[x265] [PATCH 03 of 12] asm: interp_4tap_vert_ps_4xN sse2
dtyx265 at gmail.com
dtyx265 at gmail.com
Mon May 18 04:48:54 CEST 2015
# HG changeset patch
# User David T Yuen <dtyx265 at gmail.com>
# Date 1431912252 25200
# Node ID c2624b61f4c7d894616a7dc1e8a6cc1c0a506028
# Parent 72bba6b9e99739599d040000be62c7e02a3c8faa
asm: interp_4tap_vert_ps_4xN sse2
Converted vert_pp_4xN macro to also create ps primitives. This replaces c code for ps with minimal impact on pp.
64-bit
./test/TestBench --testbench interp | grep vp | grep " 4x"
chroma_vpp[ 4x4] 2.10x 1005.00 2110.31
chroma_vps[ 4x4] 1.76x 927.49 1634.98
chroma_vpp[ 4x2] 2.17x 515.01 1117.32
chroma_vps[ 4x2] 1.72x 497.49 854.98
chroma_vpp[ 4x8] 2.28x 1928.24 4402.00
chroma_vps[ 4x8] 1.83x 1803.01 3294.14
chroma_vpp[ 4x16] 2.30x 3782.50 8710.95
chroma_vpp[ 4x8] 2.28x 1927.50 4400.15
chroma_vps[ 4x8] 1.84x 1787.48 3294.45
chroma_vpp[ 4x4] 2.11x 1000.00 2109.98
chroma_vps[ 4x4] 1.77x 924.99 1634.97
chroma_vpp[ 4x16] 2.30x 3782.50 8709.96
chroma_vps[ 4x16] 1.90x 3519.99 6698.32
chroma_vpp[ 4x32] 2.27x 7477.50 16995.56
chroma_vpp[ 4x4] 2.10x 1005.00 2109.98
chroma_vps[ 4x4] 1.76x 927.49 1634.98
chroma_vpp[ 4x8] 2.28x 1927.50 4400.14
chroma_vps[ 4x8] 1.83x 1787.48 3270.89
chroma_vpp[ 4x16] 2.30x 3782.51 8709.96
chroma_vps[ 4x16] 1.90x 3517.48 6697.70
32-bit
./test/TestBench --testbench interp | grep vp | grep " 4x"
chroma_vpp[ 4x4] 2.33x 1177.48 2747.42
chroma_vps[ 4x4] 2.47x 1092.47 2702.46
chroma_vpp[ 4x2] 2.44x 579.98 1414.90
chroma_vps[ 4x2] 2.41x 545.91 1314.92
chroma_vpp[ 4x8] 2.58x 2183.79 5640.24
chroma_vps[ 4x8] 2.08x 2020.00 4200.24
chroma_vpp[ 4x16] 2.64x 4202.49 11097.51
chroma_vpp[ 4x8] 2.58x 2187.48 5640.61
chroma_vps[ 4x8] 2.06x 2019.08 4163.00
chroma_vpp[ 4x4] 2.37x 1159.99 2747.38
chroma_vps[ 4x4] 2.47x 1092.49 2702.42
chroma_vpp[ 4x16] 2.64x 4207.48 11097.51
chroma_vps[ 4x16] 2.07x 3887.50 8034.82
chroma_vpp[ 4x32] 2.65x 8247.49 21867.51
chroma_vpp[ 4x4] 2.37x 1159.99 2747.42
chroma_vps[ 4x4] 2.48x 1088.74 2702.42
chroma_vpp[ 4x8] 2.58x 2187.48 5640.30
chroma_vps[ 4x8] 2.06x 2017.49 4162.46
chroma_vpp[ 4x16] 2.64x 4202.49 11097.51
chroma_vps[ 4x16] 2.07x 3889.43 8034.94
diff -r 72bba6b9e997 -r c2624b61f4c7 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp Sun May 17 18:13:35 2015 -0700
+++ b/source/common/x86/asm-primitives.cpp Sun May 17 18:24:12 2015 -0700
@@ -1451,7 +1451,15 @@
p.chroma[X265_CSP_I420].pu[CHROMA_420_2x4].filter_vps = x265_interp_4tap_vert_ps_2x4_sse2;
p.chroma[X265_CSP_I420].pu[CHROMA_420_2x8].filter_vps = x265_interp_4tap_vert_ps_2x8_sse2;
p.chroma[X265_CSP_I420].pu[CHROMA_420_4x2].filter_vps = x265_interp_4tap_vert_ps_4x2_sse2;
+ p.chroma[X265_CSP_I420].pu[CHROMA_420_4x4].filter_vps = x265_interp_4tap_vert_ps_4x4_sse2;
+ p.chroma[X265_CSP_I420].pu[CHROMA_420_4x8].filter_vps = x265_interp_4tap_vert_ps_4x8_sse2;
p.chroma[X265_CSP_I422].pu[CHROMA_422_2x16].filter_vps = x265_interp_4tap_vert_ps_2x16_sse2;
+ p.chroma[X265_CSP_I422].pu[CHROMA_422_4x4].filter_vps = x265_interp_4tap_vert_ps_4x4_sse2;
+ p.chroma[X265_CSP_I422].pu[CHROMA_422_4x8].filter_vps = x265_interp_4tap_vert_ps_4x8_sse2;
+ p.chroma[X265_CSP_I422].pu[CHROMA_422_4x16].filter_vps = x265_interp_4tap_vert_ps_4x16_sse2;
+ p.chroma[X265_CSP_I444].pu[LUMA_4x4].filter_vps = x265_interp_4tap_vert_ps_4x4_sse2;
+ p.chroma[X265_CSP_I444].pu[LUMA_4x8].filter_vps = x265_interp_4tap_vert_ps_4x8_sse2;
+ p.chroma[X265_CSP_I444].pu[LUMA_4x16].filter_vps = x265_interp_4tap_vert_ps_4x16_sse2;
#if X86_64
p.chroma[X265_CSP_I420].pu[CHROMA_420_6x8].filter_vpp = x265_interp_4tap_vert_pp_6x8_sse2;
p.chroma[X265_CSP_I420].pu[CHROMA_420_8x2].filter_vpp = x265_interp_4tap_vert_pp_8x2_sse2;
