Changes

[OpenMP] cmake option LIBOMPTARGET_NVPTX_MAX_SM for nvptx device RTL

It allows customizing MAX_SM for non-flagship GPU and reduces graphic memory usage.

In addition, so far the size is hard-coded up to __CUDA_ARCH__ 700 and is already a hassle for 800.
Introduce MAX_SM for 800 and protect future arch

Reviewed By: JonChesterfield

Differential Revision: https://reviews.llvm.org/D88185

[IRSim][NFC] Removing dead variables from IRSimilarityIdentifier.cpp

As informed by danielkiss.

Follow up to Differential Revision: https://reviews.llvm.org/D86972

[AArch64][GlobalISel] Implement __builtin_return_address for PAC-RET

This patch implements stripping of the PAC in the return address for GlobalISel.

Implementation for when not using GLobalISel is in
https://reviews.llvm.org/D75044 The analogous GCC patch is
https://gcc.gnu.org/git/?p=gcc.git;a=commitdiff;h=a70d5d81c41048556fd86eaa1036018a6bfba115

Differential Revision: https://reviews.llvm.org/D84502

[OPENMP]PR47606: Do not update the lastprivate item if it was captured by reference as firstprivate data member.

No need to make final copy from the firsptrivate/lastprivate copy to the original item if the item is a data memeber.
Firstprivate copy creates a copy by reference and the original item gets
updated correctly when updating the lastprivate shared variable.

Differential Revision: https://reviews.llvm.org/D88179

[AMDGPU] global-isel support for RT

Differential Revision: https://reviews.llvm.org/D87847

[AArch64] Regenerate dag-numsignbits.ll checks

To improve the codegen diff in D87502

[Scalar] ConstantHoistingPass - iterate with const references. NFCI.

Fix some clang-tidy warnings.

[OPENMP]Fix PR47621: Variable used by task inside a template function is not made firstprivate by default

Need to fix a check for the variable if it is declared in the inner
OpenMP region to be able to firstprivatize it.

Differential Revision: https://reviews.llvm.org/D88240

[SLP] Remove LHS and RHS from OperationData.

These were only really used for 2 things. One was to check if the operand matches the phi if it exists. The other was for the createOp method to build the reduction.

For the first case we still have the operation we just need to know how to index its operands. So I've modified getLHS/getRHS to just use the opcode/kind to know how to find the right operands on an instruction that is now passed in.

For the other case we had to create an OperationData object to set the LHS/RHS values and copy the opcode/kind from another object. We would then just call createOp on that temporary object. Instead I've made LHS/RHS arguments to createOp and removed all these temporary objects.

Differential Revision: https://reviews.llvm.org/D88193

[APFloat] prevent NaN morphing into Inf on conversion (PR43907)

We shift the significand right on a truncation, but that needs to be made NaN-safe:
always set at least 1 bit in the significand.
https://llvm.org/PR43907

See D88238 for the likely follow-up (but needs some plumbing fixes before it can proceed).

Differential Revision: https://reviews.llvm.org/D87835

Revert "Recommit "[CUDA][HIP] Defer overloading resolution diagnostics for host device functions""

This reverts commit e39da8ab6a286ac777d5fe7799f1eb782cf99938.

This depends on a change that needs additional design review and needs
to be reverted.

Revert "Recommit [NFC] Refactor DiagnosticBuilder and PartialDiagnostic"

This reverts commit 8e780a1653e6f87755a447e921b8f929d8b70996.

DiagnosticBuilder is a value type, created on the stack everywhere. IMO
we should not be adding a vtable to it, and making very operator<< use a
virtual interface. There are other feasible designs for implementing
this. The original review, D84362, was approved by @tra, who is
responsible for Clang's CUDA support, but it wasn't reviewed by @rsmith
or anyone responsible for clang's diagnostic library.

[llvm-exegesis] Add whitespace between words in error message

Revert "[OPENMP]Fix PR47621: Variable used by task inside a template function is not made firstprivate by default"

This reverts commit d1419c9fdab141617b6aa9f028191b9bfc8be260 to fix the
buffer overflow detected by address sanitiizer.

Basic: add an extra newline for sphinx (NFC)

This should resolve the "Bullet list ends without a blank line" warning.

[LLD][COFF] Avoid overwriting inputs in tests

Before this patch, these two tests were emitting both a .DLL and .LIB. The output .LIB file name also happens to be an input .LIB file name. This prevented the test from executing a second time when LLD is re-entrant (LLD_IN_TEST=2).

This is a support patch for https://reviews.llvm.org/D70378.

