llvmpipe: Optimize BUILD_MASK(_LINEAR) for POWER8
authorOded Gabbay <oded.gabbay@gmail.com>
Tue, 29 Dec 2015 16:12:34 +0000 (18:12 +0200)
committerOded Gabbay <oded.gabbay@gmail.com>
Wed, 6 Jan 2016 12:54:16 +0000 (14:54 +0200)
commit925c46cfc48042ec0bc5a83df962d2d7dd038394
tree5e9b575a9b45c334d59125bff921d9c91851e21c
parent3bbe16ea79bb5738109df36780cc99119a006d91
llvmpipe: Optimize BUILD_MASK(_LINEAR) for POWER8

This patch converts the SSE-optimized build_mask_32() and
build_mask_linear_32() to VMX/VSX.

I measured the results on POWER8 machine with 32 cores at 3.4GHz and
16GB of RAM.

                      FPS/Score
  Name            Before     After    Delta
------------------------------------------------
glmark2 (score)   139.8      142.7    2.07%

openarena and xonotic didn't show a significant (more than 1%)
difference.

v2: Make sure code is build only on POWER8 LE machine

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
src/gallium/drivers/llvmpipe/lp_rast_tri.c