llvm-svn: 199299

Fix the attribute enable_if example in the docs to be 100% real-world, instead of merely being close to the real world to mislead people. This now matches the test.

llvm-svn: 199298

Add OpSize16 to the two byte forms of INC/DEC that we only use in 64-bit mode and a 64-bit only LEA. Even though we'll not be in 16-bit mode when we use them it makes their tables consistent with their 32-bit counterparts.

llvm-svn: 199297

llvm-svn: 199296

Simplify x86 disassembler table handling of when to use TYPE_Rv/TYPE_R16/TYPE_R32 now that HasOpSizePrefix only means 16-bit instructions.

llvm-svn: 199295

There has been an old FIXME to find the right cut-off for when it's worth
analyzing and potentially transforming a switch to a lookup table.

The switches always have two or more cases. I could not measure any speed-up
by transforming a switch with two cases. A switch with three cases gets a nice
speed-up, and I couldn't measure any compile-time regression, so I think this
is the right threshold.

In a Clang self-host, this causes 480 new switches to be transformed,
and reduces the final binary size with 8 KB.

llvm-svn: 199294

llvm-svn: 199293

This C++ feature has been marked complete since r191549, but the documentation
claimed it wasn't supported at all and the extension check misreported it as
being available in C.

No regression test; this was a short-lived typo.

llvm-svn: 199292

strides

Fixes PR18480.

llvm-svn: 199291

control to the user anyway.  This was put in to handle monitors that would say there was no
stop reason when you first attached to them.  But it broke the case where you hit a thread specific
breakpoint on many threads, but NOT the one specified in the breakpoint.  I work around this
by only doing the junky override when the StopID is 0 - i.e. on first attach.

This commit also adds a test for thread specific breakpoints.

llvm-svn: 199290

llvm-svn: 199289

llvm-svn: 199288

The GNU as behavior is a bit different and very strange. It will mark any
label that contains an instruction. We can implement that, but using the
type looks more natural since gas will not mark a function if a .word is
used to output the instructions!

llvm-svn: 199287

llvm-svn: 199286

The line breaks around the "m:<mangling>" text in the Data Layout section
look weird. Let's see if this helps.

llvm-svn: 199285

Changes made in r192200 fixed PR16992, which requested fixit suggesting
parenthesis if sizeof is followed by type-id. However expression in form
T() followed by ')' was incorrectly considered as a type-id if 'T' is
typedef name. This change fixes this case.

Differential Revision: http://llvm-reviews.chandlerc.com/D2440

llvm-svn: 199284

Continue to accept the old name for a while to make it an easier transition
for people who rely on this.

llvm-svn: 199283

When expanding neon pseudo stores, it may miss the implicit uses of sub
regs, which may cause post RA scheduler reorder instructions that
breakes anti dependency.

For example:
  VST1d64QPseudo %R0<kill>, 16, %Q9_Q10, pred:14, pred:%noreg
  will be expanded to
    VST1d64Q %R0<kill>, 16, %D18, pred:14, pred:%noreg;

An instruction that defines %D20 may be scheduled before the store by
mistake.

This patches adds implicit uses for such case. For the example above, it
emits:
  VST1d64Q %R0<kill>, 8, %D18, pred:14, pred:%noreg, %Q9_Q10<imp-use>

llvm-svn: 199282

This is a follow-up to r199260. On ARM hosts, we were attempting to run
tests with triples such as armv7l-unknown-win32. This expands that fix to
cover all non-x86 targets since we only support MS ABI on x86.

llvm-svn: 199280

llvm-svn: 199279

llvm-svn: 199278

llvm-svn: 199277

presence of CU ranges.

llvm-svn: 199276

llvm-svn: 199275

llvm-svn: 199274

Fixes dynamically linked programs on NetBSD/i386.

llvm-svn: 199273

llvm-svn: 199269

llvm-svn: 199268

llvm-svn: 199267

The changes caused by folding an sp-adjustment into a "pop" previously
disrupted the forward search for the final real instruction in a
terminating block. This switches to a backward search (skipping debug
instrs).

This fixes PR18399.

Patch by Zhaoshi.

llvm-svn: 199266

We should set them to expand for now since there are no patterns
dealing with them. Actually, there are no instructions either so I
doubt they'll ever be acceptable.

llvm-svn: 199265

to run lld tests individually.

llvm-svn: 199264

-ffunction-sections and update comments and TODOs about other
places that we should enable this.

llvm-svn: 199263

llvm-svn: 199262

MCJIT remote execution (ChildTarget+RemoteTargetExternal) protocol was in
dire need of refactoring. It was fail-prone, had no error reporting and
implemented the same message logic on every single function.

This patch rectifies it, and makes it work on ARM, where it was randomly
failing. Other architectures shall profit from this change as well, making
their buildbots and releases more reliable.

llvm-svn: 199261

Long term we should make -triple arm7l-*-win32 do something sensible,
but for now it's broken and untested.

llvm-svn: 199260

llvm-svn: 199259

llvm-svn: 199258

llvm-svn: 199257

Someone recently wasted some time not realising that "-###" didn't
actually execute the commands it printed, and suggested a
documentation tweak.

Having made the same mistake myself on at least one occasion, I
sympathise. So here it is. Any kibitzing on an even better text
welcome.

llvm-svn: 199256