Simon Pilgrim [Mon, 31 Jan 2022 12:00:51 +0000 (12:00 +0000)]
[DAG] SimplifyDemandedBits - mul(x,x) - if only demand bit[1] then fold to zero
Simon Pilgrim [Mon, 31 Jan 2022 11:36:03 +0000 (11:36 +0000)]
[X86] Limit mul(x,x) knownbits tests with not undef/poison check
We can only assume bit[1] == zero if its the only demanded bit or the source is not undef/poison
Jay Foad [Mon, 31 Jan 2022 11:22:00 +0000 (11:22 +0000)]
[AMDGPU] AMDGPUAnnotateUniformValues: inline a single-use lambda. NFC.
Jay Foad [Fri, 28 Jan 2022 13:05:39 +0000 (13:05 +0000)]
[AMDGPU] Add test for a problem with noclobber metadata
If AMDGPUAnnotateUniformValues finds a load from a uniform pointer with
no potentially clobbering stores between the kernel entry point and the
load instruction, it adds noclobber metadata to the *address*. This is
unsafe because it can get applied to other loads in the same which do
have aliasing stores.
Differential Revision: https://reviews.llvm.org/D118458
Simon Pilgrim [Mon, 31 Jan 2022 11:06:02 +0000 (11:06 +0000)]
[X86] Add mul(x,x) tests showing miscompile
As raised by @efriedma on D117995 - the source must not be undef/poison to demand any bits in mul(x,x) other than bit[1]
https://alive2.llvm.org/ce/z/Cxkjen
Jay Foad [Fri, 28 Jan 2022 14:45:06 +0000 (14:45 +0000)]
[Local] invertCondition: try modifying an existing ICmpInst
This avoids various cases where StructurizeCFG would otherwise insert an
xor i1 instruction, and it since it generally runs late in the pipeline,
instcombine does not clean up the xor-of-cmp pattern.
Differential Revision: https://reviews.llvm.org/D118478
Paulo Matos [Mon, 31 Jan 2022 10:42:02 +0000 (11:42 +0100)]
[WebAssembly] Refactor and fix emission of external IR global decls
This patches fixes the visibility and linkage information of symbols
referring to IR globals.
Emission of external declarations is now done in the first execution
of emitConstantPool rather than in emitLinkage (and a few other
places). This is the point where we have already gathered information
about used symbols (by running the MC Lower PrePass) and not yet
started emitting any functions so that any declarations that need to
be emitted are done so at the top of the file before any functions.
This changes the order of a few directives in the final asm file which
required an update to a few tests.
Reviewed By: sbc100
Differential Revision: https://reviews.llvm.org/D118122
Andrzej Warzynski [Thu, 27 Jan 2022 16:44:44 +0000 (16:44 +0000)]
[flang][driver][nfc] Fix capitalisation
As pointed out in https://reviews.llvm.org/D93401, some methods in the
Flang driver are named inconsistently. The driver strives to follow
Flang's C++ style [1] and this patch updates these methods accordingly.
[1]
https://github.com/llvm/llvm-project/blob/main/flang/docs/C%2B%2Bstyle.md
Differential Revision: https://reviews.llvm.org/D118381
Gabor Marton [Tue, 18 Jan 2022 16:47:32 +0000 (17:47 +0100)]
[Analyzer] Add docs to StdCLibraryFunctionArgsChecker
Differential Revision: https://reviews.llvm.org/D117568
Sven van Haastregt [Mon, 31 Jan 2022 10:21:05 +0000 (10:21 +0000)]
[OpenCL] Make generic addrspace optional for -fdeclare-opencl-builtins
Currently, -fdeclare-opencl-builtins always adds the generic address
space overloads of e.g. the vload builtin functions in OpenCL 3.0
mode, even when the generic address space feature is disabled.
Guard the generic address space overloads by the
`__opencl_c_generic_address_space` feature instead of by OpenCL
version.
Guard the private, global, and local overloads using the internal
`__opencl_c_named_address_space_builtins` feature.
Differential Revision: https://reviews.llvm.org/D107769
Florian Hahn [Mon, 31 Jan 2022 10:18:28 +0000 (10:18 +0000)]
[AArch64] Fix costs of float vector compare/selects pairs.
