In preparation for implementing code generation for more @llvm.ptrauth.* intrinsics, move the expansion of blend(register, small integer) variant of @llvm.ptrauth.blend to the AArch64PointerAuth pass, where most other PAuth-related code generation takes place.

Adds raw printing of PGOAnalysisMap in llvm-readobj.

I'm leaving the fixme's for a later patch that will provide a 'pretty'
printing for BBFreq and BrProb (i.e. relative frequencies and
probabilities) that will apply to both llvm-readobj and llvm-objdump.

Much of unistd involves modifying files. The tests for these functions
need to use libc_make_test_file_path which didn't exist when they were
first implemented. This patch adds most of unistd to the bazel along
with the corresponding tests. Tests that modify directories had to be
disabled since bazel doesn't seem to handle them properly.

This functionality already exists in
cppcoreguidelines-use-default-member-init. It was deprecated from this
check in clang-tidy 17.

This allows us to fully decouple this check from the corresponding
modernize check, which has an unhealthy dependency.

Fixes https://github.com/llvm/llvm-project/issues/62169



---------

Co-authored-by: Carlos Gálvez <carlos.galvez@zenseact.com>

Write modified Linux kernel binary to disk. The output is not supposed
to be functional at the moment, but it will allow for future patches to
test the output binary.

When we adjust section sizes while rewriting a binary, we should be
using section offsets and not addresses to determine if section overlap.
NFC for existing binaries.

The intent of the test is to check that: 1. The type generic macros are
defined. 2. Those macros dispatch to the correct underlying function.

The issue is that when new functionality is added to our stdbit.h without
rolling out the new entrypoint to all targets, this test breaks because our
generated stdbit.h will not contain declarations for the underlying function.
In that case, we should just declare the underlying functions first before
including our generated stdbit.h which just contains declarations. A definition
is a declaration, but redeclarations must match, hence the additions of
noexcept and extern "C".

RegisterBank Selection for scalable vector G_ADD and G_SUB by creating
new mappings for different types of vector register banks.
Then implement Instruction Selection for the same operations by choosing
the correct RISC-V vector register class.

Based on the spec here
https://github.com/riscv/riscv-ssqosid/releases/tag/v1.0-rc1

Ssqosid extension name will be added in a separate patch.

…parse loops.

This PR implements experimental support for the RISC-V Atomic
Load-Acquire and Store-Release Extension (Zalasr). It has been approved
to be pursued as a fast track extension
(https://lists.riscv.org/g/tech-unprivileged/topic/arc_architecture_review/101951698),
but has not yet been approved by ARC or ratified. See
https://github.com/mehnadnerd/riscv-zalasr

 for draft spec.

---------

Co-authored-by: brs <turtwig@utexas.edu>
Co-authored-by: Philip Reames <preames@rivosinc.com>

Harmonizes the include guard as it is rather uncommon within the MLIR
subproject to append an underscore to include guards.

Assign `nullptr` to the pointer instead.

When printing category conflicts in the ObjC category checker, also
print the source file name of the problematic categories. Currently we
only print the object file name. This change is mostly useful only for
thinLTO builds, where the object file name will be of form
999.arm64.lto.o and thus does not reveal any information about the
original source file.

---------

Co-authored-by: Alex Borcan <alexborcan@meta.com>

GCC has supported a generic constraint "s" for a long time (since at
least 1992), which references a symbol or label with an optional
constant offset. "i" is a superset that also supports a constant
integer.

GCC's RISC-V port also supports a machine-specific constraint "S",
which cannot be used with a preemptible symbol. (We don't bother to
check preemptibility.) In PIC code, an external symbol is preemptible by
default, making "S" less useful if you want to create an artificial
reference for linker garbage collection, or define sections to hold
symbol addresses:

```
void fun();
// error: impossible constraint in ‘asm’ for riscv64-linux-gnu-gcc -fpie/-fpic
void foo() { asm(".reloc ., BFD_RELOC_NONE, %0" :: "S"(fun)); }
// good even if -fpie/-fpic
void foo() { asm(".reloc ., BFD_RELOC_NONE, %0" :: "s"(fun)); }
```

This patch adds support for "s". Modify https://reviews.llvm.org/D105254
("S") to handle multi-depth GEPs (https://reviews.llvm.org/D61560).

The header is unused now, so we can remove it.

Move the diagnostic so it fires only when doing an OpenMP capture, not
for non-OpenMP captures. This allows non-OpenMP code to work when using
OpenMP elsewhere, such as the code reported in
https://github.com/llvm/llvm-project/issues/66999.

`TargetOptions` is currently passed by value in AMDGPU targets, which
makes unnecessary copies. This PR fixes this issue.

Currently machine LICM hoist PC_ADD_REL_OFFSET out of loops, causes
register pressure when function calls are deep in loops. This is a main
cause of sgpr spill for programs containing large number of function
calls in loops.

This patch marks PC_ADD_REL_OFFSET as rematerializable, which eliminates
sgpr spills due to function calls in loops.

