We can only assume bit[1] == zero if its the only demanded bit or the source is not undef/poison

If AMDGPUAnnotateUniformValues finds a load from a uniform pointer with
no potentially clobbering stores between the kernel entry point and the
load instruction, it adds noclobber metadata to the *address*. This is
unsafe because it can get applied to other loads in the same which do
have aliasing stores.

Differential Revision: https://reviews.llvm.org/D118458

As raised by @efriedma on D117995 - the source must not be undef/poison to demand any bits in mul(x,x) other than bit[1]

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

This avoids various cases where StructurizeCFG would otherwise insert an
xor i1 instruction, and it since it generally runs late in the pipeline,
instcombine does not clean up the xor-of-cmp pattern.

Differential Revision: https://reviews.llvm.org/D118478

This patches fixes the visibility and linkage information of symbols
referring to IR globals.

Emission of external declarations is now done in the first execution
of emitConstantPool rather than in emitLinkage (and a few other
places). This is the point where we have already gathered information
about used symbols (by running the MC Lower PrePass) and not yet
started emitting any functions so that any declarations that need to
be emitted are done so at the top of the file before any functions.

This changes the order of a few directives in the final asm file which
required an update to a few tests.

Reviewed By: sbc100

Differential Revision: https://reviews.llvm.org/D118122

As pointed out in https://reviews.llvm.org/D93401, some methods in the
Flang driver are named inconsistently. The driver strives to follow
Flang's C++ style [1] and this patch updates these methods accordingly.

[1]
https://github.com/llvm/llvm-project/blob/main/flang/docs/C%2B%2Bstyle.md

Differential Revision: https://reviews.llvm.org/D118381

Differential Revision: https://reviews.llvm.org/D117568

Currently, -fdeclare-opencl-builtins always adds the generic address
space overloads of e.g. the vload builtin functions in OpenCL 3.0
mode, even when the generic address space feature is disabled.

Guard the generic address space overloads by the
`__opencl_c_generic_address_space` feature instead of by OpenCL
version.

Guard the private, global, and local overloads using the internal
`__opencl_c_named_address_space_builtins` feature.

Differential Revision: https://reviews.llvm.org/D107769

The current cost-model overestimates the cost of vector compares &
selects for ordered floating point compares. This patch fixes that by
extending the existing logic for integer predicates.

Reviewed By: dmgreen

Differential Revision: https://reviews.llvm.org/D118256

This reduces the dependencies of the MLIRVector target and makes the dialect consistent with other dialects.

Differential Revision: https://reviews.llvm.org/D118533

Do not warn on reserved identifiers resulting from expansion of system macros.
Also properly test -Wreserved-identifier wrt. system headers.

Should fix #49592

Differential Revision: https://reviews.llvm.org/D118532

Based on the output of include-what you-use.

Most notably, llvm/Remarks/Remark.h is no longer automatically included by
llvm/Remarks/RemarkParser.h, so client code may need to include explicitly.

clang++ -E  -Iinclude -I../llvm/include ../llvm/lib/Remarks/*.cpp -std=c++14 -fno-rtti -fno-exceptions | wc -l
before: 770253
after:  759347

Related discourse thread: https://llvm.discourse.group/t/include-what-you-use-include-cleanup

Differential Revision: https://reviews.llvm.org/D118506

Based on the output of include-what-you-use.
It's an utility directory, so no much impact on other code areas.

clang++ -E  -Iinclude -I../llvm/include ../llvm/utils/TableGen/*.cpp -std=c++14 -fno-rtti -fno-exceptions | wc -l
before: 4327274
after:  4316190

Related discourse thread: https://llvm.discourse.group/t/include-what-you-use-include-cleanup
Differential Revision: https://reviews.llvm.org/D118466

Display notes for a possible call chain if an unsafe function is found to be
called (maybe indirectly) from a signal handler.
The call chain displayed this way includes probably not the first calls of
the functions, but it is a valid possible (in non path-sensitive way) one.