The current cost-model overestimates the cost of vector compares &
selects for ordered floating point compares. This patch fixes that by
extending the existing logic for integer predicates.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D118256
Benjamin Kramer [Mon, 31 Jan 2022 10:17:41 +0000 (11:17 +0100)]
[clang-tidy] Make header compile standalone. NFC.
Matthias Springer [Mon, 31 Jan 2022 10:10:51 +0000 (19:10 +0900)]
[mlir][vector][NFC] Split into IR, Transforms and Utils
This reduces the dependencies of the MLIRVector target and makes the dialect consistent with other dialects.
Differential Revision: https://reviews.llvm.org/D118533
serge-sans-paille [Sat, 29 Jan 2022 08:17:53 +0000 (09:17 +0100)]
Fix -Wreserved-identifier in presence of system macro
Do not warn on reserved identifiers resulting from expansion of system macros.
Also properly test -Wreserved-identifier wrt. system headers.
Should fix #49592
Differential Revision: https://reviews.llvm.org/D118532
serge-sans-paille [Fri, 28 Jan 2022 14:03:45 +0000 (15:03 +0100)]
Cleanup LLVMRemarks includes
Based on the output of include-what you-use.
Most notably, llvm/Remarks/Remark.h is no longer automatically included by
llvm/Remarks/RemarkParser.h, so client code may need to include explicitly.
clang++ -E -Iinclude -I../llvm/include ../llvm/lib/Remarks/*.cpp -std=c++14 -fno-rtti -fno-exceptions | wc -l
before: 770253
after: 759347
Related discourse thread: https://llvm.discourse.group/t/include-what-you-use-include-cleanup
Differential Revision: https://reviews.llvm.org/D118506
serge-sans-paille [Fri, 28 Jan 2022 13:55:42 +0000 (14:55 +0100)]
Cleanup llvm/utils/TableGen headers
Based on the output of include-what-you-use.
It's an utility directory, so no much impact on other code areas.
clang++ -E -Iinclude -I../llvm/include ../llvm/utils/TableGen/*.cpp -std=c++14 -fno-rtti -fno-exceptions | wc -l
before: 4327274
after: 4316190
Related discourse thread: https://llvm.discourse.group/t/include-what-you-use-include-cleanup
Differential Revision: https://reviews.llvm.org/D118466
Balázs Kéri [Mon, 31 Jan 2022 08:47:49 +0000 (09:47 +0100)]
[clang-tidy] bugprone-signal-handler improvements: display call chain
Display notes for a possible call chain if an unsafe function is found to be
called (maybe indirectly) from a signal handler.
The call chain displayed this way includes probably not the first calls of
the functions, but it is a valid possible (in non path-sensitive way) one.
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D118224
Petr Hosek [Mon, 31 Jan 2022 08:54:17 +0000 (00:54 -0800)]
Revert "enable plugins for clang-tidy"
This reverts commit
36892727e4f19a60778e371d78f8fb09d8122c85 which
breaks the build when LLVM_INSTALL_TOOLCHAIN_ONLY is enabled with:
CMake Error at cmake/modules/AddLLVM.cmake:683 (add_dependencies):
The dependency target "clang-tidy-headers" of target "CTTestTidyModule"
does not exist.
Nikita Popov [Fri, 28 Jan 2022 11:37:28 +0000 (12:37 +0100)]
[Inline][Cloning] Reliably remove unreachable blocks during cloning (PR53206)
The pruning cloner already tries to remove unreachable blocks. The
original cloning process will simplify instructions and constant
terminators, and only clone blocks that are reachable at that point.
However, phi nodes can only be simplified after everything has been
cloned. For that reason, additional blocks may become unreachable
after phi simplification.
The code does try to handle this as well, but only removes blocks
that don't have predecessors. It misses unreachable cycles. This
can cause issues if SEH exception handling code is part of an
unreachable cycle, as the inliner is not prepared to deal with that.
This patch instead performs an explicit scan for reachable blocks,
and drops everything else.
Fixes https://github.com/llvm/llvm-project/issues/53206.
Differential Revision: https://reviews.llvm.org/D118449
Nikita Popov [Thu, 27 Jan 2022 10:15:31 +0000 (11:15 +0100)]
[RISCV] Avoid pointer element type access for masked atomicrmw intrinsics
masked.atomicrmw.*.i32 intrinsics access an i32 (and then possibly
mask it), so hardcode MVT::i32 as the access type here, rather than
determining it from the pointer element type.
