llvm-svn: 334808

Summary:
Like the following:
  // -------
  // =======
  // *******

It does not cover all the cases, but those are definitely not very
useful.

Reviewers: sammccall, ioeric, hokein

Reviewed By: sammccall

Subscribers: MaskRay, jkorous, cfe-commits

Differential Revision: https://reviews.llvm.org/D48171

llvm-svn: 334807

Test regenerated to reduce noise in further patches.

llvm-svn: 334806

Summary:
On hover, the whole asm snippet is displayed, including operands.

This requires the actual assembly output instead of just the MCInsts:
This is because some pseudo-instructions get lowered to actual target
instructions during codegen (e.g. ABS_Fp32 -> SSE or X87).

Reviewers: gchatelet

Subscribers: mgorny, tschuett, llvm-commits

Differential Revision: https://reviews.llvm.org/D48164

llvm-svn: 334805

ASTImporter tests may produce source file related warnings, the diagnostic
client should be in correct state to handle it. Added 'beginSourceFile' to set
the client state.

Patch by: Balázs Kéri

Differential Revision: https://reviews.llvm.org/D47445

llvm-svn: 334804

Revert r334802 "[X86] Prevent folding stack reloads with instructions that have an undefined register update."

There's a typo causing the build to fail.

llvm-svn: 334803

We want to keep the load unfolded so we can use the same register for both sources to avoid a false dependency.

llvm-svn: 334802

ScopedMacroState injects its own EOF token under certain conditions, and the
returned token may be modified in several different locations. If multiple
reformat operations are started in different threads, then they will both see
the same fake EOF token, and may both try to modify it. This is a data race.

This bug was caught with tsan.

Reviewers: klimek

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D47759

llvm-svn: 334801

I think this covers most of the unmasked vector instructions. We're still missing a lot of the masked instructions.

There are some test changes here because of the new folding support. I don't think these particular cases should be folded because it creates an undef register dependency. I think the changes introduced in r334175 are not handling stack folding. They're only blocking the peephole pass.

llvm-svn: 334800

llvm-svn: 334799

These tests were recently updated so it looks like gone wrong.

llvm-svn: 334786

llvm-svn: 334785

The toolchain in Xcode 10 uses a new LC_BUILD_VERSION entry to identify
simulator binaries. Add support for reading those to debugserver.

The exisitng test testing that code is currently failling when run
with Xcode 10, no need for a new test.

llvm-svn: 334784

on darwin systems and re-execing itself, to creating two
separate test programs; lldb runs the first program and it
exec's the second.  

Support for compiling for i386 is going away.

llvm-svn: 334783

Patch by Gaetano Priori

Differential Revision: https://reviews.llvm.org/D47925

llvm-svn: 334782

compnerd pointed out that the latter reads better over here.

llvm-svn: 334781

LLVM currently assumes that Apple platforms will always use ld64. In the
future, LLD Mach-O might also be supported, so add the beginnings of
linker detection support. ld64 is currently the only detected linker,
since `ld64.lld -v` doesn't yield any useful version output, but we can
add that detection later, and in the meantime it's still useful to have
the ld64 identification.

Switch clang's order file check to use this new detection rather than
just checking for the presence of an ld64 executable.

Differential Revision: https://reviews.llvm.org/D48201

llvm-svn: 334780

The locks need to be implemented in a shared library to work correctly,
so they shouldn't be part of libclang_rt.builtins.a, except in
specialized scenarios where the user can prove it will only be linked
once.

Differential Revision: https://reviews.llvm.org/D47606

llvm-svn: 334779

This code was introduced back in r178148, a change to introduce
-module-file-info - which still exists & seems like it's still tested (&
this change didn't cause any of those tests to fail).

It doesn't look like this change was necessary there - since it's about
pcm output, whereas -module-file-info looks like it's for pcm /input/.
So I'm not really sure what the original motivation was.

I'm open to ideas though, if it turns out the original change was
necessary/useful.

llvm-svn: 334778

IEEE 754 defines the expected result on overflow. As far as I know,
hardware implementations (of f16), and compiler-rt (__floatuntisf)
correctly return +-Inf on overflow. And I can't think of any useful
transform that would take advantage of overflow being undefined here.