[LLD][COFF] Cover usage of LLD-as-a-library in tests

In lit tests, we run each LLD invocation twice (LLD_IN_TEST=2), without shutting down the process in-between. This ensures a full cleanup is properly done between runs.
Only active for the COFF driver for now. Other drivers still use LLD_IN_TEST=1 which executes just one iteration with full cleanup, like before.
When the environment variable LLD_IN_TEST is unset, a shortcut is taken, only one iteration is executed, no cleanup for faster exit, like before.
A public API, lld::safeLldMain(), is also available when using LLD as a library.

Differential Revision: https://reviews.llvm.org/D70378

[PR47636] Fix tryEmitPrivate to handle non-constantarraytypes

As mentioned in the bug report, tryEmitPrivate chokes on the
MaterializeTemporaryExpr in the reproducers, since it assumes that if
there are elements, than it must be a ConstantArrayType. However, the
MaterializeTemporaryExpr (which matches exactly the AST when it is NOT a
global/static) has an incomplete array type.

This changes the section where the number-of-elements is non-zero to
properly handle non-CAT types by just extracting it as an array type
(since all we needed was the element type out of it).

Add support for setting the path to llvm-symbolizer through an environment variable

This allows to point to an executable that isn't named exactly
"llvm-symbolizer" and not necessarily in the current PATH.

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D88192

GlobalISel: Use unmerge when copying wide vectors to result registers

Avoid using G_EXTRACT and move towards a more consistent vector
legalization strategy.

[Modules] Add stats to measure performance of building and loading modules.

Measure amount of high-level or fixed-cost operations performed during
building/loading modules and during header search. High-level operations
like building a module or processing a .pcm file are motivated by
previous issues where clang was re-building modules or re-reading .pcm
files unnecessarily. Fixed-cost operations like `stat` calls are tracked
because clang cannot change how long each operation takes but it can
perform fewer of such operations to improve the compile time.

Also tracking such stats over time can help us detect compile-time
regressions. Added stats are more stable than the actual measured
compilation time, so expect the detected regressions to be less noisy.

rdar://problem/55715134

Reviewed By: aprantl, bruno

Differential Revision: https://reviews.llvm.org/D86895

[mlir][shape] Start a pass that lowers shape constraints.

This pass converts shape.cstr_* ops to eager (side-effecting)
error-handling code. After that conversion is done, the witnesses are
trivially satisfied and are replaced with `shape.const_witness true`.

Differential Revision: https://reviews.llvm.org/D87941

[NFC][regalloc] Remove unused API in AllocationOrder

Differential Revision: https://reviews.llvm.org/D88197

[SLP] clean up - use 'const' and ArrayRef constructor; NFC

Follow-on tidying suggested in the post-commit review of 6a23668.

Revert "[Modules] Add stats to measure performance of building and loading modules."

This reverts commit c4bacc3c9b333bb7032fb96f41d6f5b851623132.

Test "LLVM :: ThinLTO/X86/funcimport-stats.ll" is failing. Reverting now
and will recommit after making the test not fail with the added stats.

AArch64/GlobalISel: Add testcase for bug 47619

This is asserting on the 11 release branch, and wasn't covered by
exisiting tests at the time. This was fixed by
b98f902f1877c3d679f77645a267edc89ffcd5d6.

[IRSim] Adding wrapper pass for IRSimilarityIdentfier

This introduces an analysis pass that wraps IRSimilarityIdentifier,
and adds a printer pass to examine in what function similarities are
being found.

Test for what the printer pass can find are in
test/Analysis/IRSimilarityIdentifier.

Reviewed by: paquette, jroelofs

Differential Revision: https://reviews.llvm.org/D86973

Remove dead branch identified by @rsmith on post-commit for D88236

[flang][driver] Add missing dependency (shared library builds, NFC)

`flang-new` depends on libclangFrontend (it uses DiagnosticConsumer
classes from there). This patch adds the missing dependency in CMake.

clang::TextDiagnosticBuffer is only reported as missing when compiling
`flang-new` with BUILD_SHARED_LIBS=ON. This symbol is linked in
statically with libflangFrontend when BUILD_SHARED_LIBS=OFF.

[OPENMP]Fix PR47621: Variable used by task inside a template function is not made firstprivate by default

Need to fix a check for the variable if it is declared in the inner
OpenMP region to be able to firstprivatize it.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D88240

[CMake][AIX] Set LLVM_ENABLE_PER_TARGET_RUNTIME_DIR appropriately for AIX

AIX by default usually folds 32-bit & 64-bit arch libraries into a single
archive, a behaviour we may want for the runtime libraries in the future,
so we don't necessarily want to opt into the multlib layout introduce in
D45604, which is currently the default for runtime builds.

Reviewed By: hubert.reinterpretcast

Differential Revision: https://reviews.llvm.org/D88169

[AST] Use data-recursion when building ParentMap, avoid stack overflow.