Differential Revision: https://reviews.llvm.org/D118336
Marek Kurdej [Mon, 31 Jan 2022 08:04:26 +0000 (09:04 +0100)]
[clang-format] Use EXPECT_EQ instead of setting style to a default value. NFC.
Pavel Labath [Mon, 31 Jan 2022 07:22:31 +0000 (08:22 +0100)]
[lldb] Convert Process KDP Log to the new API
Mehdi Amini [Mon, 31 Jan 2022 07:55:00 +0000 (07:55 +0000)]
Add missing include to mlir/Target/LLVMIR/Import.h to be standalone (NFC)
Amir Ayupov [Mon, 31 Jan 2022 06:36:53 +0000 (22:36 -0800)]
[llvm] Remove redundant `;` (NFC)
Max Kazantsev [Mon, 31 Jan 2022 05:12:48 +0000 (12:12 +0700)]
[InstCombine] Generalize and-reduce pattern to handle `ne` case as well as `eq`
Following Sanjay's proposal from discussion in D118317, this patch
generalizes and-reduce handling to fold the following pattern
```
icmp ne (bitcast(icmp ne (lhs, rhs)), 0)
```
into
```
icmp ne (bitcast(lhs), bitcast(rhs))
```
https://alive2.llvm.org/ce/z/WDcuJ_
Differential Revision: https://reviews.llvm.org/D118431
Reviewed By: lebedev.ri
Uday Bondhugula [Sun, 30 Jan 2022 13:17:03 +0000 (18:47 +0530)]
Support affine.load/store ops in fold-memref-subview-ops pass
Support affine.load/store ops in fold-memref-subview ops pass. The
existing pass just "inlines" the subview operation on load/stores by
inserting affine.apply ops in front of the memref load/store ops: this
is by design always consistent with the semantics on affine.load/store
ops and the same would work even more naturally/intuitively with the
latter.
Differential Revision: https://reviews.llvm.org/D118565
Amir Ayupov [Sun, 30 Jan 2022 07:48:48 +0000 (23:48 -0800)]
[BOLT][TEST] Fix building some tests with clang-14 by passing -no-pie
Reviewed By: yota9
Differential Revision: https://reviews.llvm.org/D118556
Craig Topper [Mon, 31 Jan 2022 02:08:00 +0000 (18:08 -0800)]
[RISCV] Use existing variable intead of calling getOperand again. NFCI
This is a slight change because I'm using the ANY_EXTEND result
instead of the original operand, but getNode should constant fold.
While there, add a comment about why the code specifically checks
for a ConstantSDNode.
Uday Bondhugula [Sun, 30 Jan 2022 11:44:24 +0000 (17:14 +0530)]
[MLIR][NFC] Update SCF pass cmd line names to prefix scf
Update SCF pass cmd line names to prefix `scf`. This is consistent with
guidelines/convention on how to name dialect passes. This also avoids
ambiguity on the context given the multiple `for` operations in the
tree.
NFC.
Differential Revision: https://reviews.llvm.org/D118564
Fangrui Song [Mon, 31 Jan 2022 01:15:45 +0000 (17:15 -0800)]
[ELF] splitStrings: replace entSize==1 special case with manual loop unswitch. NFC
My x86-64 lld executable is actually smaller.
Craig Topper [Sun, 30 Jan 2022 22:32:56 +0000 (14:32 -0800)]
[RISCV] Add more pack and packw test case for Zbkb. NFC
Make sure we cover the encodings use for zext.h and other encodings
not used for zext.h.
Craig Topper [Sun, 30 Jan 2022 22:07:35 +0000 (14:07 -0800)]
[RISCV] Merge rv64zbkb-valid.s and rv64zbkb-only-valid.s. NFC
Based on the existing naming "only" tests are used for rv32 instructions
that don't exist in rv64. rv32 tests without "only" are for instructions
that are in both rv32 and rv64. The rv64 tests are for instructions
that are only in rv64.
Both of these test files have instruction encodings that are only
valid in rv64 so they can be the same file.
Craig Topper [Sun, 30 Jan 2022 22:03:45 +0000 (14:03 -0800)]
[RISCV] Rename rv64-zbkb-valid.s to rv64zbkb-valid.s. NFC
Craig Topper [Sun, 30 Jan 2022 21:58:21 +0000 (13:58 -0800)]
[RISCV] Fix bad CHECK prefix in rv32zbkb-valid.s.
