Provide fixes for inline namespaces in the same format as clang-format.

Fixes: #56804

Reviewed By: carlosgalvezp

Differential Revision: https://reviews.llvm.org/D157178

[X86] Only fold broadcast with extract_vector_elt/scalar_to_vector if the scalar type matches the vector element type

Avoid handling implicit extension/truncation with scalar<->vector transfers

Fixes #64439

Extend the existing 128-bit -> 64-bit truncation handling by widening/narrowing the src/dst vectors and use the lower half operand/result for PACKSS/PACKUS instructions.

Fixes #29973

The original test has a unused load, which is removed. Also add a
variant with a store that cannot be removed, forcing the mask for the
block to always be generated.

This patch adds a lightweight instance of Attributor that only deduces
attributes.

This is just an initial version with the goal to have a version that
only focuses on attributes to replace the function-attrs pass.

The initial version has a few open issues pending until default
enablement, the main one probably being compile time. The main
additional functionality this will provide in general is propagating
attributes to call sites.

Open issues:

* compile time
  The current version increase O3 +2.67% and ThinLTO +6.18% when replacing FunctionAttr
   https://llvm-compile-time-tracker.com/compare.php?from=c4bb3e073548cf436d5fa0406e3ae75e94684dec&to=d992630a69c79a2587d736e6a88f448850413bd1&stat=instructions%3Au
   Both are with an additional change to preserve more analysis, like FunctionAttrs CGSCC run.

* some missed attribute inference

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D152081

This stops tablegen from generating unneeded entries in
RISCVGenDAGISel.inc.

Enable record-replay for OpenMP offload kernels.  On recording the initialization
is performed on device initialization by reading env variables. (This is similar to
the way rr used to operate). The primary change takes place in the replay phase
with the replay tool explicitly initializing the record-replay functionality.

Differential Revision: https://reviews.llvm.org/D156174

Fix

The last use was removed by:

  commit b14e30f1
  Author: Sameer Sahasrabuddhe <sameer.sahasrabuddhe@amd.com>
  Date:   Thu Jul 27 09:54:50 2023 +0530

This stops tablegen from generating an unneeded pattern checking for
an i32 type.

Make sure to check type conversion results. Add missing tests.

Fix some typos in the surrounding code.

Fixes: https://github.com/llvm/llvm-project/issues/64409

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D157166

These tests were causing issues on the clang-cmake-x86_64-avx512-linux
builder due to transient failures. I'm unable to reproduce the issue
locally now, so we'll see what happens. There were a couple changes in
between disabling them and now that might have an impact.

Currently the documentation on the command line options for llvm-link is
quite sparse. This patch adds in the options that the tool understands
that aren't currently present in the documentation.

Reviewed By: tejohnson, MaskRay

Differential Revision: https://reviews.llvm.org/D155904

It's not really correct to implement this as exp(x) - 1, it was maybe
OK for the restricted float-as-double case handled here. There's not a
strong reason to special case it with the host function, as the
implementation naturally constant folds anyway. InstSimplify can fully
handle everything in it, so just running the inliner alone with a
constant argument produces the fully constant folded result. This also
had no tests.

https://reviews.llvm.org/D156892

Renames the old `DimLvlExpr::getContext` to `tryGetContext` to make it clear that the result could be a nullptr.  Also adds new delegation methods `DimSpec::tryGetContext` and `LvlSpec::getContext`.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D157163

Improves the conversion from `DimLvlMap` to STEA, in order to correct rank-mismatch issues in the roundtrip tests.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D157162

Creating stubs in JITLink require creating architecture-specific edges. In order to allow user to create stubs in cross-architecture manner, this patch exposes these stub creations functions by returning "stub creators" for given triple.

Reviewed By: lhames

Differential Revision: https://reviews.llvm.org/D155557

In file included from /data/workspace/llvm-project/lldb/unittests/ObjectFile/PECOFF/TestSectionSize.cpp:10:
/data/workspace/llvm-project/third-party/unittest/googletest/include/gtest/gtest.h:1620:28: error: comparison of integers of different signs: 'c
onst unsigned long' and 'const int' [-Werror,-Wsign-compare]
GTEST_IMPL_CMP_HELPER_(NE, !=);
~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~

/data/llvm-project/lldb/source/Interpreter/OptionValue.cpp:28:3: error: 'scoped_lock' may not intend to support class template argument deduction [-Werror,-Wctad-maybe-unsupported]
  std::scoped_lock lock(m_mutex, other.m_mutex);
  ^
/opt/rh/gcc-toolset-12/root/usr/lib/gcc/x86_64-redhat-linux/12/../../../../include/c++/12/mutex:692:11: note: add a deduction guide to suppress this warning
    class scoped_lock
          ^
1 error generated.

/data/llvm-project/lldb/source/Target/ThreadList.cpp:739:5: error: 'scoped_lock' may not intend to support class template argument deduction [-Werror,-Wctad-maybe-unsupported]
    std::scoped_lock guard(GetMutex(), rhs.GetMutex());
    ^
/opt/rh/gcc-toolset-12/root/usr/lib/gcc/x86_64-redhat-linux/12/../../../../include/c++/12/mutex:692:11: note: add a deduction guide to suppress this warning
    class scoped_lock
          ^
1 error generated.

