This merge brings in one conflicting commit from upstream,
f2f88f3e, which disables out-of-tree libomptarget builds.  This
merge makes related adjustments to Clacc's cmake code that installs
OpenMP libraries to Clang's lib directory.

This merge also brings in CI config xfail updates due to upstream
changes.

This brings in:

* A CI config change to reduce lit parallelism on leconte to avoid
  "CUDA error: out of memory", which started to appear after the
  previous merge (13529945) from upstream main.
* CI config xfail updates due to changes upstream.
* Our downstream extension of lit `--xfail` to support fully qualified
  test names.  Otherwise, we cannot precisely mark some current xfails
  in our CI config.

The test would previously disable itself using `#if TEST_STD_VER` instead
of using UNSUPPORTED markup.

The new documentation entry gives an example use case.

We'll attempt to upstream this later.

1Tb counting was the slowest test under the QEMU with MTE.

Differential Revision: https://reviews.llvm.org/D103270

Mark the `ELFRelocationEntry::dump` method as `LLVM_DUMP_METHOD` to
annotate it properly as used to prevent the function being dead stripped
away.  This allows use of `dump` in the debugger.  This is purely to
improve the developer experience.

That fix was actually incorrect and caused tests to start failing.

Determined from llvm-mca analysis (btver2 vs bdver2 vs sandybridge), the split+extends+concat sequence on AVX1 capable targets are cheaper than the #ops that the cost was previously based on.

[RISCV] Teach vsetvli insertion to use vsetvl x0, x0 form when we can tell that VLMAX and AVL haven't changed.

This can help avoid needing a virtual register for the vsetvl output
when the AVL is X0. For other register AVLs it can shorter the live
range of the AVL register if it isn't needed later.

There's probably no advantage when AVL is a 5 bit immediate that
can use vsetivli. But do it anyway for consistency.

Reviewed By: rogfer01

Differential Revision: https://reviews.llvm.org/D103215

This matches better how other gpu integration tests are done.

Differential Revision: https://reviews.llvm.org/D103099

Depends On D103109

If any of the tokens/values added to the `!async.group` switches to the error state, than the group itself switches to the error state.

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D103203

We could previously do this by accident through the later
call to getTargetConstantBitsFromNode I think, but that only worked
if N0 had a single use. This patch makes it explicit for undef and
doesn't have a use count check.

I think this is needed to move the (shl X, 1)->(add X, X)
fold to isel for PR50468. We need to be sure X won't be IMPLICIT_DEF
which might prevent the same vreg from being used for both operands.

Differential Revision: https://reviews.llvm.org/D103192

Depends On D103102

Not yet implemented:
1. Error handling after synchronous await
2. Error handling for async groups

Will be addressed in the followup PRs

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D103109

Like other sanitizers, enable __has_feature(coverage_sanitizer) if clang
has enabled at least one SanitizerCoverage instrumentation type.

Because coverage instrumentation selection is not handled via normal
-fsanitize= (and thus not in SanitizeSet), passing this information
through to LangOptions required propagating the already parsed
-fsanitize-coverage= options from CodeGenOptions through to LangOptions
in FixupInvocation().

Reviewed By: aaron.ballman

Differential Revision: https://reviews.llvm.org/D103159

Support reference counted values implicitly passed (live) only to some of the successors.

Example: if branched to ^bb2 token will leak, unless `drop_ref` operation is properly created

```
^entry:
  %token = async.runtime.create : !async.token
   cond_br %cond, ^bb1, ^bb2
^bb1:
  async.runtime.await %token
  async.runtime.drop_ref %token
  br ^bb2
^bb2:
  return
```

Reviewed By: mehdi_amini

Differential Revision: https://reviews.llvm.org/D103102

This patch changes LoopUnrollAndJamPass from FunctionPass to LoopNest pass.
The next patch will utilize LoopNest to effectively handle loop nests.

Reviewed By: Whitney

Differential Revision: https://reviews.llvm.org/D99149

We're seeing a lot of "CUDA error: out of memory" on leconte, and it's
racy.  More lit parallelism (more lit tests at once) means more CUDA
offloading at the same time, so maybe reducing lit parallelism will
help.  This attempt aims at a maximum ratio of 8 lit threads per GPU,
which has avoided these errors on at least one other system.

As discussed in PR50385, strict-fp on PowerPC SPE has not been handled
well. This patch disables it by default for SPE.

Reviewed By: nemanjai, vit9696, jhibbits

Differential Revision: https://reviews.llvm.org/D103235

In order to allow large matmul operations using the MMA ops we need to chain
operations this is not possible unless "DOp" and "COp" type have matching
layout so remove the "DOp" layout and force accumulator and result type to
match.
Added a test for the case where the MMA value is accumulated.

Differential Revision: https://reviews.llvm.org/D103023

-fgpu-sanitize is incompatible with offload arch containing xnack-.

This patch checks that.

Reviewed by: Artem Belevich

Differential Revision: https://reviews.llvm.org/D102975

@HsiangKai helped find a bug in the lowering of indirect split
scalable-vector types in our calling convention. An imminent patch will
fix this.

Adjusting the load register type is a widenScalar type action, not a
lowering. lowerLoad should be reserved for operations that change the
memory access size, such as unaligned load decomposition. With this
trying to adjust the register type, it was hard to avoid infinite
loops in the legalizer. Adds a bandaid to avoid regressing a few
AArch64 tests, but I'm not sure what the exact condition is and
there's probably a cleaner way to do this.

For AMDGPU this regresses handling of some cases for unaligned loads,
but the way this is currently working is a pretty ugly hack.

Summary:
We are going to have libc++abi.a and libunwind.a on AIX.
Add the necessary linking command to pick the libraries up.

Reviewed By: daltenty

Differential Revision: https://reviews.llvm.org/D102813

Similar to how we allow managed and asserted locks to be held and not
held in joining branches, we also allow them to be held shared and
exclusive. The scoped lock should restore the original state at the end
of the scope in any event, and asserted locks need not be released.

We should probably only allow asserted locks to be subsumed by managed,
not by (directly) acquired locks, but that's for another change.

Reviewed By: delesley

Differential Revision: https://reviews.llvm.org/D102026

It's going to become a bit more complicated, so let's have it separate.

Reviewed By: aaron.ballman

Differential Revision: https://reviews.llvm.org/D102025

Breaks check-llvm on non-linux, see comments on https://reviews.llvm.org/D85085
This reverts commit caae5709
and follow-up commit 1546c52d.

Otherwise it issues a -Werror=parentheses suggesting parentheses.

C++17 deprecated std::iterator and removed it as a base class for all
iterator adaptors. We implement that change, but we still provide a way
to inherit from std::iterator in the few cases where doing otherwise
would be an ABI break.

Supersedes D101729 and the std::iterator base parts of D103101 and D102657.

Differential Revision: https://reviews.llvm.org/D103171

This makes no real difference since we assign the same register either
way.

The SkylakeServer model (and later IceLake/TigerLake targets according to Agner) have the PMOV truncations as uops=2, rthroughput=2 instructions.

Noticed while trying to reduce the diffs between cost tables and llvm-mca analysis.