[x265] [PATCH 099 of 307] x86: AVX512 interp_4tap_horiz_pp_48x64
mythreyi at multicorewareinc.com
mythreyi at multicorewareinc.com
Sat Apr 7 04:31:37 CEST 2018
# HG changeset patch
# User Vignesh Vijayakumar
# Date 1503911788 -19800
# Mon Aug 28 14:46:28 2017 +0530
# Node ID a7bf0a24cfc8eb8edc95d340b240b91d03dac5bd
# Parent 45e4dd746cfd9380dbe2344a5754a6ff6e9feed5
x86: AVX512 interp_4tap_horiz_pp_48x64
AVX2 performance : 17.53x
AVX512 performance : 33.60x
diff -r 45e4dd746cfd -r a7bf0a24cfc8 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp Mon Aug 28 13:46:50 2017 +0530
+++ b/source/common/x86/asm-primitives.cpp Mon Aug 28 14:46:28 2017 +0530
@@ -4094,6 +4094,7 @@
p.chroma[X265_CSP_I444].pu[LUMA_16x12].filter_hpp = PFX(interp_4tap_horiz_pp_16x12_avx512);
p.chroma[X265_CSP_I444].pu[LUMA_16x4].filter_hpp = PFX(interp_4tap_horiz_pp_16x4_avx512);
p.chroma[X265_CSP_I444].pu[LUMA_16x64].filter_hpp = PFX(interp_4tap_horiz_pp_16x64_avx512);
+ p.chroma[X265_CSP_I444].pu[LUMA_48x64].filter_hpp = PFX(interp_4tap_horiz_pp_48x64_avx512);
//i422 chroma_hpp
p.chroma[X265_CSP_I422].pu[CHROMA_422_16x8].filter_hpp = PFX(interp_4tap_horiz_pp_16x8_avx512);
diff -r 45e4dd746cfd -r a7bf0a24cfc8 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm Mon Aug 28 13:46:50 2017 +0530
+++ b/source/common/x86/ipfilter8.asm Mon Aug 28 14:46:28 2017 +0530
@@ -9949,6 +9949,103 @@
vextracti32x4 [r2 + r7], m5, 3
%endmacro
+%macro PROCESS_IPFILTER_CHROMA_PP_48x4_AVX512 0
+ ; register map
+ ; m0 - interpolate coeff
+ ; m1, m2 - shuffle order table
+ ; m3 - constant word 1
+ ; m4 - constant word 512
+ movu ym5, [r0]
+ vinserti32x8 m5, [r0 + r1], 1
+ movu ym7, [r0 + 4]
+ vinserti32x8 m7, [r0 + r1 + 4], 1
+
+ pshufb m6, m5, m2
+ pshufb m5, m1
+ pshufb m8, m7, m2
+ pshufb m7, m1
+
+ pmaddubsw m5, m0
+ pmaddubsw m7, m0
+ pmaddwd m5, m3
+ pmaddwd m7, m3
+
+ pmaddubsw m6, m0
+ pmaddubsw m8, m0
+ pmaddwd m6, m3
+ pmaddwd m8, m3
+
+ packssdw m5, m7
+ packssdw m6, m8
+ pmulhrsw m5, m4
+ pmulhrsw m6, m4
+ packuswb m5, m6
+ movu [r2], ym5
+ vextracti32x8 [r2 + r3], m5, 1
+
+ movu ym5, [r0 + 2 * r1]
+ vinserti32x8 m5, [r0 + r6], 1
+ movu ym7, [r0 + 2 * r1 + 4]
+ vinserti32x8 m7, [r0 + r6 + 4], 1
+
+ pshufb m6, m5, m2
+ pshufb m5, m1
+ pshufb m8, m7, m2
+ pshufb m7, m1
+
+ pmaddubsw m5, m0
+ pmaddubsw m7, m0
+ pmaddwd m5, m3
+ pmaddwd m7, m3
+
+ pmaddubsw m6, m0
+ pmaddubsw m8, m0
+ pmaddwd m6, m3
+ pmaddwd m8, m3
+
+ packssdw m5, m7
+ packssdw m6, m8
+ pmulhrsw m5, m4
+ pmulhrsw m6, m4
+ packuswb m5, m6
+ movu [r2 + 2 * r3], ym5
+ vextracti32x8 [r2 + r7], m5, 1
+
+ movu xm5, [r0 + mmsize/2]
+ vinserti32x4 m5, [r0 + r1 + mmsize/2], 1
+ vinserti32x4 m5, [r0 + 2 * r1 + mmsize/2], 2
+ vinserti32x4 m5, [r0 + r6 + mmsize/2], 3
+ pshufb m6, m5, m2
+ pshufb m5, m1
+
+ movu xm7, [r0 + 36]
+ vinserti32x4 m7, [r0 + r1 + 36], 1
+ vinserti32x4 m7, [r0 + 2 * r1 + 36], 2
+ vinserti32x4 m7, [r0 + r6 + 36], 3
+ pshufb m8, m7, m2
+ pshufb m7, m1
+
+ pmaddubsw m5, m0
+ pmaddubsw m7, m0
+ pmaddwd m5, m3
+ pmaddwd m7, m3
+
+ pmaddubsw m6, m0
+ pmaddubsw m8, m0
+ pmaddwd m6, m3
+ pmaddwd m8, m3
+
+ packssdw m5, m7
+ packssdw m6, m8
+ pmulhrsw m5, m4
+ pmulhrsw m6, m4
+ packuswb m5, m6
+ movu [r2 + mmsize/2], xm5
+ vextracti32x4 [r2 + r3 + mmsize/2], m5, 1
+ vextracti32x4 [r2 + 2 * r3 + mmsize/2], m5, 2
+ vextracti32x4 [r2 + r7 + mmsize/2], m5, 3
+%endmacro
+
;-------------------------------------------------------------------------------------------------------------
; void interp_4tap_horiz_pp_64xN(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride, int coeffIdx
;-------------------------------------------------------------------------------------------------------------
@@ -10054,6 +10151,32 @@
IPFILTER_CHROMA_PP_16xN_AVX512 32
IPFILTER_CHROMA_PP_16xN_AVX512 64
+INIT_ZMM avx512
+cglobal interp_4tap_horiz_pp_48x64, 4,8,9
+ mov r4d, r4m
+ lea r6, [3 * r1]
+ lea r7, [3 * r3]
+%ifdef PIC
+ lea r5, [tab_ChromaCoeff]
+ vpbroadcastd m0, [r5 + r4 * 4]
+%else
+ vpbroadcastd m0, [tab_ChromaCoeff + r4 * 4]
+%endif
+
+ vbroadcasti32x8 m1, [interp4_horiz_shuf_load1_avx512]
+ vbroadcasti32x8 m2, [interp4_horiz_shuf_load2_avx512]
+ vbroadcasti32x8 m3, [pw_1]
+ vbroadcasti32x8 m4, [pw_512]
+ dec r0
+
+%rep 15
+ PROCESS_IPFILTER_CHROMA_PP_48x4_AVX512
+ lea r2, [r2 + 4 * r3]
+ lea r0, [r0 + 4 * r1]
+%endrep
+ PROCESS_IPFILTER_CHROMA_PP_48x4_AVX512
+ RET
+
%macro PROCESS_IPFILTER_CHROMA_PS_64x1_AVX512 0
movu ym6, [r0]
vinserti32x8 m6, [r0 + 4], 1
More information about the x265-devel
mailing list