jacquesguan [Tue, 3 Jan 2023 03:58:18 +0000 (11:58 +0800)]
[RISCV] Fix compile warning.
jacquesguan [Wed, 7 Dec 2022 07:41:27 +0000 (15:41 +0800)]
[RISCV] Use vfirst.m to extract the first element from mask vector.
This patch uses vfirst.m to extract the first bit of mask.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D139512
Kazu Hirata [Tue, 3 Jan 2023 03:18:46 +0000 (19:18 -0800)]
[polly] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Owen Pan [Tue, 3 Jan 2023 03:06:00 +0000 (19:06 -0800)]
[clang-format][NFC] Remove CRs at EOLs introduced in
617277e7cbda
Kazu Hirata [Tue, 3 Jan 2023 03:13:11 +0000 (19:13 -0800)]
[flang] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Kazu Hirata [Tue, 3 Jan 2023 02:56:09 +0000 (18:56 -0800)]
[mlir] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Owen Pan [Tue, 3 Jan 2023 02:49:25 +0000 (18:49 -0800)]
[clang-format][NFC] Remove a superfluous semicolon after \version
eopXD [Thu, 29 Dec 2022 09:05:54 +0000 (01:05 -0800)]
[NFC][Clang][RISCV] Reduce boilerplate when determining prototype for segment loads
No functionality change of the RVV builtin and compiler intrinsics is intended
in this patch.
This patch gathers scattered comments for the segment load builtin/intrinsics
and its variants (e.g. segment unit-stride load, segment strided load) into a
single paragraph under riscv_vector.td.
This patch also tries to reduce one level of the if-statements as the push_back
are essentially the same actions but differs in index based on the the value
of the policy attributes and whether the intrinsic is masked.
Reviewed By: kito-cheng
Differential Revision: https://reviews.llvm.org/D140662
Kazu Hirata [Tue, 3 Jan 2023 02:40:21 +0000 (18:40 -0800)]
[BOLT] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Kazu Hirata [Tue, 3 Jan 2023 02:29:04 +0000 (18:29 -0800)]
[lld] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Lang Hames [Tue, 3 Jan 2023 01:48:55 +0000 (17:48 -0800)]
[ORC] Rename ThreadSafeModule::takingModuleDo to consumingModuleDo.
Renamed to emphasize intent: this method allows the module to be consumed by the
callback (while protected by the context lock), but we don't want to imply that
the Module could be taken out of the ThreadSafeModule (where it would no longer
be protected by that lock).
Lang Hames [Tue, 3 Jan 2023 01:33:13 +0000 (17:33 -0800)]
[ORC] Add ThreadSafeModule::takingModuleDo for consuming operations on TSMs.
ThreadSafeModule::takingModuleDo passes ownership of the contained llvm::Module
into the callback, allowing ThreadSafeModules to be used with consuming
operations like Linker::link.
Kazu Hirata [Tue, 3 Jan 2023 00:17:40 +0000 (16:17 -0800)]
[clang-tools-extra] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Kazu Hirata [Mon, 2 Jan 2023 23:54:57 +0000 (15:54 -0800)]
[clang] Use std::optional instead of llvm::Optional (NFC)
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
Haojian Wu [Mon, 2 Jan 2023 23:02:15 +0000 (00:02 +0100)]
Dmitri Gribenko [Mon, 2 Jan 2023 22:34:24 +0000 (23:34 +0100)]
[clang][dataflow] Remove unused includes
Krzysztof Drewniak [Mon, 5 Dec 2022 22:17:34 +0000 (22:17 +0000)]
[mlir][ROCDL] Translate known block size attributes to ROCDL
1. When converting from the GPU dialect to the ROCDL dialect, if the
function that contains a gpu.thread_id or gpu.block_id op is annotated
with gpu.known_{block,grid}_size, use that size to set a "range"
attribute on the corresponding rocdl intrinsic so that the LLVM
frontend can optimize based on that range information.
1b. When translating from the rocdl dialect to LLVM IR, use the
"range" attribute, if present, to set !range metadata on the relevant
function call.
2. Deprecate the old rocdl.max_flat_work_group_size attribute, which
was used in a tensorflow backend. Instead, use
rocdl.flat_work_group_size going forward to allow kernel generators to
specify the minimum and maximum work group sizes a kernel may be
launched with in one attribute, thus more closely matching the backend.
