in the report. Instead, it allows to black-list specific sources via
regular expressions (e.g. now it's possible to skip all files that have "test" in its name).
This also partially fixes https://bugs.llvm.org/show_bug.cgi?id=34277
Reviewers: vsk, morehouse, liaoyuke
Reviewed By: vsk
Subscribers: kcc, mgorny, llvm-commits
Differential Revision: https://reviews.llvm.org/D43907
llvm-svn: 329581
Sam McCall [Mon, 9 Apr 2018 15:17:39 +0000 (15:17 +0000)]
[Tooling] A CompilationDatabase wrapper that infers header commands.
Summary:
The wrapper finds the closest matching compile command using filename heuristics
and makes minimal tweaks so it can be used with the header.
Subscribers: klimek, mgorny, cfe-commits
Differential Revision: https://reviews.llvm.org/D45006
llvm-svn: 329580
Alexander Kornienko [Mon, 9 Apr 2018 15:12:10 +0000 (15:12 +0000)]
[clang-tidy] Return non-zero exit code for clang errors.
Summary:
Updated tests broken by this change.
Fixes https://bugs.llvm.org/show_bug.cgi?id=27628
Reviewers: ilya-biryukov
Reviewed By: ilya-biryukov
Subscribers: klimek, xazax.hun, cfe-commits
Differential Revision: https://reviews.llvm.org/D45258
llvm-svn: 329579
Haojian Wu [Mon, 9 Apr 2018 15:09:44 +0000 (15:09 +0000)]
[clangd] Allow using customized include path in URI.
Summary:
Calculating the include path from absolute file path does not always
work for all build system, e.g. bazel uses symlink as the build working
directory. The absolute file path from editor and clang is diverged from
each other. We need to address it properly in build sysmtem integration.
This patch worksarounds the issue by providing a hook in URI which allows
clients to provide their customized include path.
Reviewers: sammccall
Subscribers: klimek, ilya-biryukov, jkorous-apple, ioeric, MaskRay, cfe-commits
Differential Revision: https://reviews.llvm.org/D45426
llvm-svn: 329578
Zaara Syeda [Mon, 9 Apr 2018 14:50:02 +0000 (14:50 +0000)]
[MachineLICM] Re-enable hoisting of constant stores
This patch fixes an issue exposed on the SystemZ build bots when committing
https://reviews.llvm.org/rL327856. The hoisting was temporarily disabled with
an option. This patch now re-enables hoisting and checks that we only hoist a
store instruction when all its operands are either constant caller preserved
registers or immediates.
Differential Revision: https://reviews.llvm.org/D45286
llvm-svn: 329577
Guansong Zhang [Mon, 9 Apr 2018 14:45:41 +0000 (14:45 +0000)]
Revert "[OpenMP] enable bc file compilation using the latest clang"
This reverts commit
6849e31c36d712d97433bca9af39b7a09c8c1207.
llvm-svn: 329576
Pavel Labath [Mon, 9 Apr 2018 14:38:53 +0000 (14:38 +0000)]
[CodeGen/AccelTable] Don't emit zero-CU name indexes
Summary:
If an input DICompileUnit is completely empty (e.g., the result of
running "clang -g" on an empty file), we don't bother emitting an empty
DWARF CU. When we do that, we must make sure we don't also emit a DWARF v5
name index, as DWARF specifies that each index must reference at least
one compilation unit.
Reviewers: JDevlieghere, aprantl, dblaikie
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D45435
llvm-svn: 329575
Marc-Andre Laperle [Mon, 9 Apr 2018 14:32:12 +0000 (14:32 +0000)]
[clangd-vscode] Update VScode dependencies
Summary:
This allows the extension to work with LSP 3.0 and is useful for testing.
Signed-off-by: Marc-Andre Laperle <marc-andre.laperle@ericsson.com>
Reviewers: ilya-biryukov
Subscribers: hokein, klimek, ilya-biryukov, jkorous-apple, ioeric, MaskRay, cfe-commits
Differential Revision: https://reviews.llvm.org/D45285
llvm-svn: 329574
Krasimir Georgiev [Mon, 9 Apr 2018 14:29:23 +0000 (14:29 +0000)]
[RuntimeDyld][PowerPC] Fix a newly added test in r329355
Summary: The bit widths are wrong.
Reviewers: bkramer, lhames, hans
Reviewed By: hans
Subscribers: hans, nemanjai, kbarton, llvm-commits
Differential Revision: https://reviews.llvm.org/D45361
llvm-svn: 329573
Xin Tong [Mon, 9 Apr 2018 14:29:13 +0000 (14:29 +0000)]
[MergeICmp] Update debug msg.NFC
llvm-svn: 329572
Sam McCall [Mon, 9 Apr 2018 14:28:52 +0000 (14:28 +0000)]
[clangd] Adapt index interfaces to D45014, and fix the old bugs.
