bixia1 [Mon, 17 Oct 2022 17:02:17 +0000 (10:02 -0700)]
[mlir][sparse] Add options to sparse-tensor-rewrite to disable rewriting rules for operators foreach and convert.
This is to help simplify FileCheck tests for sparse-tensor-rewrite.
Reviewed By: aartbik
Differential Revision: https://reviews.llvm.org/D136093
Sam McCall [Fri, 14 Oct 2022 10:56:41 +0000 (12:56 +0200)]
[include-cleaner] Add include-cleaner tool, with initial HTML report
The immediate goal is to start producing an HTML report to debug and explain
include-cleaner recommendations.
For now, this includes only the lowest-level piece: a list of the references
found in the source code.
How this fits into future ideas:
- under refs we can also show the headers providing the symbol, which includes
match those headers etc
- we can also annotate the #include lines with which symbols they cover, and
add whichever includes we're suggesting too
- the include-cleaner tool will likely have modes where it emits diagnostics
and/or applies edits, so the HTML report is behind a flag
Differential Revision: https://reviews.llvm.org/D135956
Mingming Liu [Sat, 15 Oct 2022 18:50:56 +0000 (11:50 -0700)]
[AArch64] Enhance bit-field-positioning op matcher to see through 'any_extend' for pattern 'and(any_extend(shl(val, N)), shifted-mask)'
Before this patch (and refactor patch D135843), isBitfieldPositioningOp won't handle "and(any_extend(shl(val, N), shifted-mask)" (bail out if AND op is not SHL)
After this patch, isBitfieldPositioningOp will see through "any_extend" to find "shl" to find possible bit-field-positioning nodes.
https://gcc.godbolt.org/z/3ncGKbGW6 is a four-liner LLVM IR that could be optimized to UBFIZ (see added test case test_and_extended_shift_with_imm in llvm/test/CodeGen/AArch64/bitfield-insert.ll). One existing test case also improves.
Differential Revision: https://reviews.llvm.org/D135852
Han-Kuan Chen [Tue, 18 Oct 2022 06:32:12 +0000 (23:32 -0700)]
[RISCV] Lower VECTOR_SHUFFLE to VSLIDEDOWN_VL.
Differential Revision: https://reviews.llvm.org/D136136
Han-Kuan Chen [Tue, 18 Oct 2022 04:39:20 +0000 (21:39 -0700)]
[RISCV] Pre-commit tests for lowering VECTOR_SHUFFLE to VSLIDEDOWN_VL.
Differential Revision: https://reviews.llvm.org/D136135
Anton Sidorenko [Mon, 17 Oct 2022 09:06:08 +0000 (12:06 +0300)]
[MachineCombiner][RISCV] Enable MachineCombiner for RISCV
Initial implementation to match basic FP reassociation patterns.
Differential Revision: https://reviews.llvm.org/D135264
Alexey Bataev [Thu, 18 Nov 2021 23:59:30 +0000 (15:59 -0800)]
[SLP]Generalize cost model.
Generalized the cost model estimation. Improved cost model estimation
for repeated scalars (no need to count their cost anymore), improved
cost model for extractelement instructions.
cpu2017
511.povray_r 0.57
520.omnetpp_r -0.98
521.wrf_r -0.01
525.x264_r 3.59 <+
526.blender_r -0.12
531.deepsjeng_r -0.07
538.imagick_r -1.42
Geometric mean: 0.21
Differential Revision: https://reviews.llvm.org/D115757
Arthur Eubanks [Tue, 11 Oct 2022 22:07:13 +0000 (15:07 -0700)]
Port print-cfg-sccs to new pass manager
This is actually used, see https://discourse.llvm.org/t/use-print-callgrapg-sccs-from-opt/65782.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D135718
Arthur Eubanks [Tue, 18 Oct 2022 15:46:14 +0000 (08:46 -0700)]
[win][compiler-rt] Make tests use lld-link instead of link
Git bash ships with a link.exe. We try to add git bash to the beginning
of PATH (see D84380). These tests end up executing the wrong link.exe.
As a workaround, use lld-link. Note that `REQUIRES: lld-available` tests currently aren't running, see D128567. I did manually verify that these tests pass with lld-link.
Reviewed By: rnk, hans
Differential Revision: https://reviews.llvm.org/D136108
Arthur Eubanks [Tue, 18 Oct 2022 15:45:15 +0000 (08:45 -0700)]
[ubsan][test] Make some tests have shorter file names
We're hitting path size limits on Windows on these tests. As a
workaround, make the file names shorter.
Reviewed By: vitalybuka
Differential Revision: https://reviews.llvm.org/D136113
Jordan Rupprecht [Tue, 18 Oct 2022 15:45:10 +0000 (08:45 -0700)]
[Bazel] Add target for llvm-debuginfo-analyzer binary
Arthur Eubanks [Tue, 18 Oct 2022 00:03:35 +0000 (17:03 -0700)]
[SROA] Don't speculate phis with different load user types
Fixes an SROA crash.
