Yitzhak Mandelbaum [Tue, 27 Dec 2022 14:21:29 +0000 (14:21 +0000)]
[clang][dataflow] Only model struct fields that are used in the function being analyzed.
Previously, the model for structs modeled all fields in a struct when
`createValue` was called for that type. This patch adds a prepass on the
function under analysis to discover the fields referenced in the scope and then
limits modeling to only those fields. This reduces wasted memory usage
(modeling unused fields) which can be important for programss that use large
structs.
Note: This patch obviates the need for https://reviews.llvm.org/D123032.
Differential Revision: https://reviews.llvm.org/D140694
Adrian Prantl [Thu, 5 Jan 2023 21:41:33 +0000 (13:41 -0800)]
Revert an unintentional API ABI break
lldb::LanguageType is used as a parameter in
SBExpressionOptions::SetLanguage(), which actually makes this type API
too. Commit
6eaedbb52f2a616e644e5acc7279c8b07c4cfe82 added a
`: uint16_t` to it, which broke binary compatibility for the SBAPI. This
patch reverts to the original enum.
I tried moving the entire enum into include/API, but that created a
cyclic module dependency between API and Utility. To keep things
simple, this just reverts to the original code and adds a warning.
rdar://
103415402
Differential Revision: https://reviews.llvm.org/D141087
Matt Arsenault [Mon, 28 Nov 2022 22:55:17 +0000 (17:55 -0500)]
clang/AMDGPU: Remove flat-address-space from feature map
This was only used for checking if is_shared/is_private were legal,
which we're not bothering to do anymore.
This is apparently visible to more than the target attribute (which
seems to silently ignore unrecognized features), so this has the
potential to break something (i.e. see the OpenMP test change)
Xiang Li [Thu, 5 Jan 2023 21:29:09 +0000 (16:29 -0500)]
[NFC] [Doc] Fix example for AnnotateTypeDocs
Change clang::annotate into clang::annotate_type.
The example will get error like
error: 'annotate' attribute cannot be applied to types
int* [[clang::annotate("category1", "foo", 1)]] f(int[[clang::annotate("category2")]] *);
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D139935
Vang Thao [Thu, 5 Jan 2023 20:40:12 +0000 (20:40 +0000)]
[AMDGPU] Add .uniform_work_group_size metadata to v5
Amdgpu kernel with function attribute "uniform-work-group-size"="true" requires
uniform work group size (i.e. each dimension of global size is a multiple of
corresponding dimension of work group size). hipExtModuleLaunchKernel allows to
launch HIP kernel with non-uniform workgroup size, which makes it necessary for
runtime to check and enforce uniform workgroup size if kernel requires it. To
let runtime be able to enforce that, this metadata is needed to indicate that
the kernel requires uniform workgroup size.
Reviewed By: kzhuravl, arsenm
Differential Revision: https://reviews.llvm.org/D141012
Fangrui Song [Thu, 5 Jan 2023 21:28:48 +0000 (13:28 -0800)]
[CodeGen] Default TargetOptions::RelaxELFRelocations to true
MC and lld/ELF defaults were flipped in 2016. For Clang: CMake
ENABLE_X86_RELAX_RELOCATIONS defaults to on in 2020. It makes sense for
the TargetOptions default to be true now.
R_X86_64_GOTPCRELX/R_X86_64_REX_GOTPCRELX require GNU ld newer than 2015-10
(subsumed by the current requirement of -fbinutils-version=).
This should fix `rustc -Z plt=no` PIC relocatable files with GNU ld.
(See https://github.com/rust-lang/rust/pull/106380)
Owen Pan [Mon, 2 Jan 2023 00:02:35 +0000 (16:02 -0800)]
[clang-format] Improve UnwrappedLineParser::mightFitOnOneLine()
Account for an r_brace that precedes an "else if" statement when
calculating whether the line might fit on one line if the r_brace
is removed.
Fixes #59778.
Differential Revision: https://reviews.llvm.org/D140835
Saleem Abdulrasool [Thu, 22 Dec 2022 15:46:34 +0000 (15:46 +0000)]
Parse: handle another case of invalid handling for attributes
clang would improperly disallow GNU attributes before C++ standard
attributes when a declaration had a linkage specifier. Handle this
similarly to the previous case of invalid parsing. We now better match
the parsing rules from GCC.
Differential Revision: https://reviews.llvm.org/D140507
Reviewed By: aaron.ballman
Alexander Timofeev [Wed, 7 Dec 2022 17:36:11 +0000 (18:36 +0100)]
[AMDGPU] MachineScheduler: schedule execution metric added for the UnclusteredHighRPStage
Since the divergence-driven ISel was fully enabled we have more VGPRs available.
