Rui Ueyama [Tue, 22 May 2018 02:53:11 +0000 (02:53 +0000)]
Handle --plugin-opt= options as alias options.
Previously, we had a loop to iterate over options starting with
`--plugin-opt=` and parse them by hand. But we can make OptTable
do that job for us.
Differential Revision: https://reviews.llvm.org/D47167
llvm-svn: 332935
Marshall Clow [Tue, 22 May 2018 02:19:38 +0000 (02:19 +0000)]
Missed the tests for the deduction guides for prority_queue
llvm-svn: 332931
Tom Stellard [Tue, 22 May 2018 02:03:23 +0000 (02:03 +0000)]
AMDGPU: Remove #include "MCTargetDesc/AMDGPUMCTargetDesc.h" from common headers
Summary:
MCTargetDesc/AMDGPUMCTargetDesc.h contains enums for all the instuction
and register defintions, which are huge so we only want to include
them where needed.
This will also make it easier if we want to split the R600 and GCN
definitions into separate tablegenerated files.
I was unable to remove AMDGPUMCTargetDesc.h from SIMachineFunctionInfo.h
because it uses some enums from the header to initialize default values
for the SIMachineFunction class, so I ended up having to remove includes of
SIMachineFunctionInfo.h from headers too.
Reviewers: arsenm, nhaehnle
Reviewed By: nhaehnle
Subscribers: MatzeB, kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D46272
llvm-svn: 332930
Craig Topper [Tue, 22 May 2018 02:02:13 +0000 (02:02 +0000)]
[X86] Prevent inclusion of __wmmintrin_aes.h and __wmmintrin_pclmul.h without including wmmintrin.h
llvm-svn: 332929
Marshall Clow [Tue, 22 May 2018 01:57:53 +0000 (01:57 +0000)]
Deduction guides for the container adaptors - queue, stack, and priority_queue
llvm-svn: 332927
Peter Collingbourne [Tue, 22 May 2018 01:20:46 +0000 (01:20 +0000)]
MC: Remove dead code. NFCI.
This code appears to have been copied from the mach-o streamer. It has
no effect in ELF because indirect symbols are specific to mach-o.
llvm-svn: 332926
Paul Semel [Tue, 22 May 2018 01:04:36 +0000 (01:04 +0000)]
Revert "[llvm-objcopy] Add --strip-unneeded option"
There is a use after free I didn't see. Need to investigate.
This reverts commit
f7624abeb1f0d012309baf2e78cf2499fbfe5e5f.
llvm-svn: 332925
Petr Hosek [Tue, 22 May 2018 01:01:56 +0000 (01:01 +0000)]
[CMake] Use a different source depending on C++ support
When using system C++ library, assume we have a working C++ compiler and
try to compile a complete C++ program. When using in tree C++ library,
only check the C compiler since the C++ library likely won't have been
built yet at time of running the check.
Differential Revision: https://reviews.llvm.org/D47169
llvm-svn: 332924
Petr Hosek [Tue, 22 May 2018 00:43:04 +0000 (00:43 +0000)]
[CMake] Pass Clang defaults to runtimes builds
This enables the use of Clang default options from runtimes CMake files.
Differential Revision: https://reviews.llvm.org/D47168
llvm-svn: 332923
Jim Ingham [Tue, 22 May 2018 00:06:55 +0000 (00:06 +0000)]
Work around some odd instruction single-step behavior on macOS.
We've seen some cases on macOS where you go to instruction single
step (over a breakpoint), and single step returns but the instruction
hasn't been executed (and the pc hasn't moved.) The ThreadPlanStepOverBreakpoint
used to handle this case by accident, but the patches to handle two adjacent
breakpoints broke that accident.
This patch fixes the logic of ExplainsStop to explicitly handle the case where
the pc didn't move. It also adds a WillPop that re-enables the breakpoint we
were stepping over. We never want an unexpected path through the plan to
fool us into not doing that.
I have no idea how to make this bug happen. It is very inconsistent when it
occurs IRL. We really need a full MockProcess Plugin before we can start to write
tests for this sort of system hiccup.
<rdar://problem/
38505726>
llvm-svn: 332922
Sanjay Patel [Mon, 21 May 2018 23:54:19 +0000 (23:54 +0000)]
[DAG] fold FP binops with undef operands to NaN
This is the FP sibling of D43141 with the corresponding IR change in rL327212.
We can't propagate undef here because if a variable operand is a NaN, these
binops must propagate NaN. Neither global nor node-level fast-math makes a
difference. If we have 'nnan', I think later folds can turn the NaN into undef.
The tests in X86/fp-undef.ll are meant to be the definitive verification for
these folds - everything reduces identically now.
The other test changes are collateral damage. They may need to be altered to
preserve their intent.