Fallout from opaque pointers since with typed pointers we'd bail out at the bitcast.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D136119
Arthur Eubanks [Tue, 18 Oct 2022 04:14:05 +0000 (21:14 -0700)]
[llvm-reduce] Check if reduction fails/is redundant before invoking oracle
So we don't over count the number of chunks and do unnecessary work reducing more chunks than exist.
This lowers some random reduction I tested with locally from 250s to 232s.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D136127
Chris Bieneman [Tue, 18 Oct 2022 15:43:12 +0000 (10:43 -0500)]
Make llvm-config test resilliant
Add explicit library list to match
a4b010034f57.
Arthur Eubanks [Tue, 18 Oct 2022 04:43:08 +0000 (21:43 -0700)]
[llvm-reduce] Unify pass logging
We randomly use outs() or errs(), which makes test logs confusing.
We also randomly add/don't add a line afterward.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D136130
Matthias Springer [Tue, 18 Oct 2022 15:23:42 +0000 (17:23 +0200)]
[mlir][interfaces][NFC] Move DestinationStyleOpInterface to mlir/Interfaces
This is the second (and final) step of making "destination style" usable without depending on the Linalg dialect. (The first step was D135129.)
This change allows us to provide default bufferization implementations for all destination-style ops. It also allows us to simplify `TilingInterface`. (E.g., `getDestinationOperands` can be removed.)
Differential Revision: https://reviews.llvm.org/D136179
Krzysztof Drewniak [Mon, 17 Oct 2022 17:47:56 +0000 (17:47 +0000)]
[mlir][GPU] Prevent adding duplicate async tokens
If, in the GPU async transformation, the operation being given an
async dependency already depended on the token in question, we
would add duplicate tokens, creating issues in GPU to LLVM lowering.
To resolve this issue, add a check to addAsyncDependency() to ensure
that duplicate tokens are not present in the token list.
(I'm open to a different approach here, this is just what I went with
initially)
Reviewed By: Mogball
Differential Revision: https://reviews.llvm.org/D136105
Sanjay Patel [Tue, 18 Oct 2022 15:20:53 +0000 (11:20 -0400)]
[InstCombine] fmul nnan X, 0.0 --> copysign(0.0, X)
https://alive2.llvm.org/ce/z/ybgM5F
Differential Revision: https://reviews.llvm.org/D136166
Sanjay Patel [Tue, 18 Oct 2022 14:35:12 +0000 (10:35 -0400)]
[InstCombine] reduce code duplication in visitBranchInst(); NFCI
Susana Monteiro [Tue, 18 Oct 2022 15:31:53 +0000 (16:31 +0100)]
[clang] fix typo in unit test
Troy Johnson [Tue, 18 Oct 2022 15:03:10 +0000 (08:03 -0700)]
[clang][Lexer] Speed up HeaderSearch when there are many HeaderMaps
HeaderSearch already uses a caching system to avoid duplicate searches,
but the initial cold miss can take a long time if a build system has
supplied thousands of HeaderMaps. For this case, the SearchDirs vector
begins with those HeaderMaps, so a cache miss requires testing if the
sought filename is present in each of those maps. Instead, we can
consolidate the keys of those HeaderMaps into one StringMap and then
each cache miss can skip directly to the correct HeaderMap or continue
its search beyond the initial sequence of HeaderMaps. In testing on TUs
with ~15000 SearchDirs where the initial 99% are HeaderMaps, time spent
in Clang was reduced by 15%. This patch is expected to be neutral for
SearchDir vectors that do not begin with HeaderMaps.
Differential Revision: https://reviews.llvm.org/D135801
Florian Hahn [Tue, 18 Oct 2022 15:17:49 +0000 (16:17 +0100)]
[IndVarSimplify] Clear block and loop dispositions after moving instr.
Moving an instruction can invalidate the cached block dispositions of
the corresponding SCEV. Invalidate the cached dispositions.
Also fixes a copy-paste error in forgetBlockAndLoopDispositions where
the start expression S was removed from BlockDispositions in the loop
but not the current values. This was also exposed by the new test case.
Fixes #58439.
Jannik Silvanus [Mon, 10 Oct 2022 15:39:51 +0000 (17:39 +0200)]
[ADT] Extend EnumeratedArray
EnumeratedArray is essentially a wrapper around a fixed-size
array that uses enum values instead of integers as indices.
* Add iterator support (begin/end/rbegin/rend), which enables
the use of iterator/range based algorithms on EnumeratedArrays.
* Add common container typedefs (value_type etc.), allowing
drop-in replacements of other containers in cases relying on these.
* Add a constructor that takes an std::initializer_list<T>.
* Make the size() function const.
* Add empty().
