Created using spr 1.3.6-beta.1

[skip ci]

The opcode will be determined by LMUL.

Reviewers: preames, lukel97, topperc

Reviewed By: lukel97, topperc

Pull Request: https://github.com/llvm/llvm-project/pull/84448

We have a hidden option to disable it, but I'd like to make it a
tune feature.

For some implementations, instructions with W suffix would be less
costly as they only perform on 32 bits data. Though we may lose some
chances to compress.

Fixes: #86305

Support Fastmath flag to convert trigonometric ops in the complex
dialect.

See:
https://discourse.llvm.org/t/rfc-fastmath-flags-support-in-complex-dialect/71981

Add missing constant propogation folder for [S|U]LessThan[Equal].

Implement additional folding when the operands are equal for all ops.

Allows for constant folding in the IndexToSPIRV pass.

Part of work #70704

Fixes #86109.

NFC.

`relaIplt` was added so that IRELATIVE relocations are placed at the end
of .rela.dyn (since https://reviews.llvm.org/D65651) or .rela.plt
(--pack-dyn-relocs=android[+relr]). Unfortunately, handling `relaIplt`
requires special cases all over the code base. We can extend
partitionRels/computeRels to partition both RELATIVE and IRELATIVE
relocations, rendering `relaIplt` unneeded.

The change allows IRELATIVE relocations in the DT_ANDROID_REL[A] table
(untested?!), which may be processed before other types of relocations.
This seems acceptable for Bionic's DEFINE_IFUNC_FOR use cases.

In addition, this change simplies changing .rel[a].dyn to a compact
relocation format (CREL).

SHF_INFO_LINK is removed from .rel[a].dyn with IRELATIVE relocations.
(See https://reviews.llvm.org/D89828).

The value produced by CalculateTripCountMinusVF VPInstructions is
independent of the part. Only compute it for part 0 and use that for
other parts.

Some tests in `condition_variable_any` use two `shared_lock` to guard,
which does not work.
The fix is to make the writer to use `unique_lock`

Summary:
It's possible for the OpenMP offloading plugins to be build before
tablegen is run despite the fact that we rely on it. Simply make it
depend on it currently.

This commit adds section start and stop symbol handling to ELF/aarch64, and
fixes the section symbol prefixes (using `__start_` and `__stop_`, rather than
`__start` and `__end`). It also adds a testcase for handling of these symbols.

Skip checking for references to variable in unevaluated context, like
decltype, static_assert and so on.

Closes #85838

The existing implementation didn't handle when the input text section
was some offset from the output section.

This resulted in an assert in relaxGot() with an lld built with asserts
for some large binaries, or even worse, a silently broken binary with an
lld without asserts.

Fixes #86164

Enabled vector coverage as well: i686+SSE2 and x64_64+AVX

Should improve test quality for #85722

Yes, 64-bit shifts are intended.

Reverting #85188 with follow up patches.

This reverts commit 362d2636.
This reverts commit c9bdeabd.
This reverts commit 6bc6e1ac.
This reverts commit 01fa550f.
This reverts commit ddcbab37.

Fix #61881

This commit adds the `BufferViewFlowOpInterface` to the bufferization
dialect. This interface can be implemented by ops that operate on
buffers to indicate that a buffer op result and/or region entry block
argument may be the same buffer as a buffer operand (or a view thereof).
This interface is queried by the `BufferViewFlowAnalysis`.

The new interface has two interface methods:
* `populateDependencies`: Implementations use the provided callback to
declare dependencies between operands and op results/region entry block
arguments. E.g., for `%r = arith.select %c, %m1, %m2 : memref<5xf32>`,
the interface implementation should declare two dependencies: %m1 -> %r
and %m2 -> %r.
* `mayBeTerminalBuffer`: An SSA value is a terminal buffer if the buffer
view flow analysis stops at the specified value. E.g., because the value
is a newly allocated buffer or because no further information is
available about the origin of the buffer.

Ops that implement the `RegionBranchOpInterface` or `BranchOpInterface`
do not have to implement the `BufferViewFlowOpInterface`. The buffer
dependencies can be inferred from those two interfaces.

This commit makes the `BufferViewFlowAnalysis` more accurate. For
unknown ops, it conservatively used to declare all combinations of
operands and op results/region entry block arguments as dependencies
(false positives). This is no longer the case. While the analysis is
still a "maybe" analysis with false positives (e.g., when analyzing ops
such as `arith.select` or `scf.if` where the taken branch is not known
at compile time), results and region entry block arguments of unknown
ops are now marked as terminal buffers.

This commit addresses a TODO in `BufferViewFlowAnalysis.cpp`:
```
// TODO: We should have an op interface instead of a hard-coded list of
// interfaces/ops.
```
It is no longer needed to hard-code ops.

The indexed MemProf file has a huge amount of redundancy.  In a large
internal application, 82% of call stacks, stored in
IndexedAllocationInfo::CallStack, are duplicates.

We should work toward deduplicating call stacks by referring to them
with unique IDs with actual call stacks stored in a separate data
structure, much like we refer to memprof::Frame with memprof::FrameId.

At the same time, we need to facilitate a graceful transition from the
current version of the MemProf format to the next.  We should be able
to read (but not write) the current version of the MemProf file even
after we move onto the next one.

With those goals in mind, I propose to have an integer ID next to
CallStack in IndexedAllocationInfo to refer to a call stack in a
succinct manner.  We'll gradually increase the areas of the compiler
where IDs and call stacks have one-to-one correspondence and
eventually remove the existing CallStack field.

This patch adds call stack ID, named CSId, to IndexedAllocationInfo
and teaches the raw profile reader to compute unique call stack IDs
and store them in the new field.  It does not introduce any user of
the call stack IDs yet, except in verifyFunctionProfileData.