Differential Revision: https://reviews.llvm.org/D47026
llvm-svn: 332920
Lang Hames [Mon, 21 May 2018 23:46:29 +0000 (23:46 +0000)]
[ORC] Make some more operations on VSO private. These should be done via a
MaterializationResponsibility instance.
llvm-svn: 332919
Lang Hames [Mon, 21 May 2018 23:45:40 +0000 (23:45 +0000)]
[LKH] Add a replacement RTDyldLayer.
llvm-svn: 332918
Roman Tereshin [Mon, 21 May 2018 23:28:51 +0000 (23:28 +0000)]
Reapply r332907 "[GlobalISel] Improving InstructionSelect's performance by reducing MatchTable..."
Apparently the compile time problem was caused by the fact that not
all compilers / STL implementations can automatically convert
std::unique_ptr<Derived> to std::unique_ptr<Base>. Fixed (hopefully)
by making sure it's std::unique_ptr<Derived>&& (rvalue ref) to
std::unique_ptr<Base> conversion instead.
llvm-svn: 332917
Craig Topper [Mon, 21 May 2018 23:15:00 +0000 (23:15 +0000)]
[X86] Remove 128/256-bit cvtdq2ps, cvtudq2ps, cvtqq2pd, cvtuqq2pd intrinsics.
These can all be implemented with sitofp/uitofp instructions.
llvm-svn: 332916
Paul Semel [Mon, 21 May 2018 22:50:32 +0000 (22:50 +0000)]
[llvm-objcopy] Add --strip-unneeded option
This option removes symbols that are not needed by relocations.
Differential Revision: https://reviews.llvm.org/D46896
llvm-svn: 332915
Jim Ingham [Mon, 21 May 2018 22:41:28 +0000 (22:41 +0000)]
Fix the Xcode project for the refactoring of the
DWARF reader.
llvm-svn: 332914
Raphael Isemann [Mon, 21 May 2018 22:27:22 +0000 (22:27 +0000)]
Add missing include for cstdint to Visibility.h
Summary: We use uint8_t in this header, so we need to include cstdint.
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D47164
llvm-svn: 332913
Roman Tereshin [Mon, 21 May 2018 22:21:24 +0000 (22:21 +0000)]
Revert r332907 "[GlobalISel] Improving InstructionSelect's performance by reducing MatchTable..."
There is a compile time error I didn't see locally, investigating now.
llvm-svn: 332912
Peter Collingbourne [Mon, 21 May 2018 22:14:02 +0000 (22:14 +0000)]
Add missing x86-registered-target.
llvm-svn: 332911
Diego Caballero [Mon, 21 May 2018 22:12:38 +0000 (22:12 +0000)]
Fix warning from r332654 with LLVM_ATTRIBUTE_USED
r332654 tried to fix an unused function warning with
a void cast. This approach worked for clang and gcc
but not for MSVC. This commit replaces the void cast
with the LLVM_ATTRIBUTE_USED approach.
llvm-svn: 332910
Craig Topper [Mon, 21 May 2018 22:10:02 +0000 (22:10 +0000)]
[X86] Remove a builtin that should have been removed in r332882.
llvm-svn: 332909
Peter Collingbourne [Mon, 21 May 2018 22:09:45 +0000 (22:09 +0000)]
Unbreak kaleidoscope example.
llvm-svn: 332908
Roman Tereshin [Mon, 21 May 2018 22:04:39 +0000 (22:04 +0000)]
[GlobalISel] Improving InstructionSelect's performance by reducing MatchTable, mostly NFC, perf patch 1
This patch starts a series of patches that decrease time spent by
GlobalISel in its InstructionSelect pass by roughly 60% for -O0 builds
for large inputs as measured on sqlite3-amalgamation
(http://sqlite.org/download.html) targeting AArch64.
The performance improvements are achieved solely by reducing the
number of matching GIM_* opcodes executed by the MatchTable's
interpreter during the selection by approx. a factor of 30, which also
brings contribution of this particular part of the selection process
to the overall runtime of InstructionSelect pass down from approx.
60-70% to 5-7%, thus making further improvements in this particular
direction not very profitable.
The improvements described above are expected for any target that
doesn't have many complex patterns. The targets that do should
strictly benefit from the changes, but by how much exactly is hard to
estimate beforehand. It's also likely that such target WILL benefit
from further improvements to MatchTable, most likely the ones that
bring it closer to a perfect decision tree.
This commit specifically is rather large mostly NFC commit that does
necessary preparation work and refactoring, there will be a following
series of small patches introducing a specific optimization each
shortly after.
This commit specifically is expected to cause a small compile time
regression (around 2.5% of InstructionSelect pass time), which should
be fixed by the next commit of the series.
Every commit planned shares the same Phabricator Review.