MachineScheduler trying to take advantage of that bumps up the occupancy sacrificing
the hiding of memory access latency. This really spoils the initially good schedule.
A new metric that reflects the latency hiding quality of the schedule has been created
to make it to balance between occupancy and latency. The metric is based on the latency
model which computes the bubble to working cycles ratio. Then we use this ratio to decide
if the higher occupancy schedule is profitable as follows:
Profit = NewOccupancy/OldOccupancy * OldMetric/NewMetric
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D139710
ziqingluo-90 [Thu, 5 Jan 2023 19:48:02 +0000 (11:48 -0800)]
[Fix]"[-Wunsafe-buffer-usage] Add a new `forEachDescendant` matcher that skips callable declarations"
The original patch in commit
b2ac5fd724c44cf662caed84bd8f84af574b981d
causes compilation errors which can be reproduced by the
`-fdelayed-template-parsing` flag. This commit fixes the problem.
Related differential revision: https://reviews.llvm.org/D138329
ziqingluo-90 [Thu, 5 Jan 2023 19:45:25 +0000 (11:45 -0800)]
Revert "Revert "[-Wunsafe-buffer-usage] Add a new `forEachDescendant` matcher that skips callable declarations""
This reverts commit
f58b025354ee2d3bcd7ab2399a11429ec940c1e0.
The previous revert reverts a patch that causes compilation problem on
windows which can be reproduced using `-fdelayed-template-parsing`.
I'm now to revert the patch back and commit a fix next.
Matt Arsenault [Thu, 24 Nov 2022 21:58:49 +0000 (16:58 -0500)]
Verifier: Disallow uses of intrinsic global variables
appendToGlobalCtors implicitly assumes this is the case, since it
deletes and recreates without trying to update any uses.
This ran into an interesting problem in a few linker tests. During the
link, ConstantExpr casts were speculatively created to replace any
uses that might need them for replacement. These unused ConstantExprs
would hang around and still appear on the use list. It seems like a
bug that those stick around, but I'm not sure where those are supposed
to be cleaned up. Avoid this by not creating the casts for appending
linkage.
Delete one of the casts entirely as it breaks no tests. The verifier
has enforced a specific type for these since 2011, so I don't see why
we would need to handle linking modules with a wrong types. One test
does fail without the second cast (Linker/appending-global-proto.ll,
added by D95126). This test looks contrived; it's using appending
linkage with a regular variable. The LangRef suggests this is illegal
(and suggests another missing verifier check).
Aaron Ballman [Thu, 5 Jan 2023 19:47:11 +0000 (14:47 -0500)]
Remove documentation about the Go bindings
We removed the Go bindings in https://reviews.llvm.org/D135436 but
missed documentation that talks about the bindings.
Craig Topper [Thu, 5 Jan 2023 19:18:26 +0000 (11:18 -0800)]
[RISCV] Prevent constant hoisting for or/and/xor that can use bseti/bclri/binvi.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D140928
Fangrui Song [Thu, 5 Jan 2023 19:18:17 +0000 (11:18 -0800)]
[test] Test __attribute__((noreturn)), _Noreturn, and [[return]] with conditional operator
Jez Ng [Thu, 5 Jan 2023 19:14:11 +0000 (14:14 -0500)]
[lld-macho] Don't support relocations in cstring sections
We can technically handle them, but since they shouldn't come up in any
real-world programs (since ld64 dedups strings unconditionally), there's
no reason to support them.
It's a thoroughly untested code path too -- as evidenced by the fact
that the only test this change breaks is one that verifies that we
reject relocations when dedup'ing. There is no test that covers the case
where we handle relocations in cstring sections when dedup is disabled.
Reviewed By: #lld-macho, oontvoo, keith, thakis
Differential Revision: https://reviews.llvm.org/D141025
Florian Hahn [Thu, 5 Jan 2023 19:03:33 +0000 (19:03 +0000)]
[profile] Update test to make condition non-redundant.
The tests contain a redundant condition in the else if branch that can
be simplified with
fb13dcf3431c.
Update the condition used to prevent it from getting removed.
Alive2 proof for check removal:
https://alive2.llvm.org/ce/z/iFBnsy
Luke Drummond [Thu, 5 Jan 2023 18:34:55 +0000 (18:34 +0000)]
Fix typos
I found one typo of "implemnt", then some more.
s/implemnt/implement/g
Matt Arsenault [Thu, 22 Dec 2022 16:49:00 +0000 (11:49 -0500)]
AMDGPU: Fix broken opaque pointer handling in printf pass
This was directly considering the pointee type, and also applying
special semantics to constant address space.