Iterator support slightly lowers the protection non-type-safe accesses,
because iterator arithmetic is not enum-based, and one can now use
*(begin() + IntIndex). However, it is and was also always possible to
just cast arbitrary indices to the enum type.
Differential Revision: https://reviews.llvm.org/D135594
Chris Bieneman [Tue, 18 Oct 2022 14:50:24 +0000 (09:50 -0500)]
[llvm-config] Make tests more resiliant
In some modes llvm-config issues an error if a requested library is not
available on the system. The in-tree lit tests do not add explicit
dependencies on all component libraries, instead relying on the
transitive dependencies of tools and other libraries.
This can result in some configurations where a component library may not
be used by any of the test dependencies, so it may not be generated
before the tests run. This occurs when optional components of the LLVM
build are excluded, or not included.
In the case where a component library is not generated two test caess
are currently failing. One which tests the shared library boolean flag,
and one which tests the system library requirements. Neither test
require the full set of LLVM libraries to be specifed, so in both cases
I've limited the requested library set to Support to make the tests more
robust.
bixia1 [Mon, 17 Oct 2022 21:09:44 +0000 (14:09 -0700)]
[mlir][sparse] Change getUnorderedCOOFromType to propagate the overhead types.
Reviewed By: aartbik
Differential Revision: https://reviews.llvm.org/D136112
Thomas Symalla [Tue, 18 Oct 2022 14:54:01 +0000 (16:54 +0200)]
[NFC][AMDGPU] Add tests for dependent v_bfi instructions.
This commit adds a few tests which are used to test the codegen
of nested v_bfi instructions. These instruction sequences are
being generated when using the canonical form for bitfieldInsert
and having the sequences being transformed by SimplifyDemandedBits.
This is a pre-commit for a change which enables the backend to
lower these instruction sequences into v_bfi instructions.
Kevin P. Neal [Tue, 18 Oct 2022 14:53:02 +0000 (10:53 -0400)]
[LangRef][FMF] Clarify ambiguity.
This should eliminate some ambiguous grammer that tripped me up when
reviewing D136097.
Nikita Popov [Tue, 18 Oct 2022 14:51:47 +0000 (16:51 +0200)]
[AST] Pass BatchAA to mergeSetIn() (NFCI)
Krzysztof Parzyszek [Mon, 17 Oct 2022 23:29:02 +0000 (16:29 -0700)]
[Hexagon] Fix MULHS lowering for HVX v60
The carry bit from an intermediate addition was not properly propagated.
For example mulhs(
7fffffff,
7fffffff) was evaluated as
3ffeffff, while
the correct result is
3fffffff.
Alexey Bataev [Tue, 18 Oct 2022 14:36:22 +0000 (07:36 -0700)]
[SLP][NFC]Try to fix MSVC buildbots with a workaround, NFC.
Qiongsi Wu [Tue, 18 Oct 2022 14:35:44 +0000 (10:35 -0400)]
[clang][LTO] Setting Desired Default AIX Debugging Options
On AIX, `strict-dwarf` defaults to `true`. This patch implement this default behaviour. Additionally, it adds debug tuning tests.
Reviewed By: shchenz
Differential Revision: https://reviews.llvm.org/D135908
bipmis [Tue, 18 Oct 2022 14:40:31 +0000 (15:40 +0100)]
Add test for combinations of four i8-loads spliced into a 32-bit value
Guillaume Chatelet [Tue, 18 Oct 2022 14:37:08 +0000 (16:37 +0200)]
[libc] Fix missing bazel dependency
This fixes breakage introduced in
a786096f9dd20acf29d8297e706ad96de063f612
uabkaka [Tue, 18 Oct 2022 13:48:26 +0000 (15:48 +0200)]
[SimplifyLibCalls] Add NoUndef/NonNull/Dereferenceable attributes to iprintf/siprintf
When SimplifyLibCalls fail to optimize printf and sprintf it add
NoUndef/NonNull/Dereferenceable attributes. This patch add the same attributes
if SimplifyLibCalls optimize printf/sprintf into the integer only
iprintf/siprintf.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D136140
Guillaume Chatelet [Tue, 18 Oct 2022 14:32:21 +0000 (14:32 +0000)]
[libc][NFC] Cleanup and document utils.h
Guillaume Chatelet [Tue, 18 Oct 2022 14:30:04 +0000 (14:30 +0000)]
[libc][NFC] Use ASSERT instead of EXPECT in tests
Sanjay Patel [Tue, 18 Oct 2022 12:52:34 +0000 (08:52 -0400)]
[InstCombine] add tests for fmul nnan with 0.0; NFC
Alexey Bataev [Tue, 18 Oct 2022 13:42:00 +0000 (06:42 -0700)]
[SLP][NFC]Formatting of the getEntryCost function, NFC.
