Since code-size cost doesn't scale linearly with LMUL,
this change is to separate it from throughput.

Reviewed By: reames

Differential Revision: https://reviews.llvm.org/D142068

[libc++][spaceship][NFC] Marked `operator<=>` for `move_iterator` and `counted_iterator` as implemented

Apprarently implemented as part of:

- `move_iterator` - https://reviews.llvm.org/D117656
- `counted_iterator` - https://reviews.llvm.org/D106205

Reviewed By: #libc, philnik

Spies: libcxx-commits, yaxunl

Differential Revision: https://reviews.llvm.org/D145243

The fcopysign DAG operation, unlike the IR one, allows
different types for the sign and magnitude. We can reduce
the bitwidth of the high operand since only the sign bit matters.

The default combine only introduces mixed fcopysign
operand types from fpext/fptrunc. We effectively do this
already during selection, but doing it earlier in the combiner
should expose new combine opportunities (e.g. the existing tests
now eliminate the load of the low half of the double). Unfortunately
this isn't enough to handle the case I'm interested in just yet.

Following up D127259.

Fixes https://github.com/llvm/llvm-project/issues/61028.

I initially attempted to select the source modifier from xor of
a sign mask. This proved to be more difficult since
foldBinOpIntoSelect does not consider free fneg of integers
and undoes the combine.

Reviewed By: ldionne, #libc, emaste, daltenty

Spies: fsb4000, daltenty, alvinhochun, goncharov, phosek, MaskRay, dalias, q66, thesamesam, emaste, libcxx-commits, arphaman, mstorsjo, #libc_vendors

Differential Revision: https://reviews.llvm.org/D140343

The default argument behavior test tries to make use
of the default reduced.ll, so other tests should not try to
write to the same file.

Maybe fixes issue #61132.

This is a preparation for future fixes that need more tests.

* Put all "Inferred Assignments" testing at the end
* Group together ID-dependent variable/member testing
* Group together unused inferred assignments
* Use a literal instead of `get_local_size` which may have special meaning
* Create a new `struct` for each variable because analysis is done per field,
  not per field per instance.

Depends on D145303

Reviewed By: carlosgalvezp

Differential Revision: https://reviews.llvm.org/D145304

The two macros were removed by
https://mail-index.netbsd.org/source-changes/2022/12/21/msg142441.html

Close #61151

In preparation for using the implementation of basic_string and
basic_string_view from redundant-string-cstr.cpp in other checks, let's
extract it to a header.

Using the existing <string.h> to provide size_t, using that to define
size_type and using it rather than the previous "size" type makes it
easier to provide the size() method later and makes the implementation
closer to the standard.

Reviewed By: carlosgalvezp

Differential Revision: https://reviews.llvm.org/D144216

Similar to the inserts, these are the same fp16 inserts but were missing
patterns.

These should be identical to the fp16 SVE inserts but were missing from the
tests and tablegen patterns. They are always legal (not requiring +bf16),
although there are some other issues around bf16 legalization currently.

In Clang notes are typically printed after a corresponding warning, not before.
For example, all notes issued before the first warning are ignored.

Running `clang-tidy --checks=-*,altera-id-dependent-backward-branch a.cpp` on the following code:
```
long get_local_id(int);
void error() {
  int ThreadID = get_local_id(0);
  for (int i = 0; i < ThreadID; i++) {
  }
  for (int i = 0; i < ThreadID; i++) {
  }
}
```
results in two warnings and a single note in between.

Reviewed By: carlosgalvezp

Differential Revision: https://reviews.llvm.org/D145303

scf.for loop was restricted to only operate on Index type since
splitting out from affine.for. Relax requirement to allow for signless
integer types additionally. This allows specifying explicitly different
bitwidths for different loops as well as specialize from index to iN
while still using scf.for.

Differential Revision: https://reviews.llvm.org/D145288

To have line number information in stack traces (per stack frame) it's
necessary to implement findModulesAndOffsets.

Differential Revision: https://reviews.llvm.org/D142282

This reverts commit 78e42374.

This breaks the memory sanitizer builder:
https://lab.llvm.org/buildbot/#/builders/5/builds/31959

Handle VPWidenCallRecipe in VPRecipeBase::mayHaveSideEffects by
delegating to the underlying call.

Use the existing local variable cleanup infrastructure to implement
destruction.

Differential Revision: https://reviews.llvm.org/D137070

This patch precommits more tests for:

https://github.com/llvm/llvm-project/issues/60802

This just tweaks the fix for D145232 to make the limit more precise, so
that we could actually emit a delay of 3 SALU cycles (the maximum) if we
had any SALU instructions that required it.

Reviewed By: michaelrj

Differential Revision: https://reviews.llvm.org/D145274

This applies to IsClassMethod as well.

The integration tests already have allocators so we will get all of them
from there.

While browsing the latest [clang manual](https://clang.llvm.org/docs/ReleaseNotes.html), I came across this warning and clicked, but it leads to a broken page, apparently because it's missing the trailing `/`. Seems to be the only place where it's missing in the whole of LLVM monorepo.

{F26690053}
{F26690057}

Reviewed By: brenoguim, royjacobson

Differential Revision: https://reviews.llvm.org/D145047

Resolve warning of `TARGET_NAME` macro redefinition.

Reviewed By: jhuber6

Differential Revision: https://reviews.llvm.org/D145307

Some build bots have not been updated to the new minimal CMake version.
Reverting for now and ping the buildbot owners.

This reverts commit 44c6b905.

Differential Revision: https://reviews.llvm.org/D140803

Formatting a const qualified vector<bool> fails to work on libc++. This
is due to our non-conforming type for vector<bool>::const_reference. The
type is a __bit_const_reference<vector> instead of a bool. (This is
fixed in ABI v2.)

This fixes this formatter and enables the test written for it.

Reviewed By: #libc, ldionne

Differential Revision: https://reviews.llvm.org/D144279

  LWG3833 Remove specialization template<size_t N> struct formatter<const charT[N], charT>

Depends on D143845

Reviewed By: #libc, ldionne

Differential Revision: https://reviews.llvm.org/D143850

  LWG3103 Errors in taking subview of span should be ill-formed where possible

Note that the real work was already done before, including tests.

Reviewed By: #libc, ldionne

Differential Revision: https://reviews.llvm.org/D143432

This partly undoes D137724.

This change has been discussed on discourse
https://discourse.llvm.org/t/rfc-upgrading-llvms-minimum-required-cmake-version/66193

Note this does not remove work-arounds for older CMake versions, that
will be done in followup patches.

Reviewed By: mehdi_amini, MaskRay, ChuanqiXu, to268, thieta, tschuett, phosek, #libunwind, #libc_vendors, #libc, #libc_abi, sivachandra, philnik, zibi

Differential Revision: https://reviews.llvm.org/D144509

Signed-off-by: Joyce Brum <joycebrum@google.com>

Reviewed By: tstellar

Differential Revision: https://reviews.llvm.org/D144119

Signed-off-by: Med Ismail Bennani <medismail.bennani@gmail.com>