Roman Lebedev [Thu, 5 Jan 2023 18:20:32 +0000 (21:20 +0300)]
[NFC][YAMLTraits] Use `SmallString<>`+SSO when yamlizing non-strings
This gets rid of basically all temporary allocations
during YAML serialization of llvm-exegesis snippets,
we go from ~1.8M temporary allocations to 13k.
Roman Lebedev [Thu, 5 Jan 2023 17:51:37 +0000 (20:51 +0300)]
[NFC][exegesis] `SnippetGenerator::generateConfigurations()`: don't forget to `reserve()`
Jez Ng [Thu, 5 Jan 2023 18:29:21 +0000 (13:29 -0500)]
[lld-macho][test] Simplify classrefs test
I wasn't previously aware of the `llvm-otool -o` flag; that gives us
output that's a lot easier to parse.
Reviewed By: #lld-macho, thakis
Differential Revision: https://reviews.llvm.org/D141024
Tom Honermann [Wed, 4 Jan 2023 03:41:33 +0000 (19:41 -0800)]
[clang] Correct -frewrite-includes generation of line control directives with mixed EOL forms.
Previously, if a header file and a source file used different end of line
(EOL) forms, preprocessed output generated with the -frewrite-includes option
would, in some cases, generate line control directives with the wrong line
number due to an error in how source file lines were counted.
Fixes https://github.com/llvm/llvm-project/issues/59736
Reviewed By: cor3ntin
Differential Revision: https://reviews.llvm.org/D140984
David Green [Thu, 5 Jan 2023 18:16:58 +0000 (18:16 +0000)]
[AArch64] Regenerate fp16-vector-nvcast.ll check lines. NFC
Valery N Dmitriev [Wed, 28 Dec 2022 21:22:53 +0000 (13:22 -0800)]
[SLP] Unify GEP cost modeling for load, store and GEP nodes.
Make a separate routine for GEPs cost calculation and make
the approach uniform across load, store and GEP tree nodes.
Additional issue fixed is GEP cost savings were applied twice
for ScatterVectorize nodes (aka gather load) making them look
unrealistically profitable for vectorization.
Differential Revision: https://reviews.llvm.org/D140789
Siva Chandra Reddy [Thu, 5 Jan 2023 09:26:15 +0000 (01:26 -0800)]
[libc] Make the sidebar colors have a higher constract in the white background.
Reviewed By: jeffbailey
Differential Revision: https://reviews.llvm.org/D141039
Siva Chandra Reddy [Thu, 5 Jan 2023 09:09:28 +0000 (01:09 -0800)]
[libc] Add document for a recipe to do a full standalone cross-build.
Reviewed By: jeffbailey
Differential Revision: https://reviews.llvm.org/D141037
bixia1 [Thu, 5 Jan 2023 00:01:35 +0000 (16:01 -0800)]
[mlir][sparse] Add operator sparse_tensor.indices_buffer.
Reviewed By: aartbik
Differential Revision: https://reviews.llvm.org/D140762
Sanjay Patel [Wed, 4 Jan 2023 21:00:25 +0000 (16:00 -0500)]
[InstCombine] add tests for cmp of pow2 mask; NFC
Jeff Bailey [Thu, 5 Jan 2023 17:27:16 +0000 (17:27 +0000)]
[libc] Add documentation on hacking with vscode
Reviewed By: sivachandra
Differential Revision: https://reviews.llvm.org/D140826
Roman Lebedev [Thu, 5 Jan 2023 17:11:38 +0000 (20:11 +0300)]
[NFC][exegesis] `BenchmarkCodeStreamer`: use existing RegNameToRegNo map
Roman Lebedev [Thu, 5 Jan 2023 17:10:35 +0000 (20:10 +0300)]
[NFC][AttributesTest] Fix signedness warnings
Kito Cheng [Thu, 5 Jan 2023 16:48:22 +0000 (00:48 +0800)]
[RISCV][NFC] Refine the patch of D141061
Just saw Craig's comment after I commit, he has suggest a good NFC
for that change.
Kito Cheng [Thu, 5 Jan 2023 15:10:12 +0000 (23:10 +0800)]
[RISCV][InsertVSETVLI] Using right instruction during mutate AVL of vsetvli
Fixing a crash during vsetvli insertion pass.
We have a testcase with 3 vsetvli:
1. vsetivli zero, 2, e8, m4, ta, ma
2. li a1, 32; vsetvli zero, a1, e8, m4, ta, mu
3. vsetivli zero, 2, e8, m4, ta, ma
and then we trying to optimize 2nd vsetvli since the only user is vmv.x.s, so
it could mutate the AVL operand to the AVL operand of the 3rd vsetvli.