Differential Revision: https://reviews.llvm.org/D47807

llvm-svn: 334777

llvm-svn: 334775

XFAIL's and adding %run commands.

llvm-svn: 334774

[X86] Rename __builtin_ia32_pslldqi128 to __builtin_ia32_pslldqi128_byteshift and similar for other sizes. Remove the multiply by 8 from the header files.

The previous names took the shift amount in bits to match gcc and required a multiply by 8 in the header. This creates a misleading error message when we check the range of the immediate to the builtin since the allowed range also got multiplied by 8.

This commit changes the builtins to use a byte shift amount to match the underlying instruction and the Intel intrinsic.

Fixes the remaining issue from PR37795.

llvm-svn: 334773

On macOS 10.14, debugserver needs to have an entitlement do be
allowed to debug processes. Adding this to both the Xcode and
cmake build system. This shouldn't have any impact on previous
OSs.

llvm-svn: 334772

Once a symbol has been selected for materialization it can no longer be
overridden. Stripping the weak flag guarantees this (override attempts will
then be treated as duplicate definitions and result in a DuplicateDefinition
error).

llvm-svn: 334771

This change removes a few dashes to make room for the header syntax string.

llvm-svn: 334770

Summary:
Here we relax the old constraint which utilized unsafe with the TargetOption flag HonorSignDependentRoundingFPMathOption, with the assertion that unsafe is no longer needed or never was required for correctness on FDIV/FMUL.  



Reviewers: spatel, hfinkel, wristow, arsenm, javed.absar

Reviewed By: spatel

Subscribers: efriedma, wdng, tpr

Differential Revision: https://reviews.llvm.org/D48057

llvm-svn: 334769

Differential Revision: https://reviews.llvm.org/D47296

llvm-svn: 334768

llvm-svn: 334767

Differential Revision: https://reviews.llvm.org/D47292

llvm-svn: 334766

[NFC] Generalize flags for linking in shared objects in compiler-rt tests to support using multiple shared objects at once

Differential Revision: https://reviews.llvm.org/D48156

llvm-svn: 334765

The return value of TreePatternNode::getChild is never null. This patch also
updates various places that use return values of getChild to also use
references. Those changes were suggested post-commit for D47463.

llvm-svn: 334764

not an aggregtae.

llvm-svn: 334763

llvm-svn: 334762

Fixes https://crbug.com/852882

llvm-svn: 334761

In particular, when asked to print a MemoryAccess, we'll now print where
defs are optimized to, and we'll print optimized access types.

This patch also introduces an operator<< to make printing AliasResults
easier.

Patch by Juneyoung Lee!

Differential Revision: https://reviews.llvm.org/D47860

llvm-svn: 334760

isVectorClearMaskLegal() is the TLI hook used by the generic
DAGCombiner::XformToShuffleWithZero().

We've grown to accomodate/expect this transform to shuffle
(disabling it more generally results in many regressions).
So I'm narrowly excluding the 256-bit types that clearly 
are not worthwhile for AVX1. 

I think in most cases we are able to recover by converting 
the shuffle back into 'and' ops, but the cases in:
https://bugs.llvm.org/show_bug.cgi?id=37749
...show that there are cracks.

llvm-svn: 334759

llvm-svn: 334758

Reviewers: arsenm, nhaehnle

Reviewed By: arsenm

Subscribers: kzhuravl, wdng, yaxunl, rovka, kristof.beyls, dstuttard, tpr, t-tye, llvm-commits

Differential Revision: https://reviews.llvm.org/D45907

llvm-svn: 334757

This is r334750 (which was reverted in r334754) with a fix for an
uninitialized variable that was caught by msan.

Original commit message:
> If a copy bundle happens to involve overlapping registers, we can end
> up with emitting the copies in an order that ends up clobbering some
> of the subregisters. Since instructions in the copy bundle
> semantically happen at the same time, this is incorrect and we need to
> make sure we order the copies such that this doesn't happen.

llvm-svn: 334756