llvm-svn: 339808

Summary:
In r335809, Petr Hosek lays out support for what he calls the multiarch
runtimes layout. This new way of laying out the directories for runtime
libraries is workable for all platforms. Petr did some of the common
infrastructure and made it work for Linux and Fuscia. This patch is a
cleanup to the Darwin and MachO drivers to serve as a step toward
supporting it in Darwin.

This patch does primarily two things:
(1) Changes the APIs for how the Darwin driver refers to compiler-rt
libraries to use the component names, similar to how Linux and Fuscia do

(2) Removes some legacy functionality for supporting macOS versions
before 10.6. This functionality is effectively dead code because in
r339277, the support was removed from compiler-rt for generating the 10.4
runtime support library, and Xcode 10 (currently in beta) removes
libgcc_s.10.4 and libgcc_s.10.5 from the macOS SDK.

With this patch landed a subsequent patch can modify
MachO::AddLinkRuntimeLib to support the multiarch runtimes layout.

Worth noting: None of the removed functionality was actually covered in
the test suite. So no test case updates are required.

Reviewers: phosek, bruno, arphaman

Reviewed By: phosek, arphaman

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D50618

llvm-svn: 339807

use-after-poison in check-llvm under asan

This reverts commit r339755.

llvm-svn: 339806

The compiler may produce unexpected error messages/crashes when declare
target variables were used. Patch fixes problems with the declarations
marked as declare target to or link.

llvm-svn: 339805

Mark the at_exit and at_quick_exit tests as unsupported under C++98 an 03, since those calls were introduced in C++11.  They're already guarded by an ifdef in the code, so this is a 'belt-and-suspenders' change.

llvm-svn: 339804

Before:
  message Empty {
  }

After:
  message Empty {}

llvm-svn: 339803

Summary:
The C-API supports consuming errors, converting an error to a string error
message, and querying an error's type. Other LLVM C APIs that wish to use
llvm::Error can supply error-type-id checkers and custom
error-to-structured-type converters for any custom errors they provide.

Reviewers: bogner, zturner, labath, dblaikie

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D50716

llvm-svn: 339802

Summary:
This CL changes the ExtractLane ISEL multiclass to more closely mirror
the structure of the splat and replace_lane multiclasses.

Reviewers: aheejin, dschuff

Subscribers: sbc100, jgravelle-google, sunfish, llvm-commits

Differential Revision: https://reviews.llvm.org/D50794

llvm-svn: 339801

These functions have had no-CFI annotations in the source code for
a while now.

llvm-svn: 339800

If a relocation group doesn't have the RELOCATION_GROUP_HAS_ADDEND_FLAG set, then this implies the group's addend equals zero.
In this case android packed format won't encode an explicit addend delta, instead we need to set Addend, the "previous addend" variable, to zero by ourself.

Patch by Yi-Yo Chiang!

Differential Revision: https://reviews.llvm.org/D50601

llvm-svn: 339799

Changes a comment and some whitespace to test commit access.

llvm-svn: 339798

This constructor needs to cast a pointer to uninitialized
memory to a pointer to object type in order to call
allocator_traits::construct(). This cast is not allowed when CFI cast
checks are enabled.

I did this instead of marking __addr() as _LIBCPP_NO_CFI so that we
don't lose CFI checks on get() or the dtor.

Differential Revision: https://reviews.llvm.org/D50743

llvm-svn: 339797

rdar://42473741

Differential Revision: https://reviews.llvm.org/D50775

llvm-svn: 339796

Summary: Expose VerifyMemorySSA as a debug option. If set, passes will call the MSSA->verifyMemorySSA() after calling into the updater's APIs when MemorySSA should be valid.

Reviewers: george.burgess.iv

Subscribers: sanjoy, jlebar, Prazek, llvm-commits

Differential Revision: https://reviews.llvm.org/D50749

llvm-svn: 339795

Because FreeBSD uses _M in its <sys/types.h>, and it is hard to avoid
including that header, only define _M to NASTY_MACRO for other operating
systems.  This fixes almost 2000 unexpected test failures.

Discussed with Eric Fiselier.

llvm-svn: 339794

These correspond to the x86 tests added with rL339790 / rL339791, but I widened
the non-fsin tests to v3f32 to show the problem because AArch supports v2f32 ops. 

llvm-svn: 339793

llvm-svn: 339792

This is a negative test for x86 because it has custom lowering for fabs.

llvm-svn: 339791

llvm-svn: 339790

The bit-fields of `Type` have enough space for
the `unsigned NumExpansions` of `PackExpansionType`

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D50711

llvm-svn: 339789

llvm-svn: 339788

Type has enough space for two members of
TemplateSpecializationType. Mechanical patch.

