Commit Graph

421597 Commits

Author SHA1 Message Date
Matthias Springer a3005a406e [mlir][interfaces] Fix infinite loop in insideMutuallyExclusiveRegions
This function was missing a termination condition.
2022-04-19 16:28:52 +09:00
Mehdi Amini 4e01184ad5 Apply clang-tidy fixes for performance-unnecessary-value-param in JitRunner.cpp (NFC) 2022-04-19 07:23:12 +00:00
Mehdi Amini 722a3a58e2 Apply clang-tidy fixes for performance-for-range-copy in MemRefOps.cpp (NFC) 2022-04-19 07:23:12 +00:00
Chuanqi Xu cd149dbf8e [NFC] Remove unused variable 2022-04-19 15:19:40 +08:00
Matthias Springer 0f4ba02db3 [mlir][interfaces] Add helpers for detecting recursive regions
Add helper functions to check if an op may be executed multiple times based on RegionBranchOpInterface.

Differential Revision: https://reviews.llvm.org/D123789
2022-04-19 16:13:32 +09:00
Fraser Cormack c5cac48549 [RISCV] Fix lowering of BUILD_VECTORs as VID sequences
This patch fixes a bug when lowering BUILD_VECTOR via VID sequences.
After adding support for fractional steps in D106533, elements with zero
steps may be skipped if no step has yet been computed. This allowed
certain sequences to slip through the cracks, being identified as VID
sequences when in fact they are not.

The fix for this is to perform a second loop over the BUILD_VECTOR to
validate the entire sequence once the step has been computed. This isn't
the most efficient, but on balance the code is more readable and
maintainable than doing back-validation during the first loop.

Fixes the tests introduced in D123785.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D123786
2022-04-19 07:43:38 +01:00
Fraser Cormack 00537946aa [RISCV] Add tests showing incorrect BUILD_VECTOR lowering
These tests both use vector constants misidentified as VID sequences.
Because the initial run of elements has a zero step, the elements are
skipped until such a step can be identified. The bug is that the skipped
elements are never validated, even though the computed step is
incompatible across the entire sequence.

A fix will follow in a subseqeuent patch.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D123785
2022-04-19 07:00:48 +01:00
Austin Kerbow 7f97ac94f7 Revert "[AMDGPU] Omit unnecessary waitcnt before barriers"
This reverts commit 8d0c34fd4f.
2022-04-18 21:24:08 -07:00
Konstantin Varlamov bcdb11e741 [libc++][NFC] Reindent `take_view` in accordance with the style guide. 2022-04-18 20:54:50 -07:00
Yaxun (Sam) Liu cac4e2fe25 [CUDA][HIP] Fix gpu.used.external
Rename gpu.used.external as __clang_gpu_used_external as ptxas does not
allow . in global variable name.

Fixes: https://github.com/llvm/llvm-project/issues/54934

Reviewed by: Joseph Huber, Artem Belevich

Differential Revision: https://reviews.llvm.org/D123946
2022-04-18 23:10:31 -04:00
Joseph Huber 80787213ea [Libomptarget] Fix test using old unsupported lit string
Summary:
One test had an old "unsupported" string that used the old `newDriver`
string which was removed. This test should be updated to use the
`oldDriver` one instead.
2022-04-18 23:08:12 -04:00
Chuanqi Xu f9bee35689 [Pipelines] Hoist CoroEarly as a module pass
This change could reduce the time we call `declaresCoroEarlyIntrinsics`.
And it is helpful for future changes.

Reviewed By: aeubanks

Differential Revision: https://reviews.llvm.org/D123925
2022-04-19 11:04:24 +08:00
Michael Kruse 2d92ee97f1 Reapply "[OpenMP] Refactor OMPScheduleType enum."
This reverts commit af0285122f.

The test "libomp::loop_dispatch.c" on builder
openmp-gcc-x86_64-linux-debian fails from time-to-time.
See #54969. This patch is unrelated.
2022-04-18 21:56:47 -05:00
jacquesguan 25445b94db [RISCV] Add rvv codegen support for vp.fptrunc.
This patch adds rvv codegen support for vp.fptrunc. The lowering of fp_round and vp.fptrunc share most code so use a common lowering function to handle those two, similar to vp.trunc.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D123841
2022-04-19 01:56:18 +00:00
Mehdi Amini 1881d6fc80 Apply clang-tidy fixes for performance-unnecessary-copy-initialization in MathOps.cpp (NFC) 2022-04-19 00:47:58 +00:00
Mehdi Amini 6120bd4703 Apply clang-tidy fixes for performance-for-range-copy in ElementwiseOpFusion.cpp (NFC) 2022-04-19 00:47:58 +00:00
Kirill Stoimenov 70f13bd752 Revert "[ASan] Fixed a reporting bug in (load|store)N functions which would print unknown-crash instead of the proper error message when a the data access is unaligned."
This reverts commit d81d317999.

