Add opcodes for different store instructions to the target hook that can
enable more STP pairs. This is split off from the patch that does the
same for some load instructions (#79003).

Patch co-authored by Cameron McInally.

optimizeTan has been renamed to optimizeTrigInversionPairs as a result.

Sadly, this is not mathematically true that all inverse pairs fold to x.
For example, asin(sin(x)) does not fold to x if x is over 2pi.

A simple enough op pass so we can test standard instrumentations in
future.

Removed target-triple in target-independent test case to fix failing test caused by https://github.com/llvm/llvm-project/pull/67725.

Having libc_errno outside of the namespace causes versioning issues when
trying to link the tests against LLVM-libc. Most of this patch is just
moving libc_errno inside the namespace in tests. This isn't necessary in
the function implementations since those are already inside the
namespace.

When unrolling the reduction dimension of something like a matmul for
SME, you can end up with transposed reads of illegal types, like so:

```mlir
%illegalRead = vector.transfer_read %memref[%a, %b]
                : memref<?x?xf32>, vector<[8]x4xf32>
%legalType = vector.transpose %illegalRead, [1, 0]
                : vector<[8]x4xf32> to vector<4x[8]xf32>
```

Here the `vector<[8]x4xf32>` is an illegal type, there's no way to lower
a scalable vector of fixed vectors. However, as the final type
`vector<4x[8]xf32>` is legal, we can instead lift the transpose to
memory (producing a strided memref), and eliminate all the illegal
types. This is shown below.

```mlir
%readSubview = memref.subview %memref[%a, %b] [%c8_vscale, %c4] [%c1, %c1]
                : memref<?x?xf32> to memref<?x?xf32>
%transpose = memref.transpose %readSubview (d0, d1) -> (d1, d0)
                : memref<?x?xf32> to memref<?x?xf32>
%legalType = vector.transfer_read %transpose[%c0, %c0]
                : memref<?x?xf32>, vector<4x[8]xf32>
```

Adding extern "C" to all the entry point functions to make sure that
these functions are not mangled.

Resolves #27008, #39735, #53013, #63619.

Hello, this PR adds the MainIncludeChar option to clang-format, allowing
to select which include syntax must be considered when searching for the
main header: quotes (`#include "foo.hpp"`, the default), brackets
(`#include <foo.hpp>`), or both.

The lack of support for brackets has been reported many times, see the
linked issues, so I am pretty sure there is a need for it :)

A short note about why I did not implement a regex approach as discussed
in #53013: while a regex would have allowed many extra ways to describe
the main header, the bug descriptions listed above suggest a very simple
need: support brackets for the main header. This PR answers this needs
in a quite simple way, with a very simple style option. IMHO the feature
space covered by the regex (again, for which there is no demand :)) can
be implemented latter, in addition to the proposed option.

The PR also includes tests for the option with and without grouped
includes.

Under fast-math flags it's possible to convert `sqrt(exp(X)) `into
`exp(X * 0.5)`. I suppose that this transformation is always profitable.
This is similar to the optimization existing in GCC.

This patch adds a script to automatically query the number of running
jobs and print them to the terminal as this functionality isn't
available through the Github UI (unless you are a Github administrator).

This adds the `emitc.declare_func` operation that allows to emit the
declaration of an `emitc.func` at a specific location.

Following the discussion in
https://discourse.llvm.org/t/symboltable-and-symbol-parent-child-relationship/75446

,
we should enforce that a symbol's immediate parent is a symbol table.

I changed some tests to pass the verification. In most cases, we can
wrap the func with a module, change the func to another op with regions
i.e. scf.if, or change the expected error message.

---------

Co-authored-by: Mehdi Amini <joker.eph@gmail.com>

…DOW_SCALE

As MEM_GRANULARITY represents the size of memory block mapped to a
single shadow entry, and SHADOW_SCALE represents the scale of shadow
mapping, so the single shadow entry size can be computed as
(MEM_GRANULARITY >> SHADOW_SCALE).

This patch replaces the hardcoded SHADOW_ENTRY_SIZE with
(MEM_GRANULARITY >> SHADOW_SCALE).

This patch removes a couple of redundant buffer copies in emitTable for
setting up calls to decodeULEB128. Instead, provide the Table.data
buffer directly to the calls-- where decodeULEB128 does its own buffer
overflow checking.

Factor out 7 explicit loops to emit ULEB128 bytes into emitULEB128. Also
factor out 4 copies of 24-bit numtoskip emission into emitNumToSkip.

The functionality is already covered by existing unit tests and by
virtue of most of the in-tree back-ends exercising the decoder emitter.

new

```C
while (_Generic(x, //
           long: x)(x) > x) {
}
while (_Generic(x, //
           long: x)(x)) {
}
```

old

```C
while (_Generic(x, //
       long: x)(x) > x) {
}
while (_Generic(x, //
    long: x)(x)) {
}
```

In the first case above, the second line previously aligned to the open
parenthesis.  The 4 spaces did not get added by the fallback line near
the end of getNewLineColumn because there was already some indentaton.
Now the spaces get added explicitly.

In the second case above, without the fake parentheses, the second line
did not respect the outer parentheses, because the LastSpace field did
not get set without the fake parentheses.  Now the indentation of the
outer level is used instead.

This matches the optimization pipeline's PassRegistry.def.

I ran into a bug where CONSTRUCTOR wasn't always being used (in
PassBuilder::registerMachineFunctionAnalyses()).

