[PATCH] Add direct moves to/from VSR and exploit them for FP/INT conversions

[Nemanja Ivanovic]

Hi wschmidt, hfinkel, kbarton, seurer, echristo,

This patch adds support for the direct move instructions to and from VSRs that can be used for converting floating point to fixed point values and vice versa without the slow load/store combinations to the same memory locations.
For now, exploitation is limited to explicit conversions between floating/fixed point values and only on byte/halfword/word/doubleword sizes.

REPOSITORY
  rL LLVM

http://reviews.llvm.org/D8928

Files:
  lib/Target/PowerPC/PPC.td
  lib/Target/PowerPC/PPCFastISel.cpp
  lib/Target/PowerPC/PPCISelLowering.cpp
  lib/Target/PowerPC/PPCISelLowering.h
  lib/Target/PowerPC/PPCInstrFormats.td
  lib/Target/PowerPC/PPCInstrVSX.td
  lib/Target/PowerPC/PPCSubtarget.cpp
  lib/Target/PowerPC/PPCSubtarget.h
  test/CodeGen/PowerPC/fp-int-conversions-direct-moves.ll
  test/CodeGen/PowerPC/stfiwx.ll
  test/MC/Disassembler/PowerPC/vsx.txt
  test/MC/PowerPC/vsx.s

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D8928.23501.patch
Type: text/x-patch
Size: 27254 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150409/ed522385/attachment.bin>