Over in the comments for D116821, some use-cases have cropped up where
there's a substantial increase in memory usage. A quick inspection
shows that a) it's a lot of memory and b) there are several things to
be done to reduce it. Reverting (via disabling this feature by default)
to avoid bothering people in the meantime.

Remove the offset parameter if it is zero.

Differential Revision: https://reviews.llvm.org/D117876

Wraps the expansions of TYPE_SWITCH and COMPOSITE_TYPE_SWITCH in
the constexpr interpreter with "do { ... } while (0)" so that these
macros can be used like this:

if (llvm::Optional<PrimType> T = Ctx.classify(FieldTy))
  TYPE_SWITCH(*T, Ok &= ReturnValue<T>(FP.deref<T>(), Value));
else
  Ok &= Composite(FieldTy, FP, Value);

This bug was found while testing D116316. See also review comment:
https://reviews.llvm.org/D64146?id=208520#inline-584131

Also cleaned up the macro definitions by removing the superfluous
do-while statements and removed the unused INT_TPYE_SWITCH macro.

Differential Revision: https://reviews.llvm.org/D117301

Tell the users they are specifying something without vector register.

Reviewed By: ABataev

Differential Revision: https://reviews.llvm.org/D117980

Notation: dst is `t->getThunkTargetSym()->getVA()`

On AArch64, when `src-0x8000000-r_addend <= dst < src-0x8000000`, the condition
`target->inBranchRange(rel.type, src, rel.sym->getVA(rel.addend))` may
incorrectly consider a thunk reusable.
`rel.addend = -getPCBias(rel.type)` resets the addend to 0 for AArch64/PPC
and the zero addend is used by `rel.sym->getVA(rel.addend)` to check
out-of-range relocations.

See the test for a case this computation is wrong:
`error: a.o:(.text_high+0x4): relocation R_AARCH64_JUMP26 out of range: -134217732 is not in [-134217728, 134217727]`
I have seen a real world case with r_addend=19960.

Reviewed By: peter.smith

Differential Revision: https://reviews.llvm.org/D117734

As suggested on PR53379, for all-of icmp-eq patterns, we can use ptest(sub(x,y)) on SSE41+ targets

This is a generalization of the existing allof(cmpeq(x,0)) -> ptest(x) pattern

We can probably extend this further, in particularly to handle 256-bit cases on pre-AVX2 targets, but this part of the generalization is pretty trivial

Fixes Issue #53379

Given that step_vector is practically a constant, doing this early
helps with DAGCombine folds that happen before type legalization.

There is currently no way to test this happens earlier, although existing
tests for step_vector folds continue protect the folds happening at all.

Reviewed By: david-arm

Differential Revision: https://reviews.llvm.org/D117863

Since their nested reference types are defined in terms of `iter_reference_t<T>`, which examines `decltype(*declval<T>())`.

Differential Revision: https://reviews.llvm.org/D117371

Add a transpose option to hoist padding to transpose the padded tensor before storing it into the packed tensor. The early transpose improves the memory access patterns of the actual compute kernel. The patch introduces a transpose right after the hoisted pad tensor and a second transpose inside the compute loop. The second transpose can either be fused into the compute operation or will canonicalize away when lowering to vector instructions.

Reviewed By: nicolasvasilache

Differential Revision: https://reviews.llvm.org/D117893

If the bitreverse gets expanded, it will introduce a new bswap. By
putting a bswap before the bitreverse, we can ensure it gets cancelled
out when this happens.

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D118012

No longer go through an external model. Also put BufferizableOpInterface into the same build target as the BufferizationDialect. This allows for some code reuse between BufferizationOps canonicalizers and BufferizableOpInterface implementations.

Differential Revision: https://reviews.llvm.org/D117987

This can show up during when bitreverse is expanded to bswap and
swap of bits within a byte. If the input is already a bswap, we
should cancel them out before we further transform them in a way
that makes it harder to see the redundancy.

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D118007

Add might be faster than shift. We can't do this earlier without
using a Freeze instruction.

This is the intrinsic version of D106689.

Reviewed By: frasercrmck

Differential Revision: https://reviews.llvm.org/D118013

'ln -s' isn't Windows friendly.

Both insertion points are valid. This is to make BufferizableOpInteface-based bufferization compatible with existing partial bufferization test cases. (So less changes are necessary to unit tests.)

Differential Revision: https://reviews.llvm.org/D117986

Patch originally by oktal3000: https://github.com/mikael-s-persson/templight/pull/40

When a template parameter is unnamed, the name of -templight-dump might return
an empty string. This is fine, they are unnamed after all, but it might be more
user friendly to at least describe what entity is unnamed.

Differential Revision: https://reviews.llvm.org/D115521

Add the MemoryAllocation pass into the pipeline. Add
the possibilty to pass the options directly within the tool (tco).

This patch is part of the upstreaming effort from fir-dev branch.

Reviewed By: jeanPerier

Differential Revision: https://reviews.llvm.org/D117886

This is the only op that is not supported via BufferizableOpInterfaceImpl bufferization. Once this op is supported we can switch `tensor-bufferize` over to the new unified bufferization.