Summary:
Fix bugs:
- don't count occurrences of decls where we don't spell the name
- findDefinitions at MACRO(^X) goes to the definition of MACRO
Subscribers: klimek, ilya-biryukov, jkorous-apple, ioeric, MaskRay, cfe-commits
Differential Revision: https://reviews.llvm.org/D45356
llvm-svn: 329571
Sam McCall [Mon, 9 Apr 2018 14:12:51 +0000 (14:12 +0000)]
[Index] Return SourceLocation to consumers, not FileID/Offset pair.
Summary:
The FileID/Offset conversion is lossy. The code takes the fileLoc, which loses
e.g. the spelling location in some macro cases.
Instead, pass the original SourceLocation which preserves all information, and
update consumers to match current behavior.
This allows us to fix two bugs in clangd that need the spelling location.
Reviewers: akyrtzi, arphaman
Subscribers: ilya-biryukov, ioeric, cfe-commits
Differential Revision: https://reviews.llvm.org/D45014
llvm-svn: 329570
Hans Wennborg [Mon, 9 Apr 2018 13:53:41 +0000 (13:53 +0000)]
Revert r329403 "[llvm-mca] Do not separate iterations with a newline in the timeline view."
This made AArch64/CortexA57/direct-branch.s fail on Windows, e.g.
http://lab.llvm.org:8011/builders/clang-x86-windows-msvc2015/builds/11251
> Also, update a few tests to minimize the diff in D45369.
> No functional change intended.
llvm-svn: 329569
Simon Pilgrim [Mon, 9 Apr 2018 13:52:33 +0000 (13:52 +0000)]
[X86][MMX] Fix missing itinerary for PALIGNR
llvm-svn: 329568
Simon Pilgrim [Mon, 9 Apr 2018 13:42:14 +0000 (13:42 +0000)]
[X86][MMX] Fix missing itinerary for MOVQ2DQ instruction format
llvm-svn: 329567
Philipp Stephani [Mon, 9 Apr 2018 13:31:44 +0000 (13:31 +0000)]
Improve completion experience for headers
Summary: When calling `completing-read', we should provide a default to prevent the behavior described in https://github.com/DarwinAwardWinner/ido-completing-read-plus#why-does-ret-sometimes-not-select-the-first-completion-on-the-list--why-is-there-an-empty-entry-at-the-beginning-of-the-completion-list--what-happened-to-old-style-default-selection. Also, don't use an assertion to check whether the user selected a header; raise a proper signal instead.
Reviewers: klimek
Reviewed By: klimek
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D43969
llvm-svn: 329566
Simon Pilgrim [Mon, 9 Apr 2018 13:27:47 +0000 (13:27 +0000)]
[X86][MMX] Fix missing itinerary for CVTPI2PS
llvm-svn: 329565
Xin Tong [Mon, 9 Apr 2018 13:14:06 +0000 (13:14 +0000)]
[MergeICmp] Split blocks that do other work.
Summary:
We do not try to move the instructions and split the block till we
know the blocks can be split, i.e. BCE-cmp-insts can be separated from
non-BCE-cmp-insts.
Reviewers: davide, courbet
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D44443
llvm-svn: 329564
George Rimar [Mon, 9 Apr 2018 13:13:36 +0000 (13:13 +0000)]
[ELF] - Simplify test case. NFC.
llvm-svn: 329563
Dmitry Preobrazhensky [Mon, 9 Apr 2018 13:10:33 +0000 (13:10 +0000)]
[AMDGPU][MC][GFX9] Added instructions s_mul_hi_*32, s_lshl*_add_u32
See bugs
36841: https://bugs.llvm.org/show_bug.cgi?id=36841
36842: https://bugs.llvm.org/show_bug.cgi?id=36842
Differential Revision: https://reviews.llvm.org/D45251
Reviewers: artem.tamazov, arsenm, timcorringham
llvm-svn: 329562
Simon Pilgrim [Mon, 9 Apr 2018 13:02:07 +0000 (13:02 +0000)]
[X86][MMX] Fix flipped reg/mem typo in MMX_MISC_FUNC_ITINS
The RR/RM itineraries were the wrong way around
llvm-svn: 329561
George Rimar [Mon, 9 Apr 2018 13:01:50 +0000 (13:01 +0000)]
[ELF] - Stop setting OutSecOff too early.
Currently LLD sets OutSecOff in addSection for input sections.
That is a fake offset (just a rude approximation to remember the order),
used for sorting SHF_LINK_ORDER sections
(see resolveShfLinkOrder, compareByFilePosition).
There are 2 problems with such approach:
1. We currently change and reuse Size field as a value assigned. Changing size is
not good because leads to bugs. Currently, SIZEOF(.bss) for empty .bss returns 2
because we add two empty synthetic sections and increase size twice by 1.
