GCC 7.3 cannot build 16.x releases.
```
In file included from /tmp/llvm-16/llvm/lib/Transforms/IPO/AttributorAttributes.cpp:14:0:
/tmp/llvm-16/llvm/include/llvm/Transforms/IPO/Attributor.h:1137:32: error: duplicate initialization of ‘llvm::AnalysisGetter::HasLegacyWrapper<Analysis, std::void_t<typename Analysis::Lega
cyWrapper> >’
 constexpr bool AnalysisGetter::HasLegacyWrapper<
                                ^~~~~~~~~~~~~~~~~
       Analysis, std::void_t<typename Analysis::LegacyWrapper>> = true;
       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/tmp/llvm-16/llvm/include/llvm/Transforms/IPO/Attributor.h:1137:32: error: got 1 template parameters for ‘constexpr const bool llvm::AnalysisGetter::HasLegacyWrapper< <template-parameter-1
-1>, <template-parameter-1-2> >’
/tmp/llvm-16/llvm/include/llvm/Transforms/IPO/Attributor.h:1137:32: error:   but 2 required
```

The 17.x and main branches have more failures, e.g.

```
llvm/lib/Transforms/Vectoriz...

We have multiple variations of InstrFormatR that pack different
fields into the upper 7 bits. The other 25 bits are all the same.

Add base class to capture this commonality and allow subclasses to
explicitly define Inst{31-25}.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D156590

We have multiple variations of InstrFormatI that pack different
fields into the upper 12 bits. The other 20 bits are all the same.

Add base class to capture this commonality and allow subclasses to
explicitly define Inst{31-20}.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D156589

Follow up to commit 41895843,
"[InstCombine] Only perform one iteration".

This will make sure verify-fixpoint/no-verify-fixpoint is mentioned
when doing "opt -print-passes".

For the time being, we're keeping the `Strict` versions around as deprecated synonyms so that clients can be migrated, but these synonyms will be removed soon.

Depends On D156673

Reviewed By: ymandel, xazax.hun

Differential Revision: https://reviews.llvm.org/D156674

Depends On D156672

Reviewed By: ymandel, xazax.hun

Differential Revision: https://reviews.llvm.org/D156673

This eliminates all uses of the deprecated accessors.

Reviewed By: ymandel, xazax.hun

Differential Revision: https://reviews.llvm.org/D156672

Access to individual DefinitionGenerators is serialized in order to make
generators easier to implement: serializing access means that tryToGenerate
methods don't have to handle concurrent, potentially overlapping, requests.

Prior to this patch serialization was achieved by having each lookup acquire a
lock on each generator, however this causes the lookup thread to block if the
generator is in use. In the common case where many objects reference some
common library symbol that is provided by a generator this may cause many
threads to block concurrently preventing progress on other work.

This patch changes the model so that lookups are automatically suspended if
they need to use a generator that is already in use, and then automatically
resumed once the generator is free. This is achieved by reusing the lookup
suspension machinery that was introduced in 069919c9 for optionally
asynchronous generators.

* Add KMP_CPU_EQUAL and KMP_CPU_ISEMPTY to affinity mask API

* Add printout of leader to hardware thread dump

* Allow OMP_PLACES to restrict fullMask

This change fixes an issue with the OMP_PLACES=resource(#) syntax.
Before this change, specifying the number of resources did NOT change
the default number of threads created by the runtime. e.g.,
OMP_PLACES=cores(2) would still create __kmp_avail_proc number of
threads. After this change, the fullMask and __kmp_avail_proc are
modified if necessary so that the final place list dictates which
resources are available and how thus, how many threads are created by
default.

* Introduce hybrid core attributes to OMP_PLACES and KMP_AFFINITY

For OMP_PLACES, two new features are added:
  1) OMP_PLACES=cores:<attribute> where <attribute> is either
     intel_atom, intel_core, or eff# where # is 0 - number of core
     efficiencies-1. This syntax also supports the optional (#)
     number selection of resources.
  2) OMP_PLACES=core_types|core_effs where this setting will create
     the number of core_types (or core_effs|core_efficiencies).

For KMP_AFFINITY, the granularity setting is expanded to include two new
keywords: core_type, and core_eff (or core_efficiency). This will set
the granularity to include all cores with a particular core type (or
efficiency). e.g., KMP_AFFINITY=granularity=core_type,compact will
create threads which can float across a single core type.

Differential Revision: https://reviews.llvm.org/D154547

size() > 0 -> !empty
indentation
mismatched names on parameters in decls/defs
const on value return types

Current implementation outputs the parameter name when used with @param
(or @tparam) doxygen tag but not the comment itself.

Differential Revision: https://reviews.llvm.org/D156322

This update allows constant folding for many 128 bit floating point intrinsics
through the library quadmath, which is only available on some platforms.

Differential Revision: https://reviews.llvm.org/D156435

Add codegen for llvm bitreverse elementwise builtin
The bitreverse elementwise builtin is necessary for HLSL codegen.
Tests were added to make sure that the expected errors are encountered when these functions are given inputs of incompatible types, or too many inputs.
The new builtin is restricted to integer types only.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D156357

This is a new algorithm for function layout (reordering) based on the call graph
extracted from a profile data; see diffs down the stack for more details.

This layout is very similar to the existing hfsort+, but perhaps a little better
on some benchmarks. The goals of the change is as follows:

(i) rename and replace hfsort+ with a newer (hopefully better) implementation.
I'd prefer to keep both algs together for some time to simplify evaluation and
transition, but do want to remove hfsort+ once we're confident that there are
no regressions.