Reviewers: qcolombet, dsanders, bogner, aemerson, javed.absar
Reviewed By: qcolombet
Subscribers: rovka, llvm-commits, kristof.beyls
Differential Revision: https://reviews.llvm.org/D44700
llvm-svn: 332907
Peter Collingbourne [Mon, 21 May 2018 21:48:17 +0000 (21:48 +0000)]
Fix another make_unique ambiguity.
llvm-svn: 332906
Roman Lebedev [Mon, 21 May 2018 21:41:10 +0000 (21:41 +0000)]
[DAGCombiner] isAllOnesConstantOrAllOnesSplatConstant(): look through bitcasts
Summary:
As pointed out in D46528, we errneously transform cases like `xor X, -1`,
even though we use said function.
It's because the `-1` is actually a bitcast there.
So i think we can just look through it in the function.
Differential Revision: https://reviews.llvm.org/D47156
llvm-svn: 332905
Roman Lebedev [Mon, 21 May 2018 21:41:02 +0000 (21:41 +0000)]
[DAGCombine][X86][AArch64] Masked merge unfolding: vector edition.
Summary:
This **appears** to be the last missing piece for the masked merge pattern handling in the backend.
This is [[ https://bugs.llvm.org/show_bug.cgi?id=37104 | PR37104 ]].
[[ https://bugs.llvm.org/show_bug.cgi?id=6773 | PR6773 ]] will introduce an IR canonicalization that is likely bad for the end assembly.
Previously, `andps`+`andnps` / `bsl` would be generated. (see `@out`)
Now, they would no longer be generated (see `@in`), and we need to make sure that they are generated.
Differential Revision: https://reviews.llvm.org/D46528
llvm-svn: 332904
Roman Lebedev [Mon, 21 May 2018 21:40:51 +0000 (21:40 +0000)]
[X86][AArch64][NFC] Add tests for vector masked merge unfolding
Summary:
This is [[ https://bugs.llvm.org/show_bug.cgi?id=37104 | PR37104 ]].
[[ https://bugs.llvm.org/show_bug.cgi?id=6773 | PR6773 ]] will introduce an IR canonicalization that is likely bad for the end assembly.
Previously, `andps`+`andnps` / `bsl` would be generated. (see `@out`)
Now, they would no longer be generated (see `@in`).
Differential Revision: https://reviews.llvm.org/D46008
llvm-svn: 332903
Marshall Clow [Mon, 21 May 2018 21:30:12 +0000 (21:30 +0000)]
Implement deduction guides for vector
llvm-svn: 332901
Lang Hames [Mon, 21 May 2018 21:11:22 +0000 (21:11 +0000)]
[ORC] Preserve Materializing symbol flag during resolution.
llvm-svn: 332899
Lang Hames [Mon, 21 May 2018 21:11:21 +0000 (21:11 +0000)]
[ORC] Lookup now returns an error if any symbols are not found.
Also tightens the behavior of ExecutionSession::failQuery. Queries can usually
only be failed by marking a symbol as failed-to-materialize, but
ExecutionSession::failQuery provides a second route, and both routes may be
executed from different threads. In the case that a query has already been
failed due to a materialization error, ExecutionSession::failQuery will
direct the error to ExecutionSession::reportError instead.
llvm-svn: 332898
Lang Hames [Mon, 21 May 2018 21:11:21 +0000 (21:11 +0000)]
[ORC] Remove the optional MaterializationResponsibility argument from lookup.
The lookup function provides blocking symbol resolution for JIT clients (not
layers themselves) so it does not need to track symbol dependencies via a
MaterializationResponsibility.
llvm-svn: 332897
Lang Hames [Mon, 21 May 2018 21:11:13 +0000 (21:11 +0000)]
[ORC] Add IRLayer and ObjectLayer interfaces and related MaterializationUnits.
llvm-svn: 332896
Craig Topper [Mon, 21 May 2018 21:09:18 +0000 (21:09 +0000)]
[DAGCombiner] Use computeKnownBits to match rotate patterns that have had their amount masking modified by simplifyDemandedBits
SimplifyDemandedBits can remove bits from the masks for the shift amounts we need to see to detect rotates.
This patch uses zeroes from computeKnownBits to fill in some of these mask bits to make the match work.
As currently written this calls computeKnownBits even when the mask hasn't been simplified because it made the code simpler. If we're worried about compile time performance we can improve this.
I know we're talking about making a rotate intrinsic, but hopefully we can go ahead and do this change and just make sure the rotate intrinsic also handles it.
Differential Revision: https://reviews.llvm.org/D47116
llvm-svn: 332895
Sanjay Patel [Mon, 21 May 2018 21:09:14 +0000 (21:09 +0000)]
[InstCombine] regenerate checks; NFC
llvm-svn: 332894
Reid Kleckner [Mon, 21 May 2018 21:03:19 +0000 (21:03 +0000)]
[X86] Simplify some X86 address mode folding code, NFCI
This code should really do exactly the same thing for 32-bit x86 and
64-bit small code models, with the exception that RIP-relative
addressing can't use base and index registers.
llvm-svn: 332893
Craig Topper [Mon, 21 May 2018 21:01:13 +0000 (21:01 +0000)]
[X86] Remove some unneeded check lines that I copy and pasted when I made vector tests from some scalar test cases.
llvm-svn: 332892
Craig Topper [Mon, 21 May 2018 20:58:23 +0000 (20:58 +0000)]
[X86] Remove masking from pternlog llvm intrinsics and use a select instruction instead.
