This was unconditionally folding this to the source operand, even if the access was out of bounds. Use undef instead of the extract is not the first element.

This helps with some cases where 3-vectors are legalized and avoids processing the 4th component.

Original Patch by: arsenm (Matt Arsenault)

Differential Revision: https://reviews.llvm.org/D51589

Reviewers: gchatelet

Subscribers: tschuett, mstojanovic, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D73083

Matches ISD::SRA + ISD::SRL variants.

The INPUT_SECTION_FLAGS linker script command is used to constrain the
section pattern matching to sections that match certain combinations of
flags.

There are two ways to express the constraint.
withFlags: Section must have these flags.
withoutFlags: Section must not have these flags.

The syntax of the command is:
INPUT_SECTION_FLAGS '(' sect_flag_list ')'
sect_flag_list: NAME
| sect_flag_list '&' NAME

Where NAME matches a section flag name such as SHF_EXECINSTR, or the
integer value of a section flag. If the first character of NAME is ! then
it means must not contain flag.

We do not support the rare case of { INPUT_SECTION_FLAGS(flags) filespec }
where filespec has no input section description like (.text).

As an example from the ld man page:
SECTIONS {
  .text : { INPUT_SECTION_FLAGS (SHF_MERGE & SHF_STRINGS) *(.text) }
  .text2 :  { INPUT_SECTION_FLAGS (!SHF_WRITE) *(.text) }
}
.text will match sections called .text that have both the SHF_MERGE and
SHF_STRINGS flag.
.text2 will match sections called .text that don't have the SHF_WRITE flag.

The flag names accepted are the generic to all targets and SHF_ARM_PURECODE
as it is very useful to filter all the pure code sections into a single
program header that can be marked execute never.

fixes PR44265

Differential Revision: https://reviews.llvm.org/D72756

The test in the origin patch did not create a __debug_str section.
An UBSan check triggered when the corresponding pointer was dereferenced.

Differential Revision: https://reviews.llvm.org/D72695

This reapplies fcc08aa8
which was reverted in b16f82ad.

Extends the gather/scatter pass in MVEGatherScatterLowering.cpp to
enable the transformation of masked scatters into calls to MVE's masked
scatter intrinsic.

Differential Revision: https://reviews.llvm.org/D72856

Summary:
Add a `llvm.cmpxchg` op as a counterpart to LLVM IR's `cmpxchg` instruction.
Note that the `weak`, `volatile`, and `syncscope` attributes are not yet supported.

This will be useful for upcoming parallel versions of affine.for and generally
for reduction-like semantics (especially for reductions that can't make use
of `atomicrmw`, e.g. `fmax`).

Reviewers: ftynse, nicolasvasilache

Reviewed By: ftynse

Subscribers: merge_guards_bot, jfb, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D72995

This reverts commit 0dc6c249.

The commit is reported to cause a regression in piglit/bin/glsl-vs-loop for
Mesa.

Reviewers: lenary, asb

Reviewed By: lenary

Tags: #clang

Differential Revision: https://reviews.llvm.org/D72755

This change broke the UBSan buildbots. More information available in the
original Phabricator review: https://reviews.llvm.org/D72695

This reverts commit fcc08aa8.

When triggering rename of the class name in the code with explicit destructor
calls, rename fails. Consider the following piece of code:

```
class Foo;

...

Foo f;
f.~/*...*/Foo();
```

`findExplicitReferences` will report two `ReferenceLoc` for destructor call:
one is comming from `MemberExpr` (i.e. destructor call itself) and would point
to the tilde:

```
f.~/*...*/Foo();
  ^
```

And the second one is pointing to the typename and is coming from `TypeLoc`.

```
f.~/*...*/Foo();
          ^
```

This causes rename to produce incorrect textual replacements. This patch
updates `MemberExpr` handler to detect destructor calls and prevents it
from reporting a duplicate reference.

Resolves: https://github.com/clangd/clangd/issues/236

Reviewers: kadircet, hokein

Differential Revision: https://reviews.llvm.org/D72638

Summary: This patch adds support for setting SHF_EXCLUDE flag for ELF sections.

Reviewers: jhenderson, grimar, MaskRay, mstorsjo, espindola, alexshap, rupprecht

Reviewed By: jhenderson, MaskRay

Subscribers: emaste, abrachet, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D72128

Summary:
Unlike R_RISCV_RELAX, which is a linker hint, R_RISCV_ALIGN requires the
support of the linker even when ignoring all R_RISCV_RELAX relocations.
This is because the compiler emits as many NOPs as may be required for
the requested alignment, more than may be required pre-relaxation, to
allow for the target becoming more unaligned after relaxing earlier
sequences. This means that the target is often not initially aligned in
the object files, and so the R_RISCV_ALIGN relocations cannot just be
ignored. Since we do not support linker relaxation, we must turn these
into errors.

