This is the first patch that adds support for GLSL extended
instruction set ops. These are direct conversions, apart from `spv.Tan`
that is lowered to `sin() / cos()`.

Reviewed By: antiagainst

Differential Revision: https://reviews.llvm.org/D84627

The assertion is not true anymore after D82739, this patch just removes
it, and rename related functions.

And also fixes a missing cases.

Differential Revision: https://reviews.llvm.org/D84837

Continue the change made to ParseATTOperand to take the vector by
reference. Let ParseMemOperand add its memory operand to the
vector and just return true/false to indicate error.

Pointers and SMLocs are cheap to copy. Even though the function
modifies some of these the caller doesn't use them after the call.

This change define RAII class `FileLocker` and methods `lock` and
`tryLockFor` of the class `raw_fd_stream` to facilitate using file locks.

Differential Revision: https://reviews.llvm.org/D79066

Use of BuiltinBug is replaced by BugType.
Class BuiltinBug seems to have no benefits and is confusing.

Reviewed By: Szelethus, martong, NoQ, vsavchenko

Differential Revision: https://reviews.llvm.org/D84494

- Clarify that these are extensions to DWARF 5 and not as yet a
  proposal.

Reviewed By: scott.linder

Differential Revision: https://reviews.llvm.org/D70523

[X86] Teach the assembler parser to handle a '*' between segment register and base/index/displacement part of an address

A '*' after the segment is equivalent to a '*' before the segment register. To make the AsmMatcher table work we need to place the '*' token into the operand vector before the full memory operand. To accomplish this I've modified some portions of operand parsing to expose the operand vector to ParseATTOperand so that the token can be pushed to the vector after parsing the segment register and before creating the memory operand using that segment register.

Fixes PR46879

Reviewed By: MaskRay

Differential Revision: https://reviews.llvm.org/D84895

Previously ctor homing was omitting debug info for classes if they
have both trival and nontrivial constructors, but we should only omit debug
info if the class doesn't have any trivial constructors.

retained types list.

bug: https://bugs.llvm.org/show_bug.cgi?id=46537

Differential Revision: https://reviews.llvm.org/D84870

z13 is not a target. It is a CPU.

--string-dump

Summary:
Some instructions have set the wrong [RM] flag, this patch is to fix it.

Instructions x(v|s)r(d|s)pi[zmp]? and fri[npzm] use fixed rounding
directions without referencing current rounding mode.

Also, the SETRNDi, SETRND, BCLRn, MTFSFI, MTFSB0, MTFSB1, MTFSFb,
MTFSFI, MTFSFI_rec, MTFSF, MTFSF_rec should also fix the RM flag.

Reviewed By: jsji

Differential Revision: https://reviews.llvm.org/D81360

- Clarify what context is used in DWARF expression evaluation.

- Define location descriptions to fully resolve the context and so
  include the context in their result.

- As a consequence of location descriptions being fully resoved,
  change address spaces so only a swizzled and unswizzled private
  address space is defined. The lane is now part of the location
  description context.

- Clarify how call frame information is used to fully resolve
  expressions that specify registers.

Reviewed By: scott.linder

Differential Revision: https://reviews.llvm.org/D70523

SelectionDAGBuilder just drops these, so do the same.

Move the finish_swig logic into a function in the bindings directory. By
making this a function I can reuse the logic internally where we ship
two Python versions and therefore need to finish the bindings twice.

This patch makes JumpThreading fold br(freeze(undef)) if the freeze instruction
is only used by the branch.

Reviewed By: efriedma

Differential Revision: https://reviews.llvm.org/D84818

Fixes a build failure in the Darwin kernel. Tested with:

% nm -m ./lib/libclang_rt.cc_kext_x86_64_osx.a | grep lprofSetProfileDumped

rdar://66249602

Add a fallback for `sysctl kern.osproductversion` for XNU 17 (macOS
10.13) and below, which do not provide this property.

Unfortunately, this means we have to take the detour via Darwin kernel
version again (at least for the fallback).

Reviewed By: delcypher

Differential Revision: https://reviews.llvm.org/D84892

I still think it's highly questionable that we have two intrinsics
with identical behavior and only vary by the name of the libcall used
if it happens to be lowered that way, but try to reduce the feature
delta between SDAG and GlobalISel for recently added intrinsics. I'm
not sure which opcode should be considered the canonical one, but
lower roundeven back to round.

Further abstracting the specification of a tensor, to more easily
support different types and shapes of tensor, and also to perform
initialization up-front, at TFModelEvaluator construction time.

Differential Revision: https://reviews.llvm.org/D84685

Its implementation was scrapped in the final fast16labels
instrumentation patch.

This reverts commit d8e8e32d.

This breaks the build on GreenDragon:
http://green.lab.llvm.org/green/job/clang-stage1-RA/13171/console

LLVM's make_paths_relative uses Python3_EXECUTABLE which isn't set in
standalone LLDB builds.

for device simulators

This change separates out the iOS/tvOS/watchOS simulator slices from the "libclang_rt.<os>.a"
fat archive, by moving them out to their own "libclang_rt.<os>sim.a" static archive.
This allows us to build and to link with an arm64 device simulator slice for the simulators running
on Apple Silicons, and to distribute it in one archive alongside the Intel simulator slices.

Differential Revision: https://reviews.llvm.org/D84564

This reverts commit 120e66b3.

Due to a buildbot failure.

We saw a crash recently (rdar://problem/65276489) that looks related to an invalid ValueObjectSP in a summary providers in Cocoa.cpp e.g. NSBundleSummaryProvider(...).
This adds checks before we use them usually by calling NSStringSummaryProvider.

Differential Revision: https://reviews.llvm.org/D84272

From https://reviews.llvm.org/D84872.

When attempting to build compiler-rt on a developer transition kit, the
build would fail due to `.S` files not being handled properly by the
Ninja generator.  Rather than conditionalising on Xcode, conditionalise
to Darwin.  Because we know that the system compiler is clang based, it
will always properly handle the pre-processing based on the extension.

Differential Revision: https://reviews.llvm.org/D84333

To avoid hash collisions when the only difference is in mem ops.

Differential Revision: https://reviews.llvm.org/D84782

This reduces the number of check-llvm failures by 500.

Ideally we'd have a codegen version of PassRegistry.def, or have all the
codegen passes ported and put into PassRegistry.def. But since that
doesn't exist yet, hardcode the list of codegen IR passes.

There are still codegen passes missing from this list, I'll add them
later as I stumble upon them.

Reviewed By: asbirlea, ychen

Differential Revision: https://reviews.llvm.org/D84872

The problem with 8723280b was that the
`dso_local` is *before* the void not after. Hope this works.

Reviewed By: sivachandra

Differential Revision: https://reviews.llvm.org/D84611