3. When translating from gpu.func to llvm.func within gpu-to-rocdl,
copy the known_block_size attribute as rocdl.reqd_work_group_size to
enable further translations to set the corresponding metadata on the
LLVM IR function. Also, set the rocdl.flat_work_group_size attribute
to ensure that the reqd_work_group_size metadata and the
amdgpu-flat-work-group-size metadata are consistent.
3b. Extend the ROCDL to LLVM IR translation to set the
!reqd_work_group_size metadata on LLVM functions
Also update tests and add functions to the ROCDL dialect to ensure
attribute names are used consistently.
Depends on D139865
Reviewed By: antiagainst
Differential Revision: https://reviews.llvm.org/D139866
Peter Rong [Mon, 2 Jan 2023 20:57:35 +0000 (12:57 -0800)]
precommit formatting of LowerSwitch.cpp for D140747
Signed-off-by: Peter Rong <PeterRong96@gmail.com>
Sanjay Patel [Mon, 2 Jan 2023 18:31:01 +0000 (13:31 -0500)]
[InstCombine] fold mask with not-of-sext-bool to select
~sext(A) & Op1 --> A ? 0 : Op1
With no extra uses, this pattern is already reduced,
but we would miss it in examples such as issue #59773.
https://alive2.llvm.org/ce/z/WGLcSR
Sanjay Patel [Mon, 2 Jan 2023 17:42:43 +0000 (12:42 -0500)]
[InstCombine] add tests for not-sext-bool+mask; NFC
Sanjay Patel [Fri, 30 Dec 2022 15:31:10 +0000 (10:31 -0500)]
[InstCombine] early exit to reduce indents in foldSelectIntoOp(); NFC
Roman Lebedev [Mon, 2 Jan 2023 18:15:31 +0000 (21:15 +0300)]
[InstCombine] Canonicalize math-y conditional negation into a `select`
https://alive2.llvm.org/ce/z/vPs-gZ
This is a larger pattern than would seem necessary, with minimal being:
* `and` https://alive2.llvm.org/ce/z/q9-MqK
* `or` https://alive2.llvm.org/ce/z/AUUEMZ
* `xor` https://alive2.llvm.org/ce/z/dm3Ume
.. so for all others, we canonicalize away from math to `select`,
but there we canonicalize in the opposite direction.
Fixes https://github.com/llvm/llvm-project/issues/59791
Roman Lebedev [Mon, 2 Jan 2023 17:51:51 +0000 (20:51 +0300)]
[NFC][InstCombine] Add tests for conditional negation via math
https://github.com/llvm/llvm-project/issues/59791
Florian Hahn [Mon, 2 Jan 2023 17:52:12 +0000 (17:52 +0000)]
[VPlan] Use VP_CLASSOF_IMPL for VPWidenCanonicalIVRecipe(NFC).
Replace VPWidenCanonicalIVRecipe::classof implementation with general
VP_CLASSOF_IMPL.
Florian Hahn [Mon, 2 Jan 2023 17:32:31 +0000 (17:32 +0000)]
[VPlan] Use VP_CLASSOF_IMPL for VPWidenMemoryInstructionRecipe (NFC).
Replace VPWidenMemoryInstructionRecipe ::classof implementation with general
VP_CLASSOF_IMPL.
Florian Hahn [Mon, 2 Jan 2023 17:22:34 +0000 (17:22 +0000)]
[VPlan] Use VP_CLASSOF_IMPL for VPPredInstPHI (NFC).
Replace VPPredInstPHI::classof implementation with general
VP_CLASSOF_IMPL.
Krzysztof Drewniak [Thu, 29 Dec 2022 21:36:12 +0000 (21:36 +0000)]
Re-land D139865 "Add known_block_size and known_grid_size to gpu.func"
This should fix the MSVC warning that caused the previous revert.
Reviewed By: antiagainst
Differential Revision: https://reviews.llvm.org/D140766
Nikita Popov [Mon, 2 Jan 2023 15:53:31 +0000 (16:53 +0100)]
[LoopRotate] Convert tests to opaque pointers (NFC)
Nikita Popov [Mon, 2 Jan 2023 15:58:57 +0000 (16:58 +0100)]
[LoopRotate] Make test more robust (NFC)
Use a non-zero GEP index, so it does not get dropped during opaque
pointer conversion.