Because the intrinsics in the headers are implemented as macros, we can't just use a select builtin and pternlog builtin. This would require one of the macro arguments to be used twice. Depending on what was passed to the macro we could expand an expression twice leading to weird behavior. We could maybe declare our local variable in the macro, but that would need to worry about name collisions.
To avoid that just generate IR directly in CGBuiltin.cpp.
Differential Revision: https://reviews.llvm.org/D47125
llvm-svn: 332891
Craig Topper [Mon, 21 May 2018 20:58:09 +0000 (20:58 +0000)]
[X86] Remove masking from vpternlog intrinsics. Use a select in IR instead.
This removes 6 intrinsics since we no longer need separate mask and maskz intrinsics.
Differential Revision: https://reviews.llvm.org/D47124
llvm-svn: 332890
Peter Collingbourne [Mon, 21 May 2018 20:56:28 +0000 (20:56 +0000)]
Fix a make_unique ambiguity.
llvm-svn: 332889
Walter Lee [Mon, 21 May 2018 20:43:36 +0000 (20:43 +0000)]
[asan] Make GetCurrentThread RTEMS-friendly
On RTEMS, system and user code all live in a single binary and address
space. There is no clean separation, and instrumented code may
execute before the ASan run-time is initialized (or after it has been
destroyed).
Currently, GetCurrentThread() may crash if it's called before ASan
run-time is initialized. Make it return nullptr instead.
Similarly, fix __asan_handle_no_return so that it gives up rather than
try something that may crash.
Differential Revision: https://reviews.llvm.org/D46459
llvm-svn: 332888
Sanjay Patel [Mon, 21 May 2018 20:39:35 +0000 (20:39 +0000)]
[InstCombine] remove fptrunc (select) code; NFCI
This pattern is handled within commonCastTransforms(),
so the code here is dead AFAICT.
llvm-svn: 332887
Richard Smith [Mon, 21 May 2018 20:36:58 +0000 (20:36 +0000)]
Revert r332847; it caused us to miscompile certain forms of reference initialization.
llvm-svn: 332886
Peter Collingbourne [Mon, 21 May 2018 20:31:59 +0000 (20:31 +0000)]
CodeGen, Driver: Start using direct split dwarf emission in clang.
Fixes PR37466.
Differential Revision: https://reviews.llvm.org/D47093
llvm-svn: 332885
Peter Collingbourne [Mon, 21 May 2018 20:26:49 +0000 (20:26 +0000)]
LTO: Replace split dwarf implementation that uses objcopy with one that uses direct emission.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47091
llvm-svn: 332884
Sanjay Patel [Mon, 21 May 2018 20:23:58 +0000 (20:23 +0000)]
[InstCombine] add tests for cast-of-select; NFC
In all cases, we're pulling the cast above the select.
That's not a good canonicalization if we're creating
a select that then mismatches the operand size of its
condition.
llvm-svn: 332883
Craig Topper [Mon, 21 May 2018 20:19:17 +0000 (20:19 +0000)]
[X86] Use __builtin_convertvector to implement some of the packed integer to packed float conversion intrinsics.
I believe this is safe assuming default default FP environment. The conversion might be inexact, but it can never overflow the FP type so this shouldn't be undefined behavior for the uitofp/sitofp instructions.
We already do something similar for scalar conversions.
Differential Revision: https://reviews.llvm.org/D46863
llvm-svn: 332882
Peter Collingbourne [Mon, 21 May 2018 20:16:41 +0000 (20:16 +0000)]
CodeGen: Add a dwo output file argument to addPassesToEmitFile and hook it up to dwo output.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47089
llvm-svn: 332881
Reid Kleckner [Mon, 21 May 2018 20:14:46 +0000 (20:14 +0000)]
Remove CMake workaround for LLD PR24476 which is no longer needed
llvm-svn: 332880
Richard Smith [Mon, 21 May 2018 20:10:54 +0000 (20:10 +0000)]
Revert r332028; see PR37545 for details.
llvm-svn: 332879
Brian Gesiak [Mon, 21 May 2018 19:48:27 +0000 (19:48 +0000)]
[DebugInfo] Fix typo "DWARG" in test comment (NFC)
Summary:
The correct spelling is "DWARF", the debugging format, not "DWARG".
The typo is in a (not executed by lit) comment in a test file, so
fixing it does not result in any functional change.
Test Plan: check-llvm, just in case
llvm-svn: 332878
Zachary Turner [Mon, 21 May 2018 19:47:45 +0000 (19:47 +0000)]
Enable Python API for OpenBSD.