OK, so we propagate 2 to vsetvli, BUT it's vsetvli not vsetivli, so it expect a
register rather than a immediate value, so we have to update the opcode
if needed.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D141061
Craig Topper [Thu, 5 Jan 2023 16:20:54 +0000 (08:20 -0800)]
[SelectionDAG] Improve codegen for udiv by constant if any divisors are 1.
If the divisor is 1, the magic algorithm does not return a correct
result and we end up using a select to pick the numerator for those
elements at the end.
Therefore we can use undef for that element of the earlier operations
when the divisor is 1. We sometimes get this through SimplifyDemandedVectorElts,
but not always. Definitely seems like we don't if the NPQ fixup is used.
Unfortunately, DAGCombiner is unable to fold srl X, <0, undef> to X so
I had to add flags to avoid emitting the srl unless one of the shift
amounts is non-zero.
Reviewed By: lebedev.ri
Differential Revision: https://reviews.llvm.org/D141022
Matt Arsenault [Fri, 23 Dec 2022 19:41:50 +0000 (14:41 -0500)]
AMDGPU: Switch enqueue kernel test to generated checks
Roman Lebedev [Thu, 5 Jan 2023 16:31:35 +0000 (19:31 +0300)]
[NFC][Exegesis][X86] Test that we don't crash while coming up with snippets
These tests take less than a second each,
so they are fine to have as tests.
It will be useful to add similar ones
for `prepare-and-assemble-snippet` x repetition mode,
but there we have a problem of not being truly ISA set independent...
serge-sans-paille [Thu, 5 Jan 2023 16:07:13 +0000 (17:07 +0100)]
By-pass some ArrayRef deduction guide
clang-11 doesn't fully support them and requires an explicit
instantiation in some situation. This is fixed in (at least) clang-14
Nikita Popov [Thu, 5 Jan 2023 16:12:31 +0000 (17:12 +0100)]
[Attributes] Remove trailing empty attribute sets (PR59746)
In setAttributesAtIndex(), remove any trailing empty attribute sets.
Also make sure that all the different attribute removal APIs go
through that method.
Fixes https://github.com/llvm/llvm-project/issues/59746.
OCHyams [Thu, 5 Jan 2023 14:55:41 +0000 (14:55 +0000)]
[DebugInfo][SelectionDAGISel] Do not drop all dbg.declares if one with empty metadata is found
This error was introduced in
1d1de7467c32d52926ca56b9167a2c65c451ecfa (by me)
about 1 month ago. Found while testing the D140901 patch stack.
Reviewed By: jryans
Differential Revision: https://reviews.llvm.org/D141052
Chris Cotter [Thu, 5 Jan 2023 15:28:43 +0000 (15:28 +0000)]
[clang-tidy] Fix minor bug in add_new_check.py
While rebuilding the list of checks in add_new_check.py,
check is a file is a subdirectory before traversing it.
Test plan: Ran `./add_new_check.py --update-docs` and confirmed the list.rst file was unchanged.
Reviewed By: carlosgalvezp
Differential Revision: https://reviews.llvm.org/D140772
Matt Arsenault [Thu, 5 Jan 2023 13:13:19 +0000 (08:13 -0500)]
Verifier: Move test from test/Analysis to test/Verifier
This test was purely checking the verifier error, and had no reason
to be running instcombine.
Matt Arsenault [Wed, 4 Jan 2023 22:46:03 +0000 (17:46 -0500)]
llvm-reduce: Don't delete instructions in global variable reduction
For some reason the global variable reduction was trying to delete
use instructions. This broke the verifier if the user was a terminator,
since the block now no longer has one. It doesn't make sense for this
reduction to delete the users, so just stop doing that.
Matt Arsenault [Thu, 5 Jan 2023 13:28:00 +0000 (08:28 -0500)]
IROutliner: Fix missing C++ mode comment
Alex Zinenko [Thu, 5 Jan 2023 13:37:23 +0000 (13:37 +0000)]
[mlir] fix out-of-bounds in reduction tiling
A transformation tiling a reduction dimension of a Linalg op needs a
tile size for said dimension. When an insufficient number of dimensions
was provided, it would segfault due to out-of-bounds access to a vector.
Also fix incorrect error reporting in the structured transform op
exercising this functionality.
Reviewed By: springerm, ThomasRaoux
Differential Revision: https://reviews.llvm.org/D141046
Roman Lebedev [Thu, 5 Jan 2023 14:38:35 +0000 (17:38 +0300)]
[exegesis] "Skip codegen" dry-run mode
While "skip measurements mode" is super useful for test coverage,
i've come to discover it's trade-offs. It still calls back-end
to actually codegen the target assembly, and that is what is taking
80%+ of the time regardless of whether or not we skip the measurements.
On the other hand, just being able to see that exegesis can come up
with a snippet to measure something, is already very useful,
and takes maybe a second for a all-opcode sweep.