diff -r 72bba6b9e997 -r c2624b61f4c7 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm Sun May 17 18:13:35 2015 -0700
+++ b/source/common/x86/ipfilter8.asm Sun May 17 18:24:12 2015 -0700
@@ -1111,15 +1111,15 @@
FILTER_V2_W4_H4_sse2 ps
;-----------------------------------------------------------------------------
-; void interp_4tap_vert_pp_%1x%2(pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx)
-;-----------------------------------------------------------------------------
-%macro FILTER_V4_W4_H4_sse2 1
+; void interp_4tap_vert_%1_4x%2(pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx)
+;-----------------------------------------------------------------------------
+%macro FILTER_V4_W4_H4_sse2 2
INIT_XMM sse2
%if ARCH_X86_64
-cglobal interp_4tap_vert_pp_4x%1, 4, 6, 9
+cglobal interp_4tap_vert_%1_4x%2, 4, 6, 9
pxor m8, m8
%else
-cglobal interp_4tap_vert_pp_4x%1, 4, 6, 8
+cglobal interp_4tap_vert_%1_4x%2, 4, 6, 8
%endif
mov r4d, r4m
@@ -1132,12 +1132,18 @@
movh m0, [tabw_ChromaCoeff + r4 * 8]
%endif
+%ifidn %1,pp
mova m1, [pw_32]
+%elifidn %1,ps
+ add r3d, r3d
+ mova m1, [pw_2000]
+%endif
+
lea r5, [3 * r1]
punpcklqdq m0, m0
%assign x 1
-%rep %1/4
+%rep %2/4
movd m2, [r0]
movd m3, [r0 + r1]
movd m4, [r0 + 2 * r1]
@@ -1174,12 +1180,24 @@
pshuflw m7, m3, q2301
pshufhw m7, m7, q2301
paddw m3, m7
+
+%ifidn %1,pp
psrld m2, 16
psrld m3, 16
packssdw m2, m3
-
paddw m2, m1
psraw m2, 6
+%elifidn %1,ps
+ psrldq m2, 2
+ psrldq m3, 2
+ pshufd m2, m2, q3120
+ pshufd m3, m3, q3120
+ punpcklqdq m2, m3
+
+ psubw m2, m1
+ movh [r2], m2
+ movhps [r2 + r3], m2
+%endif
movd m7, [r0 + r1]
@@ -1213,6 +1231,8 @@
pshuflw m7, m5, q2301
pshufhw m7, m7, q2301
paddw m5, m7
+
+%ifidn %1,pp
psrld m4, 16
psrld m5, 16
packssdw m4, m5
@@ -1229,19 +1249,37 @@
movd [r2], m2
psrldq m2, 4
movd [r2 + r3], m2
-
-%if x < %1/4
+%elifidn %1,ps
+ psrldq m4, 2
+ psrldq m5, 2
+ pshufd m4, m4, q3120
+ pshufd m5, m5, q3120
+ punpcklqdq m4, m5
+ psubw m4, m1
lea r2, [r2 + 2 * r3]
-%endif
+ movh [r2], m4
+ movhps [r2 + r3], m4
+%endif
+
+%if x < %2/4
+ lea r2, [r2 + 2 * r3]
+%endif
+
%assign x x+1
%endrep
RET
-%endmacro
-
- FILTER_V4_W4_H4_sse2 4
- FILTER_V4_W4_H4_sse2 8
- FILTER_V4_W4_H4_sse2 16
- FILTER_V4_W4_H4_sse2 32
+
+%endmacro
+
+ FILTER_V4_W4_H4_sse2 pp, 4
+ FILTER_V4_W4_H4_sse2 pp, 8
+ FILTER_V4_W4_H4_sse2 pp, 16
+ FILTER_V4_W4_H4_sse2 pp, 32
+
+ FILTER_V4_W4_H4_sse2 ps, 4
+ FILTER_V4_W4_H4_sse2 ps, 8
+ FILTER_V4_W4_H4_sse2 ps, 16
+ FILTER_V4_W4_H4_sse2 ps, 32
;-----------------------------------------------------------------------------
;void interp_4tap_vert_pp_6x8(pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx)
diff -r 72bba6b9e997 -r c2624b61f4c7 source/common/x86/ipfilter8.h
--- a/source/common/x86/ipfilter8.h Sun May 17 18:13:35 2015 -0700
+++ b/source/common/x86/ipfilter8.h Sun May 17 18:24:12 2015 -0700
@@ -917,6 +917,10 @@
void x265_interp_4tap_vert_ps_2x8_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
void x265_interp_4tap_vert_ps_2x16_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
void x265_interp_4tap_vert_ps_4x2_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
+void x265_interp_4tap_vert_ps_4x4_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
+void x265_interp_4tap_vert_ps_4x8_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
+void x265_interp_4tap_vert_ps_4x16_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
+void x265_interp_4tap_vert_ps_4x32_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
#ifdef X86_64
void x265_interp_4tap_vert_pp_6x8_sse2(const pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx);
void x265_interp_4tap_vert_pp_6x16_sse2(const pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx);
More information about the x265-devel
mailing list