Patch by David Carlier
llvm-svn: 332877
Kostya Serebryany [Mon, 21 May 2018 19:47:00 +0000 (19:47 +0000)]
[libFuzzer] reinstate -dump_coverage, which is still in use (reverts r332036)
llvm-svn: 332876
Peter Collingbourne [Mon, 21 May 2018 19:44:54 +0000 (19:44 +0000)]
MC: Introduce an ELF dwo object writer and teach llvm-mc about it.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47051
llvm-svn: 332875
Matt Arsenault [Mon, 21 May 2018 19:43:02 +0000 (19:43 +0000)]
AMDGPU: Update GCCBuiltin names for DS FP atomic intrinsics
llvm-svn: 332874
Jonas Devlieghere [Mon, 21 May 2018 19:36:54 +0000 (19:36 +0000)]
[DebugInfo] Use absolute addresses in location lists
Rather than relying on the user to do the address calculating in
DW_AT_location we should just dump the absolute address.
rdar://problem/
38513870
Differential revision: https://reviews.llvm.org/D47152
llvm-svn: 332873
Craig Topper [Mon, 21 May 2018 19:33:42 +0000 (19:33 +0000)]
[X86] Add test cases for D47012.
Patch by Thomasz Krupa.
llvm-svn: 332872
Petr Hosek [Mon, 21 May 2018 19:32:43 +0000 (19:32 +0000)]
Revert "[CMake] Reland "Make simple source used for CMake checks a C file""
This reverts commit rCRT332679 which doesn't quite work and will be
replaced by D47100 and D47115 which is a cleaner solution.
llvm-svn: 332871
Peter Collingbourne [Mon, 21 May 2018 19:30:59 +0000 (19:30 +0000)]
MC: Extract a derived class from ELFObjectWriter. NFCI.
This class will be used to create regular, non-split ELF files.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47049
llvm-svn: 332870
Craig Topper [Mon, 21 May 2018 19:27:50 +0000 (19:27 +0000)]
[X86] Add test cases for missed vector rotate matching due to SimplifyDemandedBits interfering with the AND masks
As requested in D47116
llvm-svn: 332869
Peter Collingbourne [Mon, 21 May 2018 19:20:29 +0000 (19:20 +0000)]
MC: Separate creating a generic object writer from creating a target object writer. NFCI.
With this we gain a little flexibility in how the generic object
writer is created.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47045
llvm-svn: 332868
Peter Collingbourne [Mon, 21 May 2018 19:18:28 +0000 (19:18 +0000)]
MC: Extract ELFObjectWriter's ELF writing functionality into an ELFWriter class. NFCI.
The idea is that we will be able to use this class to create multiple
files.
Differential Revision: https://reviews.llvm.org/D47048
llvm-svn: 332867
Peter Collingbourne [Mon, 21 May 2018 19:09:47 +0000 (19:09 +0000)]
Fix ubsan bounds check failure.
llvm-svn: 332866
Craig Topper [Mon, 21 May 2018 18:42:42 +0000 (18:42 +0000)]
[EarlyCSE] Improve EarlyCSE of some absolute value cases.
Change matchSelectPattern to return X and -X for ABS/NABS in a well defined order. Adjust EarlyCSE to account for this. Ensure the SPF result is some kind of min/max and not abs/nabs in one place in InstCombine that made me nervous.
Prevously we returned the two operands of the compare part of the abs pattern. The RHS is always going to be a 0i, 1 or -1 constant. This isn't a very meaningful thing to return for any one. There's also some freedom in the abs pattern as to what happens when the value is equal to 0. This freedom led to early cse failing to match when different constants were used in otherwise equivalent operations. By returning the input and its negation in a defined order we can ensure an exact match. This also makes sure both patterns use the exact same subtract instruction for the negation. I believe CSE should evebntually make this happen and properly merge the nsw/nuw flags. But I'm not familiar with CSE and what order it does things in so it seemed like it might be good to really enforce that they were the same.
Differential Revision: https://reviews.llvm.org/D47037
llvm-svn: 332865
Peter Collingbourne [Mon, 21 May 2018 18:28:57 +0000 (18:28 +0000)]
MC: Remove stream and output functions from MCObjectWriter. NFCI.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47043
llvm-svn: 332864
Peter Collingbourne [Mon, 21 May 2018 18:23:50 +0000 (18:23 +0000)]
MC: Have the object writers return the number of bytes written. NFCI.
This removes the last external use of the stream.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47042
llvm-svn: 332863
Stanislav Mekhanoshin [Mon, 21 May 2018 18:18:52 +0000 (18:18 +0000)]
[AMDGPU] Add divergence analysis as a dependency for ISel
AMDGPUDAGToDAGISel adds DivergenceAnalysis in getAnalysisUsage
but does not list it in pass dependencies which may lead to
crash.