The following crashes on my system before this patch, but not after:

  void foo(int i) {
    switch (i) {
      case 1:
      case 2:
      ... 100000 cases ...
        ;
    }
  }

  clang-query -c="match stmt(hasAncestor(stmt()))" deep.c

I'm not sure it's actually a sane testcase to run though, it's pretty slow :-)

Differential Revision: https://reviews.llvm.org/D88222

Remove stale assert.

This is triggered during serialization. The test is for modules, but
will occur for any serialization effort using asm goto.

Reviewed By: nickdesaulniers, jyknight

Differential Revision: https://reviews.llvm.org/D88195

[mlir][NFC] Promote memory space to BaseMemRefType

This patch moves the memory space field from MemRefType and UnrankedMemRefType
to their base class BaseMemRefType so that it can be retrieved from it without
downcasting it to the specific memref.

Reviewed By: silvas

Differential Revision: https://reviews.llvm.org/D87649

[AArch64] __builtin_return_address for PAuth.

This change adds the support for __builtin_return_address
for ARMv8.3A Pointer Authentication.
Location of the authentication code in the pointer depends on
the system configuration, therefore a dedicated instruction is used for
effectively removing the authentication code without
authenticating the pointer.

Reviewed By: chill

Differential Revision: https://reviews.llvm.org/D75044

[CUDA] Added conversion functions to builtin vars.

This is needed to compile some headers in CUDA-11 that assume that threadIdx is
implicitly convertible to dim3. With NVCC, threadIdx is uint3 and there's
dim3(uint3) constructor. Clang uses a special type for the builtin variables, so
that path does not work. Instead, this patch adds conversion function to the
builtin variable classes. that will allow them to be converted to dim3 and uint3.

Differential Revision: https://reviews.llvm.org/D88250

[CodeGen] Postprocess PHI nodes for callbr

When processing PHI nodes after a callbr, we need to make sure that the
PHI nodes on the default branch are resolved after the callbr
(inserted after INLINEASM_BR). The PHI node values on the indirect
branches are processed before the INLINEASM_BR.

Differential Revision: https://reviews.llvm.org/D86260

Fix testcase.

Revert "[CodeGen] Postprocess PHI nodes for callbr"

Accidental commit.

This reverts commit 7f4c940bd0b526f25e11c51bb4d58a85024330ae.

[NFCI][IR] ConstantRangeTest: refactor operation range gatherers

We do the same dance to acquire the "exact" range of an op via
an exhaustive approach in many places.
Let's not invent the wheel each time.

[NFCI][IR] ConstantRangeTest: add basic scaffolding for next-gen precision/correctness testing

I have long complained that while we have exhaustive tests
for ConstantRange, they are, uh, not good.

The approach of groking our own constant range
via exhaustive enumeration is, mysterious.

It neither tells us without doubt that the result is
conservatively correct, nor the precise match to the ConstantRange
result tells us that the result is precise.
But yeah, it's fast, i give it that.

In short, there are three things that we need to check:
1. That ConstantRange result is conservatively correct
2. That ConstantRange range is reasonable
3. That ConstantRange result is reasonably precise

So let's not just check the middle one, but all three.

This provides precision test coverage for D88178.

Reland [CodeGen] emit CG profile for COFF object file

This reverts commit 90242caca2074dab5a9b76e5bc36d9fafd2179a7.

Error fixed at f5435399e823746bbe1737b95c853d77a42e1ac3

Differential Revision: https://reviews.llvm.org/D87811

Temporary fix for D85085 debug_loc bug with basic block sections.

Until then, this one line fix removes the assert fail with basic block sections
with debug info. Bug tracking this: #47549
This fix does not generate loc list or DW_AT_const_value if the argument is
mentioned in a different section than the start of the function.

Temporarily fixes bugzilla : https://bugs.llvm.org/show_bug.cgi?id=47549

Differential Revision: https://reviews.llvm.org/D87787

[DebugInfo] Fix bug in constructor homing with classes with trivial
constructors.

This changes the code to avoid using constructor homing for aggregate
classes and classes with trivial default constructors, instead of trying
to loop through the constructors.

Differential Revision: https://reviews.llvm.org/D87808

[LLDB] Add a defensive check for member__f_

I only have a crash log and was not able to come up with a test case for this.

rdar://problem/69403150

[WebAssembly] Make SjLj lowering globals thread-local

Emscripten's longjump and exception mechanism depends on two global variables,
`__THREW__` and `__threwValue`, which are changed to be defined as thread-local
in https://github.com/emscripten-core/emscripten/pull/12056. This patch updates
the corresponding code in the WebAssembly backend to properly declare these
globals as thread-local as well.