Reviewed By: aaron.ballman

Differential Revision: https://reviews.llvm.org/D118224

This reverts commit 36892727 which
breaks the build when LLVM_INSTALL_TOOLCHAIN_ONLY is enabled with:

  CMake Error at cmake/modules/AddLLVM.cmake:683 (add_dependencies):
  The dependency target "clang-tidy-headers" of target "CTTestTidyModule"
  does not exist.

The pruning cloner already tries to remove unreachable blocks. The
original cloning process will simplify instructions and constant
terminators, and only clone blocks that are reachable at that point.
However, phi nodes can only be simplified after everything has been
cloned. For that reason, additional blocks may become unreachable
after phi simplification.

The code does try to handle this as well, but only removes blocks
that don't have predecessors. It misses unreachable cycles. This
can cause issues if SEH exception handling code is part of an
unreachable cycle, as the inliner is not prepared to deal with that.

This patch instead performs an explicit scan for reachable blocks,
and drops everything else.

Fixes https://github.com/llvm/llvm-project/issues/53206.

Differential Revision: https://reviews.llvm.org/D118449

masked.atomicrmw.*.i32 intrinsics access an i32 (and then possibly
mask it), so hardcode MVT::i32 as the access type here, rather than
determining it from the pointer element type.

Differential Revision: https://reviews.llvm.org/D118336

Following Sanjay's proposal from discussion in D118317, this patch
generalizes and-reduce handling to fold the following pattern
```
  icmp ne (bitcast(icmp ne (lhs, rhs)), 0)
```
into
```
  icmp ne (bitcast(lhs), bitcast(rhs))
```

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

Differential Revision: https://reviews.llvm.org/D118431
Reviewed By: lebedev.ri

Support affine.load/store ops in fold-memref-subview ops pass. The
existing pass just "inlines" the subview operation on load/stores by
inserting affine.apply ops in front of the memref load/store ops: this
is by design always consistent with the semantics on affine.load/store
ops and the same would work even more naturally/intuitively with the
latter.

Differential Revision: https://reviews.llvm.org/D118565

Reviewed By: yota9

Differential Revision: https://reviews.llvm.org/D118556

This is a slight change because I'm using the ANY_EXTEND result
instead of the original operand, but getNode should constant fold.

While there, add a comment about why the code specifically checks
for a ConstantSDNode.

Update SCF pass cmd line names to prefix `scf`. This is consistent with
guidelines/convention on how to name dialect passes. This also avoids
ambiguity on the context given the multiple `for` operations in the
tree.

NFC.

Differential Revision: https://reviews.llvm.org/D118564

My x86-64 lld executable is actually smaller.

Make sure we cover the encodings use for zext.h and other encodings
not used for zext.h.

Based on the existing naming "only" tests are used for rv32 instructions
that don't exist in rv64. rv32 tests without "only" are for instructions
that are in both rv32 and rv64. The rv64 tests are for instructions
that are only in rv64.

Both of these test files have instruction encodings that are only
valid in rv64 so they can be the same file.

"pack t0, t1, zero" disassembles to "pack t0, t1, zero" with Zbkb
not "zext.h t0, t1"

Part of the test was using a CHECK prefix that doesn't appear on
the RUN line.

Identified with readability-const-return-type.

[mlgo][regalloc] Fix -Wunused-variable in -DLLVM_ENABLE_ASSERTIONS=off builds after a8a7bf92

Previously an InputSectionBase is dead (`partition==0`) by default.
SyntheticSection calls markLive and BssSection overrides that with markDead.

It is more natural to make InputSectionBase live by default and let
--gc-sections mark InputSectionBase dead.

When linking a Release build of clang:

* --no-gc-sections:, the removed `inputSections` loop decreases markLive time from 4ms to 1ms.
* --gc-sections: the extra `inputSections` loop increases markLive time from 0.181296s to 0.188526s.
  This is as of we lose the removing one `inputSections` loop optimization (4374824c).
  I believe the loss can be mitigated if we refactor markLive.

If AllocationOrder has less than 32 elements, we were treating the extra
positions as if they were valid. This was detected by a subsequent
assert. The fix also tightens the asserts.