Using fp type in the compiler is not the best idea, here it used with
the comparison for equal to 0 and may cause undefined behavior in some
cases.

Reviewers: fhahn

Reviewed By: fhahn

Pull Request: https://github.com/llvm/llvm-project/pull/87241

Use v4 of UTC to improve regex matching of argument names to fix a filecheck matching in a future patch

Need to check that unsigned argument can be safely used in smax/smin
intrinsics by checking if at least single sign bit is cleared, otherwise
its value may be treated as negative instead of positive.

We had some instances when LLVM would not inline fixed-count memcpy and
ended up
attempting to lower it a a libcall, which would not work on AMDGPU as
the
address space doesn't meet the requirement, causing compiler crash.

The patch relaxes the threshold used for -Os/-Oz compilation so we're
always allowed
to inline memory copy functions.

This patch basically does the same thing as
https://reviews.llvm.org/D158226 for
AMDGPU.

Fix #88497.

Revert "[codegen] Emit missing cleanups for stmt-expr and coro suspensions" and related commits (#88884)

The original change caused widespread breakages in msan/ubsan tests and
causes `use-after-free`. Most likely we are adding more cleanups than
necessary.

Summary:
AIX headers define this, so we need to work around it. In the future
this will be removed but for now we should just rename it to avoid these
issues.

This patch moves OpenMP-related entities out of `Sema` to a newly
created `SemaOpenMP` class. This is a part of the effort to split `Sema`
up, and follows the recent example of CUDA, OpenACC, SYCL, HLSL.
Additional context can be found in
https://github.com/llvm/llvm-project/pull/82217,
https://github.com/llvm/llvm-project/pull/84184,
https://github.com/llvm/llvm-project/pull/87634.

On macOS, file paths start with /Users/..., which clang-cl interptrets
as the /U switch followed by a preprocessor macro name to undefine.

Put the filename after `--` to prevent this. For consistency, move %s
to the end of the regular `clang` lines (where this isn't needed) as
well.

This should ensure we explore the same VFs as before 6d66db38.

Fixes https://github.com/llvm/llvm-project/issues/88640.

Some very basic tests for a case where we could fold BLEND(PERMUTE(X),PERMUTE(Y)) -> PERMUTE(BLEND(X,Y))

These assume the permute masks are the same, and "complete" (no undefs/duplicate elements) but we could relax that depending on the blend mask

When in-place new-ing a local variable of an array of trivial type, the
generated code calls 'memset' with the correct size of the array,
earlier it was generating size (squared of the typedef array + size).

The cause: `typedef TYPE TArray[8]; TArray x;` The type of declarator is
Tarray[8] and in `SemaExprCXX.cpp::BuildCXXNew` we check if it's of
typedef and of constant size then we get the original type and it works
fine for non-dependent cases.
But in case of template we do `TreeTransform.h:TransformCXXNEWExpr` and
there we again check the allocated type which is TArray[8] and it stays
that way, so ArraySize=(Tarray[8] type, alloc Tarray[8*type]) so the
squared size allocation.

ArraySize gets calculated earlier in `TreeTransform.h` so that
`if(!ArraySize)` condition was failing.
fix: I changed that condition to `if(ArraySize)`. 


Fixes #41441

The existing lowering for tosa.max_pool2d only supports dynamic
dimensions when the dynamic dimension is the batch dimension. This
change updates the lowering to support arbitrary dynamic dimensions on
the inputs and outputs of the tosa.max_pool2d operation.

This change also fixes a bug in the implementation of implicit
broadcasting in the tosa-to-linalg pass, which was introducing uses of
constant ops that violated dominance requirements.

Co-authored-by: Frederik Harwath <fharwath@amd.com>
Co-authored-by: Mehdi Amini <joker.eph@gmail.com>

Bug fix: Handle RVV return type in calling convention correctly.
Return values are handled in a same way as function arguments.
One thing to mention is that if a type can be broken down into
homogeneous
vector types, e.g. {<vscale x 4 x i32>, {<vscale x 4 x i32>, <vscale x 4
x i32>}},
it is considered as a vector tuple type and need to be handled by tuple
type rule.

__builtin_is_aligned
__builtin_is_align_up
__builtin_is_align_down

At the moment there is no support for vector.shuffle for scalable
vectors - various hooks/helpers related to `vector.shuffle` simply
ignore the scalable flags (e.g. ` ShuffleOp::inferReturnTypes`).

This is unlikely to change any time soon (vector shuffles are known to
be tricky for scalable vectors), hence this patch restricts
`vector.shuffle` to fixed width vectors.

This adds a simple rewrite/legalization to decompose constant splats
larger than a single ArmSME tile into multiple SME virtual tile sized
splats. E.g. a constant splat to `vector<[8]x[8]xi32>` would decompose
into four `vector<[4]x[4]xi32>` splats.

This is necessary to ensure that functions declared in different
translation units whose parameter types only differ in top-level
cv-qualification generate the same USR.