(See PR37011: https://bugs.llvm.org/show_bug.cgi?id=37011)
2. Such approach simply does not work when --symbol-ordering-file is involved,
because processing of the ordering file might break the initial section order.
This fixes PR37011.
Differential revision: https://reviews.llvm.org/D45368
llvm-svn: 329560
George Rimar [Mon, 9 Apr 2018 12:45:29 +0000 (12:45 +0000)]
[ELF] - Fix cref.s test case.
The intention of -gc-sections flag was to check
that discarded is not in the output. It should be
specified in the executable command line invocation
and also, the symbol must be global as local symbols
are anyways not printed.
Differential revision: https://reviews.llvm.org/D45159
llvm-svn: 329559
Hans Wennborg [Mon, 9 Apr 2018 12:21:12 +0000 (12:21 +0000)]
Try to fix libclang reproducer tests after r329465
They were failing on Windows because the output YAML didn't parse:
YAML:1:664: error: Unrecognized escape code!
{"toolchain":"D:\\buildslave\\clang-x64-ninja-win7\\stage1",
"libclang.operation":"complete", "libclang.opts":1, "args":["clang",
"-fno-spell-checking",
"D:\buildslave\clang-x64-ninja-win7\llvm\tools\clang\test\Index\create-libclang-completion-reproducer.c",
"-Xclang", "-detailed-preprocessing-record",
"-fallow-editor-placeholders"],
"invocation-args":["-code-completion-at=D:\buildslave\clang-x64-ninja-win7\llvm\tools\clang\test\Index\create-libclang-completion-reproducer.c:10:1"],
"unsaved_file_hashes":[{"name":"D:\\buildslave\\clang-x64-ninja-win7\\llvm\\tools\\clang\\test\\Index\\create-libclang-completion-reproducer.c",
"md5":"
aee23773de90e665992b48209351d70e"}]}
This adds some more escaping to try to make it work.
llvm-svn: 329558
George Rimar [Mon, 9 Apr 2018 11:43:52 +0000 (11:43 +0000)]
[ELF] - Allow LLD to produce file symbols.
This is for PR36716 and
this enables emitting STT_FILE symbols.
Output size affect is minor:
lld binary size changes from 52,883,408 to 52,949,400
clang binary size changes from 83,136,456 to 83,219,600
Differential revision: https://reviews.llvm.org/D45261
llvm-svn: 329557
Simon Pilgrim [Mon, 9 Apr 2018 10:45:53 +0000 (10:45 +0000)]
[X86][SSE] Fix f32 mul/div itinerary groups typo
The RM folded itineraries were incorrectly using the f64 version.
llvm-svn: 329556
Pavel Labath [Mon, 9 Apr 2018 09:11:40 +0000 (09:11 +0000)]
Make the test case from r329552 more portable
- when tuning for SCE debugger (default for ps4 targets), we will not emit
the DW_AT_linkage_name, which this test needs. I explicitly set the
debugger tuning parameter to get the attribute always.
- darwin targets did not like the "section .text.startup" fragment of
the test. This is not actually needed for the test, so I remove it.
llvm-svn: 329555
Jonas Devlieghere [Mon, 9 Apr 2018 09:10:34 +0000 (09:10 +0000)]
[dsymutil] Remove trailing colon. NFC
llvm-svn: 329554
Jonas Devlieghere [Mon, 9 Apr 2018 09:09:59 +0000 (09:09 +0000)]
[dsymutil] Don't try to load Swift ASTs as objects.
With the threading refactoring, loading of object files happens before
checking whether we're dealing with a swift AST. While that's not an
issue per se, it causes a warning to be printed:
warning: /path/to/a.swiftmodule: The file was not recognized as a valid object file
note: while processing /path/to/a.swiftmodule
This suppresses the warning by checking for a Swift AST before
attempting to load is as an object file.
rdar://
39240444
llvm-svn: 329553
Pavel Labath [Mon, 9 Apr 2018 08:41:57 +0000 (08:41 +0000)]
[CodeGen/AccelTable]: Don't emit accelerator entries for functions with no names
Summary:
We were emitting accelerator entries for functions with no name, which
is contrary to the DWARF v5 spec: "All other (i.e., *not*
DW_TAG_namespace) debugging information entries without a DW_AT_name
attribute are excluded." Besides that, a name table entry with an empty
string as a key is fairly useless.
We can sometimes end up with functions which have a DW_AT_linkage_name but no
DW_AT_name. One such example is the global-constructor-initialization functions,
which C++ compilers synthesize for each compilation unit with global
constructors.
A very strict reading of the DWARF v5 spec would suggest that we should not even
emit the accelerator entry for the linkage name in this case, but I don't think
we should go that far.