Reviewed By: gchatelet
Differential Revision: https://reviews.llvm.org/D140702
Roman Lebedev [Thu, 5 Jan 2023 14:38:18 +0000 (17:38 +0300)]
[ScalarEvolutionExpanderTest] Avoid sign warning
Zahira Ammarguellat [Thu, 13 Oct 2022 18:40:21 +0000 (14:40 -0400)]
Implement support for option 'fexcess-precision'.
Differential revision: https://reviews.llvm.org/D136176
Guillaume Chatelet [Wed, 4 Jan 2023 12:08:06 +0000 (12:08 +0000)]
[libc] get_table_negative cannot be constexpr
GCC complains about if conditions not being constant expressions.
As-is get_table_negative cannot be marked constexpr.
Differential Revision: https://reviews.llvm.org/D140967
serge-sans-paille [Thu, 5 Jan 2023 14:25:47 +0000 (15:25 +0100)]
[NFC] Use std::size(x) Instead of ArrayRef(x).size
This is equivalent for all kind of `x` accepted by ArrayRef.
Matt Arsenault [Thu, 22 Dec 2022 16:09:31 +0000 (11:09 -0500)]
AMDGPU: Fix broken and permissive handling of printf format strings
This was completely broken with opaque pointers because it was
specifically looking for a constant expression with the global
variable as the first operand. Strip casts like normal, and properly
validate all of the restrictions rather than silently ignoring any
unhandled cases. Also be stricter that we aren't calling into some
unresolved or non-constant format string.
Also converts the test to opaque pointers and generated tests. There's
more broken initializer handling for strings inside the format string
processing too, but there's just no test coverage for this at all.
LLVM GN Syncbot [Thu, 5 Jan 2023 14:03:43 +0000 (14:03 +0000)]
[gn build] Port
b06b248ad9dc
Guillaume Chatelet [Thu, 5 Jan 2023 14:01:15 +0000 (14:01 +0000)]
[libc] Fix invalid constructor declaration
Chris Cotter [Thu, 5 Jan 2023 12:58:34 +0000 (12:58 +0000)]
[clang-tidy] Implement CppCoreGuideline CP.53
Implement CppCoreGuideline CP.53 to warn when a coroutine accepts
references parameters. Although the guideline mentions that it is safe
to access a reference parameter before suspension points, the guideline
recommends flagging all coroutine parameter references.
Reviewed By: carlosgalvezp
Differential Revision: https://reviews.llvm.org/D140793
Jean Perier [Thu, 5 Jan 2023 13:45:14 +0000 (14:45 +0100)]
[flang][NFC] share Constant<SomeDerived> lowering
A previous patch (https://reviews.llvm.org/D136955) already refactored
intrinsic constant lowering to place in its own file and allow using it from
both the current lowering and the new lowering to HLFIR.
This patch does the same for derived types. The core function
"genStructComponentInInitializer" is moved from ConvertExpr.cpp and
renamed "genInlinedStructureCtorLitImpl" into ConvertConstant.cpp
without significant logic change.
Then, genScalarLit, genArrayLit (and genInlinedArrayLit/genOutlinedArrayLit)
are updated to support derived types.
The core aspect of derived type constant lowering that differs between
the current lowering and the HLFIR update is the way
addresses/initial target descriptors are built when part of a derived
type constant. This part happens in ConvertVariable.cpp (since the
address of a variable is taken in an initializer and is left TODO).
The mangling of derived type global literal constant is fixed: it did not embed
the derived type name and could cause "conflicts" between unrelated
derived types containing the same data. However, the hash remains
unstable between two compilation of the same file. This is not a
correctness issue and would require a lot of work to hash the derived
type constant data without hashing some irrelevant (but not out of bound)
data in the compile time data structure that holds derived type
constants (Constant<SomeDerived>). This may have to be revisited later.
Differential Revision: https://reviews.llvm.org/D140986
Nikita Popov [Thu, 5 Jan 2023 13:16:11 +0000 (14:16 +0100)]
[Examples] Convert test to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 13:13:58 +0000 (14:13 +0100)]
[DFSan] Convert test to opaque pointers (NFC)
Retaining the unused bitcasts in this test, as they are required
to reproduce the test behavior for whatever reason.
serge-sans-paille [Wed, 4 Jan 2023 07:28:45 +0000 (08:28 +0100)]
Move from llvm::makeArrayRef to ArrayRef deduction guides - llvm/ part
Use deduction guides instead of helper functions.