Florian Hahn [Tue, 18 Oct 2022 14:12:23 +0000 (15:12 +0100)]
[LoopUnroll] Forget exit values when making changes.
When unrolling, the exit values in LCSSA phis will get updated.
Invalidate cached SCEV values for those phis in case SCEV looked through
a exit phi.
Fixes #58340.
Stefan Pintilie [Tue, 18 Oct 2022 13:45:50 +0000 (08:45 -0500)]
[NFC][PowerPC] Add a test to check power 10 features.
This patch only adds a single test for Power 10 features.
Jialun Hu [Tue, 18 Oct 2022 13:56:13 +0000 (21:56 +0800)]
[clang] Fix crash upon stray coloncolon token in C2x mode
The parser assumes that the lexer never emits coloncolon token for C code, but this assumption no longer holds in C2x attribute namespaces. As a result, stray coloncolon tokens out of attributes cause assertion failures and hangs in release build, which this patch tries to handle.
Crash input minimal example: `T n::v`
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D133248
Florian Hahn [Tue, 18 Oct 2022 13:56:44 +0000 (14:56 +0100)]
[LoopUnroll] Add test for mis-compile due to missing SCEV invalidation.
Test for #58340.
Aaron Ballman [Tue, 18 Oct 2022 13:39:17 +0000 (09:39 -0400)]
Add an additional time for monthly office hours
Office hours have gone so well for me that I've had requests to add a
second time slot which is later in the day so that folks on the US west
coast have a more reasonable time to meet. So now meeting at 10am and
2pm on the 2nd Monday of each month.
Ties Stuij [Tue, 18 Oct 2022 12:47:03 +0000 (13:47 +0100)]
[clang][ARM] follow GCC behavior for defining __SOFTFP__
GCC behavior regarding defining __SOFTFP__ when (implicitly) specifying
-mfloat-abi=softfp:
- compile without (implicit) FP: define __SOFTFP__
- compile with (implicit) FP: don't define __SOFTFP__
Currently Clang doesn't define __SOFTFP__ when softfp is specified, either with
or without FP. This patch brings Clang in line with GCC behavior.
This was raised by itaig1 over on Github:
https://github.com/llvm/llvm-project/issues/55755
Reviewed By: pratlucas
Differential Revision: https://reviews.llvm.org/D135680
Joseph Huber [Mon, 17 Oct 2022 20:53:31 +0000 (15:53 -0500)]
[OpenMP] Make device functions have hidden visibility
In OpenMP target offloading an in other offloading languages, we
maintain a difference between device functions and kernel functions.
Kernel functions must be visible to the host and act as the entry point
to the target device. Device functions however cannot be called directly
by the host and must be called by a kernel function. Currently, we make
all definitions on the device protected by default. Because device
functions cannot be called or used by the host they should have hidden
visibility. This allows for the definitions to be better optimized via
LTO or other passes.
This patch marks every device function in the AST as having `hidden`
visibility. The kernel function is generated later at code-gen and we
set its visibility explicitly so it should not be affected. This
prevents the user from overriding the visibility, but since the user
can't do anything with these symbols anyway there is no point exporting
them right now.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D136111
Nikita Popov [Tue, 18 Oct 2022 09:11:20 +0000 (11:11 +0200)]
[lldb] Fix m_hwp_regs size for ppc64le (PR54520)
The size of the m_hwp_regs array should match the default value of
m_max_hwp_supported. This ensures that no out-of-bounds accesses
occur, even if the array is accessed prior to a call to
ReadHardwareDebugInfo().
Fixes https://github.com/llvm/llvm-project/issues/54520, see also
there for additional background.
Differential Revision: https://reviews.llvm.org/D136144
Erich Keane [Tue, 11 Oct 2022 13:16:41 +0000 (06:16 -0700)]
Stop evaluating trailing requires clause after overload resolution
Reported as it showed up as a constriants failure after the deferred
instantiation patch, we were checking constraints TWICE after overload
resolution. The first is during overload resolution, the second is when
diagnosing a use.
This patch modifies DiagnoseUseOfDecl to skip the trailing requires
clause check in some cases. First, of course, after choosing a candidate
after overload resolution.
The second is when evaluating a shadow using constructor, which had its
constraints checked when picking a constructor (as this is ALWAYS an
overload situation!).
Differential Revision: https://reviews.llvm.org/D135772
Jun Zhang [Tue, 18 Oct 2022 12:38:30 +0000 (20:38 +0800)]
[Clang] Fix crash when checking misaligned member with dependent type
If the type is dependent, we should just discard it and not checking its
alignment as it doesn't exisit yet.
Fixes https://github.com/llvm/llvm-project/issues/58370
Differential Revision: https://reviews.llvm.org/D136018
Anton Afanasyev [Tue, 18 Oct 2022 12:56:39 +0000 (15:56 +0300)]
Revert "[MachineCombiner][RISCV] Enable MachineCombiner for RISCV"
This reverts commit
3112cf3b00fe45a0911ec0c2e6706ef1f8a9b972.