Reviewed By: kstoimenov

Differential Revision: https://reviews.llvm.org/D123966
2022-04-18 23:33:41 +00:00
Vasileios Porpodas b1333f03d9 Recommit "[SLP] Support internal users of splat loads"
Code review: https://reviews.llvm.org/D121940

This reverts commit 359dbb0d3d.
2022-04-18 15:58:01 -07:00
Kirill Stoimenov d81d317999 [ASan] Fixed a reporting bug in (load|store)N functions which would print unknown-crash instead of the proper error message when a the data access is unaligned.
Reviewed By: kda, eugenis

Differential Revision: https://reviews.llvm.org/D123643
2022-04-18 22:46:37 +00:00
Stanislav Mekhanoshin c1c49a3561 [AMDGPU] Fix comment type in the DSInstructions.td. NFC. 2022-04-18 14:28:12 -07:00
John McIver 3787de40de [llvm-objcopy] Make llvm-strip --only-keep-debug suppress default --strip-all
Fixes #54417

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D123798
2022-04-18 14:16:10 -07:00
Groverkss 15650b320b [MLIR][Presburger] Remove inheritence in MultiAffineFunction
This patch removes inheritence of MultiAffineFunction from IntegerPolyhedron
and instead makes IntegerPolyhedron as a member.

This patch removes virtualization in MultiAffineFunction and also removes
unnecessary functions inherited from IntegerPolyhedron.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D123921
2022-04-19 01:25:13 +05:30
Stefan Pintilie ef34442232 [NFC][PowerPC] Move the Regsiter Operands for PowerPC into PPCRegisterInfo.td
Currently the regsiter operand definitions are found in three separate files.
This patch moves all of the definitions into PPCRegisterInfo.td.

Reviewed By: amyk

Differential Revision: https://reviews.llvm.org/D123543
2022-04-18 14:50:24 -05:00
Michael Kruse af0285122f Revert "[OpenMP] Refactor OMPScheduleType enum."
This reverts commit 9ec501da76.

It may have caused the openmp-gcc-x86_64-linux-debian buildbot to fail.
https://lab.llvm.org/buildbot/#/builders/4/builds/20377
2022-04-18 14:38:31 -05:00
Sanjay Patel 3a27b51b27 [InstCombine] reduce code for freeze of undef
The description was ambiguous about the behavior
when boths select arms are constant or both arms
are not constant. I don't think there's any
evidence to support either way, but this matches
the code with a more specified description.

We can extend this to deal with vector constants
with undef/poison elements. Currently, those don't
get folded anywhere.
2022-04-18 15:14:02 -04:00
Sanjay Patel 60de144119 [InstCombine] add tests for select with frozen condition; NFC 2022-04-18 15:14:02 -04:00
Vasileios Porpodas 359dbb0d3d Revert "[SLP] Support internal users of splat loads"
This reverts commit f8e1337115.
2022-04-18 12:12:34 -07:00
Joseph Huber 15e62062c0 [Clang][Docs] Update information on the new driver now that it's default
Summary:
This patch updates some of the documentation on the new driver now that
it's the default. Also the ABI for embedding these images changed.
2022-04-18 15:05:09 -04:00
Joseph Huber ae23be84cb [OpenMP] Make the new offloading driver the default
Previously an opt-in flag `-fopenmp-new-driver` was used to enable the
new offloading driver. After passing tests for a few months it should be
sufficiently mature to flip the switch and make it the default. The new
offloading driver is now enabled if there is OpenMP and OpenMP
offloading present and the new `-fno-openmp-new-driver` is not present.

The new offloading driver has three main benefits over the old method:
- Static library support
- Device-side LTO
- Unified clang driver stages

Depends on D122683

Differential Revision: https://reviews.llvm.org/D122831
2022-04-18 15:05:09 -04:00
Michael Kruse 9ec501da76 [OpenMP] Refactor OMPScheduleType enum.
The OMPScheduleType enum stores the constants from libomp's internal sched_type in kmp.h and are used by several kmp API functions. The enum values have an internal structure, namely each scheduling algorithm (e.g.) exists in four variants: unordered, orderend, normerge unordered, and nomerge ordered.