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D50643

llvm-svn: 339787

Implement and test replace_lane instructions.

Patch by Thomas Lively

Differential Revision: https://reviews.llvm.org/D50750

llvm-svn: 339786

Older compilers don't like the implicit cast & move when returning a
unique_ptr to an llvm::Expected type.

llvm-svn: 339785

Subregister liveness applies selectively to register classes with certain
properties. Make sure that when it's enabled, it applies to a given virtual
register (in virtual register rewriter).

llvm-svn: 339784

With explicit unique_ptr casts so that bots with older compilers don't
break.

llvm-svn: 339783

Just making testnames better reflect their testing scenarios.

Differential Revision: https://reviews.llvm.org/D50641

llvm-svn: 339782

Syntactically invalid JSON payload was causing clangd to terminate because of unexpected EOF rather than exit as a response to LSP exit message.

Differential Revision: https://reviews.llvm.org/D50641

llvm-svn: 339781

llvm-svn: 339780

To make ISD::VSELECT available(legal) so long as there are altivec instruction,
otherwise it's default behavior is expanding.
Use xxsel to match vselect if vsx is open, or use vsel.

In order to do not write many patterns in td file, promote (for vector it's
bitcast) all other type into v4i32 and only pattern match vselect of v4i32 into
vsel or xxsel.

Patch by wuzish
Differential revision: https://reviews.llvm.org/D49531

llvm-svn: 339779

Change
  subreg_r32  -> subreg_h32
  subreg_r64  -> subreg_h64
  subreg_hr32 -> subreg_hh32

The subregisters subreg_r32 and subreg_r64 were added to emphasize the
fact that modifying these subregisters may clobber the entire register.
This is not necessarily the case for subreg_h32, et al.

However, the ability to compose subreg_h64 with subreg_r32, and with
subreg_h32 and subreg_l32 at the same time makes the compositions be
treated as non-overlapping (leading to problems when tracking subreg
liveness). See D50468 for more details.

Differential Revision: https://reviews.llvm.org/D50725

llvm-svn: 339778

DenseMap insert() method return a pair<iterator, bool>
not pair<iterator, char>
Noticed it and thought I might just fix it ...

llvm-svn: 339777

This option is needed to enable subreg liveness tracking during register
allocation.

Review: Ulrich Weigand
https://reviews.llvm.org/D50779

llvm-svn: 339776

This rewrites the test using yaml.

llvm-svn: 339775

This removes test that used binary input and adds a yaml based test instead.

llvm-svn: 339774

Summary:
Previously, `eraseFromParent()` calls `delete` which invalidates the value of the pointer. Copying the value of the pointer later is undefined behavior in C++11 and implementation-defined (which may cause a segfault on implementations having strict pointer safety) in C++14.

This patch removes the BasicBlock pointer from related SmallPtrSet before `delete` invalidates it in the SimplifyCFG pass.

Reviewers: kuhar, dmgreen, davide, trentxintong

Reviewed By: kuhar, dmgreen

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D50717

llvm-svn: 339773

This allows to set custom Info field value for SHT_GROUP sections.

It is useful to allow this because we would be able to replace at least one binary
object committed in LLD and replace it with the yaml2obj based test.

Differential revision: https://reviews.llvm.org/D50776

llvm-svn: 339772

The idle callback was removed from the spec as of TR7.
This removes it from the implementation.

Patch provided by Simon Convent

Reviewers: hbae, protze.joachim

Differential Revision: https://reviews.llvm.org/D48362

llvm-svn: 339771

We only try to promote types with are smaller than 16-bits, but we
also need to check that the type is not less than 8-bits.

Differential Revision: https://reviews.llvm.org/D50769

llvm-svn: 339770

When trying to combine a DAG that builds a vector out of sign-extensions of
vector extracts, the code assumes legal input types. Due to that, we have to
disable this combine prior to legalization.
In some cases, the DAG will look slightly different after legalization so
account for that in the matching code.

This is a fix for https://bugs.llvm.org/show_bug.cgi?id=38087

Differential Revision: https://reviews.llvm.org/D49080

llvm-svn: 339769