This broke some builds where GNU assembler doesn't support 'vg'.

This reverts commit 130b149b due to it
breaking nvidia build bots. Apparently, there are other users of the
conversion options that were changed in the reverted commit.

Add an nneg flag to the zext instruction, which specifies that the
argument is non-negative. Otherwise, the result is a poison value.

The primary use-case for the flag is to preserve information when sext
gets replaced with zext due to range-based canonicalization. The nneg
flag allows us to convert the zext back into an sext later. This is
useful for some optimizations (e.g. a signed icmp can fold with sext but
not zext), as well as some targets (e.g. RISCV prefers sext over zext).

Discourse thread: https://discourse.llvm.org/t/rfc-add-zext-nneg-flag/73914

This patch is based on https://reviews.llvm.org/D156444

 by
@Panagiotis156, with some implementation simplifications and additional
tests.

---------

Co-authored-by: Panagiotis K <karouzakispan@gmail.com>

This commit removes the support for lowering SPIRV to LLVM dialect with
typed pointers. Typed pointers have been deprecated for a while now and
it's planned to soon remove them from the LLVM dialect.

Related PSA:
https://discourse.llvm.org/t/psa-removal-of-typed-pointers-from-the-llvm-dialect/74502

This commit removes the support for lowering Func to LLVM dialect with
typed pointers. Typed pointers have been deprecated for a while now and
it's planned to soon remove them from the LLVM dialect.

Related PSA:
https://discourse.llvm.org/t/psa-removal-of-typed-pointers-from-the-llvm-dialect/74502

The intent of this "wording improvement" is to not change the meaning of
the text, just make it easier to process by removing "FUTURE" markers,
and slightly improving other wording.

LLVMStructTypes could be emitted with some null elements. This caused a
crash later in the LLVMDialect verifier.

We now use `convertTypes` and check that all types were successfully converted before passing them to the `LLVMStructType` constructor.

See #59990

`SubsetInsertionOpInterface` is an interface for ops that insert into a
destination tensor at a subset. It is currently used by the
bufferization framework to support efficient
`tensor.extract_slice/insert_slice` bufferization and to drive "empty
tensor elimination".

This commit moves the interface to `mlir/Interfaces`. This is in
preparation of adding a new "loop-invariant subset hoisting"
transformation to
`mlir/Transforms/Utils/LoopInvariantCodeMotionUtils.cpp`, which will
utilize `SubsetInsertionOpInterface`. (This new transform is roughly
what `Linalg/Transforms/SubsetHoisting.cpp` is doing, but in a generic
and interface-driven way.)

This patch clarifies that in case there are multiple
'-fmodule-file=<module-name>=' optins for the same `<module-name>`,
the last '-fmodule-file=<module-name>=' will override the previous
'-fmodule-file=<module-name>=' option.

Seprate registeration and CLI parsing from `MlirOptMain` to
`mlir::registrationAndParseCLIOptions` and a new `MlirOptMain` so that
external tools to drive `mlir-opt` can call these two new APIs to get CLI
option values before running `MlirOptMain`.

Looking at 855d0985 it looks like a bit was
missing. The padding variable is used further down by the KMP_ALLOCA()
function.

This reverts commit d8f5a18b.
Breaks build, see:
https://github.com/llvm/llvm-project/pull/70543#issuecomment-1784227421

These extra patterns come from the lowering of fptoi, where an extra assertzext
is present between some of the vectors.

OpenBSD does not have /proc filesystem, neither does FreeBSD (by default).

By using if init statement.

This patch enables building the polly docs in CI to catch warnings/build
regressions in the documentation.

This patch enables building the openmp documentation through Github
actions during PRs and at tip of tree to ensure that the documentation
builds and that there aren't any Sphinx warnings in a manner that
enables more rapid developer iteration without having to install
documentation tooling.

To unbreak the release build.

The commit was discussed in phabricator
(https://reviews.llvm.org/D157186).

Record replay currently fails on AMD as it conflicts with the heap
memory allocator introduced in #69806. The workaround is setting
`LIBOMPTARGET_HEAP_SIZE=0` during both record and replay run.

By associating the kernel environment with the generic kernel we can
access middle-end information easily, including the launch bounds ranges
that are acceptable. By constraining the number of threads accordingly,
we now obey the user-provided bounds that were passed via attributes.

Co-authored-by: serge-sans-paille <sguelton@mozilla.com>

We used to pass the min/max threads/teams values through different paths
from the frontend to the middle end. This simplifies the situation by
passing the values once, only when we will create the KernelEnvironment,
which contains the values. At that point we also manifest the metadata,
as appropriate. Some footguns have also been removed, e.g., our target
check is now triple-based, not calling convention-based, as the latter
is dependent on the ordering of operations. The types of the values have
been unified to int32_t.

D146121 needs to set the NSW flag, but given the result is NUW then we know that the result has leading zeros, so we don't need to call ComputeNumSignBits - just reuse the existing KnownBits value instead.

Fixes buildbots broken by e605fba3

If a shl node leaves the upper half bits zero / undemanded, then see if we can profitably perform this with a half-width shl and a free trunc/zext.

Followup to D146121

Reapplied - moved after the ShrinkDemandedOp call; reuse the existing KnownBits result; ensure that we only attempt this if all the upper bits are demanded; 547dc461 should address the remaining regressions that were noticed in the previous commit.

Differential Revision: https://reviews.llvm.org/D155472

[clang-tidy] Fix readability-avoid-const-params-in-decls - point to the correct const location (#69103)

Fixes the warning marker for redundant const from beginning of variable
to actual location of const. 

Fixes #68492

This is to fix a crash since aab8b2eb, which generates a new pattern
```
      t35: v8i32 = xor t11, t14
    t36: v8i32 = vector_shuffle<0,1,0,1,0,1,0,1> t35, undef:v8i32
```

The pattern exposed a bug introduced since f885c080, which breaks
element widen but doesn't handle the broadcast case.

The patch just solved the crash issue. I observed performance regression
cased by above patches in the test, which may need further
investigation.

template arguments

This makes the library more consistent and reduces the size of mangled
names a bit as a bonus.

fix bad `Offset` for aarch64be asan

---------

Co-authored-by: Vitaly Buka <vitalybuka@gmail.com>

This patch switched the default value of the mandatory-inlining-first
flag from true to false. This broke one of the MLGO tests that relied on
the default value of this flag. This patch explicitly sets the value to
fix the test and avoid future breakages.

This commit deprecates LLVM dialect builders that expect typed pointers.
In the process of removing typed pointers from the LLVM dialect, these
builders will eventually be removed.

Fixes #70578.

[RISCV][GISel] Directly emit X0 from getICMPOperandsForBranch instead of using buildConstant+selectConstant.

This is simpler to implement and matches what SelectionDAG emits.

Also allows us to make getICMPOperandsForBranch a static function.