Nikita Popov [Mon, 2 Jan 2023 15:57:13 +0000 (16:57 +0100)]
[LoopRotate] Name instructions in test (NFC)
Nikita Popov [Mon, 2 Jan 2023 15:51:41 +0000 (16:51 +0100)]
[LoopPredication] Convert tests to opaque pointers (NFC)
Roman Lebedev [Mon, 2 Jan 2023 15:22:54 +0000 (18:22 +0300)]
[DAGCombiner] `convertBuildVecZextToBuildVecWithZeros()`: rework split factor calculation
The original computation was both making assumptions that do not hold
in practice, and being overly pessimistic. We should just check
every possible split factor, and pick the best one.
Fixes https://github.com/llvm/llvm-project/issues/59781
Muhammad Omair Javaid [Mon, 2 Jan 2023 15:30:18 +0000 (19:30 +0400)]
[LLDB] Enable TestSignedTypes.py for AArch64/Windows
TestSignedTypes.py passes on AArch64 Windows. This patch disables this
test only for i[3-6]86 and x86_64 Windows platform.
Nikita Popov [Mon, 2 Jan 2023 14:28:10 +0000 (15:28 +0100)]
[InstCombine] Support vectors in icmp of GEP fold
EmitGEPOffset() supports vector GEPs nowadays, so we don't need
any further code changes.
compare_gep_with_base_vector1 shows a weakness in folding the
resulting comparison if an index splat has to be performed.
Nikita Popov [Mon, 2 Jan 2023 14:17:34 +0000 (15:17 +0100)]
[InstCombine] Add tests for icmp of vector GEP (NFC)
These show up as regressions with opaque pointers.
Jonas Hahnfeld [Sat, 31 Dec 2022 12:41:51 +0000 (13:41 +0100)]
[JITLink][RISCV] Order EdgeKind_riscv the same way as relocations
There were basically four different orderings: one defined by the
relocations, one by the enum definition of EdgeKind_riscv, one for
mapping the enum values to their names, and one when mapping the
relocations to edge kinds and finally processing them. Chose the
ordering defined by the relocations in the riscv-elf-psabi-doc as
the canonical one.
Differential Revision: https://reviews.llvm.org/D140802
Nikita Popov [Mon, 2 Jan 2023 13:37:02 +0000 (14:37 +0100)]
[ValueTracking] Use SmallVector for non-undef/poison ops
The way these APIs are used, there isn't really a benefit to
deduplicating the ops as part of the API. The only place that
benefits from this is PoisonChecking, and for that particular
use the assertion emission was potentially non-deterministic.
We should populate a vector for deterministic order and then
deduplicate via a separate set.
Oleksandr "Alex" Zinenko [Mon, 2 Jan 2023 13:14:40 +0000 (14:14 +0100)]
[mlir] NFC use the correct markdown header in BufferizationTransformOps.td
Daan De Meyer [Mon, 19 Dec 2022 11:53:02 +0000 (11:53 +0000)]
[llvm-objcopy] Use getNumberOfSymbols() instead of getRawNumberOfSymbols()
getRawNumberOfSymbols() assumes that a symbol table exists, which isn't
always guaranteed, while getNumberOfSymbols() handles and tolerates objects
without a symbol table. When there is a symbol table, both methods return
the same value.
Also add a test to ensure we don't regress in this regard. The test
generates a basic COFF object with symbols and overrides the symbol table
pointer with zeros to craft the input required to verify llvm-objcopy works
as expected in this scenario.
Adrian Kuegel [Mon, 2 Jan 2023 12:05:26 +0000 (13:05 +0100)]
[mlir][Bazel] Port
cf487cce6f64
Nikita Popov [Mon, 2 Jan 2023 10:16:09 +0000 (11:16 +0100)]
[ValueTracking] Treat branch on undef as UB as well
We were already treating branch on poison as UB, but branch on
undef is also UB. Move the checks into the correct function.
From LangRef for br:
> If ‘cond’ is poison or undef, this instruction has undefined behavior.
From LangRef for switch:
> If ‘value’ is poison or undef, this instruction has undefined behavior.
There is a minor regression in dont-distribute-phi.ll, apparently
we handle that pattern in logical but not bitwise form.
