When intrinsics are allowed to have mem operands, there
are two ways this can happen. First is an intrinsic
that is marked has having a mem operand, but is not handled
by getTgtMemIntrinsic.

The second way can occur even for intrinsics which do not
have a mem operand. It seems the selector table does
some kind of sorting based on the opcode, and the
mem ref recording can happen in the same scope for
intrinsics that both do and do not have mem refs.
I haven't been able to figure out exactly why this happens
(although it happens even with the matcher optimizations disabled).
I'm not sure if it's worth trying to avoid hitting this for
these nodes since I think it's still reasonable to handle
this in case getTgtMemIntrinic is not implemented.

llvm-svn: 321208

Patch by Matthew Voss!

llvm-svn: 321207

llvm-svn: 321206

The function createTailCallBranchInstr assumes that the iterator MBBI is valid.
However, only one use of MBBI is guarded in the function.
Fix this by adding an assert.

Differential Revision: https://reviews.llvm.org/D41358

llvm-svn: 321205

Prevent overlapping store elision when overlapping store is
pre-inc/dec as analysis is wrong in these cases.

llvm-svn: 321204

Summary: Very similar to AddressSanitizer, with the exception of the error type encoding.

Reviewers: kcc, alekseyshl

Subscribers: cfe-commits, kubamracek, llvm-commits, hiraditya

Differential Revision: https://reviews.llvm.org/D41417

llvm-svn: 321203

Patch by Dmitry Venikov

llvm-svn: 321202

Looking through the code, I saw a FIXME on IFunc to switch it
to a target specific attribute. In looking through it, i saw that
the no-longer-appropriately-named TargetArch didn't support ObjectFormat
checking.

This patch changes the name of TargetArch to TargetSpecific
(since it checks much more than just Arch), makes "Arch" optional, adds
support for ObjectFormat, better documents the TargetSpecific type, and
changes IFunc over to a TargetSpecificAttr.

Differential Revision: https://reviews.llvm.org/D41303

llvm-svn: 321201

This patch turns shuffles of fadd/fsub with fmul into fmsubadd.

Patch by Dmitry Venikov

Differential Revision: https://reviews.llvm.org/D40335

llvm-svn: 321200

llvm-svn: 321199

It is never null.

llvm-svn: 321198

Summary:
The function section prefix for PGO based layout (e.g. hot/unlikely)
should look at the hotness of all blocks not just the entry BB.
A function with a cold entry but a very hot loop should be placed in the
hot section, for example, so that it is located close to other hot
functions it may call. For SamplePGO it was already looking at the
branch weights on calls, and I made that code conditional on whether
this is SamplePGO since it was essentially a noop for instrumentation
PGO anyway.

Reviewers: davidxl

Subscribers: eraman, llvm-commits

Differential Revision: https://reviews.llvm.org/D41395

llvm-svn: 321197

llvm-svn: 321196

llvm-svn: 321194

Reviewers: sammccall

Subscribers: klimek, mgorny, ilya-biryukov, cfe-commits

Differential Revision: https://reviews.llvm.org/D41450

llvm-svn: 321193

llvm-svn: 321192

Reviewers: rnk, aprantl, majnemer

Reviewed By: aprantl

Differential Revision: https://reviews.llvm.org/D414

llvm-svn: 321191

The size of the result vector is currently around 4600 with
Flavor::WarningOrError, which makes std::vector a better candidate than
llvm::SmallVector.

Patch by: Andras Leitereg!

Differential Revision: https://reviews.llvm.org/D39372

llvm-svn: 321190

Summary: Adds BeforeExecute method to PrecompiledPreamble to be called
before Execute(). This method can be overriden.

Patch by William Enright.

Reviewers: malaperle, ilya-biryukov

Reviewed By: ilya-biryukov

Subscribers: arphaman, cfe-commits, ilya-biryukov

Differential Revision: https://reviews.llvm.org/D41365

llvm-svn: 321189

Fix the definitions of 'reference' and 'pointer' in string_view that no one uses :-). Thanks to K-ballo for the catch.

llvm-svn: 321188

Every shared symbol has a file, so we can use a reference.

llvm-svn: 321187

llvm-svn: 321186

llvm-svn: 321185

Summary:
The goal here is again to make it easier to read and write the tests.