Only check assertions that were meant to apply to the normal case of non-splat vector SREM expansion when we aren't hitting the special case. (#86238)

Fixes https://github.com/llvm/llvm-project/issues/84830
Introduced in https://github.com/llvm/llvm-project/pull/82706

Attach branch counters to YAML profile, covering intra-function control
flow.

Depends on: https://github.com/llvm/llvm-project/pull/86353

Test Plan: Updated bolt/test/X86/bolt-address-translation-yaml.test

Reviewers: rafaelauler, dcci, ayermolo, maksfb

Reviewed By: rafaelauler

Pull Request: https://github.com/llvm/llvm-project/pull/76911

Hide the implementations of `FuncHashes` and `BBHashMap` classes,
getting rid of `at` accessors that could throw an exception.

Test Plan: NFC

Reviewers: ayermolo, maksfb, dcci, rafaelauler

Reviewed By: rafaelauler

Pull Request: https://github.com/llvm/llvm-project/pull/86353

Move the code adding top-level cmake/Modules directory to
CMAKE_MODULE_PATH prior to including `GetDarwinLinkerVersion`, in order
to fix standalone builds.

Fixes a regression introduced by
3bc71c2a.

Instead of keeping a mapping of Inst->VPValues (of their corresponding
recipes) in VPlan's Value2VPValue mapping, keep it in VPRecipeBuilder
instead. After recently replacing the last user of this mapping after
initial construction, this mapping is only needed for recipe
construction (to map IR operands to VPValue operands).

By moving the mapping, VPlan's VPValue tracking can be simplified and
limited only to live-ins. It also allows removing disableValue2VPValue
and associated machinery & asserts.

PR: https://github.com/llvm/llvm-project/pull/84464

MIPS is different and should better off use separate code.

Commit f44db24e (2015) enabled this
simplication.

If any return from overwriteChangedFiles is true some fixes were not
applied.

Use `LINK_COMPONENTS` parameter of `add_llvm_library` rather than
passing LLVM components directly to `target_link_libraries`, in order to
ensure that LLVM dylib is linked correctly when used. Otherwise, CMake
insists on linking to static libraries that aren't present on
distributions doing pure dylib installs, such as Gentoo.

This fixes a regression introduced
in dcbddc25.

```
---------------------------------------------------
Benchmark                           old         new
---------------------------------------------------
bm_mismatch<char>/1           0.835 ns      2.37 ns
bm_mismatch<char>/2            1.44 ns      2.60 ns
bm_mismatch<char>/3            2.06 ns      2.83 ns
bm_mismatch<char>/4            2.60 ns      3.29 ns
bm_mismatch<char>/5            3.15 ns      3.77 ns
bm_mismatch<char>/6            3.82 ns      4.17 ns
bm_mismatch<char>/7            4.29 ns      4.52 ns
bm_mismatch<char>/8            4.78 ns      4.86 ns
bm_mismatch<char>/16           9.06 ns      7.54 ns
bm_mismatch<char>/64           31.7 ns      19.1 ns
bm_mismatch<char>/512           249 ns      8.16 ns
bm_mismatch<char>/4096         1956 ns      44.2 ns
bm_mismatch<char>/32768       15498 ns       501 ns
bm_mismatch<char>/262144     123965 ns      4479 ns
bm_mismatch<char>/1048576    495668 ns     21306 ns
bm_mismatch<short>/1          0.710 ns      2.12 ns
bm_mismatch<short>/2           1.03 ns      2.66 ns
bm_mismatch<short>/3           1.29 ns      3.56 ns
bm_mismatch<short>/4           1.68 ns      4.29 ns
bm_mismatch<short>/5           1.96 ns      5.18 ns
bm_mismatch<short>/6           2.59 ns      5.91 ns
bm_mismatch<short>/7           2.86 ns      6.63 ns
bm_mismatch<short>/8           3.19 ns      7.33 ns
bm_mismatch<short>/16          5.48 ns      13.0 ns
bm_mismatch<short>/64          16.6 ns      4.06 ns
bm_mismatch<short>/512          130 ns      13.8 ns
bm_mismatch<short>/4096         985 ns      93.8 ns
bm_mismatch<short>/32768       7846 ns      1002 ns
bm_mismatch<short>/262144     63217 ns     10637 ns
bm_mismatch<short>/1048576   251782 ns     42471 ns
bm_mismatch<int>/1            0.716 ns      1.91 ns
bm_mismatch<int>/2             1.21 ns      2.49 ns
bm_mismatch<int>/3             1.38 ns      3.46 ns
bm_mismatch<int>/4             1.71 ns      4.04 ns
bm_mismatch<int>/5             2.00 ns      4.98 ns
bm_mismatch<int>/6             2.43 ns      5.67 ns
bm_mismatch<int>/7             3.05 ns      6.38 ns
bm_mismatch<int>/8             3.22 ns      7.09 ns
bm_mismatch<int>/16            5.18 ns      12.8 ns
bm_mismatch<int>/64            16.6 ns      5.28 ns
bm_mismatch<int>/512            129 ns      25.2 ns
bm_mismatch<int>/4096          1009 ns       201 ns
bm_mismatch<int>/32768         7776 ns      2144 ns
bm_mismatch<int>/262144       62371 ns     20551 ns
bm_mismatch<int>/1048576     254750 ns     90097 ns
```

Remove getSizeOrUnknown call when MachineMemOperand is created.  For Scalable
TypeSize, the MemoryType created becomes a scalable_vector.

2 MMOs that have scalable memory access can then use the updated BasicAA that
understands scalable LocationSize.

Original Patch by Harvin Iriawan
Co-authored-by: David Green <david.green@arm.com>