Nikita Popov [Mon, 2 Jan 2023 11:19:39 +0000 (12:19 +0100)]
[SimpleLoopUnswitch] Perform poison query before transform
I think this doesn't make any difference right now, but once
we take into account that branch on undef is UB in
programUndefinedIfUndefOrPoison() the new position of the branch
would imply that the condition can't be poison, which would
defeat the purpose of the freeze insertion here. We need to
perform the check before the branch is moved.
Tamas Berghammer [Mon, 2 Jan 2023 11:24:00 +0000 (06:24 -0500)]
[clang-tidy] Support std::string_view in readability-redundant-string-cstr
Previously we were matching constructor calls for std::string and
llvm::StringRef and this change extends this set with including
std::string_view as well.
Reviewed By: njames93, carlosgalvezp
Differential Revision: https://reviews.llvm.org/D140018
Tobias Gysi [Mon, 2 Jan 2023 10:34:51 +0000 (11:34 +0100)]
[mlir][llvm] Make the import of LLVM IR intrinsics extensible.
The revision introduces the LLVMImportDialectInterface to make the
import of LLVM IR intrinsics extensible. It uses a dialect interface
that enables external projects to provide their own conversion functions
for custom intrinsics. These conversion functions can rely on the
ModuleImport class to perform support tasks such as mapping LLVM
values to MLIR values or for converting types between the two worlds.
The implementation largely mirrors the export implementation. One major
difference is the dispatch to the appropriate dialect interface, since
LLVM IR intrinsics have no direct association to an MLIR dialect. The
dialect interfaces thus have to publish the supported intrinsics to
ensure incoming conversion calls are dispatched to the right dialect
interface.
The revision implements the extensible intrinsic import discussed as
part of the "extensible llvm ir import" rfc:
https://discourse.llvm.org/t/rfc-extensible-llvm-ir-import/67256/6
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D140374
Nikita Popov [Mon, 2 Jan 2023 10:02:27 +0000 (11:02 +0100)]
[ValueTracking] Remove branch-on-poison-as-ub flag (NFC)
This has been enabled by default without issue for a while now,
remove the flag.
Qfrost [Mon, 2 Jan 2023 09:20:15 +0000 (17:20 +0800)]
[LLD][Windows]Feature "checksum" for Windows PE
Differential Revision: https://reviews.llvm.org/D139184
Adrian Kuegel [Mon, 2 Jan 2023 07:41:31 +0000 (08:41 +0100)]
[mlir][MemRef] Apply ClangTidy performance fix (NFC).
luxufan [Mon, 2 Jan 2023 05:48:39 +0000 (13:48 +0800)]
[JITLink][RISCV] Fix incorrectly use of uint32_t
Owen Pan [Sun, 1 Jan 2023 23:13:36 +0000 (15:13 -0800)]
[clang-format][docs] Fix indentation for the Sphinx build
See https://reviews.llvm.org/D140543#4021209.
Owen Pan [Sun, 1 Jan 2023 06:06:28 +0000 (22:06 -0800)]
[clang-format][NFC] Clean up IntegerLiteralSeparatorFixer::process
Jeff Bailey [Sun, 1 Jan 2023 20:48:24 +0000 (20:48 +0000)]
[libc] [nfc] Fix heading underline.
Differential Revision: https://reviews.llvm.org/D140825
Roman Lebedev [Sun, 1 Jan 2023 19:36:49 +0000 (22:36 +0300)]
[DAGCombiner][X86] Fold a CONCAT_VECTORS of SHUFFLE_VECTOR and it's operand into wider SHUFFLE_VECTOR
This was showing as a source of *many* regressions
with more aggressive ZERO_EXTEND_VECTOR_INREG recognition.
Roman Lebedev [Sun, 1 Jan 2023 18:34:02 +0000 (21:34 +0300)]
[NFC][X86] Add tests for concatenation of shuffle's operand to the shuffle
Florian Hahn [Sun, 1 Jan 2023 19:48:48 +0000 (19:48 +0000)]
[VPlan] Adjust mergeReplicateRegions to be in line with mergeBlock (NFC)
Adjust mergeReplicateRegions to be in line with
mergeBlocksIntoPredecessors added in
36d70a6aea6b by collecting only the
valid candidates first.
Also rename to mergeReplicateRegionsIntoSuccessors and add missing
doc-comment.