Differential Revision: https://reviews.llvm.org/D117985

A number of the filesystem tests create a directory that contains a bad
symlink. On AIX recursively setting permissions on said directory will
return a non-zero value because of the bad symlink, however the
following rm -r still completes successfully. Avoid the assertion on
AIX, and rely on the return value of the remove command to detect
problems.

Differential Revision: https://reviews.llvm.org/D112086

Reviewed By: gysit

Differential Revision: https://reviews.llvm.org/D118027

Without a langauge name it's an error (with some verisons of Sphinx
it seems) or the block is simply missing in the output.

This is in preparation of unifying the existing bufferization with One-Shot bufferization.

A subsequent commit will replace `tensor-bufferize`'s implementation with the BufferizableOpInterface-based implementation and move over missing test cases.

Differential Revision: https://reviews.llvm.org/D117984

These have negative / out of bounds frame index values and would
assert when trying to set the BitVector. Fixed stack objects can't be
colored away so ignore them.

Add test case to show missing folds for (sra (load)) -> (sextload).

Differential Revision: https://reviews.llvm.org/D116929

This reverts commit a97e20a3.

Reapply 3d2d208f, reverted in
a97e20a3

- Fixes https://github.com/llvm/llvm-project/issues/53227 that wrongly
  indents multiline comments
- Fixes wrong detection of single-line opening braces when used along
  with those only opening scopes, causing crashes due to duplicated
  replacements on the same token:
    void foo()
    {
      {
        int x;
      }
    }
- Fixes wrong recognition of first line of definition when the line
  starts with block comment, causing crashes due to duplicated
  replacements on the same token for this leads toward skipping the line
  starting with inline block comment:
    /*
      Some descriptions about function
    */
    /*inline*/ void bar() {
    }
- Fixes wrong recognition of enum when used as a type name rather than
  starting definition block, causing crashes due to duplicated
  replacements on the same token since both actions for enum and for
  definition blocks were taken place:
    void foobar(const enum EnumType e) {
    }
- Change to use function keyword for JavaScript instead of comparing
  strings
- Resolves formatting conflict with options EmptyLineAfterAccessModifier
  and EmptyLineBeforeAccessModifier (prompts with --dry-run (-n) or
  --output-replacement-xml but no observable change)
- Recognize long (len>=5) uppercased name taking a single line as return
  type and fix the problem of adding newline below it, with adding new
  token type FunctionLikeOrFreestandingMacro and marking tokens in
  UnwrappedLineParser:
    void
    afunc(int x) {
      return;
    }
    TYPENAME
    func(int x, int y) {
      // ...
    }
- Remove redundant and repeated initialization
- Do no change to newlines before EOF

Reviewed By: MyDeveloperDay, curdeius, HazardyKnusperkeks
Differential Revision: https://reviews.llvm.org/D117520

PR53379 test coverage

Explicitly name them icmp0_* - I'm intending to add PR53379 test coverage shortly

We were supposed to be testing <8 x i16> reductions

This patch introduces new intrinsics that enable the use of vsetvli in
contexts where only the returned vector length is of interest. The
pre-existing intrinsics are marked with side-effects, which prevents
even trivial optimizations on/across them.

These intrinsics are intended to be used in situations where the vector
length is fed in turn to RVV intrinsics or to vector-predication
intrinsics during loop vectorization, for example. Those codegen paths
ensure that instructions are generated with their own implicit vsetvli,
so the vector length and vtype can be relied upon to be correct.

No corresponding C builtins are planned at this stage, though that is a
possibility for the future if the need arises.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D117910

Together with the previous commit which mainly documents better LoopFlatten's
overall strategy, this addresses a concern added as a FIXME comment in D110587;
the code refactoring (NFC) introduces functions (also for the SCEV usage) to
make this clearer.

Support C++20 constinit variables for AST Matchers.

There's some unnecessary code duplication in the parser.  This
refactors that and deploys boolean variables to avoid the duplication.
These also happen to help adding module demangling (with an updated
mangling scheme).

1a) The grammar requires some lookahead concerning <template-args>. We
may discover an <unscoped-name> is actually <unscoped-template-name>
<template-args>.  (When <unscoped-name> was a substitution, there must
be a following <template-args>.)  Refactor parseName to only have one
code path looking for the 'I' indicating <template-args>.

1b) While there I altered the control flow to hold the result in a
variable, rather than tail call.  Made it easier to debug (and of
course an optimizer will DTRT here anyway).

2a) An <unscoped-name> can have an St or StL prefix.  No need for
completely separate code paths handling the following unqualified-name
though.

2b) Also no need to look for both 'St' and 'StL' separately.  Look for
'St' and then conditionally swallow an 'L'.

3) We get a similar issue as #1a when parsing a typeName.  Here I just
change the control flow slightly to bring the 'break' out to the end
of the 'S' block and embed the early return inside an if.  That's more
in keeping with the code style.

4) Although NFC, there's a new testcase as that's not covered by the
existing demangler tests and is significant in the #1a case above.

Reviewed By: ChuanqiXu

Differential Revision: https://reviews.llvm.org/D117879