The current `DebugValueManager`, which is mostly used in `RegStackify`,
simply sinks `DBG_VALUE`s along when a def instruction sinks.
(`RegStackify` only does sinks; it doesn't do hoists.)

But this simple strategy can result in incorrect combinations of
variables' values which would have not been possible in the original
program. In this case, LLVM's policy is to make the value unavailable,
so they will be shown as 'optimized out', rather than showing inaccurate
debug info. Especially, when an instruction sinks, its original
`DBG_VALUE` should be set to undef. This is well illustrated in the
third example in
https://llvm.org/docs/SourceLevelDebugging.html#instruction-scheduling.

This CL rewrites `DebugValueManager` with this principle in mind. When
sinking an instruction, it sinks its eligible `DBG_VALUE`s with it, but
also leaves undef `DBG_VALUE`s in the original place to make those
variables' values undefined.

Also, unlike the current version, we sink only an eligible subset of
`DBG_VALUE`s with a def instruction. See comments in the code for
details.

In case of cloning, because the original def is still there, we don't
set its `DBG_VALUE`s to undef. But we clone only an eligible subset of
`DBG_VALUE`s here as well.

One consequence of this change is that now we do sinking and cloning of
the def instruction itself within the `DebugValueManager`'s `sink` and
`clone` methods. This is necessary because the `DebugValueManager` needs
to know the original def's location before sinking and cloning in order
to scan other interfering `DBG_VALUE`s between the original def and the
insertion point. If we want to separate these two, we need to call
`DebugValueManager`'s `sink` and `clone` methods //before//
sinking/cloning the def instruction, which I don't think is a good
design alternative either, because the user of this class needs to pay
extra attention when using it.

Because this change is fixing the existing inaccuracy of the current
debug info, this reduces the variable info coverage in debug info, but
not by a large margin. In Emscripten core benchmarks compiled with
`-O1`, the coverage goes from 56.6% down to 55.2%, which I doubt will be
a noticeable drop. The compilation time doesn't have any meaningful
difference either with this change.

Reviewed By: dschuff

Differential Revision: https://reviews.llvm.org/D146744

The stdio test failures were due to headers potentially not being built
in the correct order. This should set up the dependencies correctly.

Reviewed By: sivachandra

Differential Revision: https://reviews.llvm.org/D146551

It is recommended to use `SmallVectorImpl`/`ArrayRef` over
`SmallVector<TypeName, N>` for function parameters:
https://llvm.org/docs/ProgrammersManual.html#llvm-adt-smallvector-h

Reviewed By: dschuff

Differential Revision: https://reviews.llvm.org/D146841

Currently calling stack locations is selected using `CALL` in ISel,
resulting in an invalid code and crashing in AsmPrinter. FastISel
correctly selects it will `CALL_INDIRECT`.

Fixes the problem reported in D146781.

Reviewed By: tlively, HerrCai0907

Differential Revision: https://reviews.llvm.org/D147033

An INLINEASM can have an implicit def of vcc. It is not appropriate for
fixImplicitOperands to change this to vcc_lo on wave32.

Differential Revision: https://reviews.llvm.org/D147157

Differential Revision: https://reviews.llvm.org/D147145

Summary:
The AMDGPU ABI isn't stable or well defined. For that reson we prefer to
rely on LTO to ensure that multiple files get linked correctly.
Currently the internal targets used for testing mix LLVM-IR and
assembly. We should be consistent here.

Mark test added in D141824 as unsupported for PS4/PS5 as those platforms require an external linker that is not present.

Some patch in the past introduce this non clang-formatted change

Adding CHECK-NOT for the MAX intrinsic implementation for which
reduction operation is not generated. Issue #61808 created to
track the same.

Move out MemRefDependenceGraph analysis structure out of LoopFusion into
the Affine Analysis library. This had been a long pending TODO. Moving
MDG out allows its use in other affine passes as well as allows building
custom affine fusion passes downstream while reusing upstream fusion
utilties. The file LoopFusion.cpp had also become lengthy and this
change makes things more modular. This change is a pure NFC and is a
code movement.

NFC.

Reviewed By: springerm

Differential Revision: https://reviews.llvm.org/D147105

This reverts commit 79116475.

Broke the ASan bots. See more information in
https://reviews.llvm.org/rG79116475124112051625b1a0665e35c861bb13fd

This reverts commit 1cfe1e73.

Depends on reverted commit 158a4312. See
https://reviews.llvm.org/rG79116475124112051625b1a0665e35c861bb13fd for
more information, this broke the ASan bots.

MSVC allows instantiations of exported or imported template classes with template
parameters that have internal linkage. Clang now allows it in Microsoft mode and for
the Playstation platform. This partially addresses issue 56068.

Note that MSVC also allows explicit dllexport/dllimport attributes on classes
with internal linkage (e.g. local classes or classes declared in anonymous name spaces).
Clang continues to reject such declarations.

Reviewed By: hans

Differential Revision: https://reviews.llvm.org/D146338

This patch fixes a crash that appear in an OpenMP MAX intrinsic
reduction when the reduction is not expressed using MAX intrinsic
function. Updated flang/test/Lower/OpenMP/wsloop-reduction-max.f90
test case with additional way of expressing MAX reduction.