This addresses post-commit suggestions by @Ayal.
Tom Praschan [Sat, 31 Dec 2022 15:24:55 +0000 (16:24 +0100)]
[clangd] Add extension for adding context (enclosing function or class) in references results
Relevant issue: https://github.com/clangd/clangd/issues/177
Reviewed By: nridge
Differential Revision: https://reviews.llvm.org/D137894
Kevin Sala [Sat, 31 Dec 2022 20:07:56 +0000 (21:07 +0100)]
[OpenMP][libomptarget] Add TargetParser as dependency in NextGen's JIT
This patch fixes an undefined reference to llvm::Triple::Triple(llvm::Twine const&).
Differential Revision: https://reviews.llvm.org/D140810
Ben Shi [Mon, 12 Dec 2022 03:40:57 +0000 (11:40 +0800)]
[lld][ELF] Support relocation R_AVR_LDS_STS_16 on AVRTiny devices
The relocation 'R_AVR_LDS_STS_16' is introduced for the compact
16-bit LDS/STS instructions on AVRTiny devices.
Reviewed By: MaskRay, aykevl
Differential Revision: https://reviews.llvm.org/D139807
Fangrui Song [Sun, 1 Jan 2023 06:39:20 +0000 (22:39 -0800)]
[X86] Revert -fno-plt __tls_get_addr workaround for old GNU ld
ENABLE_X86_RELAX_RELOCATIONS has defaulted to on in 2020.
This workaround is not exercised for a long time.
Fangrui Song [Sun, 1 Jan 2023 06:24:37 +0000 (22:24 -0800)]
[clang] Change CodeGenOptions::RelaxELFRelocations/assembler defaults to match MC default
MC default was flipped in 2016.
CMake ENABLE_X86_RELAX_RELOCATIONS defaults to on in 2020 (
c41a18cf61790fc898dcda1055c3efbf442c14c0).
It makes sense for the CodeGenOptions::RelaxELFRelocations to match, so
that most -cc1/-cc1as command lines won't have this option.
This also fixes a minor issue: -fno-plt -S will now use GOT for
__tls_get_addr calls, matching -fno-plt -c.
Fangrui Song [Sun, 1 Jan 2023 04:50:54 +0000 (20:50 -0800)]
[ELF] Support TLS GD/LD relaxations for x86-32 -fno-plt
For x86-32, {clang,gcc} -fno-plt uses `call *___tls_get_addr@GOT(%reg)` instead
of `call ___tls_get_addr@PLT`. GD to IE/LE relaxations need to shift the offset
by one while LD to LE relaxation needs to use a different code sequence.
While here, fix some comments.
Fix https://github.com/llvm/llvm-project/issues/59769
Differential Revision: https://reviews.llvm.org/D140813
LLVM GN Syncbot [Sun, 1 Jan 2023 01:57:45 +0000 (01:57 +0000)]
[gn build] Port
89aad1e6a397
Owen Pan [Thu, 22 Dec 2022 09:21:17 +0000 (01:21 -0800)]
Reland [clang-format] Add an option to format integer literal separators
Previously committed in
46c94e5067b5 which was reverted in
f0756e086010
due to a memory bug.
Closes #58949.
Differential Revision: https://reviews.llvm.org/D140543
Markus Böck [Sat, 31 Dec 2022 14:04:31 +0000 (15:04 +0100)]
[clang][NFC] Fix warning due to initializing a pointer with `false`
Florian Hahn [Sat, 31 Dec 2022 15:14:57 +0000 (15:14 +0000)]
[VPlan] Move GraphTraits definitions to separate header (NFC).
This reduces the size of VPlan.h and avoids future growth of the file
when the graph traits are extended in future patches.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D140500
Uday Bondhugula [Sat, 31 Dec 2022 14:56:40 +0000 (20:26 +0530)]
[MLIR] Fix affine LICM pass for unknown region holding ops
Fix affine LICM pass for unknown region-holding ops. The logic was
completely ignoring regions of unknown ops leading to generation of
invalid IR on hoisting. Handle affine.parallel op among those with
regions that are supported.
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D140738
Uday Bondhugula [Sat, 24 Dec 2022 10:23:36 +0000 (15:53 +0530)]
NFC. Fix/improve style in affine-licm pass
The code here appears to be out of line with proper style and
guidelines. Fix this.