"pack t0, t1, zero" disassembles to "pack t0, t1, zero" with Zbkb
not "zext.h t0, t1"
Part of the test was using a CHECK prefix that doesn't appear on
the RUN line.
Kazu Hirata [Mon, 31 Jan 2022 00:05:00 +0000 (16:05 -0800)]
[clang-tools-extra] Remove unused forward declarations (NFC)
Kazu Hirata [Mon, 31 Jan 2022 00:04:58 +0000 (16:04 -0800)]
[Analysis] Drop an unnecessary const from a return type (NFC)
Identified with readability-const-return-type.
Kazu Hirata [Mon, 31 Jan 2022 00:04:56 +0000 (16:04 -0800)]
[llvm] Use = default (NFC)
Fangrui Song [Sun, 30 Jan 2022 23:18:30 +0000 (15:18 -0800)]
[mlgo][regalloc] Fix -Wunused-variable in -DLLVM_ENABLE_ASSERTIONS=off builds after
a8a7bf922cea8af01168f7a4adf4ed0365bcc2b4
Fangrui Song [Sun, 30 Jan 2022 23:12:09 +0000 (15:12 -0800)]
[ELF] Simplify SectionBase::partition handling and make it live by default. NFC
Previously an InputSectionBase is dead (`partition==0`) by default.
SyntheticSection calls markLive and BssSection overrides that with markDead.
It is more natural to make InputSectionBase live by default and let
--gc-sections mark InputSectionBase dead.
When linking a Release build of clang:
* --no-gc-sections:, the removed `inputSections` loop decreases markLive time from 4ms to 1ms.
* --gc-sections: the extra `inputSections` loop increases markLive time from 0.181296s to 0.188526s.
This is as of we lose the removing one `inputSections` loop optimization (
4374824ccf6e7ae68264d996a9ae5bb5e3be7fc5).
I believe the loss can be mitigated if we refactor markLive.
Mircea Trofin [Sun, 30 Jan 2022 22:56:38 +0000 (14:56 -0800)]
[mlgo][regalloc] Fix register masking
If AllocationOrder has less than 32 elements, we were treating the extra
positions as if they were valid. This was detected by a subsequent
assert. The fix also tightens the asserts.
Alexandre Ganea [Sun, 30 Jan 2022 22:31:26 +0000 (17:31 -0500)]
[mlir] Silence warnings when building with MSVC
Differential Revision: https://reviews.llvm.org/D118536
Sanjoy Das [Sun, 30 Jan 2022 02:41:10 +0000 (18:41 -0800)]
Replace OwningModuleRef with OwningOpRef<ModuleOp>
This addresses a TODO in BuiltinOps.h.
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D118574
Fangrui Song [Sun, 30 Jan 2022 21:34:27 +0000 (13:34 -0800)]
[ELF] Change splitSections to objectFiles based parallelForEach. NFC
The work is more balanced.
Craig Topper [Sun, 30 Jan 2022 21:18:43 +0000 (13:18 -0800)]
[RISCV] Lower riscv_zip/unzip intrinsic to RISCVISD::SHFL/UNSHFL.
These are special versions of the more general shfli/unshfli
instructions. We can use the general ISD opcodes with the correct
immediates.
Markus Böck [Sun, 30 Jan 2022 21:08:53 +0000 (22:08 +0100)]
[Support][NFC] Fix generic `ChildrenGetterTy` of `IDFCalculatorBase`
Both IDFCalculatorBase and its accompanying DominatorTreeBase only supports pointer nodes. The template argument is the block type itself and any uses of GraphTraits is therefore done via a pointer to the node type.
However, the ChildrenGetterTy type of IDFCalculatorBase has a use on just the node type instead of a pointer to the node type. Various parts of the monorepo has worked around this issue by providing specializations of GraphTraits for the node type directly, or not been affected by using specializations instead of the generic case. These are unnecessary however and instead the generic code should be fixed instead.
An example from within Tree is eg. A use of IDFCalculatorBase in InstrRefBasedImpl.cpp. It basically instantiates a IDFCalculatorBase<MachineBasicBlock, false> but due to the bug above then goes on to specialize GraphTraits<MachineBasicBlock> although GraphTraits<MachineBasicBlock*> exists (and should be used instead).
Similar dead code exists in clang which defines redundant GraphTraits to work around this bug.
This patch fixes both the original issue and removes the dead code that was used to work around the issue.