`statistics dump` command relies on `SymbolFile::GetDebugInfoSize()` to
get total debug info size.
The current implementation is missing debug info for split dwarf
scenarios which requires getting debug info from separate dwo/dwp files.
This patch fixes this issue for split dwarf by parsing debug info from
dwp/dwo.

New yaml tests are added.

---------

Co-authored-by: jeffreytan81 <jeffreytan@fb.com>

Fixes llvm/llvm-project#80062

Run lines appear in a more natural order:
- no veclib (neon, sve)
- neon + veclib
- sve + veclib
- sve + tailfold + veclib

This patch removes on-stack `TemplateArgumentList`'s. They were primary used
to pass an `ArrayRef<TemplateArgument>` to
`Sema::getTemplateInstantiationArgs`, which had a `const
TemplateArgumentList*` parameter for the innermost template argument
list. Changing this parameter to an
`std::optional<ArrayRef<TemplateArgument>>` eliminates the need for
on-stack `TemplateArgumentList`'s, which in turn eliminates the need for
`TemplateArgumentList` to store a pointer to its template argument
storage (which is redundant in almost all cases, as it is an AST
allocated type).

The amount and format of output from `llvm-dwarfdump --verify` makes it
quite difficult to know if a change to a tool that produces or modifies
DWARF is causing new problems, or is fixing existing problems. This diff
adds a categorized summary of issues found by the DWARF verifier, on by
default, at the bottom of the error output.

The change includes a new `--error-display` option with 4 settings:

* `--error-display=quiet`: Only display if errors occurred, but no
details or summary are printed.
* `--error-display=summary`: Only display the aggregated summary of
errors with no error detail.
* `--error-display=details`: Only display the detailed error messages
with no summary (previous behavior)
* `--error-display=full`: Display both the detailed error messages and
the aggregated summary of errors (the default)

I changed a handful of tests that were failing due to new output, adding
the flag to use the old behavior for all but a couple. For those two I
added the new aggregated output to the expected output of the test.

The `OutputCategoryAggregator` is a pretty simple little class that
@clayborg suggested to allow code to only be run to dump detail if it's
enabled, while still collating counts of the category. Knowing that the
lambda passed in is only conditionally executed is pretty important
(handling errors has to be done *outside* the lambda). I'm happy to move
this somewhere else (and change/improve it) to be more broadly useful if
folks would like.

---------

Co-authored-by: Kevin Frei <freik@meta.com>

Procedure pointer lowering used `prepareUserCallActualArgument` because
it was convenient, but this helper was not meant for POINTERs when
originally written and it did not handled passing NULL to an OPTIONAL
procedure pointer correctly.

The resulting argument should be a disassociated pointer, not an absent
pointer (Fortran 15.5.2.12 point 1.).

Move the logic for procedure pointer argument "cooking" in its own
helper to avoid triggering the logic that created an absent argument in
this case.

Admittedly, its a bit ugly to pass the ABIVersion through onSymbolStart
but I'm not sure what a better place for it would be.

Add AEK_PAUTH to ARMV8_3A in TargetParser and let it propagate to
ARMV8R, as it aligns with GCC defaults.

After adding AEK_PAUTH, several tests from TargetParserTest.cpp crashed
when trying to format an error message, thus update a format string in
AssertSameExtensionFlags to account for bitmask being pre-formatted as
std::string.

The CHECK-PAUTH* lines in aarch64-target-features.c are updated to
account for the fact that FEAT_PAUTH support and pac-ret can be enabled
independently and all four combinations are possible.

According to [temp.pre] p5:
> In a template-declaration, explicit specialization, or explicit instantiation the init-declarator-list in the declaration shall contain at most one declarator. 

A member-declaration that is a template-declaration or explicit-specialization contains a declaration, even though it declares a member. This means it _will_ contain an init-declarator-list (not a member-declarator-list), so [temp.pre] p5 applies.

This diagnoses declarations such as:
```
struct A
{
    template<typename T>
    static const int x = 0, f(); // error: a template declaration can only declare a single entity

    template<typename T>
    static const int g(), y = 0; // error: a template declaration can only declare a single entity
};
```
The diagnostic messages are the same as those of the equivalent namespace scope declarations.

Note: since we currently do not diagnose declarations with multiple abbreviated function template declarators at namespace scope e.g., `void f(auto), g(auto);`, so this patch does not add diagnostics for the equivalent member declarations.

This patch also refactors `ParseSingleDeclarationAfterTemplate` (now named `ParseDeclarationAfterTemplate`) to call `ParseDeclGroup` and return the resultant `DeclGroup`.

Fixes #80284.

Calling `getASTRecordLayout` on invalid types may crash and results of
`__datasizeof` on invalid types can be arbitrary, so just use whatever
`sizeof` returns.

[X86] X86FixupVectorConstants.cpp - refactor constant search loop to take array of sorted candidates

Pulled out of #79815 - refactors the internal FixupConstant logic to just accept an array of vzload/broadcast candidates that are pre-sorted in ascending constant pool size

Properly handle fusion of loops with reductions:
* Check there are no first loop results users between loops
* Create new loop op with merged reduction init values
* Update `scf.reduce` op to contain reductions from both loops
* Update loops users with new loop results