Mark de Wever [Fri, 23 Dec 2022 17:10:00 +0000 (18:10 +0100)]
[NFC][libc++] Removes concepts tests.
Concepts should be supported by all supported compilers. This is
probably a left over from an older version of the expected patch.
Reviewed By: #libc, philnik
Differential Revision: https://reviews.llvm.org/D140632
Jie Fu [Sat, 31 Dec 2022 07:31:33 +0000 (23:31 -0800)]
[docs] TestingGuide.rst: Fix incorrect description
This patch fixes two incorrect descriptions in TestingGuide.rst.
1. test/lit.site.cfg --> test/lit.site.cfg.py
After https://reviews.llvm.org/D37838 , the `test/lit.site.cfg` had been added a .py extension.
So it should be `test/lit.site.cfg.py`.
2. $(LLVM_OBJ_ROOT)/$(BuildMode)/bin --> $(LLVM_OBJ_ROOT)/bin
The current build system doesn't create a $(BuildMode) directory any more.
So it should be removed.
Reviewed By: mehdi_amini, MaskRay
Differential Revision: https://reviews.llvm.org/D140780
Fangrui Song [Sat, 31 Dec 2022 04:33:27 +0000 (20:33 -0800)]
[test] Opaquify cgprofile.ll
Qfrost [Sat, 31 Dec 2022 03:48:58 +0000 (11:48 +0800)]
llvm-readobj COFFDumper print PEHeader CheckSum
Differential Revision:https://reviews.llvm.org/D140555
Fangrui Song [Sat, 31 Dec 2022 03:43:19 +0000 (19:43 -0800)]
[msan] Remove non-working MSAN_LINUX_X86_64_OLD_MAPPING
r249754 (2015) added this macro (not set anywhere) for an escape hatch.
This old layout has been non-working for many years (Linux
arch/x86/include/asm/elf.h ELF_ET_DYN_BASE is outside the APP range):
FATAL: Code 0x558547327980 is out of application range. Non-PIE build?
FATAL: MemorySanitizer can not mmap the shadow memory.
FATAL: Make sure to compile with -fPIE and to link with -pie.
FATAL: Disabling ASLR is known to cause this error.
FATAL: If running under GDB, try 'set disable-randomization off'.
Non-pie doesn't work either.
Yeting Kuo [Mon, 26 Dec 2022 09:52:56 +0000 (17:52 +0800)]
[RISCV] Use tail agnostic if inserting subvector/element at the end of a vector.
The patch tries to make more vslidup nodes use tail agnostic. The idea comes
from D125546 authored by Zack Chen.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D140669
Fangrui Song [Sat, 31 Dec 2022 02:23:54 +0000 (18:23 -0800)]
[CMake] Remove unused ALL_SCUDO_SUPPORTED_ARCH after D138157
Roman Lebedev [Sat, 31 Dec 2022 01:58:22 +0000 (04:58 +0300)]
[CVP] When expanding `urem`, always freeze the nominator
As per the post-commit feedback - that was not the correct precondition
to avoid it here. I think we should generally start changing mentality
about `freeze`, the fact that we have been conditioned to be afraid of it
(or of anything in LLVM in general) is the key problem here.
Roman Lebedev [Sat, 31 Dec 2022 01:43:29 +0000 (04:43 +0300)]
[NFC][CVP] Add more tests for urem expansion
As mentioned in post-commit feedback.
Roman Lebedev [Sat, 31 Dec 2022 00:23:54 +0000 (03:23 +0300)]
[DAGCombiner][TLI] Do not fuse bitcast to <1 x ?> into a load/store of a vector
Single-element vectors are legalized by splitting,
so the the memory operations would also get scalarized.
While we do have some support to reconstruct scalarized loads,
we clearly don't catch everything.
The comment for the affected AArch64 store suggests that
having two stores was the desired outcome in the first place.
This was showing as a source of *many* regressions
with more aggressive ZERO_EXTEND_VECTOR_INREG recognition.
Roman Lebedev [Sat, 31 Dec 2022 00:10:43 +0000 (03:10 +0300)]
[NFC][Codegen][x86] Add tests for load/store of a single-element vectors
Roman Lebedev [Fri, 30 Dec 2022 23:07:50 +0000 (02:07 +0300)]
[NFC][TLI] Move `isLoadBitCastBeneficial()` implementation into source file
... so any change to it does not cause 700 source files to be recompiled.