Reviewers: ruiu, MaskRay, espindola

Reviewed By: MaskRay

Subscribers: grimar, Jim, emaste, arichardson, asb, rbar, johnrusso, simoncook, sabuasal, niosHD, kito-cheng, shiva0217, zzheng, edward-jones, rogfer01, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei, psnobl, benna, lenary, s.egerton, pzheng, sameer.abuasal, apazos, luismarques, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D71820

This reverts commit a8a9c8e0.

There are multiple reported failures caused by this change.
Each failure is really weird, but it makes sense to revert
while investigating.

These weren't much different than copied output anyway.

There's no reason to introduce a new, unnaturally sized value
here. This has a chance to produce worse code with
legalization. Avoids regression in a future patch.

Revert "PR17164: Change clang's default behavior from -flax-vector-conversions=all to -flax-vector-conversions=integer."

This patch broke the Sanitizer buildbots. Please see the commit's
differential revision for more information
(https://reviews.llvm.org/D67678).

This reverts commit b72a8c65.

Add improve tests

The code doesn't apply the fix correctly to relocatable links. I could
try to fix the code that applies the fix, but it's pointless: we don't
actually know what the offset will be in the final executable. So just
ignore the flag for relocatable links.

Issue discovered building Android.

Differential Revision: https://reviews.llvm.org/D72968

Split out r600 tests, and try to be more consistent with
coverage. Cover a few more cases for 24-bit optimization and
constants.

Define pseudos mirroring the the VGPR indexing ones, and adjust the
operands in the s_movrel* instructions to avoid the result def.

a785209b switched to using a pseudos instead of manually tying
operands on the regular instruction. The VGPR indexing mode path
should have the same problems that change attempted to avoid, so these
should use the same strategy.

Use a single pseudo for the VGPR indexing mode and movreld paths, and
expand it based on the subtarget later. These have essentially the
same constraints, reading the index from m0.

Switch from using an offset to the subregister index directly, instead
of computing an offset and re-adding it back. Also add missing pseudos
for existing register class sizes.

Differential Revision: https://reviews.llvm.org/D73065

Summary:
Follow-up from https://reviews.llvm.org/D71733. Also moved an
initialization to the base class, where it belonged in the first place.

Reviewers: eraman, davidxl

Reviewed By: davidxl

Subscribers: hiraditya, haicheng, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D72949

Summary:
Generalize broadcastable trait to variadic operands. Update the
documentation that still talked about element type as part of
broadcastable trait (that bug was already fixed). Also rename
Broadcastable to ResultBroadcastableShape to be more explicit that the
trait affects the result shape (it is possible for op to allow
broadcastable operands but not have result shape that is broadcast
compatible with operands).

Doing some intermediate work to have getBroadcastedType take an optional
elementType as input and use that if specified, instead of the common
element type of type1 and type2 in this function.

Differential Revision: https://reviews.llvm.org/D72559

Document some high level strategies that should be used for register
bank selection. The constant bus restriction section hasn't actually
been implemented yet.

Summary: Fixes https://github.com/clangd/clangd/issues/256

Reviewers: kbobyrev

Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, usaxena95, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D73056

The way the IO handlers are currently managed by the debugger is wrong. The
implementation lacks proper synchronization between RunIOHandlerSync and
RunIOHandlers. The latter is meant to be run by the "main thread", while the
former is meant to be run synchronously, potentially from a different thread.

Imagine a scenario where RunIOHandlerSync is called from a different thread
than RunIOHandlers. Both functions manipulate the debugger's IOHandlerStack.
Although the push and pop operations are synchronized, the logic to activate,
deactivate and run IO handlers is not.

While investigating PR44352, I noticed some weird behavior in the Editline
implementation. One of its members (m_editor_status) was modified from another
thread. This happened because the main thread, while running RunIOHandlers
ended up execution the IOHandlerEditline created by the breakpoint callback
thread. Even worse, due to the lack of synchronization within the IO handler
implementation, both threads ended up executing the same IO handler.

Most of the time, the IO handlers don't need to run synchronously. The
exception is sourcing commands from external files, like the .lldbinit file.

I've added a (recursive) mutex to prevent another thread from messing with the
IO handlers wile another thread is running one synchronously. It has to be
recursive, because we might have to source another file when encountering a
command source in the original file.

Differential revision: https://reviews.llvm.org/D72748

- pthread affinity np has different semantic than sched affinity counterpart. On success returns strictly 0.

Reviewers: chandlerc, AndreyChurbanov, jdoerfert

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D72132

This patch introduces a small new utility (lldb-repro) to transparently
capture and replay debugger sessions through the command line driver.
Its used to test the reproducers by running the test suite twice.

During the first run, it captures a reproducer for every lldb invocation
and saves it to a well-know location derived from the arguments and
current working directory. During the second run, the test suite is run
again but this time every invocation of lldb replays the previously
recorded session.

Differential revision: https://reviews.llvm.org/D72823