I found this when running the dwarf verifier over llvm codebase compiled
with DWARF v5 accelerator tables.
Reviewers: JDevlieghere, aprantl, dblaikie
Subscribers: vleschuk, clayborg, echristo, probinson, llvm-commits
Differential Revision: https://reviews.llvm.org/D45367
llvm-svn: 329552
Sam Parker [Mon, 9 Apr 2018 08:16:11 +0000 (08:16 +0000)]
[DAGCombine] Improve ReduceLoad for SRL
Recommitting r329283, third time lucky...
If the SRL node is only used by an AND, we may be able to set the
ExtVT to the width of the mask, making the AND redundant. To support
this, another check has been added in isLegalNarrowLoad which queries
whether the load is valid.
Differential Revision: https://reviews.llvm.org/D41350
llvm-svn: 329551
Chandler Carruth [Mon, 9 Apr 2018 07:26:42 +0000 (07:26 +0000)]
Fix unused variable warning.
llvm-svn: 329550
Craig Topper [Mon, 9 Apr 2018 06:15:09 +0000 (06:15 +0000)]
[X86] Merge some of the autoupgrade handling for masked intrinsics that just need to upgrade to an unmasked version plus a select. NFCI
These are were previously grouped in small groups of similarish intrinsics. But all the intrinsics have the same number of arguments and the same order. So we can move them all into a larger group for handling.
llvm-svn: 329549
Michael Kruse [Mon, 9 Apr 2018 06:07:44 +0000 (06:07 +0000)]
Remove immediate dominator heuristic for error block detection.
This patch removes the heuristic in
- Polly :: lib/Support/ScopHelper.cpp
The heuristic forces blocks that directly follow a loop header to not to be considered error blocks.
It was introduced in r249611 with the following commit message:
> This replaces the support for user defined error functions by a
> heuristic that tries to determine if a call to a non-pure function
> should be considered "an error". If so the block is assumed not to be
> executed at runtime. While treating all non-pure function calls as
> errors will allow a lot more regions to be analyzed, it will also
> cause us to dismiss a lot again due to an infeasible runtime context.
> This patch tries to limit that effect. A non-pure function call is
> considered an error if it is executed only in conditionally with
> regards to a cheap but simple heuristic.
In the code below `CCK_Abort2()` would be considered as an error block, but not `CCK_Abort1()` due to this heuristic.
```
for (int i = 0; i < n; i+=1) {
if (ErrorCondition1)
CCK_Abort1(); // No __attribute__((noreturn))
if (ErrorCondition2)
CCK_Abort2(); // No __attribute__((noreturn))
}
```
This does not seem useful. Checking error conditions in the beginning of some work is quite common. It causes a switch default-case to be not considered an error block in SPEC's cactuBSSN. The comment justifying the heuristic mentions a "load", which does not seem to be applicable here. It has been proposed to remove the heuristic.
In addition, the patch fixes the following test cases:
- Polly :: ScopDetect/mod_ref_read_pointer.ll
- Polly :: ScopInfo/max-loop-depth.ll
- Polly :: ScopInfo/mod_ref_access_pointee_arguments.ll
- Polly :: ScopInfo/mod_ref_read_pointee_arguments.ll
- Polly :: ScopInfo/mod_ref_read_pointer.ll
- Polly :: ScopInfo/mod_ref_read_pointers.ll
The test cases failed after removing the heuristic.
Differential Revision: https://reviews.llvm.org/D45274
Contributed-by: Lorenzo Chelini <l.chelini@icloud.com>
llvm-svn: 329548
Max Kazantsev [Mon, 9 Apr 2018 06:01:22 +0000 (06:01 +0000)]
[IRCE] Relax restriction on collected range checks
In IRCE, we have a very old legacy check that works when we collect comparisons that we
treat as range checks. It ensures that the value against which the indvar is compared is
loop invariant and is also positive.
This latter condition remained there since the times when IRCE was only able to handle
signed latch comparison. As the optimization evolved, it now learned how to intersect
signed or unsigned ranges, and this logic has no reliance on the fact that the right border
of each range should be positive.
The old implementation of this non-negativity check was also naive enough and just looked
into ranges (while most of other IRCE logic tries to use power of SCEV implications), so this
check did not allow to deal with the most simple case that looks like follows:
int size; // not known non-negative
int length; //known non-negative;
i = 0;
if (size != 0) {
do {
range_check(i < size);
range_check(i < length);
++i;
} while (i < size)
}
In this case, even if from some dominating conditions IRCE could parse loop
structure, it could only remove the range check against `length` and simply
ignored the check against `size`.
In this patch we remove this obsolete check. It will allow IRCE to pick comparison
against `size` as a potential range check and then let Range Intersection logic
decide whether it is OK to eliminate it or not.
