Created using spr 1.3.4

Fixes build on Windows in C++26 mode.

Reverted in:
https://github.com/llvm/llvm-project/commit/40c07b559aa6ab4bac074c943967d3207bc07ae0
Original PR: https://github.com/llvm/llvm-project/pull/76632



---------

Co-authored-by: Zingam <zingam@outlook.com>

This will help re-use this code with the upcoming GlobalISel implementation of
this optimization.

Tiny spelling mistake fixup. please review the two lines of code below
to see the correctness of this PR.

Summary:
This accidentally included a byte past the magic, which was out of order
on big endian architectures.

The fix used macros that confuses clang-format. This is formatted with
clang-format and then excluded from formatting.

As suggested as follow-up in
https://github.com/llvm/llvm-project/pull/72164, manage inbounds via
VPRecipeWithIRFlags.

Note that in some cases we can now preserve inbounds in a few more
cases.

As %B.gep.0 executes unconditionally in the latch, inbounds could
be preserved in the vector version.

https://alive2.llvm.org/ce/z/XWbMuD

As suggested by @philnik777 this is a better fix than
02a33b72

Fixes: https://github.com/llvm/llvm-project/issues/77123

Also improve the implementation of `findCommonDominator` (skip duplicate
blocks) and extract it from `BufferPlacementTransformationBase` (so that
`BufferPlacementTransformationBase` can be retired eventually).

`BufferPlacementTransformationBase::isLoop` checks if there a loop in
the region branching graph of an operation. This algorithm is similar to
`isRegionReachable` in the `RegionBranchOpInterface`. To avoid duplicate
code, `isRegionReachable` is generalized, so that it can be used to
detect region loops. A helper function
`RegionBranchOpInterface::hasLoop` is added.

This change also turns a recursive implementation into an iterative one,
which is the preferred implementation strategy in LLVM.

Also move the `isLoop` to `BufferOptimizations.cpp`, so that we can
gradually retire `BufferPlacementTransformationBase`. (This is so that
proper error handling can be added to `BufferViewFlowAnalysis`.)

This patch relaxes the one-use constraints for `icmp pred (binop X, Z),
(binop Y, Z)`. It will enable more optimizations with pointer
arithmetic.
One example in `boost::match_results::set_size`:

```
declare void @use(i64)
define i1 @src(ptr %a1, ptr %a2, ptr %add.ptr.i66, i64 %sub.ptr.rhs.cast.i) {
  %sub.ptr.lhs.cast.i = ptrtoint ptr %a1 to i64
  %sub.ptr.rhs.cast.i = ptrtoint ptr %a2 to i64
  %sub.ptr.sub.i = sub i64 %sub.ptr.lhs.cast.i, %sub.ptr.rhs.cast.i
  %sub.ptr.div.i = sdiv exact i64 %sub.ptr.sub.i, 24
  call void @use(i64 %sub.ptr.div.i)
  %sub.ptr.lhs.cast.i.i = ptrtoint ptr %add.ptr.i66 to i64
  %sub.ptr.sub.i.i = sub i64 %sub.ptr.lhs.cast.i.i, %sub.ptr.rhs.cast.i
  %sub.ptr.div.i.i = sdiv exact i64 %sub.ptr.sub.i.i, 24
  %cmp.i.not.i.i = icmp eq i64 %sub.ptr.div.i.i, %sub.ptr.div.i
  ret i1 %cmp.i.not.i.i
}
define i1 @tgt(ptr %a1, ptr %a2, ptr %add.ptr.i66, i64 %sub.ptr.rhs.cast.i) {
  %sub.ptr.lhs.cast.i = ptrtoint ptr %a1 to i64
  %sub.ptr.rhs.cast.i = ptrtoint ptr %a2 to i64
  %sub.ptr.sub.i = sub i64 %sub.ptr.lhs.cast.i, %sub.ptr.rhs.cast.i
  %sub.ptr.div.i = sdiv exact i64 %sub.ptr.sub.i, 24
  call void @use(i64 %sub.ptr.div.i)
  %cmp.i.not.i.i = icmp eq i64 %sub.ptr.sub.i.i, %sub.ptr.sub.i
  ret i1 %cmp.i.not.i.i
}
```

This reverts commit 0903d992.

This is causing all non-Visual Studio builds fail.

Visual Studio needs the class template stuff.

C:\llvm\include\llvm/Analysis/MemoryBuiltins.h(217): error C2990: 'llvm::SizeOffsetType': non-class template has already been declared as a class template
C:\llvm\include\llvm/Analysis/MemoryBuiltins.h(193): note: see declaration of 'llvm::SizeOffsetType'

The ParamPoint datatype has each column representing an affine function.
The code for generating functions is modified to reflect this.