This patch (basically a followup to D114940) splits the "ordered" and "nomerge" bits into separate flags, as was already done for the "monotonic" and "nonmonotonic", so we can apply bit flags operations on them. It also now contains all possible combinations according to kmp's sched_type. Deriving of the OMPScheduleType enum from clause parameters has been moved form MLIR's OpenMPToLLVMIRTranslation.cpp to OpenMPIRBuilder to make available for clang as well. Since the primary purpose of the flag is the binary interface to libomp, it has been made more private to LLVMFrontend. The primary interface for generating worksharing-loop using OpenMPIRBuilder code becomes `applyWorkshareLoop` which derives the OMPScheduleType automatically and calls the appropriate emitter function.

While this is mostly a NFC refactor, it still applies the following functional changes:
 * The logic from OpenMPToLLVMIRTranslation to derive the OMPScheduleType also applies to clang. Most notably, it now applies the nonmonotonic flag for non-static schedules by default.
 * In OpenMPToLLVMIRTranslation, the nonmonotonic default flag was previously not applied if the simd modifier was used. I assume this was a bug, since the effect was due to `loop.schedule_modifier()` returning `mlir::omp::ScheduleModifier::none` instead of `llvm::Optional::None`.
 * In OpenMPToLLVMIRTranslation, the nonmonotonic default flag was set even if ordered was specified, in breach to what the comment before citing the OpenMP specification says. I assume this was an oversight.

The ordered flag with parameter was not considered in this patch. Changes will need to be made (e.g. adding/modifying function parameters) when support for it is added. The lengthy names of the enum values can be discussed, for the moment this is avoiding reusing previously existing enum value names such as `StaticChunked` to avoid confusion.

Reviewed By: peixin

Differential Revision: https://reviews.llvm.org/D123403
2022-04-18 14:03:17 -05:00
River Riddle 58ceae9561 [mlir:NFC] Remove the forward declaration of FuncOp in the mlir namespace
FuncOp has been moved to the `func` namespace for a little over a month, the
using directive can be dropped now.
2022-04-18 12:01:55 -07:00
Vasileios Porpodas f8e1337115 [SLP] Support internal users of splat loads
Until now we would only accept a broadcast load pattern if it is only used
by a single vector of instructions.

This patch relaxes this, and allows for the broadcast to have more than one
user vector, as long as all of its uses are internal to the SLP graph and
vectorized.

Differential Revision: https://reviews.llvm.org/D121940
2022-04-18 11:59:44 -07:00
chenglin.bi 222adf338a [Arch64][SelectionDAG] Add target-specific implementation of srem
1. X%C to the equivalent of X-X/C*C is not always fastest path if there is no SDIV pair exist. So check target have faster for srem only first.
2. Add AArch64 faster path for SREM only pow2 case.

Fix https://github.com/llvm/llvm-project/issues/54649

Reviewed By: efriedma

Differential Revision: https://reviews.llvm.org/D122968
2022-04-19 02:49:42 +08:00
Mehdi Amini b968c59050 Improve terminator doc in MLIR LangRef
Reviewed By: rriddle, bondhugula

Differential Revision: https://reviews.llvm.org/D123886
2022-04-18 18:43:30 +00:00
Stefan Pintilie bc9916fff2 [NFC][PowerPC] Style and ordering changes for PPCInstrP10.td
Renamed the two classes 8LS_DForm_R_SI34_RTA5 and 8LS_DForm_R_SI34_XT6_RA5 to
8LS_DForm_R_SI34_RTA5_MEM and 8LS_DForm_R_SI34_XT6_RA5_MEM because the
instructions that use the classes use memory reads/writes.

Moved the instruction defs up closer to the classes.
Removed unnecessary whitespace.
2022-04-18 13:28:48 -05:00
Mehdi Amini 21b251624b Apply clang-tidy fixes for readability-identifier-naming in GPUDialect.cpp (NFC) 2022-04-18 18:15:30 +00:00
Mehdi Amini c817628b67 Apply clang-tidy fixes for readability-simplify-boolean-expr in TypeConverter.cpp (NFC) 2022-04-18 18:15:30 +00:00
Richard Smith e43c93dd63 Don't treat 'T &forward(T&&)' as builtin.
This allows the standard library to diagnose it properly. Suppress
warning in libc++ testsuite for unused result of call to std::forward.
2022-04-18 11:11:21 -07:00
Peter Klausler b8a929cb2f [flang] Fix regression with recent work on intrinsic/generic interactions
When resolving a procedure reference, do not allow a successful
intrinsic procedure probe result to override an existing
symbol.