Differential Revision: https://reviews.llvm.org/D45362
Reviewed By: samparker
llvm-svn: 329547
Hiroshi Inoue [Mon, 9 Apr 2018 04:37:53 +0000 (04:37 +0000)]
[NFC] fix trivial typos in comments and error message
"is is" -> "is", "are are" -> "are"
llvm-svn: 329546
Petr Hosek [Mon, 9 Apr 2018 04:36:04 +0000 (04:36 +0000)]
Revert "[CMake] Use custom command and target to install libc++ headers"
This reverts commit r329544 which is failing on libcxx standalone bots.
llvm-svn: 329545
Petr Hosek [Mon, 9 Apr 2018 04:23:04 +0000 (04:23 +0000)]
[CMake] Use custom command and target to install libc++ headers
Using file(COPY FILE...) has several downsides. Since the file command
is only executed at configuration time, any changes to headers made
after the initial CMake execution are ignored. This can lead to subtle
errors since the just built Clang will be using stale libc++ headers.
Furthermore, since the headers are copied prior to executing the build
system, this may hide missing dependencies on libc++ from other LLVM
components.
This changes replaces the use of file(COPY FILE...) command with a
custom command and target which addresses all aforementioned issues and
matches the implementation already used by other LLVM components that
also install headers like Clang builtin headers.
Differential Revision: https://reviews.llvm.org/D44773
llvm-svn: 329544
Dean Michael Berris [Mon, 9 Apr 2018 04:02:09 +0000 (04:02 +0000)]
[XRay][llvm+clang] Consolidate attribute list files
Summary:
This change consolidates the always/never lists that may be provided to
clang to externally control which functions should be XRay instrumented
by imbuing attributes. The files follow the same format as defined in
https://clang.llvm.org/docs/SanitizerSpecialCaseList.html for the
sanitizer blacklist.
We also deprecate the existing `-fxray-instrument-always=` and
`-fxray-instrument-never=` flags, in favour of `-fxray-attr-list=`.
This fixes http://llvm.org/PR34721.
Reviewers: echristo, vlad.tsyrklevich, eugenis
Reviewed By: vlad.tsyrklevich
Subscribers: llvm-commits, cfe-commits
Differential Revision: https://reviews.llvm.org/D45357
llvm-svn: 329543
Michael Zolotukhin [Mon, 9 Apr 2018 00:54:47 +0000 (00:54 +0000)]
Remove MachineLoopInfo dependency from AsmPrinter.
Summary:
Currently MachineLoopInfo is used in only two places:
1) for computing IsBasicBlockInsideInnermostLoop field of MCCodePaddingContext, and it is never used.
2) in emitBasicBlockLoopComments, which is called only if `isVerbose()` is true.
Despite that, we currently have a dependency on MachineLoopInfo, which makes
pass manager to compute it and MachineDominator Tree. This patch removes the
use (1) and makes the use (2) lazy, thus avoiding some redundant
recomputations.
Reviewers: opaparo, gadi.haber, rafael, craig.topper, zvi
Subscribers: rengolin, javed.absar, hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D44812
llvm-svn: 329542
Eric Fiselier [Sun, 8 Apr 2018 21:57:35 +0000 (21:57 +0000)]
[test] Fix Container::insert(value_type const&) tests
Patch from Joe Loser.
Several unit tests meaning to test the behavior of lvalue insertion incorrectly
pass rvalues. Fixes bug PR # 27394
Reviewed as https://reviews.llvm.org/D44411
llvm-svn: 329541
Sanjay Patel [Sun, 8 Apr 2018 19:56:04 +0000 (19:56 +0000)]
[TargetSchedule] shrink interface for init(); NFCI
The TargetSchedModel is always initialized using the TargetSubtargetInfo's
MCSchedModel and TargetInstrInfo, so we don't need to extract those and
pass 3 parameters to init().
Differential Revision: https://reviews.llvm.org/D44789
llvm-svn: 329540
Craig Topper [Sun, 8 Apr 2018 17:53:18 +0000 (17:53 +0000)]
[X86] Add SchedWrites for CMOV and SETCC. Use them to remove InstRWs.
Summary:
Cmov and setcc previously used WriteALU, but on Intel processors at least they are more restricted than basic ALU ops.
This patch adds new SchedWrites for them and removes the InstRWs. I had to leave some InstRWs for CMOVA/CMOVBE and SETA/SETBE because those have an extra uop relative to the other condition codes on Intel CPUs.
The test changes are due to fixing a missing ZnAGU dependency on the memory form of setcc.
