Created using spr 1.3.6-beta.1

Created using spr 1.3.6-beta.1

This is a followup to 0f52f4dd
It breaks dialects that don't use properties yet.

It has been noticed that the arguments are being passed twice to ptxas.
This also has been fixed by filtering out the arguments before appending
them to the new DAL created by CudaToolChain::TranslateArgs.

github:https://github.com/llvm/llvm-project/pull/86807

The cmake macro documentation had once again become unsorted. For names
of the form 'PREFIX_{list-or-category}_SUFFIX' I collate as 'PREFIX_SUFFIX'.

Convert the relocation routines to switch on architecture and have an explicit unreachable default.

Disallows initialization of scalable vectors with an attribute of
arbitrary values, e.g.:
```mlir
  %c = arith.constant dense<[0, 1]> : vector<[2] x i32>
```

Initialization using vector splats remains allowed (i.e. when all the
init values are identical):
```mlir
  %c = arith.constant dense<[1, 1]> : vector<[2] x i32>
```

Currently only machine function passes support `insertPass`, but it
seems to be enough, all targets tune their pipelines when adding machine
function passes.

PR
[https://github.com/llvm/llvm-project/pull/84461](https://github.com/llvm/llvm-project/pull/84461)


disabled `clang/unittests/Interpreter/InterpreterExtensionsTest.cpp` for
AIX by turning on `CLANG_INTERPRETER_PLATFORM_CANNOT_CREATE_LLJIT`.
This PR turns `CLANG_INTERPRETER_PLATFORM_CANNOT_CREATE_LLJIT` on for
zOS as well, since LLJIT cannot be created on zOS either.

Co-authored-by: Bahareh <bahareh.farhadi@ibm.com>

Summary:
The device linking phase only wants to create the necessary commands to
emit the device binary. There were issues where the user's default
config file was being used and passing incompatible arguments to the
device compilation step. Simply disable this since we do not want any
additional arguments to these clang invocations.

llvm-project/mlir/lib/Target/LLVMIR/DebugImporter.cpp:377:10:
error: unused variable '[_, inserted]' [-Werror,-Wunused-variable]
    auto [_, inserted] = dependentCache.try_emplace(
         ^
1 error generated.

The checker may create failure branches for all stream write operations
only if the new option "pedantic" is set to true.
Result of the write operations is often not checked in typical code. If
failure branches are created the checker will warn for unchecked write
operations and generate a lot of "false positives" (these are valid
warnings but the programmer does not care about this problem).

This fixes a regression where building against an installation without
the static libraries would fail. This just reinstates the old behaviour
for out-of-tree builds, assuming that in-tree builds (which still aren't
officially supported) will have the static libraries available. We can
refine this as we move towards supporting in-tree builds.

This adds support for complex type to the OpenMP reductions. 

Note that some more work would be needed to give decent error messages when complex 
is used in ways that need client supplied functions (e.g. MAX or MIN). It does fail these with
a not so user friendly message at present.

Remove them from experimental.

See also:
https://github.com/riscv/riscv-isa-manual/blob/main/src/zimop.adoc

Reviewers: kito-cheng

Reviewed By: kito-cheng

Pull Request: https://github.com/llvm/llvm-project/pull/87966

Flang supports source allocation to allocatable or pointers with a non
deferred length that do not match the source length. This documented at:
https://github.com/llvm/llvm-project/blob/9708d0900311503aa4685d6810d8caf0412e15d7/flang/docs/Extensions.md?plain=1#L312

The current lowering code was bugged when such explicit length allocate
object appeared after a deferred length object in the source allocation
list:

Since "lenParams" had been computed when generating allocation of the
deferred length object, the call to genSetDeferredLengthParameters was
not a no-op on when lowering the explicit length allocation, and the
explicit length was overridden with the source length.

The output of the program added in test was:

```
ZZheZZ
ZZhelloZZ
ZZhelloZZ
```

Instead of:

```
ZZheZZ
ZZhelloZZ
ZZhello  ZZ
```

Skip genSetDeferredLengthParameters when the allocate object has non
deferred length.

It is highly ambiguous to what type BOZ should be resolved in DIM, MOD,
MODULO, and SIGN intrinsic arguments. Some other compilers accept them,
but none agree. See table below.

List them explicitly as non supported extensions (semantics already
reject them, this is an NFC).