(ii) unify the implementation of code layout algorithms across LLVM. Currently
Passes/HfsortPlus.cpp and Utils/CodeLayout.cpp share many implementation-specific
details; this diff unifies the code.

Reviewed By: Amir

Differential Revision: https://reviews.llvm.org/D153039

RoundedSize is supposed to be used on directly mapping. To determine if
a cached block is feasible, the size doesn't need to be rounded in
advance. As a result, the use of RoundedSize may miss some chance of
using cached blocks.

This reverts commit 4c6b8bb8.

Reviewed By: frs513

Differential Revision: https://reviews.llvm.org/D156583

This commit does a few related things:
- Removes unused function `uuid_is_null`
- Removes unneeded includes of UuidCompatibility.h
- Renames UuidCompatibility to AppleUuidCompatibility and adds a comment
  to clarify intent of header.
- Moves AppleUuidCompatibility to the include directory

Differential Revision: https://reviews.llvm.org/D156562

Fix typo in ScalarEvolution public method.

Differential Revision: https://reviews.llvm.org/D156621

Improve memory usage by reducing the lifetime of CTF types. Once a CTF
type has been converted to a (complete) LLDB type, there's no need to
keep it in memory anymore. For most types, we can free them right after
creating the corresponding LLDB types. The only exception is record
types, which are only completed lazily.

Differential revision: https://reviews.llvm.org/D156606

Instead of creating psuedo source files for each stack frame this change adopts the new DAP “disassemble” request, allowing clients to inspect assembly instructions of files with debug info in addition to files without debug info.

[[ https://microsoft.github.io/debug-adapter-protocol/specification#Requests_Disassemble | spec ]]

See attached screenshot of the disassembly view. {F28473848}

Reviewed By: wallace

Differential Revision: https://reviews.llvm.org/D156493

Adds tests to guard against the issue solved in D156592.

Depends on D156592

Reviewed By: ChuanqiXu, #libc, philnik

Differential Revision: https://reviews.llvm.org/D156594

@ChuanqiXu noticed std::atomic was not properly exported in the std module.
Investigation showed other named declarations were not exported either. This
fixes the issue.

Depends on D156550

Reviewed By: #libc, philnik

Differential Revision: https://reviews.llvm.org/D156592

Importing modules in the test suite failed due to linker errors. This
fixes the issue and adds a minimal test to validate importing the std
module works.

Reviewed By: #libc, philnik

Differential Revision: https://reviews.llvm.org/D156550

This is the first of transition tapi-diff to be readtapi.
This tool will eventually replace functionality for Xcode's `xcrun tapi stubify` and
`xcrun tapi archive`.
This patch updates the tool name and is a minor refactor for the driver to handle more options.

Reviewed By: zixuw

Differential Revision: https://reviews.llvm.org/D153045

Switch to LLVM-18 as ToT version.
Use the latest official CMake 3.27 release.

Reviewed By: #libc, philnik

Differential Revision: https://reviews.llvm.org/D156593

The docs were still talking about Visual Studio 2013, which we've not
supported in a while.

I made the documentation a bit more vague because NATVIS is sometimes
automatically picked up when building and sometimes requires manually
installing the files. (For me, LLVM NATVIS is automatically picked up
while Clang NATVIS needs manual installation; this happens because the
way the monorepo is organized causes confusion when used with Visual
Studio's built-in CMake functionality. Building from a .sln file
generated by CMake may have better behavior in this case.)

[AMDGPU] AMDGPUTargetStreamer::getArchNameFromElfMach - use GPUKind::GK_NONE for unknown AMDGPU ElfMach flags

Avoids issue reported on ea2c159f where objects with unknown subtarget ids were causing llvm_unreachable crashes

[X86] Allow pre-SSE41 targets to extract multiple v16i8 elements coming from the same DWORD/WORD super-element

Pre-SSE41 targets tended to have weak (serial) GPR<->VEC moves, meaning we only allowed a single v16i8 extraction before spilling the vector to stack and loading the i8 elements instead. But this didn't make use of the DWORD/WORD extraction we had to use could extract multiple i8 elements at the same time.

This patch attempts to determine if all uses of a vector are element extractions, and works out whether all the extractions share the same WORD or (lowest) DWORD, in which case we can perform a single extraction and just shift/truncate the individual elements.

Differential Revision: https://reviews.llvm.org/D156350

Update comment in previous commit from D156603.

This was adding confusing global state which was shadowed most of the
time.

https://reviews.llvm.org/D156680

Provide Python bindings for transform ops defined in the vector dialect.
All of these ops are sufficiently simple that no mixins are necessary
for them to be nicely usable.

Reviewed By: ingomueller-net

Differential Revision: https://reviews.llvm.org/D156554

Add an ODS (tablegen) backend to generate Python enum classes and
attribute builders for enum attributes defined in ODS. This will allow
us to keep the enum attribute definitions in sync between C++ and
Python, as opposed to handwritten enum classes in Python that may end up
using mismatching values. This also makes autogenerated bindings more
convenient even in absence of mixins.

Use this backend for the transform dialect failure propagation mode enum
attribute as demonstration.

Reviewed By: ingomueller-net

Differential Revision: https://reviews.llvm.org/D156553

Differential Revision: https://reviews.llvm.org/D156274

Exist gracefully instead of triggering an assertions.

Fixes: https://github.com/llvm/llvm-project/issues/64257

Reviewed By: springerm

Differential Revision: https://reviews.llvm.org/D156694

Reviewed By: aeubanks

Differential Revision: https://reviews.llvm.org/D153107

https://reviews.llvm.org/D156678