Reviewers: RKSimon, andreadb, GGanesh
Reviewed By: RKSimon
Subscribers: GGanesh, llvm-commits
Differential Revision: https://reviews.llvm.org/D45380
llvm-svn: 329539
Craig Topper [Sun, 8 Apr 2018 17:53:15 +0000 (17:53 +0000)]
[X86][Znver1] Remove InstRWs for BLENDVPS/PD
Summary:
This removes the InstRWs for BLENDVPS/PD in favor of WriteFVarBlend. The latency listed was 3 cycles but WriteFVarBlend is defined as 1 cycle latency. The 1 cycle latency matches Agner Fog's data.
The patterns were missing the VEX forms which is why there are no test changes. We don't test "-mcpu=znver1 -mattr=-avx"
Reviewers: RKSimon, GGanesh
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D44841
llvm-svn: 329538
Jonas Devlieghere [Sun, 8 Apr 2018 17:35:17 +0000 (17:35 +0000)]
[dsymutil] Don't crash on empty CU
Add some additional checks so we don't crash on empty compile units.
llvm-svn: 329537
Mandeep Singh Grang [Sun, 8 Apr 2018 16:46:22 +0000 (16:46 +0000)]
[Support] Change std::sort to llvm::sort in response to r327219
Summary:
r327219 added wrappers to std::sort which randomly shuffle the container before sorting.
This will help in uncovering non-determinism caused due to undefined sorting
order of objects having the same key.
To make use of that infrastructure we need to invoke llvm::sort instead of std::sort.
Note: This patch is one of a series of patches to replace *all* std::sort to llvm::sort.
Refer the comments section in D44363 for a list of all the required patches.
Reviewers: chandlerc, jordan_rose, bkramer
Reviewed By: bkramer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D45140
llvm-svn: 329536
Mandeep Singh Grang [Sun, 8 Apr 2018 16:45:04 +0000 (16:45 +0000)]
[PowerPC] Change std::sort to llvm::sort in response to r327219
Summary:
r327219 added wrappers to std::sort which randomly shuffle the container before sorting.
This will help in uncovering non-determinism caused due to undefined sorting
order of objects having the same key.
To make use of that infrastructure we need to invoke llvm::sort instead of std::sort.
Note: This patch is one of a series of patches to replace *all* std::sort to llvm::sort.
Refer the comments section in D44363 for a list of all the required patches.
Reviewers: hfinkel, RKSimon
Reviewed By: RKSimon
Subscribers: nemanjai, kbarton, llvm-commits
Differential Revision: https://reviews.llvm.org/D44870
llvm-svn: 329535
Mandeep Singh Grang [Sun, 8 Apr 2018 16:42:52 +0000 (16:42 +0000)]
[X86] Change std::sort to llvm::sort in response to r327219
Summary:
r327219 added wrappers to std::sort which randomly shuffle the container before sorting.
This will help in uncovering non-determinism caused due to undefined sorting
order of objects having the same key.
To make use of that infrastructure we need to invoke llvm::sort instead of std::sort.
Note: This patch is one of a series of patches to replace *all* std::sort to llvm::sort.
Refer the comments section in D44363 for a list of all the required patches.
Reviewers: chandlerc, craig.topper, RKSimon
Reviewed By: chandlerc, craig.topper
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D44874
llvm-svn: 329534
Piotr Padlewski [Sun, 8 Apr 2018 16:04:09 +0000 (16:04 +0000)]
NFC: Update NewGVN invariant.group test
llvm-svn: 329533
Andrea Di Biagio [Sun, 8 Apr 2018 15:10:19 +0000 (15:10 +0000)]
[llvm-mca] Simplify code. NFC
llvm-svn: 329532
Piotr Padlewski [Sun, 8 Apr 2018 13:53:04 +0000 (13:53 +0000)]
Mark invariant.group as experimental
Differential Revision: https://reviews.llvm.org/D33235
llvm-svn: 329531
Xin Tong [Sun, 8 Apr 2018 13:19:53 +0000 (13:19 +0000)]
[LIR] Reorder header. NFC
llvm-svn: 329530
Simon Pilgrim [Sun, 8 Apr 2018 12:31:52 +0000 (12:31 +0000)]
[X86] Regenerate and + immediate mask tests
Added i686 checks
llvm-svn: 329529
Simon Pilgrim [Sun, 8 Apr 2018 12:30:30 +0000 (12:30 +0000)]
[X86][PKU] Regenerate rdpkru/wrpkru intrinsic tests
Added i686 checks
llvm-svn: 329528
Simon Pilgrim [Sun, 8 Apr 2018 12:29:11 +0000 (12:29 +0000)]
[X86][SSE3] Regenerate mwait/monitor intrinsic tests
Added i686 checks
llvm-svn: 329527
Piotr Padlewski [Sun, 8 Apr 2018 12:23:58 +0000 (12:23 +0000)]
NFC: delete ValueMap move ctor
llvm-svn: 329526
Zvi Rackover [Sun, 8 Apr 2018 11:35:20 +0000 (11:35 +0000)]
DAGCombiner: Combine SDIV with non-splat vector pow2 divisor
Summary:
Extend existing SDIV combine for pow2 constant divider to handle
non-splat vectors of pow2 constants.
