[x265] [PATCH] asm : chroma_hpp[16x4, 16x8, 16x12, 16x32] for i420 avx2 - improved 444c-> 276c, 824c->504c, 1144c->730c, 2913c->1871c

[aasaipriya at multicorewareinc.com]

# HG changeset patch
# User Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
# Date 1426739879 -19800
#      Thu Mar 19 10:07:59 2015 +0530
# Node ID 45b0f63a0fb89a26974ae3b458a319707c258220
# Parent  cef1fc70dbf24acadc3423d56128a7e481795b95
 asm : chroma_hpp[16x4, 16x8, 16x12, 16x32] for i420 avx2 - improved 444c-> 276c, 824c->504c, 1144c->730c, 2913c->1871c

diff -r cef1fc70dbf2 -r 45b0f63a0fb8 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Thu Mar 19 09:59:01 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp	Thu Mar 19 10:07:59 2015 +0530
@@ -1610,6 +1610,11 @@
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x8].filter_hpp = x265_interp_4tap_horiz_pp_4x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x16].filter_hpp = x265_interp_4tap_horiz_pp_4x16_avx2;
 
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x4].filter_hpp = x265_interp_4tap_horiz_pp_16x4_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x8].filter_hpp = x265_interp_4tap_horiz_pp_16x8_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x12].filter_hpp = x265_interp_4tap_horiz_pp_16x12_avx2;
+        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_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;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_32x8].filter_hpp = x265_interp_4tap_horiz_pp_32x8_avx2;
diff -r cef1fc70dbf2 -r 45b0f63a0fb8 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm	Thu Mar 19 09:59:01 2015 +0530
+++ b/source/common/x86/ipfilter8.asm	Thu Mar 19 10:07:59 2015 +0530
@@ -18679,3 +18679,73 @@
     dec               r4d
     jnz              .loop
     RET
+
+;-------------------------------------------------------------------------------------------------------------
+; void interp_4tap_horiz_pp_16xN(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride, int coeffIdx
+;-------------------------------------------------------------------------------------------------------------
+%macro IPFILTER_CHROMA_PP_16xN_AVX2 2
+INIT_YMM avx2
+cglobal interp_4tap_horiz_pp_%1x%2, 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              m6,           [pw_512]
+    mova              m1,           [interp4_horiz_shuf1]
+    vpbroadcastd      m2,           [pw_1]
+
+    ; register map
+    ; m0 - interpolate coeff
+    ; m1 - shuffle order table
+    ; m2 - constant word 1
+
+    dec               r0
+    mov               r4d,          %2/2
+
+.loop:
+    ; 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
+    vbroadcasti128    m4,           [r0 + 4]                    ; [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 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
+    vbroadcasti128    m5,           [r0 + r1 + 4]               ; [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
+    vpermq            m3,           m3,      11011000b
+
+    vextracti128      xm4,          m3,       1
+    movu              [r2],         xm3
+    movu              [r2 + r3],    xm4
+    lea               r2,           [r2 + r3 * 2]
+    lea               r0,           [r0 + r1 * 2]
+    dec               r4d
+    jnz               .loop
+    RET
+%endmacro
+
+IPFILTER_CHROMA_PP_16xN_AVX2 16 , 8
+IPFILTER_CHROMA_PP_16xN_AVX2 16 , 32
+IPFILTER_CHROMA_PP_16xN_AVX2 16 , 12
+IPFILTER_CHROMA_PP_16xN_AVX2 16 , 4