Differential Revision: https://reviews.llvm.org/D118386
Craig Topper [Sun, 30 Jan 2022 20:36:15 +0000 (12:36 -0800)]
[RISCV] Custom lower brev8 intrinsic to RISCVISD::GREV.
We can use the RISCVISD::GREV encoding that swaps the bits in
each byte. This allows it to use the existing computeKnownBits
support for RISCVISD::GREV.
Kazu Hirata [Sun, 30 Jan 2022 20:32:59 +0000 (12:32 -0800)]
[OpenMP] Use nullptr instead of NULL (NFC)
Identified with modernize-use-nullptr.
Kazu Hirata [Sun, 30 Jan 2022 20:32:57 +0000 (12:32 -0800)]
[Analysis] Use != to compare strings (NFC)
Identified with readability-string-compare.
Kazu Hirata [Sun, 30 Jan 2022 20:32:55 +0000 (12:32 -0800)]
[clang] Remove redundant string initialization (NFC)
Identified with readability-redundant-string-init.
Kazu Hirata [Sun, 30 Jan 2022 20:32:53 +0000 (12:32 -0800)]
[lldb] Forward-declare ClangExpressionParser (NFC)
ClangUserExpression.h is relying on the forward declaration of
ClangExpressionParser in ClangFunctionCaller.h. This patch moves the
forward declaration to ClangUserExpression.h.
Kazu Hirata [Sun, 30 Jan 2022 20:32:51 +0000 (12:32 -0800)]
[CodeGen] Use default member initialization (NFC)
Identified with modernize-use-default-member-init.
Simon Pilgrim [Sun, 30 Jan 2022 20:06:59 +0000 (20:06 +0000)]
[X86] combineVectorTruncation - use PACKUSDW(BLENDW(X,0),BLENDW(Y,0)) for v8i32->v8i16 truncation
Limit this to SSE41 - AVX1 targets to avoid UNPCKL(PSHUFB,PSHUFB), pre-SSE41 we don't have PACKUSDW/BLENDW and with AVX2 we can perform this as PERMQ(PSHUFB()).
Anatoly Trosinenko [Sun, 30 Jan 2022 18:23:00 +0000 (21:23 +0300)]
[compiler-rt][builtins] Use c[tl]zsi macro instead of __builtin_c[tl]z
`__builtin_c[tl]z` accepts `unsigned int` argument that is not always the
same as uint32_t. For example, `unsigned int` is uint16_t on MSP430.
Reviewed By: aykevl
Differential Revision: https://reviews.llvm.org/D86547
Matthias Springer [Sun, 30 Jan 2022 20:01:24 +0000 (05:01 +0900)]
[mlir][bufferize] Clean up remaining references to `tensor-constant-bufferize`
Philip Sigillito [Sun, 30 Jan 2022 19:45:32 +0000 (20:45 +0100)]
[clang-format] Handle C variables with name that matches c++ access specifier
Reviewed By: MyDeveloperDay, curdeius, HazardyKnusperkeks
Differential Revision: https://reviews.llvm.org/D117416
Mehdi Amini [Sun, 30 Jan 2022 11:24:36 +0000 (11:24 +0000)]
Apply clang-tidy fixes for llvm-include-order in AttrOrTypeFormatGen.cpp (NFC)
Mehdi Amini [Sun, 30 Jan 2022 10:59:42 +0000 (10:59 +0000)]
Apply clang-tidy fixes for performance-move-const-arg in TestLinalgCodegenStrategy.cpp (NFC)
Mehdi Amini [Sun, 30 Jan 2022 10:17:41 +0000 (10:17 +0000)]
Apply clang-tidy fixes for readability-identifier-naming in TosaOps.cpp (NFC)
Mehdi Amini [Sun, 30 Jan 2022 09:56:28 +0000 (09:56 +0000)]
Apply clang-tidy fixes for readability-identifier-naming in PolynomialApproximation.cpp (NFC)
Mehdi Amini [Sun, 30 Jan 2022 09:41:27 +0000 (09:41 +0000)]
Apply clang-tidy fixes for performance-for-range-copy in SCFInterfaceImpl.cpp (NFC)
Mehdi Amini [Sun, 30 Jan 2022 09:34:39 +0000 (09:34 +0000)]
Apply clang-tidy fixes for modernize-use-equals-default in BufferizableOpInterface.cpp (NFC)
Mehdi Amini [Sun, 30 Jan 2022 09:03:19 +0000 (09:03 +0000)]
Apply clang-tidy fixes for readability-identifier-naming in AsyncToLLVM.cpp (NFC)
Anatoly Trosinenko [Sun, 30 Jan 2022 16:28:08 +0000 (19:28 +0300)]
[compiler-rt][builtins] Use explicitly-sized integer types for LibCalls
Use s[iu]_int instead of `(unsigned) int` and d[ui]_int instead of
`(unsigned) long long` for LibCall arguments.