The only non-automatic changes have been:
1. ArrayRef(some_uint8_pointer, 0) needs to be changed into ArrayRef(some_uint8_pointer, (size_t)0) to avoid an ambiguous call with ArrayRef((uint8_t*), (uint8_t*))
2. CVSymbol sym(makeArrayRef(symStorage)); needed to be rewritten as CVSymbol sym{ArrayRef(symStorage)}; otherwise the compiler is confused and thinks we have a (bad) function prototype. There was a few similar situation across the codebase.
3. ADL doesn't seem to work the same for deduction-guides and functions, so at some point the llvm namespace must be explicitly stated.
4. The "reference mode" of makeArrayRef(ArrayRef<T> &) that acts as no-op is not supported (a constructor cannot achieve that).
Per reviewers' comment, some useless makeArrayRef have been removed in the process.
This is a follow-up to https://reviews.llvm.org/D140896 that introduced
the deduction guides.
Differential Revision: https://reviews.llvm.org/D140955
Nikita Popov [Thu, 5 Jan 2023 13:04:32 +0000 (14:04 +0100)]
[LoopSimplifyCFG] Convert test to opaque pointers (NFC)
Luke Drummond [Thu, 5 Jan 2023 13:01:28 +0000 (13:01 +0000)]
[NFC][NVPTX] Move a comment back to its proper place
The comment introduced in
b94bd05b952a5 was misplaced during
f14af1621942 and no longer comments on the relevant bit of code; move it
back so it makes sense.
Christian Ulmann [Thu, 5 Jan 2023 12:21:57 +0000 (13:21 +0100)]
[mlir] Add function_entry_count to LLVMFuncOp
This commit introduces the function_entry_count metadata field to the
LLVMFuncOp and adds both the corresponding import and export
funtionalities.
The import of the function metadata uses the same infrastructure as the
instruction metadata, i.e., it dispatches through a dialect interface.
Reviewed By: gysit
Differential Revision: https://reviews.llvm.org/D141001
Freddy Ye [Thu, 5 Jan 2023 12:27:13 +0000 (20:27 +0800)]
[X86] Support -march=emeraldrapids
Reviewed By: pengfei, skan
Differential Revision: https://reviews.llvm.org/D140950
Nikita Popov [Thu, 5 Jan 2023 12:13:06 +0000 (13:13 +0100)]
[CodeGen] Convert some tests to opaque pointers (NFC)
These are mostly MIR tests, which I did not handle during previous
conversions.
Nikita Popov [Thu, 5 Jan 2023 12:10:32 +0000 (13:10 +0100)]
[Verifier] Convert some tests to opaque pointers (NFC)
Owen Pan [Wed, 4 Jan 2023 07:07:15 +0000 (23:07 -0800)]
[clang-format] Add an option for breaking after C++11 attributes
Fixes #45968.
Fixes #54265.
Fixes #58102.
Differential Revision: https://reviews.llvm.org/D140956
Freddy Ye [Thu, 5 Jan 2023 12:07:58 +0000 (20:07 +0800)]
[X86] Modify model value of meteorlake.
Reviewed By: pengfei
Differential Revision: https://reviews.llvm.org/D140943
Nikita Popov [Thu, 5 Jan 2023 12:01:24 +0000 (13:01 +0100)]
[SCEVExpanderTest] Convert to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 11:50:39 +0000 (12:50 +0100)]
[PredicateInfo] Enable test with broken REQUIRES condition (NFC)
Add some extra uses of the comparisons, so that these do get
visited.
Ties Stuij [Thu, 5 Jan 2023 11:19:56 +0000 (11:19 +0000)]
[AArch64][GlobalISel] implement GPR (U/S)(MIN/MAX) instr support
Lower umin, umax, smin, smax intrinsics to corresponding UMIN, UMAX, SMIN, SMAX
instructions when feat CSSC is available.
Reviewed By: aemerson
Differential Revision: https://reviews.llvm.org/D139420
Nikita Popov [Thu, 5 Jan 2023 11:35:52 +0000 (12:35 +0100)]
[Transforms] Convert some tests to opaque pointers (NFC)
These are all tests where conversion worked automatically, and
required no manual fixup.
David Green [Thu, 5 Jan 2023 11:37:37 +0000 (11:37 +0000)]
[LoopVectorizer] Fix inloop reductions mask placement
The validation of vplans could fail if an inloop reduction was created
with a block-in mask that did not dominate the reduction. This makes
sure that the insert point is set when creating the mask, to ensure it
dominates the reduction.
Differential Revision: https://reviews.llvm.org/D141003
Nikita Popov [Thu, 5 Jan 2023 11:34:36 +0000 (12:34 +0100)]
[CVP] Convert tests to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 11:32:44 +0000 (12:32 +0100)]
[SLPVectorizer] Convert test to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 11:26:39 +0000 (12:26 +0100)]
[llvm-reduce] Remove typed pointer test (NFC)
There already is a test using opaque pointers instead.
