<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Nov 14, 2013 at 5:00 AM, <span dir="ltr"><<a href="mailto:nabajit@multicorewareinc.com" target="_blank">nabajit@multicorewareinc.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"># HG changeset patch<br>
# User Nabajit Deka<br>
# Date 1384426801 -19800<br>
# Thu Nov 14 16:30:01 2013 +0530<br>
# Node ID 373fa609f3309420e5d5a9b3227d41757d315ac5<br>
# Parent 5683ee5b793cca5956f1e44e4e0bb3d6be70e942<br>
asm: routines for chroma vps filter functions for 2x4 and 2x8 block sizes<br></blockquote><div><br></div><div>already queued, thanks</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
diff -r 5683ee5b793c -r 373fa609f330 source/common/x86/ipfilter8.asm<br>
--- a/source/common/x86/ipfilter8.asm Wed Nov 13 13:53:13 2013 +0000<br>
+++ b/source/common/x86/ipfilter8.asm Thu Nov 14 16:30:01 2013 +0530<br>
@@ -1401,6 +1401,164 @@<br>
<br>
RET<br>
<br>
+;------------------------------------------------------------------------------------------------------------<br>
+;void interp_4tap_vert_ps_2x4(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride, int coeffIdx)<br>
+;------------------------------------------------------------------------------------------------------------<br>
+INIT_XMM sse4<br>
+cglobal interp_4tap_vert_ps_2x4, 4, 7, 8<br>
+<br>
+ mov r4d, r4m<br>
+ sub r0, r1<br>
+ add r3d, r3d<br>
+<br>
+%ifdef PIC<br>
+ lea r5, [tab_ChromaCoeff]<br>
+ movd m0, [r5 + r4 * 4]<br>
+%else<br>
+ movd m0, [tab_ChromaCoeff + r4 * 4]<br>
+%endif<br>
+<br>
+ pshufb m0, [tab_Cm]<br>
+<br>
+ mova m1, [tab_c_8192]<br>
+<br>
+ movd m2, [r0]<br>
+ movd m3, [r0 + r1]<br>
+ movd m4, [r0 + 2 * r1]<br>
+ lea r5, [r0 + 2 * r1]<br>
+ movd m5, [r5 + r1]<br>
+<br>
+ punpcklbw m2, m3<br>
+ punpcklbw m6, m4, m5<br>
+ punpcklbw m2, m6<br>
+<br>
+ pmaddubsw m2, m0<br>
+<br>
+ movd m6, [r0 + 4 * r1]<br>
+<br>
+ punpcklbw m3, m4<br>
+ punpcklbw m7, m5, m6<br>
+ punpcklbw m3, m7<br>
+<br>
+ pmaddubsw m3, m0<br>
+ phaddw m2, m3<br>
+ psubw m2, m1<br>
+<br>
+ movd [r2], m2<br>
+ movhlps m2, m2<br>
+ movd [r2 + r3], m2<br>
+<br>
+ lea r5, [r0 + 4 * r1]<br>
+ movd m2, [r5 + r1]<br>
+<br>
+ punpcklbw m4, m5<br>
+ punpcklbw m3, m6, m2<br>
+ punpcklbw m4, m3<br>
+<br>
+ pmaddubsw m4, m0<br>
+<br>
+ movd m3, [r5 + 2 * r1]<br>
+<br>
+ punpcklbw m5, m6<br>
+ punpcklbw m2, m3<br>
+ punpcklbw m5, m2<br>
+<br>
+ pmaddubsw m5, m0<br>
+ phaddw m4, m5<br>
+ psubw m4, m1<br>
+<br>
+ movd [r2 + 2 * r3], m4<br>
+ movhlps m4, m4<br>
+ lea r6, [r2 + 2 * r3]<br>
+ movd [r6 + r3], m4<br>
+<br>
+ RET<br>
+<br>
+;-------------------------------------------------------------------------------------------------------------<br>
+; void interp_4tap_vert_ps_2x8(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride, int coeffIdx)<br>
+;-------------------------------------------------------------------------------------------------------------<br>
+INIT_XMM sse4<br>
+cglobal interp_4tap_vert_ps_2x8, 4, 7, 8<br>
+<br>
+ mov r4d, r4m<br>
+ sub r0, r1<br>
+ add r3d, r3d<br>
+<br>
+%ifdef PIC<br>
+ lea r5, [tab_ChromaCoeff]<br>
+ movd m0, [r5 + r4 * 4]<br>
+%else<br>
+ movd m0, [tab_ChromaCoeff + r4 * 4]<br>
+%endif<br>
+<br>
+ pshufb m0, [tab_Cm]<br>
+<br>
+ mova m1, [tab_c_8192]<br>
+<br>
+ mov r4d, 2<br>
+.loop<br>
+ movd m2, [r0]<br>
+ movd m3, [r0 + r1]<br>
+ movd m4, [r0 + 2 * r1]<br>
+ lea r5, [r0 + 2 * r1]<br>
+ movd m5, [r5 + r1]<br>
+<br>
+ punpcklbw m2, m3<br>
+ punpcklbw m6, m4, m5<br>
+ punpcklbw m2, m6<br>
+<br>
+ pmaddubsw m2, m0<br>
+<br>
+ movd m6, [r0 + 4 * r1]<br>
+<br>
+ punpcklbw m3, m4<br>
+ punpcklbw m7, m5, m6<br>
+ punpcklbw m3, m7<br>
+<br>
+ pmaddubsw m3, m0<br>
+<br>
+ phaddw m2, m3<br>
+ psubw m2, m1<br>
+<br>
+<br>
+ movd [r2], m2<br>
+ movhlps m2, m2<br>
+ movd [r2 + r3], m2<br>
+<br>
+ lea r5, [r0 + 4 * r1]<br>
+ movd m2, [r5 + r1]<br>
+<br>
+ punpcklbw m4, m5<br>
+ punpcklbw m3, m6, m2<br>
+ punpcklbw m4, m3<br>
+<br>
+ pmaddubsw m4, m0<br>
+<br>
+ movd m3, [r5 + 2 * r1]<br>
+<br>
+ punpcklbw m5, m6<br>
+ punpcklbw m2, m3<br>
+ punpcklbw m5, m2<br>
+<br>
+ pmaddubsw m5, m0<br>
+<br>
+ phaddw m4, m5<br>
+<br>
+ psubw m4, m1<br>
+<br>
+ movd [r2 + 2 * r3], m4<br>
+ lea r6, [r2 + 2 * r3]<br>
+ movhlps m4, m4<br>
+ movd [r6 + r3], m4<br>
+<br>
+ lea r0, [r0 + 4 * r1]<br>
+ lea r2, [r2 + 4 * r3]<br>
+<br>
+ dec r4d<br>
+ jnz .loop<br>
+<br>
+RET<br>
+<br>
;-------------------------------------------------------------------------------------------------------------<br>
; void interp_4tap_vert_ps_4x2(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride, int coeffIdx)<br>
;-------------------------------------------------------------------------------------------------------------<br>
_______________________________________________<br>
x265-devel mailing list<br>
<a href="mailto:x265-devel@videolan.org">x265-devel@videolan.org</a><br>
<a href="https://mailman.videolan.org/listinfo/x265-devel" target="_blank">https://mailman.videolan.org/listinfo/x265-devel</a><br>
</blockquote></div><br><br clear="all"><div><br></div>-- <br>Steve Borho
</div></div>