Differential Revision: https://reviews.llvm.org/D88262

[Instruction] Add dropLocation and updateLocationAfterHoist helpers

Introduce a helper which can be used to update the debug location of an
Instruction after the instruction is hoisted. This can be used to safely
drop a source location as recommended by the docs.

For more context, see the discussion in https://reviews.llvm.org/D60913.

Differential Revision: https://reviews.llvm.org/D85670

[lld-macho][re-land] Initial support for common symbols

Fix earlier build break via a static_cast.

This reverts commit 8112d494d344dc0935d5c078f066a43d7c984e0c.

Differential Revision: https://reviews.llvm.org/D86909

[lld-macho][re-land] Implement and test resolution of common symbols

Earlier build break fixed in c32e69b2ce7abfb151a87ba363ac9e25abf7d417.

This reverts commit c367f93e8539c4d0bcdc86ad7ea7923e06231a93.

[lld] Make -z keep-text-section-prefix recognize .text.split. as a prefix.

".text.split." holds symbols which are split out from functions in
other input sections. For example, with -fsplit-machine-functions,
placing the cold parts in .text.split instead of .text.unlikely mitigates
against poor profile inaccuracy. Techniques such as hugepage remapping can
make conservative decisions at the section granularity.

Differential Revision: https://reviews.llvm.org/D87840

Fix regex in test.

[OpenMP] OpenMPOpt Support for Globalization Remarks

Summary:
This patch add support for printing analysis messages relating to data
globalization on the GPU. This occurs when data is shared between the
threads in a GPU context and must be pushed to global or shared memory.

Reviewers: jdoerfert

Subscribers: guansong hiraditya llvm-commits ormris sstefan1 yaxunl

Tags: #OpenMP #LLVM

Differential Revision: https://reviews.llvm.org/D88243

[llvm] Add -bbsections-cold-text-prefix to emit cold clusters to a different section.

This change adds an option to basic block sections to allow cold
clusters to be assigned a custom text prefix. With a custom prefix such
as ".text.split." (D87840), lld can place them in a separate output section.
The benefits are -

* Empirically shown to improve icache and itlb metrics by 3-5%
(absolute) compared to placing split parts in .text.unlikely.
* Mitigates against poor profiles, eg samplePGO profiles used with the
machine function splitter. Optimizations such as hugepage remapping can
make different decisions at the section granularity.
* Enables section granularity hotness monitoring (checking on the
decisions made during compilation vs sample data from production).

Differential Revision: https://reviews.llvm.org/D87813

[RISCV] Merge the pipeline models for Rocket

Merge the 32 and 64 bit pipeline models for Rocket into a single file.

Differential Revision: https://reviews.llvm.org/D87873

[AMDGPU] Fixes typo in the test. NFC.

denormal-fp-math-fp32 -> denormal-fp-math-f32

[MS] Simplify rules for passing C++ records

Regardless of the target architecture, we should always use the C rules
(RAA_Default) for records that "canBePassedInRegisters". Those are
trivially copyable things, and things marked with [[trivial_abi]].

This should be NFC, although it changes where the final decision about
x86_32 overaligned records is made. The current x86_32 C rules say that
overaligned things are passed indirectly, so there is no functional
difference.

[MS] For unknown ISAs, pass non-trivially copyable arguments indirectly

Passing them directly is likely to be non-conforming, since it usually
involves copying the bytes of the record. For unknown architectures, we
don't know what MSVC does or will do, but we should at least try to
conform as well as we can.

[intel-pt] Refactor the JSON parsing

Recently https://reviews.llvm.org/D88103 introduced a nice API for
converting a JSON object into C++ types, which include nice error
messaging.

I'm using that new functioniality to perform the parsing in a much more
elegant way. As a result, the code looks simpler and more maintainable,
as we aren't parsing anymore individual fields manually.

I updated the test cases accordingly.

Differential Revision: https://reviews.llvm.org/D88264

Add a static_assert confirming that DiagnosticBuilder is small

Revert "Add a static_assert confirming that DiagnosticBuilder is small"

This reverts commit a32feed0dbeac7606d042d0d7e041c9eaf12cd51.

This assert doesn't hold in 32-bit builds, I didn't do the math right.

Revert "[NFCI][IR] ConstantRangeTest: add basic scaffolding for next-gen precision/correctness testing"

This reverts commit 9bcf7b1c7a139a455400df109d81c638b9e75150.

Breaks build with MSVC.

[ValueTracking] Check uses of Argument if it is given to isGuaranteedNotToBeUndefOrPoison

This is a patch that allows isGuaranteedNotToBeUndefOrPoison to return more precise result
when an argument is given, by looking through its uses at the entry block (and following blocks as well, if it is checking poison only).

This is useful when there is a function call with noundef arguments at the entry block.

Reviewed By: nikic

Differential Revision: https://reviews.llvm.org/D88207