Nikita Popov [Thu, 5 Jan 2023 11:19:16 +0000 (12:19 +0100)]
[UpdateTestChecks] Convert tests to opaque pointers (NFC)
Brad Smith [Thu, 5 Jan 2023 11:08:42 +0000 (06:08 -0500)]
[Driver] move NetBSD header search path management to the driver
This matches OpenBSD and FreeBSD. https://reviews.llvm.org/D138183
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D140817
Jay Foad [Tue, 3 Jan 2023 17:06:59 +0000 (17:06 +0000)]
[GlobalISel] New combine to commute constant operands to the RHS
Differential Revision: https://reviews.llvm.org/D140907
Nikita Popov [Thu, 5 Jan 2023 11:11:12 +0000 (12:11 +0100)]
[LinkModulesTest] Convert to opaque pointers (NFC)
This requires switching to the ssa.copy intrinsic, otherwise
we don't preserve test intent.
Mehdi Amini [Wed, 4 Jan 2023 21:52:39 +0000 (21:52 +0000)]
Use --pass-pipeline syntax for mlir/test/Dialect/LLVMIR/canonicalize.mlir (NFC)
This is just a cleanup to make the scheduling of the pass pipeline
explicit.
Nikita Popov [Thu, 5 Jan 2023 10:14:31 +0000 (11:14 +0100)]
[OperationsTest] Convert to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 10:08:23 +0000 (11:08 +0100)]
[RandomIRBuilderTest] Convert to opaque pointers (NFC)
The FirstClassTypes test is no longer relevant with opaque pointers,
or at least not in that form.
Nikita Popov [Thu, 5 Jan 2023 09:59:37 +0000 (10:59 +0100)]
[CodeGenOpenCLCXX] Convert tests to opaque pointers (NFC)
Guillaume Chatelet [Wed, 4 Jan 2023 12:48:01 +0000 (12:48 +0000)]
[libc][NFC] Use a function instead of templated static member functions for TYPE_DESC
I'm surprised that clang accepts the current code.
It seems odd to me to specialize templated static member variables.
GCC rejects them: https://godbolt.org/z/3ecE9Ps7T
This patch is in the context of https://github.com/llvm/llvm-project/issues/59368
Differential Revision: https://reviews.llvm.org/D140981
Mehdi Amini [Thu, 5 Jan 2023 00:58:52 +0000 (00:58 +0000)]
Apply clang-tidy fixes for performance-unnecessary-value-param in AffineCanonicalizationUtils.cpp (NFC)
Nikita Popov [Thu, 5 Jan 2023 09:57:30 +0000 (10:57 +0100)]
[CodeGenOpenCL] Convert some tests to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 09:53:37 +0000 (10:53 +0100)]
[AsmParserTest] Convert to opaque pointers (NFC)
Nikita Popov [Thu, 5 Jan 2023 09:48:29 +0000 (10:48 +0100)]
[IVDescriptorsTest] Convert to opaque pointers (NFC)
The LoopWithScalableTypes test is removed, because it is testing
a situation that cannot occur with opaque pointers. If we convert
this test to opaque pointers, it will not contain any scalable
types at all.
Brad Smith [Thu, 5 Jan 2023 09:39:46 +0000 (04:39 -0500)]
[Hexagon][VE][WebAssembly] Define __GCC_HAVE_SYNC_COMPARE_AND_SWAP macros
Define __GCC_HAVE_SYNC_COMPARE_AND_SWAP macros
Reviewed By: kparzysz, aheejin, MaskRay
Differential Revision: https://reviews.llvm.org/D140757
Dawid Jurczak [Sun, 1 Jan 2023 12:18:58 +0000 (13:18 +0100)]
[NFC][Coroutines] Build DominatorTree only once before collecting frame allocas (PR58650)
Assuming that collecting frame allocas doesn't modify CFG we can safely
move DominatorTree construction outside loop and avoid expensive computations.
Differential Revision: https://reviews.llvm.org/D140818
Diana Picus [Wed, 4 Jan 2023 10:58:58 +0000 (11:58 +0100)]
[GlobalISel] Add G_BUILD_VECTOR[_TRUNC] to CSE
Add G_BUILD_VECTOR and G_BUILD_VECTOR_TRUNC to the list of opcodes in
`shouldCSEOpc`. This simplifies the code generated for vector splats.
