[x265] [PATCH] asm: chroma_hpp[32x16, 32x24, 32x8] for i420 - improved 2966c->1647c, 4514c->2627c, 1494c->870c

[aasaipriya at multicorewareinc.com]

# HG changeset patch
# User Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
# Date 1426482389 -19800
#      Mon Mar 16 10:36:29 2015 +0530
# Node ID 3c310c18cc2026f464c7ed1fa4369e1df6105843
# Parent  cdac951240a3f9a174733d5c26418d2414352e31
asm: chroma_hpp[32x16, 32x24, 32x8] for i420 - improved 2966c->1647c, 4514c->2627c, 1494c->870c

diff -r cdac951240a3 -r 3c310c18cc20 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Mon Mar 16 10:32:43 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp	Mon Mar 16 10:36:29 2015 +0530
@@ -1583,6 +1583,10 @@
 
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x2].filter_hpp = x265_interp_4tap_horiz_pp_4x2_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;
+
         p.chroma[X265_CSP_I420].pu[CHROMA_420_32x32].filter_hps = x265_interp_4tap_horiz_ps_32x32_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x16].filter_hps = x265_interp_4tap_horiz_ps_16x16_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x4].filter_hps = x265_interp_4tap_horiz_ps_4x4_avx2;
diff -r cdac951240a3 -r 3c310c18cc20 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm	Mon Mar 16 10:32:43 2015 +0530
+++ b/source/common/x86/ipfilter8.asm	Mon Mar 16 10:36:29 2015 +0530
@@ -16725,3 +16725,68 @@
     movd              [r2],         xm2
     pextrd            [r2+r3],      xm2,     2
     RET
+
+;-------------------------------------------------------------------------------------------------------------
+; void interp_4tap_horiz_pp_32xN(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride, int coeffIdx
+;-------------------------------------------------------------------------------------------------------------
+%macro IPFILTER_CHROMA_PP_32xN_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              m1,           [interp4_horiz_shuf1]
+    vpbroadcastd      m2,           [pw_1]
+    mova              m6,           [pw_512]
+    ; register map
+    ; m0 - interpolate coeff
+    ; m1 - shuffle order table
+    ; m2 - constant word 1
+
+    dec               r0
+    mov               r4d,          %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]
+    pshufb            m4,           m1
+    pmaddubsw         m4,           m0
+    pmaddwd           m4,           m2
+    packssdw          m3,           m4
+    pmulhrsw          m3,           m6
+
+    vbroadcasti128    m4,           [r0 + 16]
+    pshufb            m4,           m1
+    pmaddubsw         m4,           m0
+    pmaddwd           m4,           m2
+    vbroadcasti128    m5,           [r0 + 20]
+    pshufb            m5,           m1
+    pmaddubsw         m5,           m0
+    pmaddwd           m5,           m2
+    packssdw          m4,           m5
+    pmulhrsw          m4,           m6
+
+    packuswb          m3,           m4
+    vpermq            m3,           m3,      11011000b
+
+    movu              [r2],         m3
+    add               r2,           r3
+    add               r0,           r1
+    dec               r4d
+    jnz               .loop
+    RET
+%endmacro
+
+IPFILTER_CHROMA_PP_32xN_AVX2 32, 16
+IPFILTER_CHROMA_PP_32xN_AVX2 32, 24
+IPFILTER_CHROMA_PP_32xN_AVX2 32, 8