Roman Lebedev [Fri, 30 Dec 2022 00:49:48 +0000 (03:49 +0300)]
[DAG] BUILD_VECTOR: absorb ZERO_EXTEND of a single first operand if all other ops are zeros
This kind of pattern seems to come up as regressions
with better ZERO_EXTEND_VECTOR_INREG recognition.
For initial implementation, this is quite restricted
to the minimal viable transform, otherwise there are
too many regressions to be dealt with.
Valery N Dmitriev [Fri, 30 Dec 2022 19:07:04 +0000 (11:07 -0800)]
[NFC] Test case intended to cover SLP cost for chain with masked gather loads.
SLP produces two gather loads (one feeds another).
For the first set of scalar loads GEP indices are all constant.
The result of the second load is then fed into reduction (as a seed).
Differential Revision: https://reviews.llvm.org/D140785
Eugene Zhulenev [Thu, 29 Dec 2022 21:24:45 +0000 (13:24 -0800)]
[mlir] Add a test for default valued dictionary attributes
+ Replace special placeholders ($_builder, etc...) in default attribute string
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D140765
Roman Lebedev [Fri, 30 Dec 2022 18:25:14 +0000 (21:25 +0300)]
[NFC][X86] Add exhaustive-ish coverage for broadcast of implicitly aext/zext element
Some of these even crash instruction selection for AVX512.
This is one of the patterns that comes up as regressions
with more aggressive ZERO_EXTEND_VECTOR_INREG recognition.
https://godbolt.org/z/x88aqfrT5
Vitaly Buka [Fri, 30 Dec 2022 18:57:12 +0000 (10:57 -0800)]
Revert "[CodeGen] Temporarily disable-lsr in HWASAN build"
We can do the same with cmake on the bot.
This reverts commit
8f70b848d339cabfaa8f1379d41dae11b9b75014.
Benjamin Kramer [Fri, 30 Dec 2022 18:07:19 +0000 (19:07 +0100)]
[Sema] Avoid double hash lookup. NFCI
Florian Hahn [Fri, 30 Dec 2022 18:24:34 +0000 (18:24 +0000)]
[VPlan] Also consider operands of sink candidates in same block.
Even if the the sink candidate is already in the target block, its
operands can be candidates for sinking. Queue them up as well. Also
moves the queuing logic to a helper.
Thomas Symalla [Fri, 30 Dec 2022 13:25:51 +0000 (14:25 +0100)]
[NFC][AMDGPU] Make method declarations in SIInstrInfo equivalent to their definitions.
Some functions from SIInstrInfo have their operands named different in
their declarations vs. their defs. This was caught by cppcheck.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D140778
Benjamin Kramer [Fri, 30 Dec 2022 17:02:26 +0000 (18:02 +0100)]
[AArch64] Stringref'ize AArch64Subtarget constructor. NFCI
Ivan Butygin [Thu, 22 Dec 2022 19:50:52 +0000 (20:50 +0100)]
[mlir][gpu][spirv] Lower gpu reduction ops to spirv
Supports only "add" and "mul" ops for now. More ops will be added later.
Differential Revision: https://reviews.llvm.org/D140576
Roman Lebedev [Fri, 30 Dec 2022 16:06:59 +0000 (19:06 +0300)]
[CVP] Expand bound `urem`s
This kind of thing happens really frequently in LLVM's very own
shuffle combining methods, and it is even considered bad practice
to use `%` there, instead of using this expansion directly.
Though, many of the cases there have variable divisors,
so this won't help everything.
Simple case: https://alive2.llvm.org/ce/z/PjvYf-
There's alternative expansion via `umin`:
https://alive2.llvm.org/ce/z/hWCVPb
BUT while we can transform the first expansion
into the `umin` one (e.g. for SCEV):
https://alive2.llvm.org/ce/z/iNxKmJ
... we can't go in the opposite direction.
Also, the non-`umin` expansion seems somewhat more codegen-friendly:
https://godbolt.org/z/qzjx5bqWK
https://godbolt.org/z/a7bj1axbx
There's second variant of precondition:
https://alive2.llvm.org/ce/z/zE6cbM
but there the numerator must be non-undef / must be frozen.
