The required target features for VMOVSR instruction (integer -> MVE
FP register copy) are HasFPRegs and UseVMOVSR (see
https://reviews.llvm.org/D60694).

The instruction selection pattern must use the same set of
requirements (i.e., HasFPRegs instead of HasVFP2). The discrepancy
can cause instruction selection failure when compiling code for
execute-only position-independence mode.

Reviewed By: dmgreen

Differential Revision: https://reviews.llvm.org/D147523

Assumed type are represented differently in the ActualArgument
class. Correctly handle them in intrinsic arg lowering.

Reviewed By: PeteSteinfeld

Differential Revision: https://reviews.llvm.org/D147487

This rarely comes up because most `vselect` are lowered with actually
avx512 mask instructions, but is an improvement in the rare cases.

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D145221

Original location of the unittests was Transforms/InstCombine/X86
which caused an issue as non-x86 targets ended up with a dependency on
X86 which isn't right.

Moved the unit tests to Target/X86 so should be good now.

Differential Revision: https://reviews.llvm.org/D145220

Since clang started emitting roundeven intrinsics in a7d6593a, they would
cause a crash in the WebAssembly backend because it did not know the roundeven
library function signatures. Fix the crash by adding the signatures.

Differential Revision: https://reviews.llvm.org/D147476

While we eventually want to remove the mapping from .Private to _Private
modules, until we do, ensure that it behaves the same for explicit
modules.

rdar://107449872

Differential Revision: https://reviews.llvm.org/D147477

Differential Revision: https://reviews.llvm.org/D147354

/data/llvm-project/flang/lib/Semantics/check-call.cpp:1234:29: error: comparison of integers of different signs: 'int' and 'size_type' (aka 'unsigned long') [-Werror,-Wsign-compare]
  CHECK(index >= 0 && index < actuals.size());
                      ~~~~~ ^ ~~~~~~~~~~~~~~
/data/llvm-project/flang/include/flang/Common/idioms.h:89:20: note: expanded from macro 'CHECK'
                   ^
1 error generated.

We added the option `VerilogBreakBetweenInstancePorts` to put ports on
separate lines in module instantiations.  We made it default to true
because style guides mostly recommend it that way for example:

https://github.com/lowRISC/style-guides/blob/master/VerilogCodingStyle.md#module-instantiation

Reviewed By: HazardyKnusperkeks

Differential Revision: https://reviews.llvm.org/D147327

Improves fix for #61683 to handle cases where the legalization concatenation isn't just 2 subops.

Verify the SlotIndexes analysis after a pass that claims to preserve it,
even if there are no further passes (apart from the verifier itself)
that would use the analysis.

Differential Revision: https://reviews.llvm.org/D129201

If the legalized type is a legal interleaved access type (i.e. there's a
supported vlseg/vsseg instruction for it), the interleaved access pass
will pick any interleaved memory op (wide load + shuffles) and lower it
into a vlseg/vsseg intrinsic.

Reviewed By: reames

Differential Revision: https://reviews.llvm.org/D146522

Given just how many arguments we pass to
preferPredicateOverEpilogue and considering this list may
grow over time I've decided to pass in a pointer to a new
TailFoldingInfo structure instead, similar to what we do
with IntrinsicCostAttributes, etc. In addition, many of the
arguments we pass in are actually available in the
LoopVectorizationLegality class so I've managed to
reduce the set of pointers that we need to pass in the
TailFoldingInfo struct.

Differential Revision: https://reviews.llvm.org/D146127

This reverts commit 5bff5237. The
bf16->f32 conversion is incorrect. This can't be on by default, if you
want this behavior make it a separate pass.

This will be necessary once we merge with combineVectorSizedSetCCEquality and we need to support KORTEST handling for 128/256-bit comparisons

Implements the PowerPC mtfsf and mtfsfi intrinsics as well as introduces semantic error checking code for PowerPC intrinsics

Reviewed By: klausler

Differential Revision: https://reviews.llvm.org/D144876

Show failure for other vector size concatenations

For functions that use DWARF encodings, their compact unwind entry will
contain a hint about the offset of their DWARF entry from the start of
the `__eh_frame` section. The encoding only has 3 bytes to encode this
hint.

Previously, I neglected to check for overflow (and didn't realize that
the value was merely a hint without needing to be exact.) So for large
`__eh_frame` sections, the hint would overflow and cause the compact
unwind MODE flag to be corrupted, leading to uncaught exceptions at
runtime.

This diff fixes things by encoding zero as the hint for offsets that are
too large. The unwinder will start a linear search at the hint location
for the matching CFI record. The only requirement is that the hint
points to a valid CFI record start, and the start of the section is
always the start of a CFI record (in well-formed programs).