Test breakage: https://lab.llvm.org/buildbot/#/builders/16/builds/36631
Andrew Ng [Mon, 26 Sep 2022 14:10:13 +0000 (15:10 +0100)]
[CMake] Add Python script to generate version script symbol exports
Using a Python script instead of the various shell commands means that
it is now possible to cross compile LLVM for Linux on Windows.
Differential Revision: https://reviews.llvm.org/D136092
Weining Lu [Tue, 18 Oct 2022 09:33:50 +0000 (17:33 +0800)]
[LoongArch] Fix codegen of atomicrmw nand
Fix invalid RISCV-like MI being emitted for performing the `not`
operation: the LoongArch `xori` zero-extends the immediate, hence is
not equivalent to RISCV `xori`. The LoongArch `not` is a `nor` with
zero.
Differential Revision: https://reviews.llvm.org/D136021
Jannik Silvanus [Tue, 18 Oct 2022 11:26:32 +0000 (13:26 +0200)]
[NFC][ADT] Clean up EnumeratedArray.h
As discussed in D135594, remove superfluous inline
attributes, and remove top-level consts on function arguments.
Anton Sidorenko [Mon, 17 Oct 2022 09:06:08 +0000 (12:06 +0300)]
[MachineCombiner][RISCV] Enable MachineCombiner for RISCV
Initial implementation to match basic FP reassociation patterns.
Differential Revision: https://reviews.llvm.org/D135264
Carlos Alberto Enciso [Tue, 18 Oct 2022 11:02:36 +0000 (12:02 +0100)]
[llvm-debuginfo-analyzer] Fix linking errors in buildbots.
The tool used the 'old' LLVM build information (LLVMBuild.txt),
which caused linking errors in:
https://lab.llvm.org/buildbot/#/builders/177/builds/10125
https://lab.llvm.org/buildbot/#/builders/196/builds/19699
Update the CMake configuration to support the new LLVM build
system that uses only CMakeLists.txt.
Reviewed By: jryans
Differential Revision: https://reviews.llvm.org/D136159
Ben Dunbobbin [Tue, 18 Oct 2022 11:47:56 +0000 (12:47 +0100)]
[LLD][ELF] --wrap: __real_foo references should trigger archive extraction for foo
A reference to __real_foo should trigger archive extraction of the input file that defines foo, otherwise a link using --wrap=foo might fail to link with an undefined reference to foo.
This matches bfd linker behaviour.
Differential Revision: https://reviews.llvm.org/D135897
Andrew Ng [Fri, 14 Oct 2022 15:06:34 +0000 (16:06 +0100)]
[CMake] Fix Findzstd module for shared DLL on Windows
Differential Revision: https://reviews.llvm.org/D136065
chenglin.bi [Tue, 18 Oct 2022 11:18:23 +0000 (19:18 +0800)]
[AArch64] add test case for pattern ((X >> C) - Y) + Z; NFC
Nico Weber [Mon, 17 Oct 2022 13:26:23 +0000 (09:26 -0400)]
[gn build] port
fe7a3cedf771 (llvm-debuginfo-analyzer)
Valentin Clement [Tue, 18 Oct 2022 10:12:11 +0000 (12:12 +0200)]
[flang] Reconstruct binding tables for fir.dispatch codegen
Binding tables are needed to perform the fir.dispatch code generation.
The binding tables are defined as fir.global from the initial lowering.
This patch adds the ability to reconstruct the binding tables information
and store the procedure name and binding index for later use by the
fir.dispatch code generation.
Tests will come with follow up patch that makes full use of this information.
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D136141
Anton Sidorenko [Thu, 15 Sep 2022 12:00:05 +0000 (15:00 +0300)]
[MachineCombiner][RISCV] Precommit test for D135264
Jean Perier [Tue, 18 Oct 2022 09:07:47 +0000 (11:07 +0200)]
[flang][NFC] Simplify mapSymbolAttributes in symbol lowering
mapSymbolAttributes currently has a lot of very similar code for
each kind of explicit shape and scalar symbols.
Refactor it so that the change to lower symbols with fir.declare
can be added in centralized places instead of being scattered.
This is a preparation patch and fir.declare is not yet added.
Differential Revision: https://reviews.llvm.org/D136061
Daniel Kiss [Tue, 18 Oct 2022 08:35:10 +0000 (10:35 +0200)]
[AArch64] Make ACLE intrinsics always available part MTE
Make MTE intrinsics available in function scope too.
Followup from D133359.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D136062
Martin Storsjö [Tue, 18 Oct 2022 08:41:49 +0000 (08:41 +0000)]
[LLDB] Fix the build for ppc64le linux
812ad2167bd2e27f5d0dee07bb03a5910616e0b6 changed the signature of
RegisterValue::SetFromMemoryData.