Roman Lebedev [Fri, 30 Dec 2022 16:15:43 +0000 (19:15 +0300)]
[NFC][CVP] `processURem()`: add statistic and increase readability
Roman Lebedev [Fri, 30 Dec 2022 16:03:11 +0000 (19:03 +0300)]
[NFC][CVP] Add tests for urem expansion
Roman Lebedev [Fri, 30 Dec 2022 14:12:21 +0000 (17:12 +0300)]
[NFC][PhaseOrdering] Re-autogenerate check lines in one test
Matt Arsenault [Fri, 30 Dec 2022 16:08:33 +0000 (11:08 -0500)]
ValueTracking: Fix canCreateUndefOrPoison for saturating shifts
These need to consider the shift amount.
Ivan Kosarev [Fri, 30 Dec 2022 15:05:21 +0000 (15:05 +0000)]
[AMDGPU][AsmParser] Refine parsing cache policy modifiers.
Reviewed By: dp, arsenm
Differential Revision: https://reviews.llvm.org/D140108
Teresa Johnson [Thu, 29 Dec 2022 20:11:38 +0000 (12:11 -0800)]
[MemProf] Fix inline propagation of memprof metadata
It isn't correct to always remove memprof metadata MIBs from the
original allocation call after inlining.
Let's say we have the following partial call graph:
C D
\ /
v v
B E
| /
v v
A
where A contains an allocation call. If both contexts including B have
the same allocation behavior, the context in the memprof metadata on the
allocation will be pruned, and we will have 2 MIBs with contexts:
A,B and A,E.
Previously, if we inlined A into B we propagate the matching MIBs onto
the inlined allocation call in B' (A,B in this case), and remove it from
the original out of line allocation in A. This is correct if we have a
single round of bottom up inlining.
However, in the compiler we can have multiple invocations of the inliner
pass (e.g. LTO). We may also inline non-bottom up with an alternative
inliner such as the ModuleInliner. In that case, we could end up first
inlining B into C, without having inlined A into B. The call graph then
looks like:
D
|
v
C' B E
\ | /
v v v
A
If we subsequently (perhaps on a later invocation of bottom up inlining)
inline A into B, the previous handling would propagate the memprof MIB
context A,B up into the inlined allocation in B', and remove it from the
original allocation in A. The propagation into B' is fine, however, by
removing it from A's allocation, we no longer reflect the context coming
from C'.
To fix this, simply prevent the removal of MIB from the original
allocation callsites.
Note that the memprof_inline.ll test has some changes to existing
checking to replace "noncold" with "notcold" in the metadata. The
corresponding CHECK was accidentally commented out in the old version
and thus this mistake was not previously detected.
Differential Revision: https://reviews.llvm.org/D140764
Alexey Bataev [Thu, 22 Dec 2022 20:07:44 +0000 (12:07 -0800)]
[SLP]Do not emit many extractelements, reuse the single one emitted.
We do not need to emit many extractelements for each particular use, we
can reuse the only one, just need to adjust it to make it dominate on
all uses.
Differential Revision: https://reviews.llvm.org/D140580
Sanjay Patel [Fri, 30 Dec 2022 13:47:51 +0000 (08:47 -0500)]
[InstSimplify] fold selects where true/false arm is the same as condition
We managed to fold related patterns in issue #59704,
but we were missing these more basic folds:
https://alive2.llvm.org/ce/z/y6d7SN
Sanjay Patel [Thu, 29 Dec 2022 19:51:30 +0000 (14:51 -0500)]
[InstSimplify] add tests for select-of-bool; NFC
Matt Arsenault [Fri, 30 Dec 2022 13:40:31 +0000 (08:40 -0500)]
IROutliner: Convert tests to opaque pointers
Some of these show improvements. outlining-bitcasts.ll might not be
relevant anymore (or should be rewritten to test some other type of
non-pointer bitcast).
Dmitry Preobrazhensky [Fri, 30 Dec 2022 13:29:47 +0000 (16:29 +0300)]
[AMDGPU][GFX11] Correct tied src2 of v_fmac_f16_e64
src2 was incorrectly defined as VSrc_f16 but it is tied to dst which is VGPR_32. As a result, disassembler failed to decode src2.
Differential Revision: https://reviews.llvm.org/D140299