For example:
```
// A.cpp
void f(const int x); // c:@F@f#1I#

// B.cpp
void f(int x);       // c:@F@f#I#
``` 
With this patch, the USR for both functions will be
`c:@F@f#I#`.

[DAG] Fold extract_subvector(insert_subvector(x,y,c1),c2) --> extract_subvector(y,c2-c1) (#87925) (REAPPLIED)

If the extract_subvector is cheap, attempt to extract directly from an inserted subvector

Reapplied with a check to ensure we only attempt this for fixed vectors

Fixes:
/buildbot/worker/arc-folder/llvm-project/clang/lib/AST/Interp/Disasm.cpp:143:25: warning: cast from type 'const clang::interp::Block*' to type 'void*' casts away qualifiers [-Wcast-qual]
/buildbot/worker/arc-folder/llvm-project/clang/lib/AST/Interp/Disasm.cpp:271:23: warning: cast from type 'const clang::interp::Block*' to type 'void*' casts away qualifiers [-Wcast-qual]

This reverts commit f4960da6.
Includes a fix for the MLIR test case.

For G_LOAD and G_STORE we want this information during regbankselect.
Today we treat load dest as integer and insert converts.

---------

Co-authored-by: Evgenii Kudriashov <evgenii.kudriashov@intel.com>

Fixes #76609
This patch does:
- relax the phis constraint in `CanRedirectPredsOfEmptyBBToSucc`
- guarantee the BB has multiple different predecessors to redirect, so
that we can handle the case without phis in BB. Without this change and
phi constraint, we may redirect the CommonPred.

The motivation is consistent with JumpThreading. We always want the
branch to jump more direct to the destination, without passing the
middle block. In this way, we can expose more other optimization
opportunities.

An obivous example proposed by @dtcxzyw is like:
```llvm
define i32 @test(...) {
entry:
   br i1 %c, label %do.end, label %if.then

if.then:                                          ; preds = %entry
   %call2 = call i32 @dummy()
   %tobool3.not = icmp eq i32 %call2, 0
   br i1 %tobool3.not, label %do.end, label %return

do.end:                                           ; preds = %entry, %if.then
   br label %return

return:                                           ; preds = %if.then, %do.end
   %retval.0 = phi i32 [ 0, %do.end ], [ %call2, %if.then ]
   ret i32 %retval.0
}
```
`entry` can directly jump to return, without passing `do.end`, and then
the if-else pattern can be simplified further:
```llvm
define i32 @test(...) {
entry:
   br i1 %c, label %return, label %if.then

if.then:                                          ; preds = %entry
   %call2 = call i32 @dummy()
   br label %return

return:                                           ; preds = %if.then
   %retval.0 = phi i32 [ 0, %entry ], [ %call2, %if.then ]
   ret i32 %retval.0
}
```

Currently neither the SPIR nor the SPIRV targets specify the AS for
globals in their datalayout strings. This is problematic because
CodeGen/LLVM will default to AS0 in this case, which produces Globals
that end up in the private address space for e.g. OCL, HIPSPV or SYCL.
This patch addresses it by completing the datalayout string.

In LoongArch psABI v2.30, the R_LARCH_ALIGN requires symbol index to
support the third parameter of alignment directive. Create symbol for
each section is redundant because they have section symbol which can
also be used as symbol index. So use section symbol directly for
R_LARCH_ALIGN.

This reverts commit 61717c1a. Failed a MLIR test

Fixes #62449

Check if the non-null function pointer is even valid before calling
the function.

This patch performs several cleanups with the main purpose of
normalizing the code patterns used to trigger codegen for MLIR OpenMP
operations and making the processing of clauses and constructs
independent. The following changes are made:

- Clean up unused `directive` argument to
`ClauseProcessor::processMap()`.
- Move general helper functions in OpenMP.cpp to the appropriate section
of the file.
- Create `gen<OpName>Clauses()` functions containing the clause
processing code specific for the associated OpenMP construct.
- Update `gen<OpName>Op()` functions to call the corresponding
`gen<OpName>Clauses()` function.
- Sort calls to `ClauseProcessor::process<ClauseName>()` alphabetically,
to avoid inadvertently relying on some arbitrary order. Update some
tests that broke due to the order change.
- Normalize `genOMP()` functions so they all delegate the generation of
MLIR to `gen<OpName>Op()` functions following the same pattern.
- Only process `nowait` clause on `TARGET` constructs if not compiling
for the target device.

A later patch can move the calls to `gen<OpName>Clauses()` out of
`gen<OpName>Op()` functions and passing completed clause structures
instead, in preparation to supporting composite constructs. That will
make it possible to reuse clause processing for a given leaf construct
when appearing alone or in a combined or composite construct, while
controlling where the associated code is produced.

…return only Load since other output is chain.

Added testcase that showed mismatched expected arity when Load and chain
were returned as separate items after
003b58f6

We don't need to consider the offset here anymore since we now
have proper integral pointers.

This patch updates the definition of `omp.taskloop` to enforce the
restrictions of a wrapper operation.

Instead of iterating over potential induction var uses looking for
suitable `arith.addi`, try to trace it back from yield argument.