Table listing the resolved types of the intrinsic results when there is
a BOZ argument:

|                     | gfortran | nvfortran | ifort | nagfor | xlf    |
| ------------------- | -------- | --------- | ----- | ------ | ------ |
|  DIM(INT4, BOZ)     |  INT16   |  INT4     |  INT8 | crash  |  INT4  |
|  DIM(BOZ, REAL4)    | error    |  INT8     | error | error  |  REAL4 |
|  DIM(REAL4, BOZ)    | error    |  REAL4    | error | error  |  REAL4 |
|  DIM(BOZ, INT4)     |  INT16   |  INT8     |  INT8 | INT8   |  INT4  |
|  DIM(BOZ, BOZ)      |  INT16   |  INT8     |  INT8 | INT8   |  REAL4 |
|  MOD(INT4, BOZ)     |  INT16   |  INT4     |  INT8 | crash  |  INT4  |
|  MOD(BOZ, REAL4)    | error    |  INT8     | error | error  |  REAL4 |
|  MOD(REAL4, BOZ)    | error    |  REAL4    | error | error  |  REAL4 |
|  MOD(BOZ, INT4)     |  INT16   |  INT8     |  INT8 | INT8   |  INT4  |
|  MOD(BOZ, BOZ)      |  INT16   |  INT8     |  INT8 | INT8   |  INT4  |
|  MODULO(INT4, BOZ)  |  INT16   |  INT4     |  INT8 | crash  |  INT4  |
|  MODULO(BOZ, REAL4) | error    |  INT8     | error | error  |  REAL4 |
|  MODULO(REAL4, BOZ) | error    |  REAL4    | error | error  |  REAL4 |
|  MODULO(BOZ, INT4)  |  INT16   |  INT8     |  INT8 | INT8   |  INT4  |
|  MODULO(BOZ, BOZ)   |  INT16   |  INT8     |  INT8 | INT8   |  INT8  |
|  SIGN(INT4, BOZ)    | error    |  INT4     |  INT8 | INT4   |  INT4  |
|  SIGN(BOZ, REAL4)   | error    |  INT8     | error | error  |  REAL4 |
|  SIGN(REAL4, BOZ)   | error    |  REAL4    | error | error  |  REAL4 |
|  SIGN(BOZ, INT4)    | error    |  INT8     |  INT8 | INT8   |  INT4  |
|  SIGN(BOZ, BOZ)     |  INT16   |  INT8     |  INT8 | INT8   |  REAL4 |

The all one masks was not properly created for i128 types because
builder.createIntegerConstant ended-up truncating -1 to something
positive.

Add a builder.createAllOnesInteger/createMinusOneInteger helpers and use
them where createIntegerConstant(..., -1) was used.
Add an assert in createIntegerConstant to catch negative numbers for
i128 type.

Followup to this discussion:
https://github.com/llvm/llvm-project/pull/80251#discussion_r1535599920.

The previous debug importer was correct but inefficient. For cases with
mutual recursion that contain more than one back-edge, each back-edge
would result in a new translated instance. This is because the previous
implementation never caches any translated result with unbounded
self-references. This means all translation inside a recursive context
is performed from scratch, which will incur repeated run-time cost as
well as repeated attribute sub-trees in the translated IR (differing
only in their `recId`s).

This PR refactors the importer to handle caching inside a recursive
context.
- In the presence of unbound self-refs, the translation result is cached
in a separate cache that keeps track of the set of dependent unbound
self-refs.
- A dependent cache entry is valid only when all the unbound self-refs
are in scope. Whenever a cached entry goes out of scope, it will be
removed the next time it is looked up.

SVE has some non-temporal masked loads and stores. The metadata coming
from the nodes is not copied to the MMO at the moment though, meaning it
will generate a normal instruction. This patch ensures that the right
flags are set if the instruction has non-temporal metadata.

This attempts to standardize and extend some of the insert vector
element lowering. Most notably:
- More types are handled by splitting illegal vectors.
- The index type for G_INSERT_VECTOR_ELT is canonicalized to
  TLI.getVectorIdxTy(), similar to extact_vector_element.
- Some of the existing patterns now have the index type specified to
  make sure they can apply to GISel too.
- The C++ selection code has been removed, relying on tablegen patterns.
- G_INSERT_VECTOR_ELT with small GPR input elements are pre-selected to
  use a i32 type, allowing the existing patterns to apply.
- Variable index inserts are lowered in post-legalizer lowering,
  expanding into a stack store and reload.

This reverts commit c91254db.

It was throwing error:
 g++: error: unrecognized command line option ‘-resource-dir=

The conversion expansions did not properly handle bfloat types.

I'm not certain that these expansions are completely correct;
I don't have any experience with AMDGPU or the ability to run
anything to test it.

Note that it doesn't seem like AMDGPU with GlobalISel can
handle fptrunc of float to bfloat, which is needed for itofp.
I've omitted the GISEL run for the bfloat case.

This fixes #85379.

This reverts commit 006aaf32 while I
investigate some bot failures (See e.g.
https://lab.llvm.org/buildbot/#/builders/109/builds/86659).

This TSFlags was introduced by https://reviews.llvm.org/D108767.

A base class of all RISCV RegisterClass is added and we store
IsVRegClass/VLMul/NF into TSFlags and add helpers to get them.

This can reduce some lines and I think there will be more usages.

Reviewers: preames, topperc

Reviewed By: topperc

Pull Request: https://github.com/llvm/llvm-project/pull/84894

This reverts commit 739fa1c8.

This introduces a layering violation by using IR in Support headers.

This allows us to remove a lot of boilerplate .first and .second references
and improve readability.

Coding my way home: 1.58814S, 91.93889W