Note: the `*vfp` LibCall versions were NOT touched.
Reviewed By: aykevl
Differential Revision: https://reviews.llvm.org/D86546
Keith Smiley [Sat, 29 Jan 2022 02:46:51 +0000 (18:46 -0800)]
[lld-macho] Add support for -add_empty_section
This is a ld64 option equivalent to `-sectcreate seg sect /dev/null`
that's useful for creating sections like the RESTRICT section.
Differential Revision: https://reviews.llvm.org/D117749
Simon Pilgrim [Sun, 30 Jan 2022 17:59:05 +0000 (17:59 +0000)]
[X86][AVX] matchUnaryShuffle - avoid creation of on-the-fly nodes (PR45974)
Don't extract the ANY/ZERO_EXTEND_VECTOR_INREG subvector source until we're definitely combining to a new node.
Keith Smiley [Sat, 29 Jan 2022 01:03:02 +0000 (17:03 -0800)]
[test][lld-macho] Improve LC_FUNCTION_STARTS test coverage
Previously functions that aren't included in the symtab were also
excluded from the function starts. Symbols missing from function starts
degrades the debugger experience in the case you don't have debug info
for them.
Differential Revision: https://reviews.llvm.org/D114275
Simon Pilgrim [Sun, 30 Jan 2022 15:53:21 +0000 (15:53 +0000)]
[X86] Attempt to fold MOVMSK(CMPEQ(AND(X,C1),0)) -> MOVMSK(NOT(SHL(X,C2)))
Allows pow2 mask tests to avoid an unnecessary constant load.
Noticed while investigating how to extend MatchVectorAllZeroTest to support more allof/anyof patterns.
Simon Pilgrim [Sun, 30 Jan 2022 15:25:07 +0000 (15:25 +0000)]
[X86] Add tests showing failure to fold MOVMSK(CMPEQ(AND(X,C1),0)) -> MOVMSK(NOT(SHL(X,C2)))
This would allow pow2 mask tests to avoid an unnecessary constant load.
Noticed while investigating how to extend MatchVectorAllZeroTest to support more allof/anyof patterns.
Ricky Zhou [Sun, 30 Jan 2022 15:05:11 +0000 (10:05 -0500)]
[InstCombine] Do not combine atomic and non-atomic loads
Before this change, InstCombine was willing to fold atomic and
non-atomic loads through a PHI node as long as the first PHI argument
is not an atomic load. The combined load would be non-atomic, which is
incorrect.
Fix this by only combining the loads in a PHI node when all of the
arguments are non-atomic loads.
Thanks to Eli Friedman for pointing out the bug at
https://github.com/llvm/llvm-project/issues/50777#issuecomment-
981045342!
Fixes #50777
Differential Revision: https://reviews.llvm.org/D115113
Ricky Zhou [Sun, 30 Jan 2022 15:03:49 +0000 (10:03 -0500)]
[InstCombine] Add test reproducing PR51435 (NFC)
Baseline test for D115113
Differential Revision: https://reviews.llvm.org/D118554
Sanjay Patel [Sun, 30 Jan 2022 15:02:20 +0000 (10:02 -0500)]
Revert "[InstCombine] Do not combine atomic and non-atomic loads"
This reverts commit
a2bb7313e62daa8aaa9b06fe57386ed7221a990a.
The baseline test was not pre-committed as expected.
Ricky Zhou [Sun, 30 Jan 2022 14:28:56 +0000 (09:28 -0500)]
[InstCombine] Do not combine atomic and non-atomic loads
Before this change, InstCombine was willing to fold atomic and
non-atomic loads through a PHI node as long as the first PHI argument
is not an atomic load. The combined load would be non-atomic, which is
incorrect.
Fix this by only combining the loads in a PHI node when all of the
arguments are non-atomic loads.
