Reverts llvm/llvm-project#130013 due to failures with expensive checks
on.

Updates a test that I forgot to update in #136699. Failing bot:
  * https://lab.llvm.org/buildbot/#/builders/143/builds/7166

This PR is modified cherry-pick of
intel/llvm@cba338e5fb1c
This PR sets OpenCL language version to be the same, which is 3.0,
for every target and device, in order to unify the build process.
Target should define supported extensions and features via
setSupportedOpenCLOpts API.

llvm-diff shows one change to amdgcn--amdhsa.bc:
* ctz symbols are added since they are now enabled for amdgcn.

…'t known scale factors (#136680)

…#130933)

Add lowering in tablegen for PARTIAL_REDUCE_U/SMLA ISD nodes. Only
happens when the combine has been performed on the ISD node. Also adds
in check to only do the DAG combine when the node can then eventually be
lowered, so changes neon tests too.

---------

Co-authored-by: James Chesterman <james.chesterman@arm.com>

…36313)

Make the FAIL clause contain OmpMemoryOrderType enumeration instead of
OmpClause. This simplifies the semantic checks of the FAIL clause.

Implement all mf8 FMOP4A instructions in clang and llvm following the
acle in https://github.com/ARM-software/acle/pull/381/files.

It also updates previous mop4 instructions from IntrNoMem to
IntrInaccessibleMemOnly

Summary:
We treated the missing kernel environment as a unique mode, but it was
kind of this random bool that was doing the same thing and it explicitly
expects the kernel environment to be zero. It broke after the previous
change since it used to default to SPMD and didn't handle zero in any of
the other cases despite being used. This fixes that and queries for it
without needing to consume an error.

Summary:
This was accidentally kept in the old location when we moved to the
new `lib/<triple>/` location for the DeviceRTL. Move this to reduce the
delta with llvm/llvm-project#136729.

Avoid performing AA queries on non-pointers.

In order for the instruction result to alias with the pointer it
needs to be a pointer.

No longer needed with opaque pointers.

Not needed with opaque pointers.

Fixes #130469 

Now uses REV32/REV64 instructions to complete operation.

New Output:
```
G1:
        rev64   v0.4s, v0.4s
        ret
G2:
        rev32   v0.8h, v0.8h
        ret
G3:
        rev16   v0.16b, v0.16b
        ret
G4:
        rev32   v0.4h, v0.4h
        ret
G5:
        rev16   v0.8b, v0.8b
        ret
```

Old Output:

```
G1:
        shl     v1.2d, v0.2d, #32
        usra    v1.2d, v0.2d, #32
        mov     v0.16b, v1.16b
        ret

G2:
        shl     v1.4s, v0.4s, #16
        usra    v1.4s, v0.4s, #16
        mov     v0.16b, v1.16b
        ret

G3:
        rev16   v0.16b, v0.16b
        ret

G4:
        shl     v1.2s, v0.2s, #16
        usra    v1.2s, v0.2s, #16
        fmov    d0, d1
        ret

G5:
        rev16   v0.8b, v0.8b
        ret
```

These are all pointer bitcasts, which are no longer necessary
with opaque pointers.

`acc.set`, `acc.init`, and `acc.shutdown` take a `device_num` operand.
However, this was named inconsistently. Give it the same consistent name
for all aforementioned operations.

---------

Co-authored-by: erichkeane <ekeane@nvidia.com>

With opaque pointers, these are all the same type. Consolidate to
just PtrTy.

Not needed with opaque pointers.

…galization (#134904)

Precommit tests for vector load/store legalization changes. This exposes
a miscompile in LegalizerHelper::reduceLoadStoreWidth for non-byte-sized
vector elements, which will be fixed in a follow-up patch.
The other tests are potential miscompilations due to unclear semantics
of vector load/stores, which will be addressed in a follow-up
discussion.

…er parameters

I noticed that there are test-cases that are commented out. But the
manglings for them seem to be impossible to generate from valid C++. I
added two test-cases generated from following C++ program:
```
struct X {
    int func() const && { return 5; }
    const int &&func2() { return 5; }
    const int &&func3(const int &x) volatile { return 5; }
};

void f(int (X::*)() const &&, int const && (X::*)(),
       int const && (X::*)(const int &) volatile) {}

int main() {
    f(&X::func, &X::func2, &X::func3);
    return 0;
}
```

Patch corrects output_zp in case of usigned output

let constructor is legacy (do not use in tree!) since the tableGen
backend emits most of the glue logic to build a pass.

Note: The following constructor has been retired:

```cpp
std::unique_ptr<Pass> createExpandReallocPass(bool emitDeallocs = true);
```
    
To update your codebase, replace it with the new options-based API:
    
```cpp
memref::ExpandReallocPassOptions expandAllocPassOptions{
          /*emitDeallocs=*/false};
pm.addPass(memref::createExpandReallocPass(expandAllocPassOptions));
```

Commit 8292e05 which switched the OpenCL C version to 3.0 exposed this
issue, which wasn't caught in pre-commit CI.

libclc is now clean of code that triggers the
bitwise-conditional-parentheses warning, so we can finally remove the
workaround.

… channel

Currently the data-formatters log to either DataFormatters or Types. The
former is probably more sensible, so log there consistently from all
formatters.

…(#137211)

This should be fine as long as we're not reading from it.

Note that this regresses
CXX/special/class.init/class.inhctor.init/p1.cpp, which used to work
fine with the bytecode interpreter.

That's because this code now fails:

```c++
  struct Param;
  struct A {
    constexpr A(Param);
    int a;
  };

  struct B : A { B(); using A::A; int b = 2; };
  struct Wrap1 : B { constexpr Wrap1(); };
  struct Wrap2 : Wrap1 {};
  extern const Wrap2 b;

  struct Param {
    constexpr Param(int c) : n(4 * b.a + b.b + c) {}
    int n;
  };
```
and reports that the Param() constructor is never a valid constant
expression. But that's true and the current interpeter should report
that as well. It also fails when calling at compile time.

…ags (#137131)

RFC link:
https://discourse.llvm.org/t/rfc-clarify-the-behavior-of-select-with-fp-poison-generating-flags/85974

Actually, it does not conflict with the definition of FMF if we
interpret a select as `applyFMF(select cond, applyFMF(TrueArm),
applyFMF(FalseArm))`.

Support for Ziccamoc is added in this pr.
Specification link:
https://drive.google.com/file/d/12QKRm92cLcEk8-5J9NI91m0fAQOxqNAq/view

---------

Co-authored-by: Tie <Tie@llvm.com>

…9c2190eb5dacfb7991f616c83a9e87740c1c6d49

Following "[mlir] Remove deprecated cast member functions (#135556)
(0078cf7)",
 `Type::cast` and `Attribute::cast` are no longer available.

Update ProcessDerivedPDLValue to use `mlir::cast<T>(baseValue)` when
adapting PDL values for native rewrite functions with derived argument
types. Add a regression test for a rewrite function taking `ShapedType`
and `IntegerAttr`.

Signed-off-by: Jonas Rickert <jonas.rickert@amd.com>

Following "[mlir] Remove deprecated cast member functions (#135556)
(0078cf7)",
 `Type::cast` and `Attribute::cast` are no longer available.

Update ProcessDerivedPDLValue to use `mlir::cast<T>(baseValue)`.
Add a regression test for a rewrite function taking `ShapedType`
and `IntegerAttr`.

Signed-off-by: Jonas Rickert <jonas.rickert@amd.com>