Differential Revision: https://reviews.llvm.org/D140965
Diana Picus [Wed, 4 Jan 2023 10:59:49 +0000 (11:59 +0100)]
[GlobalISel] Don't switch opcodes in MIRBuilder::buildInstr
At the moment, `MachineIRBuilder::buildInstr` may build an instruction
with a different opcode than the one passed in as parameter. This may
cause confusion for its consumers, such as `CSEMIRBuilder`, which will
memoize the instruction based on the new opcode, but will search
through the memoized instructions based on the original one (resulting
in missed CSE opportunities). This is all the more unpleasant since
buildInstr is virtual and may call itself recursively both directly
and via buildCast, so it's not always easy to follow what's going on.
This patch simplifies the API of `MachineIRBuilder` so that the `buildInstr`
method does the least surprising thing (i.e. builds an instruction with
the specified opcode) and only the convenience `buildX` methods
(`buildMerge` etc) are allowed freedom over which opcode to use. This can
still be confusing (e.g. one might write a unit test using
`buildBuildVectorTrunc` but instead get a plain `G_BUILD_VECTOR`), but at
least it's explained in the comments.
In practice, this boils down to 3 changes:
* `buildInstr(G_MERGE_VALUES)` will no longer call itself with
`G_BUILD_VECTOR` or `G_CONCAT_VECTORS`; this functionality is moved to
`buildMerge` and replaced with an assert;
* `buildInstr(G_BUILD_VECTOR_TRUNC)` will no longer call itself with
`G_BUILD_VECTOR`; this functionality is moved to `buildBuildVectorTrunc`
and replaced with an assert;
* `buildInstr(G_MERGE_VALUES)` will no longer call `buildCast` and will
instead assert if we're trying to merge a single value; no change is
needed in `buildMerge` since it was already asserting more than one
source operand.
This change is NFC for users of the `buildX` methods, but users that
call `buildInstr` with relaxed parameters will have to update their code
(such instances will hopefully be easy to find thanks to the asserts).
Differential Revision: https://reviews.llvm.org/D140964
Diana Picus [Wed, 4 Jan 2023 13:15:31 +0000 (14:15 +0100)]
[GlobalISel] Precommit a test for D140965
Add a test for CSE-ing G_BUILD_VECTOR. This will be enabled in D140965.
Benjamin Chetioui [Thu, 5 Jan 2023 08:26:13 +0000 (09:26 +0100)]
[NVPTX] Fix NVPTX lowering of frem when denominator is infinite.
`frem x, {+,-}inf` must return x to match the specification of LLVM's frem.
Reviewed By: tra
Differential Revision: https://reviews.llvm.org/D140846
Johannes Doerfert [Wed, 28 Dec 2022 07:50:42 +0000 (23:50 -0800)]
Introduce environment variables to deal with JIT IR
We can now dump the IR before and after JIT optimizations into the
files passed via `LIBOMPTARGET_JIT_PRE_OPT_IR_MODULE` and
`LIBOMPTARGET_JIT_POST_OPT_IR_MODULE`, respectively.
Similarly, users can set `LIBOMPTARGET_JIT_REPLACEMENT_MODULE` to
replace the IR in the image with a custom IR module in a file.
All options take file paths, documentation was added.
Reviewed by: tianshilei1992
Differential revision: https://reviews.llvm.org/D140945
Joshua Cao [Fri, 16 Dec 2022 09:00:56 +0000 (04:00 -0500)]
[LoopUnrollAndJam] Visit phi operand dependencies in post-order
Fixes https://github.com/llvm/llvm-project/issues/58565
The previous implementation visits operands in pre-order, but this does
not guarantee an instruction is visited before its uses. This can cause
instructions to be copied in the incorrect order. For example:
```
a = ...
b = add a, 1
c = add a, b
d = add b, a
```
Pre-order visits does not guarantee the order in which `a` and `b` are
visited. LoopUnrollAndJam may incorrectly insert `b` before `a`.
This patch implements post-order visits. By visiting dependencies first,
we guarantee that an instruction's dependencies are visited first.
Differential Revision: https://reviews.llvm.org/D140255
Chen Zheng [Thu, 5 Jan 2023 07:53:59 +0000 (07:53 +0000)]
1: use class instead of MVT
2: minor fix for the comments
Chen Zheng [Tue, 3 Jan 2023 09:44:03 +0000 (09:44 +0000)]
[PowerPC][GISel]fcmp support
This patch also includes:
1: CRRegBank support
2: Some workarounds in PPC table gen for anyext/setcc patterns
selection.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D140878
luxufan [Thu, 5 Jan 2023 06:38:08 +0000 (14:38 +0800)]
[InstCombine][NFC] Regenerate test cases by update_test_checks.py
luxufan [Thu, 5 Jan 2023 06:20:41 +0000 (14:20 +0800)]
[InstCombine][NFC] Regenerate test cases by update_test_checks.py
Fangrui Song [Thu, 5 Jan 2023 06:05:40 +0000 (22:05 -0800)]
[lldb] Remove unneeded .get() NFC