Make it a compile-time error to pass an int/unsigned/etc to
fromRawInteger.

Hopefully this prevents errors of the form:

```
for (unsigned ID : getVarLocs()) {
  auto VL = LocMap[LocIndex::fromRawInteger(ID)];
  ...
```

When an implicitly generated decl was the first entry in the group, we
attempted to lookup comments with an empty FileID, leading to crashes. Avoid
this by trying to use the other declarations in the group, and then bailing out
if none are valid.

rdar://59919733
Differential revision: https://reviews.llvm.org/D75483

Summary:
It is not safe for ARMConstantIslands to undoLRSpillRestore. PrologEpilogInserter is
the one to ensure stack alignment, taking into consideration LR is spilled or not.

For noreturn function with StackAlignment 8 (function contains call/alloc),
undoLRSpillRestore cause stack be mis-aligned. Fixing stack alignment in
ARMConstantIslands doesn't give us much benefit, as undo LR spill/restore only
occur in large function with near branches only, also doesn't have callee-saved LR spill.

Reviewers: t.p.northover, rengolin, efriedma, apazos, samparker, ostannard

Reviewed By: ostannard

Subscribers: dmgreen, ostannard, kristof.beyls, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D75288

This is a follow up to D74108. This option is not supported by older
versions of libtool so remove it. We keep the -s option to build the
index.

Use emplace_back to allow for in-place construction of the
BroadcastEventSpec instead of copying it.

YAML files were not being run during lit testing as there was no lit.local.cfg file. Once this was fixed, some buildbots would fail due to a StringRef that pointed to a std::string inside of a temporary llvm::Triple object. These issues are fixed here by making a local triple object that stays around long enough so the StringRef points to valid data. Also fixed an issue where strings for files in the file table could be added in opposite order due to parameters to function calls not having a strong ordering, which caused tests to fail. Added new arch specfic directories so when targets are not enabled, we continue to function just fine.

Differential Revision: https://reviews.llvm.org/D75390

Summary:
Currently `SymbolFileDWARF::TypeSet` is a typedef to a `std::set<Type *>`.
In `SymbolFileDWARF::GetTypes` we iterate over a TypeSet variable when finding
types so that logic is non-deterministic as it depends on the actual pointer address values.

This patch changes the `TypeSet` to a `llvm::UniqueVector` which always iterates in
the order in which we inserted the types into the list.

Reviewers: JDevlieghere, aprantl

Reviewed By: JDevlieghere

Subscribers: mgrang, abidh, lldb-commits

Tags: #lldb

Differential Revision: https://reviews.llvm.org/D75481

While here, explain that this is not really used. Probably NFC.

as it broke the Werror build:

.../sources/llvm-project/mlir/include/mlir/ExecutionEngine/CRunnerUtils.h:85:16: error: zero size arrays are an extension [-Werror,-Wzero-length-array]
  char padding[detail::nextPowerOf2<sizeof(T[Dim])>() - sizeof(T[Dim])];
               ^~~~~~~~~~~~~~~

This reverts commit 78f9e5d0.

X86 has several instructions which are documented as enabling interrupts exactly one instruction *after* the one which changes the SS segment register. Inserting a nop between these two instructions allows an interrupt to arrive before the execution of the following instruction which changes semantic behaviour.

The list of instructions is documented in "Table 24-3. Format of Interruptibility State" in Volume 3c of the Intel manual. They basically all come down to different ways to write to the SS register.

Differential Revision: https://reviews.llvm.org/D75359

One of the checks has been removed as it seem invalid.
The LoopStep size is always almost a 32-bit.

Differential Revision: https://reviews.llvm.org/D75079

[TargetLowering] Fix what look like copy/paste mistakes in compare with infinity handling SimplifySetCC.

I expect that the isCondCodeLegal checks should match that CC of
the node that we're going to create.

Rewriting to a switch to minimize repeated mentions of the same
constants.

This reverts commit 80d0a137, and the
follow on fix in 873c0d07. It is
causing test failures after a multi-stage clang bootstrap. See
discussion on D73242 and D75201.

Depending on the OS used, a module-enabled build can fail due to the
special handling <cassert> gets as textual header.

This reverts commit 398b4ed8.
As requested in https://bugs.llvm.org/show_bug.cgi?id=43465#c37 .

