[x265] [PATCH 1 of 2] asm: interp_4tap_vert_ps_4x2 sse2
dtyx265 at gmail.com
dtyx265 at gmail.com
Wed May 20 01:33:13 CEST 2015
# HG changeset patch
# User David T Yuen <dtyx265 at gmail.com>
# Date 1432070824 25200
# Node ID 3e07cba4b2034db2b819b2e11e98ee4b851d52b5
# Parent d7b100e51e828833eee006f1da93e499ac161d28
asm: interp_4tap_vert_ps_4x2 sse2
Removed unneeded add instruction. In theory this should provide a small
performance improvement but the testbench shows no change.
diff -r d7b100e51e82 -r 3e07cba4b203 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm Mon May 18 18:24:08 2015 -0500
+++ b/source/common/x86/ipfilter8.asm Tue May 19 14:27:04 2015 -0700
@@ -1098,10 +1098,9 @@
pshufd m3, m3, q3120
punpcklqdq m2, m3
- add r3d, r3d
psubw m2, [pw_2000]
movh [r2], m2
- movhps [r2 + r3], m2
+ movhps [r2 + 2 * r3], m2
%endif
RET
More information about the x265-devel
mailing list