I've extracted `parseTextMarker` from CodeCompleteTests into an `Annotations`
class, adding features to it:
  - as well as points `^s` it allows ranges `[[...]]`
  - multiple points and ranges are supported
  - points and ranges may be named: `$name^` and `$name[[...]]`

These features are used for the xrefs tests. This also paves the way for
replacing the lit diagnostics.test with more readable unit tests, using named
ranges.

Alternative considered: `TestSelectionRange` in clang-refactor/TestSupport
Main problems were:
 - delimiting the end of ranges is awkward, requiring counting
 - comment syntax is long and at least as cryptic for most cases
 - no separate syntax for point vs range, which keeps xrefs tests concise
 - Still need to convert to Position everywhere
 - Still need helpers for common case of expecting exactly one point/range

(I'll probably promote the extra `PrintTo`s from some of the core Protocol types
into `operator<<` in `Protocol.h` itself in a separate, prior patch...)

Reviewers: ioeric

Subscribers: klimek, mgorny, ilya-biryukov, cfe-commits

Differential Revision: https://reviews.llvm.org/D41432

llvm-svn: 321184

Summary: Plant an inline version of "((ac+bd)/(cc+dd)) + i((bc-ad)/(cc+dd))" instead.

Patch by Paul Walker.

Reviewed By: hfinkel

Differential Revision: https://reviews.llvm.org/D40299

llvm-svn: 321183

This commits the non-controversial part of https://reviews.llvm.org/D41029
(making the queries virtual). The PPC-specific portion of this will be
committed in a subsequent patch once some of the finer points are ironed out.

llvm-svn: 321182

llvm-svn: 321181

These functions simply call their counterparts in the associated SDNode,
which do take an optional SelectionDAG. This change makes the legalization
debug trace a little easier to read, since target-specific nodes will
now have their names shown instead of "Unknown node #123".

llvm-svn: 321180

llvm-svn: 321179

Summary:
Initial changes in interfaces of MCAsmStreamer/MCTargetStreamer for
correct debug info emission for Cuda.
1. PTX foramt does not support `.ascii` directives. Added the ability to
nullify it.
2. The initial function label must follow the first debug `.loc`
directive, not be followed by.
3. DWARF sections must be enclosed in braces.

Reviewers: hfinkel, probinson, jlebar, rafael, echristo

Subscribers: sdardis, nemanjai, llvm-commits, aprantl

Differential Revision: https://reviews.llvm.org/D40033

llvm-svn: 321178

llvm-svn: 321177

NFC for currently supported targets. This resolves a problem encountered by
targets such as RISCV that reference `Subtarget` in ImmLeaf predicates.

llvm-svn: 321176

Reviewers: mehdi_amini, ioeric, klimek

Reviewed By: ioeric

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D41393

llvm-svn: 321175

Summary:
- Fixed an assert in Sema::InstantiateFunctionDefinition and added
  support for instantiating a function template with skipped body.
- Properly call setHasSkippedBody for FunctionTemplateDecl passed to
  Sema::ActOnSkippedFunctionBody.

Reviewers: sepavloff, bkramer

Reviewed By: sepavloff

Subscribers: klimek, cfe-commits

Differential Revision: https://reviews.llvm.org/D41237

llvm-svn: 321174

llvm-svn: 321173

canVectorize is only checking if the loop has a normalized pre-header if DoExtraAnalysis is true.
This doesn't make sense to me because reporting analysis information shouldn't alter legality
checks. This is probably the result of a last minute minor change before committing (?).

Patch by Diego Caballero.

Reviewed By: fhahn

Differential Revision: https://reviews.llvm.org/D40973

llvm-svn: 321172

llvm-svn: 321171

a change to the AsmMatcherEmitter, and should fix the buildbot
failure on llvm-clang-x86_64-expensive-checks-win.

The issue is also described here:
http://lists.llvm.org/pipermail/llvm-dev/2017-December/119617.html

llvm-svn: 321170

llvm-svn: 321169

Examples:
* Always evaluates to 0:

```
  int X;
  if (0 & X) return;
```

* Always evaluates to ~0:

```
  int Y;
  if (Y | ~0) return;
```

* The symbol is unmodified:

```
  int Z;
  Z &= ~0;
```

Patch by: Lilla Barancsuk!

Differential Revision: https://reviews.llvm.org/D39285

llvm-svn: 321168