HCFG builder doesn't correctly handle cases when non-outermost loop is
requested to be vectorized

[Original] Differential Revision: https://reviews.llvm.org/D150700

No need to emit bitcat op for integer operands if it is detected that
after minbitwidth analysis the type is the same.

These were placed in LegalizeForLLVMExport.cpp, which is the wrong stage
for these, as these lower to high-level ArmSME ops, not intrinsics.

When storing a subvector, too many element were written when the
size of the alloca is smaller than the size of the vector store.
This patch checks for the minimum of the alloca vector and the
store vector to determine the number of elements to store.

See https://github.com/llvm/llvm-project/pull/68879

Parameters marked as uniform take a scalar value, assuming the value is
invariant in the scalar loop. In order to support this, we need to stop
asking for a vector function variant with a default shape assuming that all
arguments will become vector arguments, and instead consider all available
variants and their parameter types.

Use the function value type instead of the element type of the
function pointer type.

Fixes https://github.com/llvm/llvm-project/issues/72798.

This moves the SROA implementation from SROAPass into a separate SROA
class that is defined in the cpp file, and reduces the SROAPass class to
a thin NewPM wrapper. This allows to remove all implementation details
from the SROA header, and the SROALegacyPass can wrap the SROA class
instead of the NewPM SROAPass.

The trigger for this change is a GCC warning about visibility of
implementation details in the SROA header after D138238. Credits to
Nikita Popov for suggesting this reorganization.

Add live-through register set printing, assuming live-through register
is in live-in and live-out sets, has no redefinitions but may have uses
in the block.

It's a user-facing tool, so including it makes sense.

Fixes https://github.com/llvm/llvm-project/issues/72754

The previous code was just an open-coded version of isArrayRoot().

Reduces diff in #72227

Adds missing check prefixes

We have STURHi store patterns but would fail to select from unscaled
offsets. This adds the missing pattern.

Add an additional test case where we currently incorrectly identify a
dependence as Foward instead of ForwardButPreventsForwarding.

Also cleans up the names in the tests a bit to improve readability.

s/makred/marked

[DAG] narrowExtractedVectorBinOp - ensure we limit late node creation to LegalOperations only (#72130)

Avoids infinite issues in some upcoming patches to help D152928 - x86 sees a number of regressions that are addressed by extending SimplifyDemandedVectorEltsForTargetNode to cover more binop opcodes

When folding urem instructions we can end up not recognizing that
the output will always be 0 due to Value*s being different, despite
generating the same data (in this case, 2 different calls to vscale).

This patch recognizes the (x << N) & (add (x << M), -1) pattern that
instcombine replaces urem with after the two vscale calls have been
reduced to one via CSE, then replaces with 0 when x is a power of 2
and N >= M.

* Declare arguments/results with `let` statements.
* Rename `transp` to `permutation`.
* Change type of `transp` from `I64ArrayAttr` to `DenseI64ArrayAttr`
(provides direct access to `ArrayRef<int64_t>` instead of `ArrayAttr`).

Fixes https://github.com/llvm/llvm-project/issues/65884

Unless specified as a "raw" string, Python will try to interpret
backslashes, which means they don't get passed on correctly to the
underlying regex-engine, unless escaped manually (`\\`).

Use raw strings in `llvm/test/lit.cfg.py` to avoid a `SyntaxWarning`:
```
llvm/test/lit.cfg.py:275: SyntaxWarning: invalid escape sequence '\d'
  match = re.search("release (\d+)\.(\d+)", ptxas_out)
```

Other such warning present in 17.0.x were already fixed in
7ed0f5b6.

When there is a COPY instruction in the loop with other uses, we want to
hoist the COPY, which in turn leads to the users being hoisted as well.

Co-authored-by David Green : David.Green@arm.com

The svldr_vnum and svstr_vnum builtins always modify the base register
and tile slice and provide immediate offsets of zero, even when the
offset provided to the builtin is an immediate. This patch optimises the
output of the builtins when the offset is an immediate, to pass it
directly to the instruction and to not need the base register and tile
slice updates.

Currently, when making the ioctl system call, we're not preserving rcx
and r11. The system call will clobber these registers, meaning that the
values of the registers in the snippet will be different than expected.
This patch fixes that be preserving the registers around the system
call, similar to how the other registers involved in the making the
system call get preserved.

Fixes #72741.

If there were some scheduler effects where something like the parent
process got interrupted while the child process continued to run, there
would be nothing blocking it from exiting before the parent process
issued a PTRACE_ATTACH call. This would cause transient failures as this
occurred pretty rarely. This patch removes the possibility of a
transient failure by ensuring that the parent process attaches to the
child process before sending the counter file descriptor through the
socket, ensuring that the child process has at most progressed to being
blocked in the read call for the counter file descriptor.

This pass isn't used anywhere upstream and thus has no test coverage.
Because of these reasons, remove it.

Diff Revision: https://reviews.llvm.org/D123235

This legacy pass isn't used anywhere upstream and thus has no test
coverage, so remove it.

This pass isn't used upstream anywhere and doesn't have have a
create...() function, thus isn't tested anywhere. Because of this,
remove it.

UnifyFunctionExitNodesLegacyPass isn't used anywhere in upstream and
thus isn't tested at all. For these reasons, remove it.

For chain functions, PAL uses a `backend_stack_size` metadata item,
which at the moment has the same meaning as `stack_frame_size_in_bytes`.
We emit both for now in order to simplify coordination with PAL.

The new item must be emitted in the `shader_functions` section, just as
the metadata for other module entry functions. For simplicity, we mark
chain functions as module entry functions and emit the same metadata for
all of them.