Reviewers: RKSimon, craig.topper, spatel, hfinkel, efriedma
Reviewed By: RKSimon
Subscribers: magabari, llvm-commits
Differential Revision: https://reviews.llvm.org/D42479
llvm-svn: 329525
Simon Pilgrim [Sun, 8 Apr 2018 11:26:26 +0000 (11:26 +0000)]
[X86][Btver2] Add vector extract costs
llvm-svn: 329524
Eric Fiselier [Sun, 8 Apr 2018 08:48:58 +0000 (08:48 +0000)]
[ADT] Fix MapVector when 'Map::mapped_type != unsigned'.
Previously MapVector assumed `Map::mapped_type` was `unsigned`.
This caused problems when using MapVector with a user-specified
map where this didn't hold (For example StringMap<unsigned>).
This patch adjusts MapVector to use the same type as the underlying
map, avoiding reference binding errors in functions like `insert`.
llvm-svn: 329523
Michal Gorny [Sun, 8 Apr 2018 06:49:17 +0000 (06:49 +0000)]
[LLVMTestingSupport] Add explicit linkage to LLVMSupport
Explicitly link LLVMTestingSupport library against LLVMSupport. This
is necessary to fix linking errors when LLVMTestingSupport is built
as a shared library (with BUILD_SHARED_LIBS=ON) and -Wl,-z,defs is used.
Differential Revision: https://reviews.llvm.org/D45408
llvm-svn: 329522
Eric Fiselier [Sun, 8 Apr 2018 06:21:33 +0000 (06:21 +0000)]
[Sema] Fix PR35832 - Ambiguity accessing anonymous struct/union with multiple bases.
Summary:
Currently clang doesn't do qualified lookup when building indirect field decl references. This causes ambiguity when the field is in a base class to which there are multiple valid paths even though a qualified name is used.
For example:
```
class B {
protected:
int i;
union { int j; };
};
class X : public B { };
class Y : public B { };
class Z : public X, public Y {
int a() { return X::i; } // works
int b() { return X::j; } // fails
};
```
Reviewers: rsmith, aaron.ballman, rjmccall
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D45411
llvm-svn: 329521
Eric Fiselier [Sun, 8 Apr 2018 06:05:33 +0000 (06:05 +0000)]
Revert "[Sema] Fix PR35832 - Ambiguity accessing anonymous struct/union with multiple bases."
This reverts commit r329519. There are some unaddressed test failures.
llvm-svn: 329520
Eric Fiselier [Sun, 8 Apr 2018 05:50:01 +0000 (05:50 +0000)]
[Sema] Fix PR35832 - Ambiguity accessing anonymous struct/union with multiple bases.
Summary:
Currently clang doesn't do qualified lookup when building indirect field decl references. This causes ambiguity when the field is in a base class to which there are multiple valid paths even though a qualified name is used.
For example:
```
class B {
protected:
int i;
union { int j; };
};
class X : public B { };
class Y : public B { };
class Z : public X, public Y {
int a() { return X::i; } // works
int b() { return X::j; } // fails
};
```
Reviewers: rsmith, aaron.ballman, rjmccall
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D45411
llvm-svn: 329519
Eric Fiselier [Sun, 8 Apr 2018 05:12:55 +0000 (05:12 +0000)]
[Sema] Remove dead code in BuildAnonymousStructUnionMemberReference. NFCI
Summary:
This patch cleans up a bunch of dead or unused code in BuildAnonymousStructUnionMemberReference.
The dead code was a branch that built a new CXXThisExpr when we weren't given a base object expression or base variable.
However, BuildAnonymousFoo has only two callers. One of which always builds a base object expression first, the second only calls when the IndirectFieldDecl is not a C++ class member. Even within C this branch seems entirely unused.
I tried diligently to write a test which hit it with no success.
This patch removes the branch and replaces it with an assertion that we were given either a base object expression or a base variable.
Reviewers: rsmith, aaron.ballman, majnemer, rjmccall
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D45410
llvm-svn: 329518
Eric Fiselier [Sun, 8 Apr 2018 05:11:59 +0000 (05:11 +0000)]
[Sema] Fix PR22637 - IndirectFieldDecl's discard qualifiers during template instantiation.
Summary:
Currently Clang fails to propagate qualifiers from the `CXXThisExpr` to the rebuilt `FieldDecl` for IndirectFieldDecls. For example:
```
template <class T> struct Foo {
struct { int x; };
int y;
void foo() const {
static_assert(__is_same(int const&, decltype((y))));
static_assert(__is_same(int const&, decltype((x)))); // assertion fails
}
};
template struct Foo<int>;
```
The fix is to delegate rebuilding of the MemberExpr to `BuildFieldReferenceExpr` which correctly propagates the qualifiers.
