llvm-svn: 110653

llvm-svn: 110652

llvm-svn: 110650

llvm-svn: 110649

llvm-svn: 110648

Remove replicate intrinsics, clang will generate shufflevector for those. The shuffles can't be matched by x86 codegen yet, but will soon

llvm-svn: 110647

llvm-svn: 110645

llvm-svn: 110643

Refactor and simplify code to avoid redundant checks.

llvm-svn: 110642

if-else statements, to hopefully support older pythons (PR7850).

llvm-svn: 110638

is the default.  Patch by NAKAMURA Takumi!

llvm-svn: 110636

llvm-svn: 110633

register at a time. This turns out to be slightly faster than iterating over
instructions, but more importantly, it allows us to compute spill weights for
new registers created after the spill weight pass has run.

Also compute the allocation hint at the same time as the spill weight. This
allows us to use the spill weight as a cost metric for copies, and choose the
most profitable hint if there is more than one possibility.

The new hints provide a very small (< 0.1%) but universal code size improvement.

llvm-svn: 110631

llvm-svn: 110630

llvm-svn: 110628

pass. This pass should expand with all of the small, fine-grained optimization
passes to reduce compile time and increase happiment.

llvm-svn: 110627

llvm-svn: 110623

llvm-svn: 110621

llvm-svn: 110618

previously collected info from the .file directives and outputs the encoded
bytes for it.  For now this is only in the Mach-O streamer but at some point
will move to a more generic place.

llvm-svn: 110617

This will always be false before PEI:
(DisableFramePointerElim(MF) && MFI->adjustsStack())
Which means it's going to make r11 available as a general purpose register even
if -disable-fp-elim is specified. It's working on Darwin only because r7 is
always reserved. But it's obviously broken for other targets.

llvm-svn: 110614

Add VCVTPD2PS, VCVTPS2DQ, VCVTPS2PDY, VCVTTPD2DQY, VCVTTPS2DQ and VCVTPD2DQ 256-bit conversion intrinsics

llvm-svn: 110608

llvm-svn: 110607

Add patterns to AVX conversions instructions. Do that instead of declaring more intructions whenever is possible, more coming

llvm-svn: 110605

gold plugin.

llvm-svn: 110604

llvm-svn: 110601

llvm-svn: 110600

llvm-svn: 110598

it doesn't regress again.

llvm-svn: 110597

llvm-svn: 110594

Next time the build is broken due to wrong library dependencies, just
try building again (if you are on some Unix and are building all LLVM
targets) or ask someone to commit the regenerated LLVMLibDeps.cmake.

llvm-svn: 110593

llvm-svn: 110590

If we are emitting COPY instructions for the REG_SEQUENCE, make sure the kill
flag goes on the last COPY. Otherwise we may be using a killed register.

<rdar://problem/8287792>

llvm-svn: 110589

llvm-svn: 110587

llvm-svn: 110586

Change -prefer-32bit-thumb to attribute -mattr=+32bit instead to disable more 32-bit to 16-bit optimizations.

llvm-svn: 110584

llvm-svn: 110582

llvm-svn: 110580

llvm-svn: 110579

Map pointers of any contained ValueMapCallbackVH's) and unused.

llvm-svn: 110577