For ODS generated operations enable querying whether there is a derived
attribute with a given name.

This reverts commit 57688350.

Need to conditionalize for ARM targets, this is failing on machines that don't have ARM targets.

YAML files were not being run during lit testing as there was no lit.local.cfg file. Once this was fixed, some buildbots would fail due to a StringRef that pointed to a std::string inside of a temporary llvm::Triple object. These issues are fixed here by making a local triple object that stays around long enough so the StringRef points to valid data. Also fixed an issue where strings for files in the file table could be added in opposite order due to parameters to function calls not having a strong ordering, which caused tests to fail.

Differential Revision: https://reviews.llvm.org/D75390

Summary: This follows up D67377 that added the isHot side.

Reviewers: davidxl

Subscribers: eraman, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D75283

A two-stage ThinLTO build previously failed the clang/test/Driver/hurd.c test because of a static_cast in "tools::gnutools::Linker::ConstructJob()" which wrongly converted an instance of "clang::driver::toolchains::Hurd" into that of "clang::driver::toolchains::Linux". ThinLTO would later devirtualize the "ToolChain.getDynamicLinker(Args)" call and use "Linux::getDynamicLinker()" instead, causing the test to generate a wrong "-dynamic-linker" linker flag (/lib/ld-linux.so.2 instead of /lib/ld.so)

Fixes PR45061.

Differential Revision: https://reviews.llvm.org/D75373

Summary: fixed the warning from gcc since prios 0-100 are reserved for the internal use.

Reviewers: grokos

Subscribers: kkwli0, caomhin, openmp-commits

Tags: #openmp

Differential Revision: https://reviews.llvm.org/D75458

This is needed for D74873, AMDGPU going to have 16 bit subregs
and the largest tuple is 32 VGPRs, which results in 64 lanes.

Differential Revision: https://reviews.llvm.org/D75378

This reverts commit 63776766.
Need to fix the testcase.

It looks like the suppression file isn't being copied over to the
device.

After the format change from D69471, there can be more than one section
in an object that contains coverage function records. Look up each of
these sections and concatenate all the records together.

This re-enables the instrprof-merging.cpp test, which previously was
failing on OSes which use comdats.

Thanks to Jeremy Morse, who very kindly provided object files from the
bot I broke to help me debug.

Added basic support (parsing/sema/serialization) for 'destroy' clause in
depobj directives.

Pickup the default crt and libs when the target is musl.

Differential Revision: https://reviews.llvm.org/D75139

Specifically, this fixes a false-positive in -Wobjc-signed-char-bool.
rdar://57372317

Differential revision: https://reviews.llvm.org/D75387

We always want to build the table of contents. Additionally, we also
set the flag to make the output deterministic which is already the
default for llvm-ar.

Differential Revision: https://reviews.llvm.org/D74108

Summary:
This function is (supposed) to be a list of asserts that just do a generic sanity check
on declarations we return. Right now this function is hidden behind the
LLDB_CONFIGURATION_DEBUG macro which means it will *only* be run in
debug builds (but not Release+assert builds and so on).

As we have not a single CI running in Debug build, failures in VerifyDecl are hidden
from us until someone by accident executes the tests in Debug mode on their own machine.

This patch removes the `ifdef`'s for LLDB_CONFIGURATION_DEBUG and puts
the `getAccess()` call in `VerifyDecl` behind a `#ifndef NDEBUG` to make sure
that this function is just an empty function with internal linkage when NDEBUG
is defined (so compilers should just optimize away the calls to it).

Reviewers: aprantl

Reviewed By: aprantl

Subscribers: shafik, abidh, JDevlieghere, lldb-commits

Tags: #lldb

Differential Revision: https://reviews.llvm.org/D75330

CFI instructions can only safely be outlined when the outlined call is a tail
call, or when the outlined frame is fixed up.

For the sake of correctness, disable outlining from CFI instructions.

Add machine-outliner-cfi.mir to test this.

Added basic parsing/sema/serialization support for depobj directive.

Previously we would also accept DISubprograms that matched in name
only, but this doesn't appear to be necessary any more.

I did a Full and Thin LTO build of Clang and it completed without a warning.

Differential Revision: https://reviews.llvm.org/D75213