Reviewed By: kiranchandramohan

Differential Revision: https://reviews.llvm.org/D146967

The intent of this attribute is for it to be applied to a module and
then hold information on runtime library (RTL) flags given to
Flang (or other OpenMP frontend) that should be lowered down to
LLVM-IR for devices as LLVM globals. The following related
flags are:

-fopenmp-target-debug
-fopenmp-assume-threads-oversubscription
-fopenmp-assume-teams-oversubscription
-fopenmp-assume-no-nested-parallelism
-fopenmp-assume-no-thread-state

These exist within Clang and are lowered into the IR when
offloading for device. This attribute allows this infromation
to be carried down from the Flang frontend to the
LLVM/OpenMP Dialect to LLVM-IR translation phase
and then be lowered to LLVM-IR.

Reviewers: kiranchandramohan

Differential Revision: https://reviews.llvm.org/D144896

Failing due to forgetting the .exe suffix the clang-offload-packager
receives on windows.

Differential Revision: https://reviews.llvm.org/D147098

This reverts commit 3b4cb1e9.

Reland D145718 but unconditionally define __sanitizer_mallinfo which is
exposed as part of the hwasan interface.

Differential Revision: https://reviews.llvm.org/D145718

mallinfo is platform-specific and not specified by either posix or the C
standard, but the hwasan interface unconditionally exposes
__sanitizer_mallinfo which returns a struct __sanitizer_struct_mallinfo
which is defined in sanitizer_platform_limits_posix.h, so this should
also be available for fuchsia to provide __sanitizer_mallinfo. Fuchsia
doesn't need the rest of what's in sanitizer_platform_limits_posix.h so
we can just move it to its own header.

Exposing this and not forcing it to hide behind
SANITIZER_INTERCEPT_MALLOPT_AND_MALLINFO fixes the test failures found
after landing D145718.

Differential Revision: https://reviews.llvm.org/D147092

Handle the splat and dense case.

I saw this pattern show up in a couple recent SPIR-V-specific bug
report.

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D147109

Currently LLVM fails to determine that conditional loads in
@accesses_to_struct_dereferenceable are dereferenceable unconditionally.

circumstances when parsing ASTs

Fix ArgsAsWritten being null for ConceptSpecializationExpr in certain
circumstances when parsing ASTs

ASTStmtWriter::VisitConceptSpecializationExpr specifically expects
getTemplateArgsAsWritten() to return true, which it wasn't when parsed
by ASTContext.cpp in certain edge cases.

Fixes: #61486

Differential Revision: https://reviews.llvm.org/D146678

Pass the correct option to the linker when "-flto-jobs=" is passed to
the driver.

Error only applies if it is a whole array. Page 157 in OpenMP 5.2
standard.

Reviewed By: TIFitis

Differential Revision: https://reviews.llvm.org/D147142

Fixes: https://github.com/llvm/llvm-project/issues/61710

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D147163

When  depend on template parameter,
compiler can use NullStmt instead of CompoundStmt.
This causes issues as we losing information about
end location of that Stmt. To avoid this issue
check now ignores ifStmt with NullStmt on true-branch.

Fixes: https://github.com/llvm/llvm-project/issues/61435

Reviewed By: carlosgalvezp

Differential Revision: https://reviews.llvm.org/D146887

This reverts commit e4918d43.

Reason: Broke the sanitizer bots. More information in
https://reviews.llvm.org/D146920 and should be re-landed with
https://reviews.llvm.org/D147164 (or similar)

This patches chagnes the type of map_operands to OpenMP_PointerLikeType for OMP Target Data directives.

Differential Revision: https://reviews.llvm.org/D147050

Commit 3d7946c5 implemented a DR that allowed us to error in a case
where an ill-formedness in a RequiresExpr is diagnosed as a satisfaction
failure.  However, it failed to cover cases where the RequiresExpr had
Requirements that failed for similar reasons.

This patch propagates the RecoveryExpr "containsErrors" correctly
through RequiresExpr.

Fixes: #61776

This reverts commit ccdcfad0.

Reason: Introduced an ODR that broke the ASan bots. See more information
in Phabricator: https://reviews.llvm.org/D146935

As pointed out by @fhahn in https://reviews.llvm.org/D146876.

Reviewed By: tejohnson

Differential Revision: https://reviews.llvm.org/D147133

Ensure that the current symbol is added to the parent contexts in the
output of libclang function for generating symbol graphs for single symbols.

Differential Revision: https://reviews.llvm.org/D147138

Move the conversion of DILineInfo to JSON into a separate function, so
it can be used in other places too.

This is a prerequisite patch for implementation of symbol+offset lookup.

Differential Revision: https://reviews.llvm.org/D147112

llvm-symbolizer echoed input if it was not recognized as a valid address.
This behavior was extended to llvm-addr2line as well. GNU addr2line in
this case optputs "??:0". This difference prevents implementation of
symbol+offset lookup available in the recent versions of GNU binutils.
In that case a string that is not an address may be a symbol.

This change make reaction of llvm-addr2line on unrecognized input closer
to GNU addr2line.

On AIX, the build-id can be embedded in a binary using the -mxcoff-build-id
compiler option. When present, the build id is stored as an ascii string at the
beginning of the string table in the loader section of the XCOFF file.

Reviewed By: stephenpeckham, daltenty

Differential Revision: https://reviews.llvm.org/D146976