Differential Revision: https://reviews.llvm.org/D132744

This was requested in https://reviews.llvm.org/D132437.

This reverts commit d23df9c9.
Revert due to missing review link.

The use of std::clamp should be safe here.  MinRZ is at most 32, while
kMaxRZ is 1 << 18, so we have MinRZ <= kMaxRZ, avoiding the undefind
behavior of std::clamp.

Reviewed By: bcl5980
Differential Revision: https://reviews.llvm.org/D132820

The motivation of this patch is to lower the IR pattern
(vp.merge mask, (add x, y), false, vl) to
(PseudoVADD_VV_<LMUL>_MASK false, x, y, mask, vl).

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D131841

mlir::TypedValue is a wrapper class for mlir::Values with a known type
getType will return the known type and all assignements will be checked

Also the tablegen Operation generator was adapted to use mlir::TypedValue
when appropriate

This patch adds cost model coverage for fp arithmetic instructions. Some is not exact, I am working on a revision to implement that.

Differential Revision: https://reviews.llvm.org/D132537

Change fadd test case in D122563 to match the fold base case.

Differential Revision: https://reviews.llvm.org/D132722

Reviewed By: benshi001

Differential Revision: https://reviews.llvm.org/D132281

Reviewed By: reames

Differential Revision: https://reviews.llvm.org/D132724

Differential Revision: https://reviews.llvm.org/D132718

This reverts commit ebf574f5.

This patch seems to cause build failures on Windows.

combinartorial
       v
combinatorial

This patch removes greatestCommonDivisor and GreatestCommonDivisor64
as I've migrated all the uses to std::gcd.

When an l_brace is wrapped and the line above it ends with a
comment, the annotator adds ColumnLimit to the TotalLength of the
l_brace, so the actual column position of the l_brace must be
adjusted accordingly.

Fixes #57376.

Differential Revision: https://reviews.llvm.org/D132805

Differential Revision: https://reviews.llvm.org/D132562

To align with many Linux distributions which configure GCC with
--enable-default-hash-style=gnu or binutils with --with-linker-hash-style=gnu.
ld.lld does not support changed default, so passing the option in the driver is
better.

Note: there was a bug that Debian specified both --hash-style=gnu and
--hash-style=both. Actually its GCC has defaulted to --hash-style=gnu since 2007.

https://reviews.llvm.org/D93267 implemented handling more than 127
compact unwind encodings, and https://reviews.llvm.org/D123435 and
https://reviews.llvm.org/D124561 implemented stripping redundant
__eh_frame entries.

This has been the default for a very long time.

Reviewed By: vitalybuka

Differential Revision: https://reviews.llvm.org/D131343

Reviewed By: vitalybuka

Differential Revision: https://reviews.llvm.org/D131574

Reviewed By: vitalybuka

Differential Revision: https://reviews.llvm.org/D131342

To avoid changing semantics inadvertently, this patch casts arguments
to uint64_t before calling std::gcd.

This patch replaces calls to greatestCommonDivisor with std::gcd where
two arguments are of the same type.  This means that
std::common_type_t of the argument type is the same as the argument
type.

We could drop calls to std::abs in some cases, but that's left for
another patch.