Daniel Kiss [Tue, 18 Oct 2022 08:45:32 +0000 (10:45 +0200)]
Revert "[AArch64] Make ACLE intrinsics always available part MTE"
This reverts commit
09aaf190d93393d9e29d29a033cc3979589c5e84.
Daniel Kiss [Tue, 18 Oct 2022 08:35:10 +0000 (10:35 +0200)]
[AArch64] Make ACLE intrinsics always available part MTE
Make MTE intrinsics available in function scope too.
Followup from D133359.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D136062
Tobias Gysi [Tue, 18 Oct 2022 08:01:07 +0000 (11:01 +0300)]
[mlir][llvm] Import matrix, vector, and assume intrinsics from LLVM IR.
The revision adds support to import:
- matrix intrinsics
- vector reduce fadd/fmul intrinsics
- assume intrinsics
from LLVM IR.
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D136137
Nikita Popov [Tue, 18 Oct 2022 08:05:12 +0000 (10:05 +0200)]
[CFG] Add const qualifier to isPotentiallyReachableFromMany() (NFC)
Accept a const pointer for StopBB. Unfortunately the worklist has
to use non-const pointers due to LoopInfo interaction.
Carlos Alberto Enciso [Tue, 18 Oct 2022 07:39:26 +0000 (08:39 +0100)]
Recommit [llvm-debuginfo-analyzer] (02/09) - Driver and documentation
Originally committed in
fe7a3cedf77125a6309150d85cecbc20b1a31775
Reverted in
26dd64ba9cfabe5474bb207f3b7099965f81fed7
Buildbot failures:
https://lab.llvm.org/buildbot#builders/139/builds/29663
- unittest trigger an invalid assertion.
https://lab.llvm.org/buildbot#builders/196/builds/19665
- 'has virtual functions but non-virtual destructor' warning as error.
Recommitted with fix:
- Removed the assertion.
- Added virtual destructor.
Dominik Adamski [Thu, 8 Sep 2022 11:39:18 +0000 (06:39 -0500)]
[OpenMP][OMPIRBuilder] Add generation of SIMD align assumptions to OMPIRBuilder
Currently generation of align assumptions for OpenMP simd construct is done
outside OMPIRBuilder for C code and it is not supported for Fortran.
According to OpenMP 5.0 standard (2.9.3) only pointers and arrays can be
aligned for C code.
If given aligned variable is pointer, then Clang generates the following set
of the LLVM IR isntructions to support simd align clause:
; memory allocation for pointer address:
%A.addr = alloca ptr, align 8
; some LLVM IR code
; Alignment instructions (alignment is equal to 32):
%0 = load ptr, ptr %A.addr, align 8
call void @llvm.assume(i1 true) [ "align"(ptr %0, i64 32) ]
If given aligned variable is array, then Clang generates the following set
of the LLVM IR isntructions to support simd align clause:
; memory allocation for array:
%B = alloca [10 x i32], align 16
; some LLVM IR code
; Alignment instructions (alignment is equal to 32):
%arraydecay = getelementptr inbounds [10 x i32], ptr %B, i64 0, i64 0
call void @llvm.assume(i1 true) [ "align"(ptr %arraydecay, i64 32) ]
OMPIRBuilder was modified to generate aligned assumptions. It generates only
llvm.assume calls. Frontend is responsible for generation of aligned pointer
and getting the default alignment value if user does not specify it in aligned
clause.
Unit and regression tests were added to check if aligned clause was handled correctly.
Differential Revision: https://reviews.llvm.org/D133578
Reviewed By: jdoerfert
Max Kazantsev [Tue, 18 Oct 2022 06:26:00 +0000 (13:26 +0700)]
[NFC] Reuse NonTrivialUnswitchCandidate instead of std::pair
Clement Courbet [Tue, 18 Oct 2022 06:47:21 +0000 (08:47 +0200)]
[doc] Fix invalid reference to `hasReturnArgument` matcher.
The matcher is called `hasReturnValue`.
Yuanfang Chen [Tue, 18 Oct 2022 06:05:35 +0000 (23:05 -0700)]
[Clang] add DR tests for D128745
Sheng [Tue, 18 Oct 2022 05:18:09 +0000 (05:18 +0000)]
[NFC][mlir] Remove redundant wording
"that provides" repeats two times.
Emilia Dreamer [Tue, 11 Oct 2022 19:16:48 +0000 (22:16 +0300)]
[clang-format] Correctly annotate star/amp in function pointer params
Inside the arguments part of a function pointer declaration,
`determineStarAmpUsage` results in a binary operator rather than
pointers, because said parens are assumed to be an expression.
This patch correctly marks the argument parens of a function
pointer type as not an expression. Note that this fix already
existed for Objective-C blocks as part of
f1f267b447f60528440d2c066b29ab014ae7f90f.
