[LegalizeTypes][RISCV][WebAssembly] Expand ABS in PromoteIntRes_ABS if it will expand to sra+xor+sub later.

If we promote the ABS and then Expand in LegalizeDAG, then both the
sra and the xor will have their inputs sign extended. This generates
extra code on RISCV which lacks an i8 or i16 sign extend instructon.
If we expand during type legalization, then only the sra will get its
input sign extended. RISCV is able to combine this with the sra by
doing a shift left followed by an sra.

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D121664

[DAGCombiner][RISCV] Adjust (aext (and (trunc x), cst)) -> (and x, cst) to sext cst based on target preference

RISCV strong prefers i32 values be sign extended to i64. This combine
was always zero extending the constant using APInt methods.

This adjusts the code so that it calls getNode using ISD::ANY_EXTEND instead.
getNode will call TLI.isSExtCheaperThanZExt to decide how to handle
the constant.

Tests were copied from D121598 where I noticed that we were creating
constants that were hard to materialize.

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D121650

This reverts commit dddf4ce0. It breaks
a couple of tests on macos.

This code handles fixed vector SPLAT_VECTOR, but is never called in
any tests.

We only form fixed vector splat vectors for vXi64 on RV32 as part
of DAGCombine. This will be type legalized to SPLAT_VECTOR_PARTS.
So the Custom handling for SPLAT_VECTOR is never needed.

This patch makes SPLAT_VECTOR for vXi64 'Legal' on RV32 so that
DAGCombine will create it, but there's no need for Custom handler.
It will still be type legalized to SPLAT_VECTOR_PARTS.

Reviewed By: frasercrmck

Differential Revision: https://reviews.llvm.org/D121673

Reviewed By: sivachandra

Differential Revision: https://reviews.llvm.org/D121656

Terminators are handled specially in the transfer functions so we need an
additional check on whether the analysis has disabled built-in transfer
functions.

Differential Revision: https://reviews.llvm.org/D121694

This can be viewed as swapping the select arms:
https://alive2.llvm.org/ce/z/jUvFMJ
...so we don't have the 'nsz' problem with the more general fold.

This unlocks other folds for the motivating fabs example.
This was discussed in issue #38828.

Prior to this patch, there was no distinction between tests that check
basic assertions and tests that check full-fledged iterator debugging
assertions. Both were disabled when support for the debug mode is not
provided in the dylib, which is stronger than it needs to be.

Furthermore, all of the tests using "debug_macros.h" that contain more
than one assertion in them were broken -- any code after the first
assertion would never be executed.

This patch refactors all of our assertion-related tests to:
1. Be enabled whenever they can, i.e. basic assertions tests are run
   even when the debug mode is disabled.
2. Use the superior `check_assertion.h` (previously `debug_mode_helper.h`)
   instead of `debug_macros.h`, which allows multiple assertions in the
   same program.
3. Coalesce some tests into the same file to make them more readable.
4. Use consistent naming for test files -- no more db{1,2,3,...,10} tests.

This is a large but mostly mechanical patch.

Differential Revision: https://reviews.llvm.org/D121462

ISel for experimental.vp.strided.load|store for v256.32 types via
lowering to vvp_load|store SDNodes.

Reviewed By: kaz7

Differential Revision: https://reviews.llvm.org/D121616

When using `--convert-func-to-llvm=emit-c-wrappers` the attribute arguments of the wrapper would not be created correctly in some cases.
This patch fixes that and introduces a set of tests for (hopefully) all corner cases.

See https://github.com/llvm/llvm-project/issues/53503

Author: Sam Carroll <sam.carroll@lmns.com>
Co-Author: Laszlo Kindrat <laszlo.kindrat@lmns.com>

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D119895

Implement expm1f function that is correctly rounded for all rounding modes.  This is based on expf implementation.

From exhaustive testings, using expf implementation, and subtract 1.0 before rounding the final result to single precision
gives correctly rounded results for all |x| > 2^-4 with 1 exception.  When |x| < 2^-25, we use x + x^2 (implemented with a
single fma).  And for 2^-25 <= |x| <= 2^-4, we use a single degree-8 minimax polynomial generated by Sollya.

Reviewed By: sivachandra, zimmermann6

Differential Revision: https://reviews.llvm.org/D121574

[InstCombine] Add general constant support to eq/ne icmp(add(X,C1),add(Y,C2)) -> icmp(add(X,C1-C2),Y) fold

A further extension for Issue #32161

For eq/ne comparisons - the sign mismatch and bounds constraints are redundant, so if the that fold fails, fallback and just fold the constants directly.

https://alive2.llvm.org/ce/z/cdodNQ

The loop rotation test change looks mostly benign - the backend doesn't seem to suffer? https://gcc.godbolt.org/z/dErMY78To

Differential Revision: https://reviews.llvm.org/D121551

Perform the mask inside parentheses before applying the offset

Reviewed By: amilendra

Differential Revision: https://reviews.llvm.org/D121206

Patch adds a new operation for the SIMD construct.  The op is designed to be very similar to the existing `wsloop` operation, so that the `CanonicalLoopInfo` of `OpenMPIRBuilder` can be used.

