Fixes https://github.com/llvm/llvm-project/issues/67761
Trying `getDimSize()` before checking for 0-ranked-tensors throws assert
errors. This PR ensures that it is checked for.
Or should we throw an error if we have a 0-ranked-tensor in a tosa
operation?

Drop code inserting pointer casts. Check pointer types instead of
address spaces.

A recent change modified the parameter tileSize from Value to
OpFoldResult. Therefore we should call getAsOpFoldResult before passing
on the tileSize.
Adjust a test regarding this new behavior.

Fix two issues:
* If a constant is used in another constant, we need to insert newly
  created instructions to worklist so that constant used in them will
  be converted.
* Set debug info of original instruction to newly created instructions.

Add AfterPlacementNew option to SpaceBeforeParensOptions to have more
control on placement new expressions.

Fixes #41501
Relates to #54703

Differential Revision: https://reviews.llvm.org/D127270

This uses the fast-check allowlist added in the previous commit.
This is behind a config option to allow users/developers to enable checks
we haven't timed yet, and to allow the --check-tidy-time flag to work.

Fixes https://github.com/clangd/clangd/issues/1337

Differential Revision: https://reviews.llvm.org/D138505

This adds a flag to the `TransformDialectInterpreter` that relaxes the
requirement for only a single top-level transform op.
This is useful for supporting transforms that take transform IR as
payload.

This also aligns the function `findTopLevelTransform`
[here](https://github.com/llvm/llvm-project/commit/7b0f4c9db55c355bffddf94d7710f40ee2c1e9db#diff-551f92bb609487ccf981daf9571f0f1b1703ab2330560a388a5f0d133e520be4L59)
with its documentation:
In the presence of multiple top-level transform ops it now correctly
returns the first of them after reporting the error instead of returning
a `nullptr`.

Type extension is currently handled in FIR by inlining the parents
components as the first member of the record type.

This is not correct from a memory layout point of view since the storage
size of the parent type may be bigger than the sum of the size of its
component (due to alignment requirement). To avoid making FIR types
target dependent and fix this issue, make the parent component a single
component with the parent type at the beginning of the record type.

This also simplifies addressing since parent component is now a "normal"
component that can be designated with hlfir.designate.

StructureComponent lowering however is a bit more complex since the
symbols in the structure component may refer to subcomponents of parent
types.

Notes:
1. The fix is only done in HLFIR for now, a similar fix should be done
in ConvertExpr.cpp to fix the path without HLFIR (I will likely still do
it in a new patch since it would be an annoying bug to investigate for
people testing flang without HLFIR).
2. The private component extra mangling is useless after this patch. I
will remove it after 1.
3. The "parent component" TODO in constant CTOR is free to implement for
HLFIR after this patch, but I would rather remove it and test it in a
different patch.

As requested in (#66521)

I confirmed a crash with "return" instead of "continue" in
setVectorizedCallDecision's fmuladd reduction recognition.

Power10 does not support Hardware Transactional Memory instructions.
Remove to keep consistency.

The code in `copyHostAssociateVar` is using `createSomeArrayAssignment`
for arrays which is using the soon legacy expression lowering. Update
the copy to use hlfir.assign instead.

I used the temporary_lhs flag to mimic the current behavior, but maybe
user defined assignment should be called when needed .This flag also
prevents any finalizers to be called on the LHS if the LHS type has
finalizers (which would occur otherwise in normal intrinsic assignment).
Again, I am not sure what the OpenMP spec wants here.

Also, I added special handling for ALLOCATABLE, the current code seems
broken to me since it is basically copying the descriptor which would
lead to memory leak given the TEMP was previously allocated with the
shape of the variable in createHostAssociateVarClone. So copying the
DATA instead seemed like the right thing to do.

In TOSA MLIR dialect, fix the definition of the Clamp op to
accept fp16 & bf16 datatype for the min_fp and max_fp attributes.
Add ClampOp verifier to check attributes types compatibility.
Add related test cases in Tosa/ops.mlir.

Signed-off-by: Fabrizio Indirli <Fabrizio.Indirli@arm.com>

This makes the docs a little nicer to read, as these otherwise show up
as "«unnamed»".

The extra include is needed as naming means getters are generated, and
the getters use the LLVM types.

As described in: https://github.com/ARM-software/acle/pull/257

Patch by: Rosie Sumpter <rosie.sumpter@arm.com>

Reviewed By: dtemirbulatov

Differential Revision: https://reviews.llvm.org/D151709

ISO_Fortran_binding.h was only added to in gcc 10.0. Flang should be
buildable with older versions. Remove the test until a safe way to
check that the compiler can run the test (that it is clang from the
build for instance).

Fix bots failure https://lab.llvm.org/buildbot/#/builders/181/builds/24526
Also in:

https://lab.llvm.org/buildbot/#/builders/160
https://lab.llvm.org/buildbot/#/builders/268
https://lab.llvm.org/buildbot/#/builders/181

perf2bolt launches a few perf script commands and stores the output in
temporary files before processing the output and cleaning them up before
it exits.

The command `perf script --show-mmap-events` outputs PERF_RECORD_MMAP2
and instruction tracing data but when processed it only looks for
PERF_RECORD_MMAP2 and the instruction tracing data is ignored. This is
fine for small amounts of instruction trace data but when I've recorded
Arm ETM or Intel PT AUX I get lots of it

By adding `--no-itrace` is will just show the PERF_RECORD_MMAP2 records
and will save on time running the `perf script`, disk space storing the
output & time parsing the output.

It is the same for `perf script --show-task-events` where BOLT is only
interested in the PERF_RECORD_COMM & PERF_RECORD_FORK records.

### Data

| Perf Record | Perf Data Size  | MMap Size | MMap No Itrace Size |
|---|---|---|---|
| perf record -e cs_etm/@tmc_etr0/u | 137K | 4468K | 0.632K |
| perf record -e intel_pt//u | 890K | 33378K | 0.673K |

On Linux this contains a single register that determines memory tagging
and tagged address ABI settings.

Adding an additional parameter to run_clang_tidy.py to accept a
directory where the clang-tidy fixes are saved to. This directory can
then be used to run `clang-apply-replacements`.

Closes #69450

This adds implementations for `getSingleIterationVar`,
`getSingleLowerBound`, `getSingleUpperBound`, `getSingleStep` of
`LoopLikeOpInterface` to `scf::ParallelOp`. Until now, the
implementations for these methods defaulted to returning `std::nullopt`,
even in the special case where the parallel Op only has one dimension.

Related: https://github.com/llvm/llvm-project/pull/67883

This patch fixes the nesting of TosaValidation pass added in
TosaToLinalg pipeline.

When inferring readonly/writeonly on arguments, if the argument is
passed to a call, we should only check the ArgMem effects implied by the
call -- we don't care whether the call reads/writes non-arg memory
(captured pointers are not relevant here, because they will abort the
analysis entirely).

This also fixes a regression that was introduced when moving to
MemoryEffects: The code was still checking the old WriteOnly attribute
on functions, which no longer exists.

This is primarily only useful when debugging. It's generally assumed
that users
will have their custom flags applied if it's specified in their CMake
cache files.

Addresses
https://github.com/llvm/llvm-project/pull/68393#discussion_r1363399029