As Objective-C blocks and C/C++ function pointers share a lot
of the same logic, that fix also makes sense here.
Fixes https://github.com/llvm/llvm-project/issues/31659
Differential Revision: https://reviews.llvm.org/D135707
Chuanqi Xu [Tue, 18 Oct 2022 05:15:52 +0000 (13:15 +0800)]
[NFC] Add missing ABI requirement from the previous patch
bixia1 [Tue, 18 Oct 2022 04:41:50 +0000 (21:41 -0700)]
Add ::mlir:: prefix to Attribute type in the generated code.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D136129
YunQiang Su [Tue, 18 Oct 2022 04:49:25 +0000 (04:49 +0000)]
[CMake] Fix MIPSr6 build for compiler-rt
The current version pass -mips64r2 or -mips32r2 options,
which make it failed to build on r6 platform.
In this patch: we detect whether we are MIPSr6 by
_MIPS_ARCH_MIPS32R6/_MIPS_ARCH_MIPS64R6
The out and install path is set to the default triple instead of
hardcoded one, since the clang ask for it.
Differential Revision: https://reviews.llvm.org/D135735
Jez Ng [Tue, 18 Oct 2022 04:17:34 +0000 (00:17 -0400)]
[lld-macho][test] Rework map-file.s
Merge the checks done on the `c-string-literals` binary with those on
the main `test` binary. Also switch some checks to `-DAG` instead of
`-NEXT`. I'm about to extend this test, and this is easier to work with.
Reviewed By: #lld-macho, oontvoo
Differential Revision: https://reviews.llvm.org/D135999
Chuanqi Xu [Tue, 18 Oct 2022 03:32:34 +0000 (11:32 +0800)]
[NFC] [C++20] [Modules] Test if the functions in importee are generated
In O0, all the functions (except the always-inline-functions) in the importee
shouldn't be imported for compilation speeds.
But with optimizations, all the potentially called function in the
importee should be imported to not prevent any inter-procedural
optimizations (primarily inline), which is pretty important for runtime
performances.
This patch adds the tests for the feature.
Jeff Bailey [Tue, 18 Oct 2022 03:12:33 +0000 (03:12 +0000)]
[libc]Identify which processors are completed for Math
Switch from green checkmarks to the following legend:
X = x86_64
A = aarch64
a = arm32
Reviewed By: lntue
Differential Revision: https://reviews.llvm.org/D136020
Volodymyr Sapsai [Tue, 18 Oct 2022 01:48:24 +0000 (18:48 -0700)]
[ODRHash] Hash `ObjCMethodDecl` and diagnose discovered mismatches.
Differential Revision: https://reviews.llvm.org/D130325
Volodymyr Sapsai [Tue, 18 Oct 2022 01:23:28 +0000 (18:23 -0700)]
[ODRHash] Rename `isDeclToBeProcessed` to `isSubDeclToBeProcessed`. NFC intended.
The method is used only for sub-Decls, so reflect that in the name.
LiaoChunyu [Tue, 18 Oct 2022 01:21:47 +0000 (09:21 +0800)]
[RISCV] Optimize SELECT_CC when the true value of select is Constant
(select (setcc lhs, rhs, CC), constant, falsev) -> (select (setcc lhs, rhs, InverseCC), falsev, constant)
This patch removes unnecessary copies
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D129757
Aart Bik [Tue, 18 Oct 2022 00:02:25 +0000 (17:02 -0700)]
[mlir][sparse] implement simple codegen for insertion (and related ops)
This is a proof of concept insertion implementation that sets up
the basic framework and implements it with push backs for just
sparse vectors. It adds insertion/compression through SSA values,
so that we properly update the memref after after pushback operation.
Note that properly using SSA values in sparsification is still TBD
but I will wait until Peiming's loop emitter is in to avoid conflicts.
Reviewed By: wrengr
Differential Revision: https://reviews.llvm.org/D136008
Xiang Li [Tue, 18 Oct 2022 00:47:48 +0000 (17:47 -0700)]
[NFC] use llvm_unreachable instead of return on switch which all cases are covered.
Adrian Prantl [Fri, 14 Oct 2022 23:45:01 +0000 (16:45 -0700)]
Make sure Target::EvaluateExpression() passes up an error instead of silently dropping it.
When UserExpression::Evaluate() fails and doesn't return a ValueObject there is no vehicle for returning the error in the return value.
This behavior can be observed by applying the following patch:
diff --git a/lldb/source/Target/Target.cpp b/lldb/source/Target/Target.cpp
index
f1a311b7252c..