Differential Revision: https://reviews.llvm.org/D47151
llvm-svn: 332862
Peter Collingbourne [Mon, 21 May 2018 18:17:42 +0000 (18:17 +0000)]
MC: Change object writers to use endian::Writer. NFCI.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47040
llvm-svn: 332861
Diego Caballero [Mon, 21 May 2018 18:14:23 +0000 (18:14 +0000)]
[VPlan] Reland r332654 and silence unused func warning
r332654 was reverted due to an unused function warning in
release build. This commit includes the same code with the
warning silenced.
Differential Revision: https://reviews.llvm.org/D44338
llvm-svn: 332860
Rui Ueyama [Mon, 21 May 2018 18:12:46 +0000 (18:12 +0000)]
Fix typo.
llvm-svn: 332859
Peter Collingbourne [Mon, 21 May 2018 18:11:35 +0000 (18:11 +0000)]
MC: Change MCAssembler::writeSectionData and writeFragmentPadding to take a raw_ostream. NFCI.
Also clean up a couple of hacks where we were writing the section
contents to another stream by setting the object writer's stream,
writing and setting it back.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47038
llvm-svn: 332858
Peter Collingbourne [Mon, 21 May 2018 17:57:19 +0000 (17:57 +0000)]
MC: Change MCAsmBackend::writeNopData() to take a raw_ostream instead of an MCObjectWriter. NFCI.
To make this work I needed to add an endianness field to MCAsmBackend
so that writeNopData() implementations know which endianness to use.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47035
llvm-svn: 332857
Tom Stellard [Mon, 21 May 2018 17:49:31 +0000 (17:49 +0000)]
AMDGPU/GlobalISel: Address post-commit review comments for r332379
MCRegisterInfo::getPhysRegSize() will be deprecated.
llvm-svn: 332856
Alexey Bataev [Mon, 21 May 2018 17:46:34 +0000 (17:46 +0000)]
[InstCombine] Fix PR37526: MinMax patterns produce an infinite loop.
Summary:
This patch fixes PR37526 by simplifying the newly generated LoadInst
instructions. If the pointer address is a bitcast from the pointer to
the NewType, we can just remove this extra bitcast instead of creating
the new one. This fixes the PR37526 + may speed up the whole compilation
process.
Reviewers: spatel, RKSimon, hfinkel
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D47144
llvm-svn: 332855
Mark Searles [Mon, 21 May 2018 17:29:08 +0000 (17:29 +0000)]
[Clang Tablegen] Add llvm_unreachable() to getModifierName()
Fix internal build failure:
../../../ClangDiagnosticsEmitter.cpp -o ClangDiagnosticsEmitter.o
../../../ClangDiagnosticsEmitter.cpp: In function 'llvm::StringRef
{anonymous}::getModifierName({anonymous}::ModifierType)':
../../../ClangDiagnosticsEmitter.cpp:495:1: error: control reaches end of non-void function [-Werror=return-type]
}
^
Build failure triggered by git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@332799
91177308-0d34-0410-b5e6-
96231b3b80d8
Differential Revision: https://reviews.llvm.org/D47150
llvm-svn: 332854
Andrea Di Biagio [Mon, 21 May 2018 17:11:56 +0000 (17:11 +0000)]
[llvm-mca] Removed an empty line generated by the timeline view. NFC.
Also, regenerate all tests.
llvm-svn: 332853
Alexey Bataev [Mon, 21 May 2018 16:40:32 +0000 (16:40 +0000)]
[OPENMP-SIMD] Fix PR37536: Fix definition of _OPENMP macro.
if `-fopenmp-simd` is specified alone, `_OPENMP` macro should not be
defined. If `-fopenmp-simd` is specified along with the `-fopenmp`,
`_OPENMP` macro should be defined with the value `201511`.
llvm-svn: 332852
Andrea Di Biagio [Mon, 21 May 2018 16:30:26 +0000 (16:30 +0000)]
[X86][BtVer2] Add a 'J' prefix to the PRF/RCU defs. NFC
This is to keep the Jaguar model's naming convention. Processor resources all
have a 'J' prefix in the BtVer2 scheduling model.
llvm-svn: 332851
Robert Widmann [Mon, 21 May 2018 16:27:35 +0000 (16:27 +0000)]
[LLVM-C] Add DIBuilder Bindings For ObjC Classes
Summary: Add LLVMDIBuilderCreateObjCIVar, LLVMDIBuilderCreateObjCProperty, and LLVMDIBuilderCreateInheritance to allow declaring metadata for Objective-C class hierarchies and their associated properties and instance variables.