Reviewed By: shraiysh

Differential Revision: https://reviews.llvm.org/D118065

Fix darwin interface test after D121464. asan_rtl_x86_64.S is not
available on Darwin.

Reviewed By: kstoimenov

Differential Revision: https://reviews.llvm.org/D121636

Revert "Load pass plugins during option processing, so that plugin options are registered and live."

This reverts commit 5e8700ce.

Revert rG9c542a5a "Lower `@llvm.global_dtors` using `__cxa_atexit` on MachO"

Mane of the build bots are complaining: Unknown command line argument '-lower-global-dtors'

Avoids polluting the namespace of all files including the header.

The decision which categories are relevant for a particular test run
happen very early in the test setup process. They use the SBPlatform
object to determine which categories should be skipped. The platform
object created for this purpose transcends individual test runs.

This setup is not compatible with the direction discussed in
<https://discourse.llvm.org/t/multiple-platforms-with-the-same-name/59594>
-- when platform objects are tied to a specific (SB)Debugger, they need
to be created alongside it, which currently happens in the test setUp
method.

This patch is the first step in that direction -- it rewrites the
category skipping logic to avoid depending on a global SBPlatform
object. Fortunately, the skipping logic is fairly simple (and I believe
it outght to stay that way) and mainly consists of comparing the
platform name against some hardcoded lists. This patch bases this
comparison on the platform name instead of the os part of the triple (as
reported by the platform).

Differential Revision: https://reviews.llvm.org/D121605

@mul_may_overflow_var_nonzero_minabsvarindex_one_index shows BasicAA
incorrectly determining noalias for (%gep.917, i8* %gep.idx).
If %v == 10581764700698480926, %idx == 917 and the GEPs alias.
https://alive2.llvm.org/ce/z/yzDgnn

This improves the modularity of the bufferization.

From now on, all ops that do not implement BufferizableOpInterface are considered hoisting barriers. Previously, all ops that do not implement the interface were not considered barriers and such ops had to be marked as barriers explicitly. This was unsafe because we could've hoisted across unknown ops where it was not safe to hoist.

As a side effect, this allows for cleaning up AffineBufferizableOpInterfaceImpl. This build unit no longer needed and can be deleted.

Differential Revision: https://reviews.llvm.org/D121519

Fixes https://github.com/llvm/llvm-project/issues/54257.

Reviewed By: MyDeveloperDay, HazardyKnusperkeks, owenpan

Differential Revision: https://reviews.llvm.org/D121456

[X86] combineSelect - canonicalize (vXi1 bitcast(iX Cond)) with combineToExtendBoolVectorInReg before legalization

This replaces the attempt in 20af71f8 to use combineToExtendBoolVectorInReg to create X86ISD::BLENDV masks directly, instead we use it to canonicalize the iX bitcast to a sign-extended mask and then truncate it back to vXi1 prior to legalization breaking it apart.

Fixes #53760

Fixes https://github.com/llvm/llvm-project/issues/54374.

Also update test to not use branch on undef.

This method just calls LazyValueInfoImpl::clear

Fixes https://github.com/llvm/llvm-project/issues/54245.

Reviewed By: MyDeveloperDay, HazardyKnusperkeks, owenpan

Differential Revision: https://reviews.llvm.org/D121584

Introduce OpenCL-style global_id op and corresponding spirv lowering.

Differential Revision: https://reviews.llvm.org/D121548

Differential Revision: https://reviews.llvm.org/D121601

This makes bufferization more modular. This is in preparation of future refactorings.

Differential Revision: https://reviews.llvm.org/D121362

Hardcode the function type as ParallelTask, which is the guaranteed
pointee type of this runtime function argument (if pointee types
exist). The elimination of the callee bitcast is left for InstCombine.

Differential Revision: https://reviews.llvm.org/D120885

Differential Revision: https://reviews.llvm.org/D121361

`semantics::IsSaved()` was not applying -Msave/-fno-automatic for main programs.
This caused issues since lowering relies on it to allocate static
variables. This did not match nvfortran/gfortran behaviors where
-fno-automatic/-Msave control the static allocation of scalars in
main programs.

Some program may rely on main program scalars to be statically allocated in
bss (and therefore initialized to zero) with -Msave/-fno-automatic
flags.

Differential Revision: https://reviews.llvm.org/D121603

Also add a TODO to switch to a custom walk instead of the GreedyPatternRewriter, which should be more efficient. (The bufferization pattern is guaranteed to apply only a single time for every op, so a simple walk should suffice.)

We currently specify a top-to-bottom walk order. This is important because other walk orders could introduce additional casts and/or buffer copies. These canonicalize away again, but it is more efficient to never generate them in the first place.

Note: A few of these canonicalizations are not yet implemented.

Differential Revision: https://reviews.llvm.org/D121518