58c03ccdb068 100644
--- a/lldb/source/Target/Target.cpp
+++ b/lldb/source/Target/Target.cpp
@@ -2370,6 +2370,7 @@ UserExpression *Target::GetUserExpressionForLanguage(
Expression::ResultType desired_type,
const EvaluateExpressionOptions &options, ValueObject *ctx_obj,
Status &error) {
+ error.SetErrorStringWithFormat("Ha ha!"); return nullptr;
auto type_system_or_err = GetScratchTypeSystemForLanguage(language);
if (auto err = type_system_or_err.takeError()) {
error.SetErrorStringWithFormat(
and then running
$ lldb -o "p 1"
(lldb) p 1
(lldb)
This patch fixes this by creating an empty result ValueObject that wraps the error.
Differential Revision: https://reviews.llvm.org/D135998
Adrian Prantl [Tue, 18 Oct 2022 00:27:10 +0000 (17:27 -0700)]
Revert "Make sure Target::EvaluateExpression() passes up an error instead of silently dropping it."
This reverts commit
a31a5da3c7d7393749a43dbc678fd28fb94d07f6.
Nathan Ridge [Tue, 18 Oct 2022 00:15:40 +0000 (20:15 -0400)]
[clangd] Update 'using enum' semantic highlighting testcase to use the 'definition' modifier
Che-Yu Wu [Tue, 18 Oct 2022 00:03:47 +0000 (00:03 +0000)]
[MLIR] Vectorize tensor.extract on 1-d tensor
This patch implements the vectorization of tensor.extract for the
basic 1-d lookup case. It only vectorizes the tensor.extract to a
vector.gather when the op extracts value from an 1-d tensor.
Related discussion: https://github.com/iree-org/iree/issues/9198
Reviewed By: dcaballe
Differential Revision: https://reviews.llvm.org/D133786
Koakuma [Tue, 18 Oct 2022 00:01:55 +0000 (00:01 +0000)]
[SPARC] Make calls to function with big return values work
Implement CanLowerReturn and associated CallingConv changes for SPARC/SPARC64.
In particular, for SPARC64 there's new `RetCC_Sparc64_*` functions that handles the return case of the calling convention.
It uses the same analysis as `CC_Sparc64_*` family of funtions, but fails if the return value doesn't fit into the return registers.
This makes calls to functions with big return values converted to an sret function as expected, instead of crashing LLVM.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D132465
Nathan Ridge [Mon, 17 Oct 2022 22:53:43 +0000 (18:53 -0400)]
[clangd] Update testcase for issue 1222 to use the 'definition' modifier
Jordan Rupprecht [Mon, 17 Oct 2022 23:54:37 +0000 (16:54 -0700)]
Dan Gohman [Mon, 17 Oct 2022 20:36:19 +0000 (13:36 -0700)]
[wasm-ld] Define a `__heap_end` symbol marking the end of allocated memory.
Define a `__heap_end` symbol that marks the end of the memory region
that starts at `__heap_base`. This will allow malloc implementations to
know how much memory they can use at `__heap_base` even if someone has
done a `memory.grow` before they can initialize their state.
Differential Revision: https://reviews.llvm.org/D136110
Arthur Eubanks [Mon, 17 Oct 2022 23:37:05 +0000 (16:37 -0700)]
[NFC][SROA] Update comment to use opaque pointers for clarity
Volodymyr Sapsai [Mon, 17 Oct 2022 23:21:59 +0000 (16:21 -0700)]
[ODRHash] Hash `ObjCProtocolDecl` and diagnose discovered mismatches.
Differential Revision: https://reviews.llvm.org/D130324
Michael Jones [Fri, 14 Oct 2022 21:03:46 +0000 (14:03 -0700)]
[libc] add putc, fputc, and putchar
These three functions are simple, but needed for libc build testing.
Reviewed By: lntue
Differential Revision: https://reviews.llvm.org/D135990
Jacques Pienaar [Mon, 17 Oct 2022 23:28:46 +0000 (16:28 -0700)]
[mlir][quant] Initial bytecode encoding for quantized types
Add bytecode encoding for quantized types. These mostly follow the
storage representation of these.
Differential Revision: https://reviews.llvm.org/D136004
Maksim Panchenko [Mon, 17 Oct 2022 21:15:52 +0000 (14:15 -0700)]
[BOLT][NFCI] Avoid calling registerName() twice
Calling registerName() for the same symbol twice, even with a different
size, has no effect other than the lookup overhead. Avoid the
redundancy.
Fixes facebookincubator/BOLT#299
Reviewed By: Amir
Differential Revision: https://reviews.llvm.org/D136115
Xiang Li [Mon, 17 Oct 2022 22:50:49 +0000 (15:50 -0700)]
[NFC] Fix warning on no return after switch.
Artem Belevich [Wed, 12 Oct 2022 23:41:28 +0000 (16:41 -0700)]
Do not append terminating NUL to the binary string with embedded fatbin.
Extra NUL does not impact functionality of the generated code, but it confuses
various NVIDIA tools used to examine embedded GPU binaries.
Differential Revision: https://reviews.llvm.org/D135832