Reviewers: rsmith, lebedev.ri, aaron.ballman, bkramer, rjmccall
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D45412
llvm-svn: 329517
Guozhi Wei [Sat, 7 Apr 2018 23:36:10 +0000 (23:36 +0000)]
[DAGCombiner] Fold (zext (and/or/xor (shl/shr (load x), cst), cst))
In our real world application, we found the following optimization is missed in DAGCombiner
(zext (and/or/xor (shl/shr (load x), cst), cst)) -> (and/or/xor (shl/shr (zextload x), (zext cst)), (zext cst))
If the user of original zext is an add, it may enable further lea optimization on x86.
This patch add a new function CombineZExtLogicopShiftLoad to do this optimization.
Differential Revision: https://reviews.llvm.org/D44402
llvm-svn: 329516
Fangrui Song [Sat, 7 Apr 2018 20:50:35 +0000 (20:50 +0000)]
[libclang] Add clang_File_tryGetRealPathName
Summary:
clang_getFileName() may return a path relative to WorkingDir.
On Arch Linux, during clang_indexTranslationUnit(), clang_getFileName() on
CXIdxIncludedIncludedFileInfo::file may return
"/../lib64/gcc/x86_64-pc-linux-gnu/7.3.0/../../../../include/c++/7.3.0/string",
for `#include <string>`.
I presume WorkingDir is somehow changed to /usr/lib or /usr/include and
clang_getFileName() returns a path relative to WorkingDir.
clang_File_tryGetRealPathName() returns "/usr/include/c++/7.3.0/string"
which is more useful for the indexer in this case.
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D42893
llvm-svn: 329515
Philip Pfaffe [Sat, 7 Apr 2018 20:22:38 +0000 (20:22 +0000)]
Revert "Followup for r329293: Temporarily disable the breaking test on windows."
This reverts commit r329393 /
b52ba35e7759cd4002221be1dbb63ec80fde21ec.
llvm-svn: 329514
John McCall [Sat, 7 Apr 2018 20:16:47 +0000 (20:16 +0000)]
Generalize the swiftcall API since being passed indirectly isn't
C++-specific anymore.
llvm-svn: 329513
Manoj Gupta [Sat, 7 Apr 2018 19:59:58 +0000 (19:59 +0000)]
[Driver] Update GCC libraries detection logic for Gentoo.
Summary:
1. Find GCC's LDPATH from the actual GCC config file.
2. Avoid picking libraries from a similar named tuple if the exact
tuple is installed.
Reviewers: mgorny, chandlerc, thakis, rnk
Reviewed By: mgorny, rnk
Subscribers: cfe-commits, mgorny
Differential Revision: https://reviews.llvm.org/D45233
llvm-svn: 329512
Simon Pilgrim [Sat, 7 Apr 2018 19:50:09 +0000 (19:50 +0000)]
[X86] Regenerate atom pshufb test
llvm-svn: 329511
Craig Topper [Sat, 7 Apr 2018 19:09:52 +0000 (19:09 +0000)]
[X86] Combine vXi64 multiplies to MULDQ/MULUDQ during DAG combine instead of lowering.
Previously we used a custom lowering for this because of the AVX1 splitting requirement. But we can do the split during DAG combine if we check the types and subtarget
llvm-svn: 329510
Craig Topper [Sat, 7 Apr 2018 19:09:50 +0000 (19:09 +0000)]
[DAGCombiner] Add a combine to turn a build vector of zero extends of extract vector elts into a vector zero extend and possibly an extract subvector.
llvm-svn: 329509
John McCall [Sat, 7 Apr 2018 17:42:06 +0000 (17:42 +0000)]
Allow equality comparisons between block pointers and
block-pointer-compatible ObjC object pointer types.
Patch by Dustin Howett!
llvm-svn: 329508
Robert Widmann [Sat, 7 Apr 2018 16:26:59 +0000 (16:26 +0000)]
[llgo] Move SetSubprogram
Summary: Fixes the bots - I moved LLVMSetSubprogram into the DIBuilder bindings, so the Go bindings need to move as well.
Reviewers: whitequark
Reviewed By: whitequark
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D45402
llvm-svn: 329505
Simon Pilgrim [Sat, 7 Apr 2018 14:20:10 +0000 (14:20 +0000)]
[CostModel][X86] Regenerate vector reduction cost tests with update_analyze_test_checks.py
NOTE: We're only really interested in the extractelement cost (which represents the entire reduction).
llvm-svn: 329504
Sanjay Patel [Sat, 7 Apr 2018 14:14:23 +0000 (14:14 +0000)]
[InstCombine] simplify code that propagates FMF; NFC
llvm-svn: 329503