Thanks to Eli Friedman for pointing out the bug at
https://github.com/llvm/llvm-project/issues/50777#issuecomment-
981045342!
Fixes #50777
Differential Revision: https://reviews.llvm.org/D115113
Ricky Zhou [Sun, 30 Jan 2022 14:08:42 +0000 (09:08 -0500)]
[InstCombine] Use range for loops (NFC)
Preliminary clean-up for D115113
Differential Revision: https://reviews.llvm.org/D116086
Ricky Zhou [Sun, 30 Jan 2022 13:57:48 +0000 (08:57 -0500)]
[InstCombine] Uppercase some variable names (NFC)
Uppercase some variable names, per LLVM coding standards. This change
intentionally does not rename every miscased variable, as a follow-up
change ( D116086 ) intends to eliminate many of those by switching
loops to range for loops.
Differential Revision: https://reviews.llvm.org/D118553
Matthias Springer [Sun, 30 Jan 2022 13:19:06 +0000 (22:19 +0900)]
[mlir][linalg][bufferize] Fix insertion point InitTensorElimination
There was a bug where some of the OpOperands needed in the replacement op were not in scope.
It does not matter where the replacement op is inserted. Any insertion point is OK as long as there are no dominance errors. In the worst case, the newly inserted op will bufferize out-of-place. This is no worse than not eliminating the InitTensorOp at all.
Differential Revision: https://reviews.llvm.org/D117685
Mark de Wever [Sun, 30 Jan 2022 13:21:13 +0000 (14:21 +0100)]
[libc++][format][nfc] Move tests.
The formatter specialization tests were placed in the wrong
subdirectory. This moves them to the proper place.
Mark de Wever [Sat, 22 Jan 2022 11:34:44 +0000 (12:34 +0100)]
[libc++][doc] Update the release notes.
I had a look at the changes since the last release and updated the
release notes with interesting changes.
It seems this time the release notes were already rather up to date :-)
If there are more interesting changes, please let me know and I'll
update the patch. I'd like to commit these changes latest next weekend
so they land before branching the 14.0 release.
I've added most active libc++ contributors. If I forgot anybody please add them.
Reviewed By: Quuxplusone, ldionne, philnik, #libc
Differential Revision: https://reviews.llvm.org/D117948
Matthias Springer [Sun, 30 Jan 2022 13:02:22 +0000 (22:02 +0900)]
[mlir][bufferize][NFC] Move std BufferizableOpInterfaceImpl to std dialect
Also reimplement `std-bufferize` in terms of BufferizableOpInterface-based bufferization. The old `std.select` bufferization pattern is no longer needed and deleted.
Differential Revision: https://reviews.llvm.org/D118559
Florian Hahn [Sun, 30 Jan 2022 13:07:29 +0000 (13:07 +0000)]
[VPlan] Use VPlan to check if only the first lane is used.
This removes the remaining dependence on LoopVectorizationCostModel from
buildScalarSteps and is required so it can be moved out of ILV.
It also improves allows us to remove a few unneeded instructions.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D116554
Matthias Springer [Sun, 30 Jan 2022 12:53:02 +0000 (21:53 +0900)]
[mlir][bufferize][NFC] Move scf BufferizableOpInterface impl to scf dialect
Differential Revision: https://reviews.llvm.org/D118557
Matthias Springer [Sun, 30 Jan 2022 12:48:37 +0000 (21:48 +0900)]
Update BUILD.bazel.
This should have been done as part of D118483.
Matthias Springer [Sun, 30 Jan 2022 12:23:28 +0000 (21:23 +0900)]
[mlir][bufferize] Merge tensor-constant-bufferize into arith-bufferize
The bufferization of arith.constant ops is also switched over to BufferizableOpInterface-based bufferization. The old implementation is deleted. Both implementations utilize GlobalCreator, now renamed to just `getGlobalFor`.
GlobalCreator no longer maintains a set of all created allocations to avoid duplicate allocations of the same constant. Instead, `getGlobalFor` scans the module to see if there is already a global allocation with the same constant value.
For compatibility reasons, it is still possible to create a pass that bufferizes only `arith.constant`. This pass (createConstantBufferizePass) could be deleted once all users were switched over to One-Shot bufferization.
