[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