I'm not adding a test for this because generating the test inputs takes
a bit too much time. However, I have been testing locally with this lit
file, which takes about 15s to run on my machine:

```
# RUN: rm -rf %t; mkdir %t
# RUN: llvm-mc -filetype=obj -triple=x86_64-apple-macos11.0 %s -o %t/test.o
# RUN: %lld -dylib -lSystem %t/test.o -o %t/test

.subsections_via_symbols
.text
.p2align 2

_f:
  .cfi_startproc
.rept 0x7fffff
  .cfi_escape 0x2e, 0x10
.endr
  ret
  .cfi_endproc

_g:
  .cfi_startproc
  .cfi_escape 0x2e, 0x10
  ret
  .cfi_endproc
```

Reviewed By: #lld-macho, smeenai

Differential Revision: https://reviews.llvm.org/D147505

Leftover from eb435b7f

Differential Revision: https://reviews.llvm.org/D147461

Fixes #61683

Add is_in_system_header property for Location class.

Corresponding unit test was also added.

Differential Revision: https://reviews.llvm.org/D147414

They are no longer used, which leads to the -Wunused-function warning.

Reviewed By: uweigand

Differential Revision: https://reviews.llvm.org/D147521

Test is related to vectors so probably belongs in the rvv directory

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D147412

This reverts commit 18c594c1.
Build bots broke

1. Remove redundant definition of constructor
2. Move the array in .inc to .def file
3. Add a licence for the .def file

The default code can detect what width of atomic instructions are supported
based on the targeted architecture profile, version etc so there's no need to
hard-code 64 on Darwin targets (especially as it's wrong in most M-class
cases).

/data/llvm-project/llvm/lib/Target/SystemZ/Disassembler/SystemZDisassembler.cpp:209:21: error: unused function 'decodeU6ImmOperand' [-Werror,-Wunused-function]
static DecodeStatus decodeU6ImmOperand(MCInst &Inst, uint64_t Imm,
                    ^
1 error generated.

Differential Revision: https://reviews.llvm.org/D145622

The build of scudo was failing on armv7l, with undefined references
to unwinder symbols, such as __aeabi_unwind_cpp_pr0. These are
needed by RTGwpAsan and thus, on ARM, scudo must also be linked
against an unwind library.

The cmake command that caused the build failure was:

cmake --fresh -S "$PWD/llvm/" -B "$PWD/build/" -G Ninja \
  -DCMAKE_INSTALL_PREFIX="$PWD/install" \
  -DCMAKE_BUILD_TYPE=Release \
  -DLLVM_ENABLE_PROJECTS="clang;lld;lldb;clang-tools-extra;polly" \
  -DLLVM_ENABLE_RUNTIMES="compiler-rt;libcxx;libcxxabi;libunwind" \
  -DLLVM_TOOLCHAIN_TOOLS="llvm-ar;llvm-ranlib;llvm-objdump;\
llvm-rc;llvm-cvtres;llvm-nm;llvm-strings;llvm-readobj;\
llvm-dlltool;llvm-pdbutil;llvm-objcopy;llvm-strip;llvm-cov;\
llvm-profdata;llvm-addr2line;llvm-symbolizer;llvm-windres;llvm-ml;\
llvm-readelf;llvm-size" \
  -DLLVM_INSTALL_BINUTILS_SYMLINKS=OFF -DLLVM_PARALLEL_LINK_JOBS=1

Fixes #60115

Reviewed By: hctim

Differential Revision: https://reviews.llvm.org/D142888

There does not seem to be any purpose to allocating this object, which
is Metadata, so LLVM will simply leak it until the context is destroyed
(the subprogram metadata it contains are distinct, so there is little
chance of it being reused later). This should not have a functional
change.

This seems to be left over from an earlier design, when it used to call
```
CUNode->replaceSubprograms(SPs.get());
```
here also. (removed in http://reviews.llvm.org/D19034)

Reviewed By: aprantl

Differential Revision: https://reviews.llvm.org/D147469

For all RotateSelect* instructions, PoP says:

    Bits 0-1 of the I5 field (bits 32-33 of the instruction) are
    ignored.

LLVM, however, completely prohibits using them, e.g.:

    error: invalid operand for instruction
    asm("rxsbg %[r1],%[r2],177,43,228\n"

Lift this unnecessary restriction.

Reviewed By: uweigand

Differential Revision: https://reviews.llvm.org/D146185

Retaining bitcasts to keep test intent.

addrsig_sym is no longer emitted for metedata_f1, but this
improvement is ultimately inconsequential, as it will get dropped
at a later stage anyway.

Reviewed By: pengfei

Differential Revision: https://reviews.llvm.org/D147301

The original typed pointer version of the tests (matrix-intrinsics.ll)
is now obsolete. Replace matrix-intrinsics.ll with
matrix-intrinsics-opaque-ptrs.ll.

Certain standard library functions (e.g. std::move) are also implemented
as builtins. This patch moves filtering logic to the symbol->header
mapping phase to rather generate these references without any providers
only when we don't have a mapping.
That way we can also map them to header names mentioned in the builtin
mappings.

Differential Revision: https://reviews.llvm.org/D147449