[PATCH] D30533: [GlobalISel][X86] Support float/double and vector types.

[Igor Breger via Phabricator via llvm-commits]

igorb created this revision.
Herald added subscribers: kristof.beyls, rovka, dberris.

[GlobalISel][X86] Add support for f32/f64 and vector types in RegisterBank and InstructionSelector.


Repository:
  rL LLVM

https://reviews.llvm.org/D30533

Files:
  lib/Target/X86/X86GenRegisterBankInfo.def
  lib/Target/X86/X86InstructionSelector.cpp
  lib/Target/X86/X86InstructionSelector.h
  lib/Target/X86/X86LegalizerInfo.cpp
  lib/Target/X86/X86LegalizerInfo.h
  lib/Target/X86/X86RegisterBankInfo.cpp
  lib/Target/X86/X86RegisterBanks.td
  test/CodeGen/X86/GlobalISel/X86-regbankselect.mir
  test/CodeGen/X86/GlobalISel/binop-isel.ll
  test/CodeGen/X86/GlobalISel/irtranslator-callingconv.ll
  test/CodeGen/X86/GlobalISel/legalize-add.mir
  test/CodeGen/X86/GlobalISel/legalize-sub.mir
  test/CodeGen/X86/GlobalISel/x86_64-instructionselect.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D30533.90322.patch
Type: text/x-patch
Size: 58160 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170302/b6f3fb8b/attachment.bin>