Differential Revision: https://reviews.llvm.org/D123905
2022-04-18 11:10:41 -07:00
Joseph Huber ba01306009 [Libomptarget] Fix LIBOMPTARGET_INFO test
Summary:
A patch added a new line to one of the info outputs without updating
this test. This patch adds the new text to the existing test.
2022-04-18 14:09:02 -04:00
Wenlei He 526af13eba Fix llvm-profgen breakage 2022-04-18 11:02:29 -07:00
Arthur Eubanks e0736e7429 [ocaml bindings] Remove LTO bindings
Followup to D123882.
2022-04-18 10:22:26 -07:00
Arthur Eubanks 2e6ac54cf4 [LegacyPM] Remove ThinLTO/LTO pipelines
Using the legacy PM for the optimization pipeline was deprecated in 13.0.0.
Following recent changes to remove non-core features of the legacy
PM/optimization pipeline, remove the (Thin)LTO pipelines.

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D123882
2022-04-18 10:09:41 -07:00
Arthur Eubanks a7e20a8a7a [CallPrinter] Port CallPrinter passes to new pass manager
Port the legacy CallGraphViewer and CallGraphDOTPrinter to work with the new pass manager.

Addresses issue https://github.com/llvm/llvm-project/issues/54323
Adds back related tests that were removed in commits d53a4e7b4a and 9e9d9aba14

Reviewed By: aeubanks

Differential Revision: https://reviews.llvm.org/D122989
2022-04-18 10:02:18 -07:00
Wenlei He 17f6cba30d [llvm-profgen] Add process filter for perf reader
For profile generation, we need to filter raw perf samples for binary of interest. Sometimes binary name along isn't enough as we can have binary of the same name running in the system. This change adds a process id filter to allow users to further disambiguiate the input raw samples.

Differential Revision: https://reviews.llvm.org/D123869
2022-04-18 09:50:16 -07:00
Jacques Pienaar f4085c57dd [mlir] Fix two AttributeParser aborts
Reproducers that resulted in triggering the following asserts

mlir::NamedAttribute::NamedAttribute(mlir::StringAttr, mlir::Attribute)
mlir/lib/IR/Attributes.cpp:29:3
consumeToken
mlir/lib/Parser/Parser.h:126

Differential Revision: https://reviews.llvm.org/D122240
2022-04-18 09:30:35 -07:00
Blue Gaston a6efcf1c31 Format fix in recent change. Removed whitespace. [NFC] 2022-04-18 09:23:01 -07:00
Zarko Todorovski ce87133120 [llvm][IPO] Inclusive language: Rename mergefunc-sanity to mergefunc-verify and remove other instances of sanity in MergeFunctions.cpp
This patch renames the mergefunc-sanity to mergefunc-verify and renames the related functions to use more
inclusive language

Reviewed By: cebowleratibm

Differential Revision: https://reviews.llvm.org/D114374
2022-04-18 11:50:08 -04:00
Eli Friedman 13fc178173 Force GHashCell to be 8-byte-aligned.
Otherwise, with recent versions of libstdc++, clang can't tell that the
atomic operations are properly aligned, and generates calls to
libatomic.  (Actually, because of the use of reinterpret_cast, it wasn't
guaranteed to be aligned, but I think it ended up being aligned in
practice.)

Fixes https://github.com/llvm/llvm-project/issues/54790 , the part where
LLVM failed to build.

Differential Revision: https://reviews.llvm.org/D123872
2022-04-18 08:46:03 -07:00
Corentin Jabot daa6d7b250 [Clang] Use of decltype(capture) in parameter-declaration-clause
Partially implement the proposed resolution to CWG2569.

D119136 broke some libstdc++ code, as P2036R3, implemented as a DR to
C++11 made ill-formed some previously valid and innocuous code.

We resolve this issue to allow decltype(x) - but not decltype((x)
to appear in the parameter list of a lambda that capture x by copy.

Unlike CWG2569, we do not extend that special treatment to
sizeof/noexcept yet, as the resolution has not been approved yet
and keeping the review small allows a quicker fix of impacted code.

Reviewed By: aaron.ballman

Differential Revision: https://reviews.llvm.org/D123909
2022-04-18 15:58:25 +02:00