Fix failure at https://lab.llvm.org/buildbot/#/builders/216/builds/25114

due to missing ROCm

Removes dependency of the DataflowWorklistBase type on the PostOrderCFGView. The
base had no particular connection to this order -- looks like it was included as
a convenience to the two derived classes rather than because it belonged there
conceptually. Removing it cleans up the base class a bit and makes it suitable
to extension with the new WTODataflowWorklist class.

Differential Revision: https://reviews.llvm.org/D153071

ThreadList::Update is being caught by thread sanitizer. There's even a
comment on the function that both mutexes should be lock (even though
only the "this" mutex was actually being locked). Fix this by locking
both mutexes.

Differential Revision: https://reviews.llvm.org/D157153

Thread sanitizer is catching data races in OptionValue, protect accesses
to OptionValue with a mutex.

Differential Revision: https://reviews.llvm.org/D157041

When doing combined compilation/link for HIP source
files, clang should link the HIP runtime library
automatically without --hip-link.

Reviewed by: Siu Chi Chan, Joseph Huber

Differential Revision: https://reviews.llvm.org/D156426

PSTL contains many pragmas that request loop vectorization, which would
produce a warning when the compiler is unable to fulfill the request (if
`-Wpass-failed` is enabled). This is normal and expected in some cases,
and we don't want `-Werror` to turn that into a compilation failure.

Differential Revision: https://reviews.llvm.org/D157145

The v variants of the printf functions take their variadic arguments as
a va_list instead of as individual arguments. They are otherwise
identical to the corresponding printf variants. This patch adds them
(vprintf, vfprintf, vsprintf, and vsnprintf) as well as tests.

Reviewed By: phosek

Differential Revision: https://reviews.llvm.org/D157138

The import/export maps, and the ModuleToDefinedGVSummaries map, are all
indexed by module paths, which are StringRef obtained from the module
summary index, which already has a data structure than owns these
strings (the ModulePathStringTable). Because these other maps are also
StringMap, which makes a copy of the string key, we were keeping
multiple extra copies of the module paths, leading to memory overhead.

Change these to DenseMap keyed by StringRef, and document that the
strings are owned by the index.

The only exception is the llvm-link tool which synthesizes an import list
from command line options, and I have added a string cache to maintain
ownership there.

I measured around 5% memory reduction in the thin link of a large
binary.

Differential Revision: https://reviews.llvm.org/D156580

OpenACC and OpenMP directive are
emitted in the module file. Just reorganized the code
so this is well separated and do not pollute the main
Fortran part.

Reviewed By: kiranchandramohan

Differential Revision: https://reviews.llvm.org/D157143

GCC complains FieldParser returns the value that does not have an
obvious initial value. Set an initial return value for the FieldParser
function.

The change is a no-op functionality-wise. The return value should be
updated within the function.

This should fix https://github.com/llvm/llvm-project/issues/64312

Differential Revision: https://reviews.llvm.org/D156848

All Python files in the LLVM repository were reformatted with Black [1].
Files inside the LLDB subproject were reformatted in 2238dcc3. This
patch updates a handful of tests that were added or modified since then
and weren't formatted with Black.

[1] https://discourse.llvm.org/t/rfc-document-and-standardize-python-code-style/68257

SWIG 4 was released in 2019 and has been the de-facto standard for a
while now. All bots are running SWIG 4.0 or later.

This was motivated by #64279 which discovered that 662548c8 broke the
LLDB build with SWIG 3 on Windows.

Differential revision: https://reviews.llvm.org/D156804

Broke sanitizer build bots

This reverts commit b82c2b9a.

Reorganize options and group them as "Common Offloading options",
"CUDA options" and "HIP options", so that they will be grouped
together like "OpenCL options" and "SYCL options" in
https://clang.llvm.org/docs/ClangCommandLineReference.html

Reviewed by: Joseph Huber, Fangrui Song

Differential Revision: https://reviews.llvm.org/D156886

Allow users to run a dependency scan with a cc1 command line in addition to a driver command line. DependencyScanningAction was already being run with a cc1 command line, but DependencyScanningWorker::computeDependencies assumed that it was always provided a driver command line. Now DependencyScanningWorker::computeDependencies can handle cc1 command lines too.

Reviewed By: jansvoboda11

Differential Revision: https://reviews.llvm.org/D156234

[flang] Add -fppc-native-vector-element-order option to control the element order in PowerPC vector types

This patch also adds a LIT test for the vec_cvf intrinsic that
can be affected by the option.

Co-authored-by: Mark Danial <Mark.Danial@ibm.com>
Co-authored-by: Daniel Chen <cdchen@ca.ibm.com>

Differential Revision: https://reviews.llvm.org/D155852

This fixes 2 bugs and adds corresponding tests. Both related to unreachable
blocks. One occured in the `WTOCompare` construction, which assumed the size of
the order was the same as the number of blocks in the CFG, which isn't true when
some blocks are unreachable.  The other assumed predecessor pointers were
non-null, which can be false for blocks with unreachable predecessors.

Differential Revision: https://reviews.llvm.org/D157033

F18 produce a module file with a declaration part. This patch
propagates the OpenACC declare information in the module file along the variable
declaration.

Depends on D157042

Reviewed By: razvanlupusoru

Differential Revision: https://reviews.llvm.org/D157121