diff options
| author | Elena Demikhovsky <elena.demikhovsky@intel.com> | 2012-09-04 12:49:02 +0000 |
|---|---|---|
| committer | Elena Demikhovsky <elena.demikhovsky@intel.com> | 2012-09-04 12:49:02 +0000 |
| commit | 32510207382720a97c682e098d45cb1371c48639 (patch) | |
| tree | 6d5f6007ba1667333625c3d331851a160cf3be10 /test/CodeGen/Thumb2 | |
| parent | 7765492a7a7e6eab36bc43558ea7c1f91e57cfec (diff) | |
| download | llvm-32510207382720a97c682e098d45cb1371c48639.tar.gz llvm-32510207382720a97c682e098d45cb1371c48639.tar.bz2 llvm-32510207382720a97c682e098d45cb1371c48639.tar.xz | |
This patch optimizes shuffle instruction - generates 2 instructions instead of 4.
Since this specific shuffle is widely used in many workloads we have ~10% performance on them.
shufflevector <8 x float> %A, <8 x float> %B, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
vmovaps (%rdx), %ymm0
vshufps $8, %ymm0, %ymm0, %ymm0
vmovaps (%rcx), %ymm1
vshufps $8, %ymm0, %ymm1, %ymm1
vunpcklps %ymm0, %ymm1, %ymm0
vmovaps (%rcx), %ymm0
vmovsldup (%rdx), %ymm1
vblendps $85, %ymm0, %ymm1, %ymm0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@163134 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test/CodeGen/Thumb2')
0 files changed, 0 insertions, 0 deletions