[x265] [PATCH] chroma_hpp[6x16] for i422 - improved 1035c->868c
aasaipriya at multicorewareinc.com
aasaipriya at multicorewareinc.com
Wed May 6 06:35:13 CEST 2015
# HG changeset patch
# User Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
# Date 1430825722 -19800
# Tue May 05 17:05:22 2015 +0530
# Node ID d6fa24967f2313f8cd6ac5a2ea2626f00b23c91a
# Parent 6eca9bacad162fb215d5f209c072e8444f48dd4d
chroma_hpp[6x16] for i422 - improved 1035c->868c
diff -r 6eca9bacad16 -r d6fa24967f23 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp Tue May 05 16:55:51 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp Tue May 05 17:05:22 2015 +0530
@@ -2140,6 +2140,7 @@
p.chroma[X265_CSP_I420].pu[CHROMA_420_16x32].filter_hpp = x265_interp_4tap_horiz_pp_16x32_avx2;
p.chroma[X265_CSP_I420].pu[CHROMA_420_6x8].filter_hpp = x265_interp_4tap_horiz_pp_6x8_avx2;
+ p.chroma[X265_CSP_I422].pu[CHROMA_422_6x16].filter_hpp = x265_interp_4tap_horiz_pp_6x16_avx2;
p.chroma[X265_CSP_I420].pu[CHROMA_420_32x16].filter_hpp = x265_interp_4tap_horiz_pp_32x16_avx2;
p.chroma[X265_CSP_I420].pu[CHROMA_420_32x24].filter_hpp = x265_interp_4tap_horiz_pp_32x24_avx2;
diff -r 6eca9bacad16 -r d6fa24967f23 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm Tue May 05 16:55:51 2015 +0530
+++ b/source/common/x86/ipfilter8.asm Tue May 05 17:05:22 2015 +0530
@@ -24335,3 +24335,69 @@
movd [r2 + r3 * 2], xm2
.end
RET
+
+INIT_YMM avx2
+cglobal interp_4tap_horiz_pp_6x16, 4, 6, 7
+ mov r4d, r4m
+
+%ifdef PIC
+ lea r5, [tab_ChromaCoeff]
+ vpbroadcastd m0, [r5 + r4 * 4]
+%else
+ vpbroadcastd m0, [tab_ChromaCoeff + r4 * 4]
+%endif
+
+ mova m1, [tab_Tm]
+ mova m2, [pw_1]
+ mova m6, [pw_512]
+ lea r4, [r1 * 3]
+ lea r5, [r3 * 3]
+ ; register map
+ ; m0 - interpolate coeff
+ ; m1 - shuffle order table
+ ; m2 - constant word 1
+
+ dec r0
+%rep 4
+ ; Row 0
+ vbroadcasti128 m3, [r0] ; [x x x x x A 9 8 7 6 5 4 3 2 1 0]
+ pshufb m3, m1
+ pmaddubsw m3, m0
+ pmaddwd m3, m2
+
+ ; Row 1
+ vbroadcasti128 m4, [r0 + r1] ; [x x x x x A 9 8 7 6 5 4 3 2 1 0]
+ pshufb m4, m1
+ pmaddubsw m4, m0
+ pmaddwd m4, m2
+ packssdw m3, m4
+ pmulhrsw m3, m6
+
+ ; Row 2
+ vbroadcasti128 m4, [r0 + r1 * 2] ; [x x x x x A 9 8 7 6 5 4 3 2 1 0]
+ pshufb m4, m1
+ pmaddubsw m4, m0
+ pmaddwd m4, m2
+
+ ; Row 3
+ vbroadcasti128 m5, [r0 + r4] ; [x x x x x A 9 8 7 6 5 4 3 2 1 0]
+ pshufb m5, m1
+ pmaddubsw m5, m0
+ pmaddwd m5, m2
+ packssdw m4, m5
+ pmulhrsw m4, m6
+
+ packuswb m3, m4
+ vextracti128 xm4, m3, 1
+ movd [r2], xm3
+ pextrw [r2 + 4], xm4, 0
+ pextrd [r2 + r3], xm3, 1
+ pextrw [r2 + r3 + 4], xm4, 2
+ pextrd [r2 + r3 * 2], xm3, 2
+ pextrw [r2 + r3 * 2 + 4], xm4, 4
+ pextrd [r2 + r5], xm3, 3
+ pextrw [r2 + r5 + 4], xm4, 6
+ lea r2, [r2 + r3 * 4]
+ lea r0, [r0 + r1 * 4]
+%endrep
+ RET
More information about the x265-devel
mailing list