Differential Revision: https://reviews.llvm.org/D118483
Nuno Lopes [Sun, 30 Jan 2022 11:34:32 +0000 (11:34 +0000)]
[InstCombine] remove incorrect gep(x, undef) -> undef optimization
gep(x, undef) carries the provenance of x, so we can't replace it with any
pointer like undef.
This leaves room for improvement for the poison case, but that's currently
not possible as the demanded bits API doesn't distinguish between undef &
poison bits.
Fixes #44790
Nuno Lopes [Sun, 30 Jan 2022 10:46:54 +0000 (10:46 +0000)]
[InstSimplify] fold 'xor X, poison' and 'div/rem X, poison' to poison
Nuno Lopes [Sun, 30 Jan 2022 10:04:00 +0000 (10:04 +0000)]
[NewGVN][NFC] add poison tests
Fangrui Song [Sun, 30 Jan 2022 09:18:41 +0000 (01:18 -0800)]
[ELF] copyShtGroup: replace unordered_set<uint32_t> with DenseSet<uint32_t>. NFC
We don't need to support the empty/tombstone key section index.
Fangrui Song [Sun, 30 Jan 2022 09:11:19 +0000 (01:11 -0800)]
[ELF] crtbegin/crtend test: replace std::regex with hand-written matcher. NFC
My x86-64 lld executable is 18KiB smaller.
Fangrui Song [Sun, 30 Jan 2022 09:08:09 +0000 (01:08 -0800)]
[ELF][test] Test {crtbegin,crtend}{S,T}.o
Fangrui Song [Sun, 30 Jan 2022 08:44:19 +0000 (00:44 -0800)]
[ELF] Simplify maybeCompress with lld::split. NFC
Fangrui Song [Sun, 30 Jan 2022 08:10:52 +0000 (00:10 -0800)]
[ELF] Optimize MergeInputSection::splitNonStrings with resize_for_overwrite. NFC
Fangrui Song [Sun, 30 Jan 2022 07:55:29 +0000 (23:55 -0800)]
[ELF] Add some Mips*Section to InStruct and change make<Mips*Section> to std::make_unique
Similar to D116143. My x86-64 lld executable is 20+KiB smaller.
Fangrui Song [Sun, 30 Jan 2022 07:35:15 +0000 (23:35 -0800)]
[ELF] Remove make<std::unique_ptr<MemoryBuffer>>. NFC
Fangrui Song [Sun, 30 Jan 2022 07:05:54 +0000 (23:05 -0800)]
[ELF] De-template getErrorPlace. NFC
Craig Topper [Sun, 30 Jan 2022 06:19:15 +0000 (22:19 -0800)]
[RISCV] Remove RISCVISD::BREV8 and use RISCVISD::GREV instead.
We already have an ISD opcode for the more general GREV/GREVI
instructon. We can just use it with the encoding that corresponds
to the behavior of brev8. This is similar to what we do for orc.b
where we use the GORC ISD opcode.
Craig Topper [Sun, 30 Jan 2022 05:15:11 +0000 (21:15 -0800)]
[RISCV] Rerrange RISCVInstrInfoZB.td to better group related wthings. NFC
Especially placing W instructions/patterns near their non-W versions.
Fangrui Song [Sun, 30 Jan 2022 04:11:59 +0000 (20:11 -0800)]
[ELF] De-template getAndFeatures. NFC
Richard [Sat, 29 Jan 2022 01:14:33 +0000 (18:14 -0700)]
[clang-tidy] Organize the release notes a little better
- Sort new checks by check name
- Sort changes to existing checks by check name
- Add docs for changes to readability-simplify-boolean-expr
- Move check changes from "Improvements to clang-tidy" to
"Changes in existing checks" section
Differential Revision: https://reviews.llvm.org/D118519
Fangrui Song [Sun, 30 Jan 2022 03:31:09 +0000 (19:31 -0800)]
[ELF] Make errorOrWarn opaque to decrease code size. NFC
In my x86-64 lld, .text is -3.08Ki smaller.
Craig Topper [Sun, 30 Jan 2022 02:26:20 +0000 (18:26 -0800)]
[RISCV] Use RVBUnary to simplify ZEXT_H_RV32/ZEXT_H_RV64 definitions. NFC
Ben Shi [Fri, 28 Jan 2022 10:09:12 +0000 (10:09 +0000)]
[clang][AVR] Set '-fno-use-cxa-atexit' to default
AVR is baremetal environment, so the avr-libc does not support
'__cxa_atexit()'.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D118445