Reviewers: whitequark, deadalnix
Reviewed By: whitequark
Subscribers: harlanhaskins, llvm-commits
Differential Revision: https://reviews.llvm.org/D47123
llvm-svn: 332850
Lama Saba [Mon, 21 May 2018 16:23:16 +0000 (16:23 +0000)]
[X86] - Avoid SFB pass - fix bug in updating the offsets for newly created copies
Change-Id: I169ab6fe7e187727c0298c2a1e2868a683f3e688
llvm-svn: 332849
Daniil Fukalov [Mon, 21 May 2018 16:18:07 +0000 (16:18 +0000)]
[AMDGPU] fixes for lds f32 builtins
1. added restrictions to memory scope, order and volatile parameters
2. added custom processing for these builtins - currently is not used code,
needed to switch off GCCBuiltin link to the builtins (ongoing change to llvm
tree)
3. builtins renamed as requested
Differential Revision: https://reviews.llvm.org/D43281
llvm-svn: 332848
Serge Pavlov [Mon, 21 May 2018 16:09:54 +0000 (16:09 +0000)]
[CodeGen] Recognize more cases of zero initialization
If a variable has an initializer, codegen tries to build its value. If
the variable is large in size, building its value requires substantial
resources. It causes strange behavior from user viewpoint: compilation
of huge zero initialized arrays like:
char data_1[2147483648u] = { 0 };
consumes enormous amount of time and memory.
With this change codegen tries to determine if variable initializer is
equivalent to zero initializer. In this case variable value is not
constructed.
This change fixes PR18978.
Differential Revision: https://reviews.llvm.org/D46241
llvm-svn: 332847
James Henderson [Mon, 21 May 2018 15:31:23 +0000 (15:31 +0000)]
[ELF] Update due to API change in .debug_line parsing
See r332845.
Reviewed by: grimar
Differential Revision: https://reviews.llvm.org/D46832
llvm-svn: 332846
James Henderson [Mon, 21 May 2018 15:30:54 +0000 (15:30 +0000)]
[DWARF] Refactor callback usage for .debug_line error handling
Change the "recoverable" error callback to take an Error instaed of a
string.
Reviewed by: JDevlieghere
Differential Revision: https://reviews.llvm.org/D46831
llvm-svn: 332845
Simon Pilgrim [Mon, 21 May 2018 15:17:23 +0000 (15:17 +0000)]
[X86][SSE] Add an assert to ensure that rotation amount is converted to a scale
Missed in rL332832 where we added SSE v4i32 rotations for PR37426.
llvm-svn: 332844
Momchil Velikov [Mon, 21 May 2018 14:28:43 +0000 (14:28 +0000)]
[Sema] Fix incorrect packed aligned structure layout
Handle attributes before checking the record layout (e.g. underalignment check
during `alignas` processing), as layout may be cached without taking into
account attributes that may affect it.
Differential Revision: https://reviews.llvm.org/D46439
llvm-svn: 332843
Greg Clayton [Mon, 21 May 2018 14:14:36 +0000 (14:14 +0000)]
Fix PathMappingList for relative and empty paths after recent FileSpec normalization changes
PathMappingList was broken for relative and empty paths after normalization changes in FileSpec. There were also no tests for PathMappingList so I added those.
Changes include:
Change PathMappingList::ReverseRemapPath() to take FileSpec objects instead of ConstString. The only client of this was doing work to convert to and from ConstString objects for no reason.
Normalize all paths prefix and replacements that are added to the PathMappingList vector so they match the paths that have been already normalized in the debug info
Unify code in the two forms of PathMappingList::RemapPath() so only one contains the actual functionality. Prior to this, there were two versions of this code.
Use FileSpec::AppendPathComponent() and remove a long standing TODO so paths are correctly appended to each other.
Added tests for absolute, relative and empty paths.
Differential Revision: https://reviews.llvm.org/D47021
llvm-svn: 332842
Pavel Labath [Mon, 21 May 2018 14:12:52 +0000 (14:12 +0000)]
Reland "[DWARF] Extract indexing code into a separate class hierarchy"
After this commit, the xcode project will need to be updated to include
the new files added here.
llvm-svn: 332841
Tim Northover [Mon, 21 May 2018 12:43:54 +0000 (12:43 +0000)]
ARM: be conservative when asked load/store alignment of weird type.
Chances are we'll be asked again after type legalization, but before that point
it's better to claim misaligned accesses aren't allowed than to assert.
llvm-svn: 332840
Pavel Labath [Mon, 21 May 2018 11:47:45 +0000 (11:47 +0000)]
[CodeGen] Disable aggressive structor optimizations at -O0, take 2
The first version of the patch (r332228) was flawed because it was
putting structors into C5/D5 comdats very eagerly. This is correct only
if we can ensure the comdat contains all required versions of the
structor (which wasn't the case). This version uses a more nuanced
approach:
- for local structor symbols we use an alias because we don't have to
worry about comdats or other compilation units.
- linkonce symbols are emitted separately, as we cannot guarantee we
will have all symbols we need to form a comdat (they are emitted
lazily, only when referenced).
- available_externally symbols are also emitted separately, as the code
seemed to be worried about emitting an alias in this case.
- other linkage types are not affected by the optimization level. They
either get put into a comdat (weak) or get aliased (external).
