PR14040.

llvm-svn: 165415

llvm-svn: 165414

llvm-svn: 165412

llvm-svn: 165411

llvm-svn: 165410

This patch add altivec support for v4i32 to v4f32 and for v4f32 to
v4i32 vector rounding conversion.

llvm-svn: 165409

llvm-svn: 165408

llvm-svn: 165406

llvm-svn: 165404

llvm-svn: 165403

llvm-svn: 165402

llvm-svn: 165401

llvm-svn: 165397

llvm-svn: 165396

llvm-svn: 165395

Review: http://llvm-reviews.chandlerc.com/D30
llvm-svn: 165392

llvm-svn: 165391

Some regression tests which are testing the old jit and are exercising functionality which is both known to be broken and not expected to be fixed in the old jit. To remove these from the regression test output, I've marked them XFAIL (for lit tests) and ifdef'd them out (unit tests). These modifications remove the last long-standing regression test failures from the buildbots (though updating the triple to reflect new ubuntu configuration has temporarily caused some new failures). Tested on x86-64 and ARM Linux.

Patch by David Tweed!

llvm-svn: 165390

llvm-symbolizer for Mac: add support for reading DWARF from separate object file in resource directory generated by dsymutil. Fix minor symbol table lookup issues.

llvm-svn: 165389

Contributed by: Sameer Sahasrabuddhe <sameer.sahasrabuddhe@amd.com>

llvm-svn: 165387

llvm-svn: 165386

llvm-svn: 165385

llvm-svn: 165384

llvm-svn: 165383

llvm-svn: 165382

llvm-svn: 165381

Found the fix on this page:
http://permalink.gmane.org/gmane.comp.python.sphinx.devel/112

llvm-svn: 165380

Scoplib only supports access functions, but not the more generic
access relations. This commit now also supports access functions
that where not directly expresses as A[sub] with sub = i + 5b,
but with A[sub] with -sub = -i + (-5b).

Test case to come.

Contributed by: Dustin Feld <d3.feld@gmail.com>

llvm-svn: 165379

llvm-svn: 165378

Otherwise it will try to use SSE patterns and fail horribly if sse is disabled.
Fixes PR14035.

llvm-svn: 165377

llvm-svn: 165376

llvm-svn: 165375

Patch by Leo Liu, test case by me.

llvm-svn: 165374

contents of the Attributes class over to an AttributesImpl.

llvm-svn: 165373

llvm-svn: 165372

llvm-svn: 165371

llvm-svn: 165370

llvm-svn: 165369

The darwin change should be a nop since Triple::getArchTypeForDarwinArchName
doesn't know about amd64.

If things like amd64-mingw32 are to be rejected, we should print a error
earlier on instead of silently using the wrong abi.

Remove old comment that looks out of place, this is "in clang".

llvm-svn: 165368

Without this change, when the estimated cost for inlining a function with
an "alwaysinline" attribute was lower than the inlining threshold, the
getInlineCost function was returning that estimated cost rather than the
special InlineCost::AlwaysInlineCost value. That is fine in the normal
inlining case, but it can fail when the inliner considers the opportunity
cost of inlining into an internal or linkonce-odr function. It may decide
not to inline the always-inline function in that case. The fix here is just
to make getInlineCost always return the special value for always-inline
functions. I ran into this building clang with libc++. Tablegen failed to
link because of an always-inline function that was not inlined. I have been
unable to reduce the testcase down to a reasonable size.

llvm-svn: 165367