We add some basic type aliases and function definitions relating to
cones for Barvinok's algorithm.
These include functions to get the dual of a cone and find its index.

This should have been checking that the transform was valid, but used
incorrect conditions letting through invalid combinations of lo/hi
extracts.

Hopefully fixes #76769

StackSafetyAnalysis determines whether stack-allocated variables are
safe from memory access bugs and allows removing certain unneeded
instrumentations.
(hwasan enables StackSafetyAnalysis in https://reviews.llvm.org/D108381)

Test updates:

* asan-stack-safety.ll: test the -asan-use-stack-safety=1 default
* lifetime-uar-uas.ll: switch to an indexed store to prevent
  StackSafetyAnalysis from optimizing out instrumentation for %c
* alloca_vla_interact.cpp: add a load to prevent StackSafetyAnalysis
  from optimizing out `__asan_alloca_poison` for the VLA `array`
* scariness_score_test.cpp: add -asan-use-stack-safety=0 to make a load
  of a `__asan_poison_memory_region`-poisoned local variable fail as
  intended.
* other .ll tests: add -asan-use-stack-safety=0

Pull Request: https://github.com/llvm/llvm-project/pull/77210

see #73359

Declarative assemblyFormat ODS is more concise and requires less
boilerplate than filling out CPP interfaces.

Changes:
* updates the Ops defined in `SPIRVAtomicOps.td` to use assemblyFormat.
* Removes print/parse from`AtomcOps.cpp` which is now generated by
assemblyFormat
* Adds `Trait` to verify that a pointer operand `foo`'s pointee type
matches operand `bar`'s type
* * Updates error message expected in tests from new Trait
* Updates tests to updated format (largely using <operand> in place of
"operand")

This reverts commit b5de136e.

Based on post commit feedback, I need to some other work before
this makes sense.

Closes #76106

---------

Co-authored-by: Lei Zhang <antiagainst@gmail.com>

-Rename sub_32_hi to sub_gpr_odd
-Add dedicated sub_gpr_even.
-Rename sub_32 and sub_16 to sub_fpr32 and sub_fpr16.
-Remove start offset from sub_gpr_odd. AArch64 doesn't use non-zero offset for GPR
tuples so I don't think we need to.

This is preparation for a RV64 GPRPair for Zacas.

This reverts commit 16875555.

BTW, we relax the condition for EVEX compression from
ST.hasAVX512() to ST.hasEGPR() || ST.hasAVX512(). It does not have any
effect now b/c no APX instruction is in the EVEX compression table so
far.

This patch is to extract NFC in #77065 into a separate commit.

This fixes the bug introduced by

https://github.com/llvm/llvm-project/commit/6db007a0654ed7a6ed5c3aa3b61a937c19a6bc6b.

We construct placeholder template arguments for template-template
parameters to avoid mismatching argument substitution since they have
different depths with their corresponding template arguments. In this
case,

```cpp
template <template <Concept C> class T> void foo(T<int>);
```

T lies at the depth 0, and C lies at 1. The corresponding argument, of
which there is exactly one, int, is at depth 0. If we consider the
argument as the outermost one, then we would end up substituting 'int'
into the wrong parameter T.

We used to perform such placeholder construction during the context
walk-up. In the previous patch, we slipped through that inadvertently
because we would walk up to the parent, which is precisely a FileContext
for template-template parameters, after adding innermost arguments.

Besides, this patch moves the sanity check up to the context switch.
That way, we avoid dereferencing null pointers if ND is unspecified.

Closes https://github.com/llvm/llvm-project/issues/57410.
Closes https://github.com/llvm/llvm-project/issues/76604. (The case is
slightly different than that in #57410. We should *not* assume the
surrounding context to be a file-scope one.)

Those "_REV" instructions should not appear before encoding
optimization, while macro fusion and flag-copy lowering are before
encoding optimization.

1. Simplify getValueFromBitsInit about cast and return type
2. Remove out-of-date comments and allow memory ops in function
   object `IsMatch` so that we can reuse it for EVEX2Legacy compression.

This patch is to extract NFC in #77065 into a separate commit.

With #70253 landed, selects for reduction results are explicitly used by
ComputeReductionResult and Selects can be marked as not having
side-effects again.

This reverts the revert commit 17303290.

Parameter variable which is forwarded in lambda capture list or in body
by reference is reasonable and current version of this check produces
false positive on these cases. This patch try to fix the
[issue](https://github.com/llvm/llvm-project/issues/68105

)

Co-authored-by: huqizhi <836744285@qq.com>

We shift the GeneratingFunction.h header file to the include/ directory
and wrap it in a `detail` namespace.

This patch is to address my review comments in #77065 to simplify the
implemention of EVEX2Legacy compression.