Reviewers: rjmccall, aprantl
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D46685
llvm-svn: 332839
Nico Weber [Mon, 21 May 2018 11:44:39 +0000 (11:44 +0000)]
revert r332610, it breaks cfi, see D46326
llvm-svn: 332838
Aleksandar Beserminji [Mon, 21 May 2018 11:38:52 +0000 (11:38 +0000)]
[mips] Revert Merge MipsLongBranch and MipsHazardSchedule passes
Revert this patch due buildbot failure.
Differential Revision: https://reviews.llvm.org/D46641
llvm-svn: 332837
David Green [Mon, 21 May 2018 11:06:28 +0000 (11:06 +0000)]
[CVP] Require DomTree for new Pass Manager
We were previously using a DT in CVP through SimplifyQuery, but not requiring it in
the new pass manager. Hence it would crash if DT was not already available. This now
gets DT directly and plumbs it through to where it is used (instead of using it
through SQ).
llvm-svn: 332836
Eric Christopher [Mon, 21 May 2018 10:27:36 +0000 (10:27 +0000)]
Fix up a few grammar issues.
llvm-svn: 332835
Aleksandar Beserminji [Mon, 21 May 2018 10:20:02 +0000 (10:20 +0000)]
[mips] Merge MipsLongBranch and MipsHazardSchedule passes
MipsLongBranchPass and MipsHazardSchedule passes are joined to one pass
because of mutual conflict. When MipsHazardSchedule inserts 'nop's, it
potentially breaks some jumps, so they have to be expanded to long
branches. When some branch is expanded to long branch, it potentially
creates a hazard situation, which should be fixed by adding nops.
New pass is called MipsBranchExpansion, it combines these two passes,
and runs them alternately until one of them reports no changes were made.
Differential Revision: https://reviews.llvm.org/D46641
llvm-svn: 332834
Pavel Labath [Mon, 21 May 2018 10:09:22 +0000 (10:09 +0000)]
Improve coverage of the apple-tables test
After closer examination, it turns out we mis-classify one of the
methods only if two of the structs have the same name. Since this was
meant to be a basic test, I rename one of the structs in the test so
that we have at least some coverage for the apple tables lookup.
Instead, I create an XFAILed test which specifically targets the
same-name case (and file a bug to track it).
llvm-svn: 332833
Simon Pilgrim [Mon, 21 May 2018 09:45:59 +0000 (09:45 +0000)]
[X86][SSE] Support v4i32 rotations (PR37426)
As suggested by Fabian on PR37426, we can use PMULUDQ to perform v4i32 vector rotations as the upper 32bits of the multiply will contain the 'wrapped' bits of the rotation.
v8i16/v16i8 rotations would be straightforward to add to lowerRotate in the future - ideally we'd mostly share code with the vector shifts lowering.
Differential Revision: https://reviews.llvm.org/D46954
llvm-svn: 332832
Pavel Labath [Mon, 21 May 2018 09:27:16 +0000 (09:27 +0000)]
Add some apple-tables lookup tests
Summary:
Now that we are able to parse MachO files everywhere, we can write some
cross-platform tests for handling of apple accelerator tables. This
reruns the same lookup tests we have for manual indexes on MachO files
which will use the accelerator tables instead. This makes sure we return
the same results regardless of the method we used to access the debug
info.
The tests confirm we return the same results for looking up types,
namespaces and variables, but have found an inconsistency in the
treatment of function lookup. In the function case we mis-classify the
method "foo" declared in the local struct sbar (inside function ffbar).
We classify it as a function whereas it really is a method. Preliminary
analysis suggests this is because
DWARFASTParserClang::GetClangDeclContextForDIE returns null when given
the local "struct sbar" DIE. This causes us to get the wrong
CompilerDeclContext when we ask for the context of the inner foo, which
means CompilerDeclContext::ISStructUnionOrClass returns false.
Until this is fixed, I do not include the darwin versions of the "base"
and "method" function lookup tests.
Reviewers: JDevlieghere, clayborg
Subscribers: aprantl, ilya-biryukov, ioeric, lldb-commits
Differential Revision: https://reviews.llvm.org/D47064
llvm-svn: 332831
Craig Topper [Mon, 21 May 2018 06:07:49 +0000 (06:07 +0000)]
[X86] Remove some preprocessor feature checks from intrinsic headers
Summary:
These look to be a couple things that weren't removed when we switched to target attribute.
The popcnt makes including just smmintrin.h also include popcntintrin.h. The popcnt file itself already contains target attrributes.
The prefetch ones are just wrappers around __builtin_prefetch which we have graceful fallbacks for in the backend if the exact instruction isn't available. So there's no reason to hide them. And it makes them available in functions that have the write target attribute but not a -march command line flag.
Reviewers: echristo, RKSimon, spatel, DavidKreitzer
Reviewed By: echristo
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D47029
llvm-svn: 332830
Craig Topper [Mon, 21 May 2018 03:36:57 +0000 (03:36 +0000)]
[X86] Remove some unused builtins.
These were upgraded to native shufflevectors months ago.
llvm-svn: 332829