[x265] [PATCH] asm: filter_vpp, filter_vps for 12x32 in avx2
Divya Manivannan
divya at multicorewareinc.com
Thu Apr 30 12:09:55 CEST 2015
# HG changeset patch
# User Divya Manivannan <divya at multicorewareinc.com>
# Date 1430387179 -19800
# Thu Apr 30 15:16:19 2015 +0530
# Node ID f000408d0243a286ed0a56b65779fdc0c040db5c
# Parent 5c9b9856de29c079c268dfe070dd9ece8fd96f56
asm: filter_vpp, filter_vps for 12x32 in avx2
filter_vpp[12x32]: 2307c->1885c
filter_vps[12x32]: 1884c->1612c
diff -r 5c9b9856de29 -r f000408d0243 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp Wed Apr 29 17:11:03 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp Thu Apr 30 15:16:19 2015 +0530
@@ -2537,6 +2537,7 @@
p.chroma[X265_CSP_I422].pu[CHROMA_422_8x64].filter_vps = x265_interp_4tap_vert_ps_8x64_avx2;
p.chroma[X265_CSP_I422].pu[CHROMA_422_32x64].filter_vps = x265_interp_4tap_vert_ps_32x64_avx2;
p.chroma[X265_CSP_I422].pu[CHROMA_422_32x48].filter_vps = x265_interp_4tap_vert_ps_32x48_avx2;
+ p.chroma[X265_CSP_I422].pu[CHROMA_422_12x32].filter_vps = x265_interp_4tap_vert_ps_12x32_avx2;
//i444 for chroma_vps
p.chroma[X265_CSP_I444].pu[LUMA_4x4].filter_vps = x265_interp_4tap_vert_ps_4x4_avx2;
@@ -2578,6 +2579,7 @@
p.chroma[X265_CSP_I422].pu[CHROMA_422_8x64].filter_vpp = x265_interp_4tap_vert_pp_8x64_avx2;
p.chroma[X265_CSP_I422].pu[CHROMA_422_32x64].filter_vpp = x265_interp_4tap_vert_pp_32x64_avx2;
p.chroma[X265_CSP_I422].pu[CHROMA_422_32x48].filter_vpp = x265_interp_4tap_vert_pp_32x48_avx2;
+ p.chroma[X265_CSP_I422].pu[CHROMA_422_12x32].filter_vpp = x265_interp_4tap_vert_pp_12x32_avx2;
//i444 for chroma_vpp
p.chroma[X265_CSP_I444].pu[LUMA_4x4].filter_vpp = x265_interp_4tap_vert_pp_4x4_avx2;
diff -r 5c9b9856de29 -r f000408d0243 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm Wed Apr 29 17:11:03 2015 +0530
+++ b/source/common/x86/ipfilter8.asm Thu Apr 30 15:16:19 2015 +0530
@@ -7560,9 +7560,9 @@
FILTER_VER_CHROMA_AVX2_16x4 pp
FILTER_VER_CHROMA_AVX2_16x4 ps
-%macro FILTER_VER_CHROMA_AVX2_12x16 1
-INIT_YMM avx2
-cglobal interp_4tap_vert_%1_12x16, 4, 7, 8
+%macro FILTER_VER_CHROMA_AVX2_12xN 2
+INIT_YMM avx2
+cglobal interp_4tap_vert_%1_12x%2, 4, 7, 8
mov r4d, r4m
shl r4d, 6
@@ -7582,7 +7582,7 @@
vbroadcasti128 m7, [pw_2000]
%endif
lea r6, [r3 * 3]
-
+%rep %2 / 16
movu xm0, [r0] ; m0 = row 0
movu xm1, [r0 + r1] ; m1 = row 1
punpckhbw xm2, xm0, xm1
@@ -7868,11 +7868,15 @@
vextracti128 xm5, m5, 1
movq [r2 + r6 + 16], xm5
%endif
- RET
-%endmacro
-
- FILTER_VER_CHROMA_AVX2_12x16 pp
- FILTER_VER_CHROMA_AVX2_12x16 ps
+ lea r2, [r2 + r3 * 4]
+%endrep
+ RET
+%endmacro
+
+ FILTER_VER_CHROMA_AVX2_12xN pp, 16
+ FILTER_VER_CHROMA_AVX2_12xN ps, 16
+ FILTER_VER_CHROMA_AVX2_12xN pp, 32
+ FILTER_VER_CHROMA_AVX2_12xN ps, 32
;-----------------------------------------------------------------------------
;void interp_4tap_vert_pp_24x32(pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx)
More information about the x265-devel
mailing list