The linter complains about the formatting in subsequent changes.
Fixing that now.

Reviewed By: iains

Differential Revision: https://reviews.llvm.org/D120117

This fold is done in IR:
https://alive2.llvm.org/ce/z/jWyFrP

There is an x86 test that shows an improvement
from the added flexibility of using add (commutative).

The other diffs are presumed neutral.

Note that this could also be folded to an 'xor',
but I'm not sure if that would be universally better
(eg, x86 can convert adds more easily into LEA).

This helps prevent regressions from a potential fold for
issue #53829.

Add `llvm_unreachable` to prevent warnings/errors in gcc and msvc.

Differential Revision: https://reviews.llvm.org/D119737

Take advantage of D117117 to simplify all {{\[}} to [ and {{\]}} to ].

Differential Revision: https://reviews.llvm.org/D117298

[OpenMP][libomptarget] Delay restore of shadow pointers in structs to after H2D memory copies are completed

When using asynchronous plugin calls, shadow pointer restore could happen before the D2H copy for the entire struct has completed, effectively leaving a device pointer in a host struct.
This patch fixes the problem by delaying restore's to after a synchronization happens (target regions) and by calling early synchronization (target update).

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D119968

MLIR has the notion of allocation scopes which specify that stack allocations (e.g. memref.alloca, llvm.alloca) should be freed or equivalently aren't available at the end of the corresponding region.
Currently neither OpenMP parallel nor SCF parallel regions have the notion of such a scope.

This clearly makes sense for an OpenMP parallel as this is implemented in with a new function which outlines the region, and clearly any allocations in that newly outlined function have a lifetime that ends at the return of the function, by definition.

While SCF.parallel doesn't have a guaranteed runtime which it is implemented with, this similarly makes sense for SCF.parallel since otherwise an allocation within an SCF.parallel will needlessly continue to allocate stack memory that isn't cleaned up until the function (or other allocation scope op) which contains the SCF.parallel returns. This means that it is impossible to represent thread or iteration-local memory without causing a stack blow-up. In the case that this stack-blow-up behavior is intended, this can be equivalently represented with an allocation outside of the SCF.parallel with a size equal to the number of iterations.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D119743

At the moment, this just groups comments with a reasonably named predicate, but I plan to add other cases to this in the near future.

This is a follow-up patch of D119378.

Reviewed By: ABataev

Differential Revision: https://reviews.llvm.org/D119392

The benchmark framework synthesizes fake "aggregate" Samples representing mean, median and cv.
We're only interested in "iteration" samples.

Differential Revision: https://reviews.llvm.org/D120062

This patch adds the support for `atomic compare capture` in parser and part of
sema. We don't create an AST node for this because the spec doesn't say `compare`
and `capture` clauses should be used tightly, so we cannot look one more token
ahead in the parser.

Reviewed By: ABataev

Differential Revision: https://reviews.llvm.org/D116261

Previously the name of the class (AMDGPUCFGStructurizer) did not
match the name of the file (AMDILCFGStructurizer).

Standardize on the name R600MachineCFGStructurizer by analogy with
AMDGPUMachineCFGStructurizer.

Differential Revision: https://reviews.llvm.org/D120128

Revert rGc24e197a "[clangd] getHover - pass FormatStyle argument by const reference"

There are a number of buildbot build failures on non MSVC compilers

Lower simple binary operation (+, -, *, /) for scalars.

This patch is part of the upstreaming effort from fir-dev branch.

Depends on D120058

Reviewed By: PeteSteinfeld

Differential Revision: https://reviews.llvm.org/D120063



Co-authored-by: Jean Perier <jperier@nvidia.com>
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>

Reported by coverity

This patch simplifies constraint handling by removing the
ConstraintListTy wrapper struct and moving the Preconditions directly
into ConstraintTy. This reduces the amount of memory needed for managing
constraints.

The only use case for ConstraintListTy was adding 2 constraints to model
ICMP_EQ conditions. But this can be handled by adding an IsEq flag. When
adding an equality constraint, we need to add the constraint and the
inverted constraint.

Clean up code that worked around this limitation.

Differential Revision: https://reviews.llvm.org/D120119

Differential Revision: https://reviews.llvm.org/D99134

Fix some typos in the amdgpu backend.

Differential Revision: https://reviews.llvm.org/D119235

This makes the called functions independent of the container type.

Same as on vgpr copies, we cannot kill the source register if it
overlaps with the destination register. Otherwise, the kill of the
source register will also count as a kill for the destination register.

Differential Revision: https://reviews.llvm.org/D120042

One of the optimizations performed in OpenMPOpt pushes globalized
variables to static shared memory. This is preferable to keeping the
runtime call in all cases, however if too many variables are pushed to
hared memory the kernel will crash. Since this is an optimization and
not something the user specified explicitly, there should be an option
to limit this optimization in those cases. This path introduces the
`-openmp-opt-shared-limit=` option to limit the amount of bytes that
will be placed in shared memory from HeapToShared.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D120079

The runtime uses thread state values to indicate when we use an ICV or
are in nested parallelism. This is done for OpenMP correctness, but it
not needed in the majority of cases. The new flag added is
`-fopenmp-assume-no-thread-state`.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D120106

If the alternate cmp instruction is a swapped predicate of the main cmp
instruction, need to generate alternate instruction, not the one with
the swapped predicate. Also, the lane with the alternate opcode should
be selected only, if the corresponding operands are not compatible.

Correctness confirmed:
https://alive2.llvm.org/ce/z/94BG66

Differential Revision: https://reviews.llvm.org/D119855

The broadcast patterns for all-true|false masks are available now.
Enable the true|fast fcmp predicate tests that use them.

Reviewed By: kaz7

Differential Revision: https://reviews.llvm.org/D119936

vmp0 is the all-ones v512i1 register and does not break down into
subregisters.

Reviewed By: kaz7

Differential Revision: https://reviews.llvm.org/D120054

This subsumes the builder and verifier.

This patch adds introduces a new kind of an lldbinit file. Unlike the
lldbinit in the home directory (useful for customizing lldb to the needs
of a particular user), or the cwd lldbinit file (useful for
project-specific settings), this file can be used to customize an entire
lldb installation to a particular environment.

The feature is enabled at build time, by setting the
LLDB_GLOBAL_INIT_DIRECTORY variable to a path to a directory which
should contain an "lldbinit" file. Lldb will then load the file at
startup, if it exists, and if automatic init loading has not been
disabled. Relative paths will be resolved (at runtime) relative to the
location of the lldb library (liblldb or LLDB.framework).

The system-wide lldbinit file will be loaded first, before any
$HOME/.lldbinit and $CWD/.lldbinit files are processed, so that those
can override any system-wide settings.

More information can be found on the RFC thread at
<https://discourse.llvm.org/t/rfc-system-...

insert is soft deprecated, so remove all references so it's less likely
to be used and can be easily removed in the future.

Differential Revision: https://reviews.llvm.org/D120021