This patch adds support for lowering intrinsics that have no dynamic
optionality aspects to handle and that requires argument to be lowered
to value, addr, or box.

It uses the current intrinsic lowering framework that can be re-used in
HLFIR to start with. HLFIR operations for charater/transformational
intrinsics will be added as needed from an optimization point of view.
The current approach will still create temporary variables for their
value directly in lowering.

Later patch will still need to add:
- support for dynamically optional arguments
- inquires
- "moving" the in memory computation of character and transformational
intrinsics to hlfir.expr. This is not needed from a semantic point of
view, but will help optimizing and will probably be required inside
hlfir.elemental returning characters so that the returned element
type is an hlfir.expr and match the result type of later hlfir.apply.

Differential Revision: https://reviews.llvm.org/D139613

Differential Revision: https://reviews.llvm.org/D139274

Adding support for ZExt lowering for destination types beyond the existing support for (8|16) x i32

Patch for lowering zext instructions to 'tbl' for (8|16)xi8 -> (8|16)xi32 conversions in https://reviews.llvm.org/D120571 is extended to support zext to 'tbl' lowering for Y x i8 to Y x i8X where X > 2 and X < 8, that is, any number of vector elements & any destination element type whose size is a multiple of 8 and lies between 16 & 64 is allowed for this transformation.

Related microbenchmarks are in https://reviews.llvm.org/D136274 & https://reviews.llvm.org/D138059

Differential Revision: https://reviews.llvm.org/D136722

These are unit tests for code generated for zext lowering of vectors.
There are tests for different types of vectors, as well as cases where multiple 'zext' instructions of same type occur back-to-back.
Testing of Global-ISel path is added.
This will help in comparing changes in zext lowering in https://reviews.llvm.org/D136722.

Reviewed By: fhahn

Differential Revision: https://reviews.llvm.org/D137993

When call an elemental subroutine with a monomorphic or polymorphic
passed object, the iteration shape could not be computed. Use the
passed object to infer the implicit iteration shape so the loop
can be constructed.

Reviewed By: PeteSteinfeld

Differential Revision: https://reviews.llvm.org/D139635

These source files no longer use None, so they do not need to include
None.h.

This is part of an effort to migrate from llvm::Optional to
std::optional:

https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716

This is part of an effort to migrate from llvm::Optional to
std::optional:

https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716

This is part of an effort to migrate from llvm::Optional to
std::optional:

https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716

Reviewed By: david-arm

Differential Revision: https://reviews.llvm.org/D134413

Missing two extra parameters.

AAPointerInfoFloating::updateImpl
  Move the member function definition out of its class before modifying it.

translateAndAddState
  Split the function definition into two along the FromCallee boolean argument.

handleAccess
  This should only be called when Size is not known. In other cases, replaced
  with a direct call to addAccess.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D138645

Reviewed By: rriddle

Differential Revision: https://reviews.llvm.org/D139664

This reverts commit e89e8dcf.

Some configurations still have references to discarded sections. I'm
reverting until these can be addressed.

This commit resolves the FIXME around the behavior of
`Lexer::makeFileCharRange` that `getRangeForEdit` inherits around
source locations in macro expansions.

We add a flag to `getRangeForEdit` that allows a caller to disable the
behavior, and instead uses the spelling location instead, with checks
to ensure that the source locations are not within a macro definition.

Differential Revision: https://reviews.llvm.org/D139676

Differential Revision: https://reviews.llvm.org/D139572

Corresponding to CMake change in llvmorg-16-init-13673-g138d5389

Breakpoint option `-t` checks that `option_arg` is empty by checking `option_arg[0] == '\0'`. This is unnecessary: the next two checks for comparing against "current" and calling `getAsInteger` already gracefully handle an empty StringRef. If the `option_arg` string is empty, this crashes (or triggers an assertion failure with asserts enabled). Also, this sets the thread id to `LLDB_INVALID_THREAD_ID` if the thread id is invalid -- it should just not set the thread id.

Likewise of `-x` which checks `option_arg[0] == '\n'` unnecessarily.

I believe both of these bugs are unreachable via normal LLDB usage, and are only accessible via the fuzzer -- most likely some other CLI parsing is trimming whitespace and rejecting empty inputs. Still, it doesn't hurt to simplify this bit.

In many cases, we can use an alias to avoid a symbolic relocations,
instead of using the public, interposable symbol. When the instrumented
function is in a COMDAT, we can use a hidden alias, and still avoid
references to discarded sections.

This version makes the new runtime test a Linux only test.

Reviewed By: phosek

Differential Revision: https://reviews.llvm.org/D137982

LoongArch doesn't use GlobalISel for now.

The breakpoint list is a list of raw pointers. When breakpoints are removed, the memory is not deleted. Switch to unique pointers. I did some minor cleanup while making this change.

Found by the LLDB command interpreter fuzzer. The input is `br	m G`.

This reverts commit bc061779.

Be more careful with #includes to avoid linker errors on some buildbots.

`m_options.Append(new OptionPermissions())` leaks because the pointer passed in is not owned. Use a class member to ensure lifetime, which is the common pattern used for this API.

Found by the LLDB command interpreter fuzzer. The fuzz input is running `ap $` twice.

Set KCFI type metadata for the sanitizer constructors to prevent
runtime failures when these functions are indirectly called in
instrumented code. This fixes a compatibility issue with KASAN and
-fsanitize=kcfi in the Linux kernel.

Link: https://github.com/ClangBuiltLinux/linux/issues/1742

Reviewed By: nickdesaulniers, MaskRay

Differential Revision: https://reviews.llvm.org/D138945

This reverts commit 3cfaea2d.

The runtime test fails on Mac.

The presence of dladdr on Linux depends on whether or not _GNU_SOURCE is
set.

Reviewed By: phosek

Differential Revision: https://reviews.llvm.org/D139575

This change exposes the floor library function for HLSL,
excluding long, int, and long long doubles.
Floor is supported for all scalar, vector, and matrix types.

Long and long long double support is missing in this patch because those types
don't exist in HLSL. Int is missing because the floor function only works on floating type arguments.

The full documentation of the HLSL floor function is available here:
https://docs.microsoft.com/en-us/windows/win32/direct3dhlsl/dx-graphics-hlsl-floor

Reviewed By: beanz

Differential Revision: https://reviews.llvm.org/D139137

To address https://github.com/llvm/llvm-project/issues/59394.

Reduction on negation of the output tensor is a non-sparsifiable kernel, it creates cyclic dependency.

This patch reject those cases instead of crashing.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D139659

This reverts commit 107ee261 to facilitate
investigating and fixing the root cause.

Differential Revision: https://reviews.llvm.org/D135269