Make DUMMY_* just accept a pass name, there's no point in having proper
constructors if the generated dummy class has a templated constructor
accepting arbitrary arguments.

Remove unused getPassNameFromLegacyName() as it was using this but for
no purpose.

Remove DUMMY_MACHINE_FUNCTION_ANALYSIS, we can just add those as we port
them.

This for some reason exposed missing mock calls in existing unittests.

This is part of
https://github.com/llvm/llvm-project/commit/66347e516e22f9159b86024071fb92f364ac4418

The regression in downstream projects is about transfer_read patterns,
which needs more investigation. Add the support for transfer_write for
now.

Fixes https://github.com/llvm/llvm-project/issues/73285.

llvm-project/mlir/test/CAPI/sparse_tensor.c:50:43:
error: format specifies type 'unsigned long long' but the argument has type 'MlirSparseTensorLevelType' (aka 'unsigned long') [-Werror,-Wformat]
    fprintf(stderr, "level_type: %llu\n", lvlTypes[l]);
                                 ~~~~     ^~~~~~~~~~~
                                 %lu
1 error generated.

debugserver on arm64 devices can manage both Byte Address Select
watchpoints (1-8 bytes) and MASK watchpoints (8 bytes-2 gigabytes). This
adds a SupportedWatchpointTypes key to the QSupported response from
debugserver with a list of these, so lldb can take full advantage of
them when creating larger regions with a single hardware watchpoint.

Also add documentation for this, and two other lldb extensions, to the
lldb-gdb-remote.txt documentation.

Re-enable TestLargeWatchpoint.py on Darwin systems when testing with the
in-tree built debugserver. I can remove the "in-tree built debugserver"
in the future when this new key is handled by an Xcode debugserver.

This patch adjusts the build process for building the toolchain for the
CI container to perform more rigorous perf-training for PGO,
particularly building the entirety of LLVM as that is what showed the
best results while benchmarking. This patch also splits the job into two
stages to avoid timeouts due to the large increase in buildtime. There
are a couple other hacks added in here to make things work that we can
do away with eventually once we're able to run jobs like this on more
powerful self-hosted runners.

llvm-project/mlir/test/CAPI/sparse_tensor.c:50:42:
error: format specifies type 'unsigned long' but the argument has type 'MlirSparseTensorLevelType' (aka 'unsigned long long') [-Werror,-Wformat]
   50 |     fprintf(stderr, "level_type: %lu\n", lvlTypes[l]);
      |                                  ~~~     ^~~~~~~~~~~
      |                                  %llu
1 error generated.

[concepts] Extract function template pack arguments from the current instantiation if possible (#80594)

Before the constraint substitution, we employ
`getTemplateInstantiationArgs`, which in turn attempts to inspect
`TemplateArgument`s from the function template. For parameter packs from
their parent contexts, we used to extract the arguments from the
specialization type, in which could result in non-canonical argument
types e.g. `PackExpansionType`.

This may break the contract that, during a tree transformation, in
`TreeTransform::TryExpandParameterPacks`, the corresponding
`TemplateArgument`s for an `UnexpandedParameterPack` are expected to be
of `Pack` kinds if we're expanding template parameters.

Fixes https://github.com/llvm/llvm-project/issues/72557.

We have a Python script that needs to locate coredump path during
debugging so that we can retrieve certain metadata files associated with
it. Currently, there is no API for this.

This patch adds a new `SBProcess::GetCoreFile()` to retrieve target dump
file spec used for dump debugging. Note: this is different from the main
executable module spec. To achieve this, the patch hoists m_core_file
into PostMortemProcess for sharing.

---------

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

Fixes #80533

A set of microbenchmarks (https://github.com/llvm/llvm-test-suite/pull/26) showed that loop interleaving can be beneficial for loops with low trip count as well. Loop interleaving count computation is updated accordingly in prior patches while this patch removes the loop trip count threshold for interleaving. 

[Sanitizer][NFC] Replaces a few `InternalScopedString::AppendF` with `InternalScopedString::Append` (#80574)

Refer to commit 6611d58f ("Relax R_RISCV_ALIGN"), we can relax
R_LARCH_ALIGN by same way. Reuse `SymbolAnchor`, `RISCVRelaxAux` and
`initSymbolAnchors` to simplify codes. As `riscvFinalizeRelax` is an
arch-specific function, put it override on `TargetInfo::finalizeRelax`,
so that LoongArch can override it, too.

The flow of relax R_LARCH_ALIGN is almost consistent with RISCV. The
difference is that LoongArch only has 4-bytes NOP and all executable
insn is 4-bytes aligned. So LoongArch not need rewrite NOP sequence.
Alignment maxBytesEmit parameter is supported in psABI v2.30.

When promoted value, it is meaningless to copy value from reg to another
reg with the same type.
This PR add additional check for this cases to reduce the code size.
Fixes: #80053.

[LoopUnroll] Introduce PragmaUnrollFullMaxIterations as a hard cap on how many iterations we try to unroll (#78648)

Fixes [PR77842](https://github.com/llvm/llvm-project/issues/77842

) where
UBSAN causes pragma full unroll to try and unroll INT_MAX times. This
sets a cap to make sure we don't attempt this and crash the compiler.

Testing:
ninja check-all with new test

---------

Co-authored-by: Nikita Popov <github@npopov.com>

Reverts llvm/llvm-project#74515

Broke build: https://lab.llvm.org/buildbot/#/builders/75/builds/42512