[x265] [PATCH] asm-avx2: filter_vpp[8x32], filter_vps[8x32]: 1028c->937c, 902c->860c

[Divya Manivannan]

# HG changeset patch
# User Divya Manivannan <divya at multicorewareinc.com>
# Date 1425551071 -19800
#      Thu Mar 05 15:54:31 2015 +0530
# Node ID be5f8a18fd5038212147f1446ea4f4dd309651e7
# Parent  dc38f1755137e00af00e358bffa9dedc710a2916
asm-avx2: filter_vpp[8x32], filter_vps[8x32]: 1028c->937c, 902c->860c

diff -r dc38f1755137 -r be5f8a18fd50 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Thu Mar 05 15:08:04 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp	Thu Mar 05 15:54:31 2015 +0530
@@ -1644,6 +1644,7 @@
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x4].filter_vpp = x265_interp_4tap_vert_pp_8x4_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x6].filter_vpp = x265_interp_4tap_vert_pp_8x6_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x16].filter_vpp = x265_interp_4tap_vert_pp_8x16_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_8x32].filter_vpp = x265_interp_4tap_vert_pp_8x32_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x8].filter_vpp = x265_interp_4tap_vert_pp_16x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x4].filter_vpp = x265_interp_4tap_vert_pp_16x4_avx2;
 
@@ -1655,6 +1656,7 @@
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x6].filter_vps = x265_interp_4tap_vert_ps_8x6_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x8].filter_vps = x265_interp_4tap_vert_ps_8x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x16].filter_vps = x265_interp_4tap_vert_ps_8x16_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_8x32].filter_vps = x265_interp_4tap_vert_ps_8x32_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x8].filter_vps = x265_interp_4tap_vert_ps_16x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x4].filter_vps = x265_interp_4tap_vert_ps_16x4_avx2;
 
diff -r dc38f1755137 -r be5f8a18fd50 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm	Thu Mar 05 15:08:04 2015 +0530
+++ b/source/common/x86/ipfilter8.asm	Thu Mar 05 15:54:31 2015 +0530
@@ -4499,6 +4499,38 @@
 FILTER_VER_CHROMA_AVX2_8x16 pp
 FILTER_VER_CHROMA_AVX2_8x16 ps
 
+%macro FILTER_VER_CHROMA_AVX2_8x32 1
+INIT_YMM avx2
+cglobal interp_4tap_vert_%1_8x32, 4, 7, 8
+    mov             r4d, r4m
+    shl             r4d, 6
+
+%ifdef PIC
+    lea             r5, [tab_ChromaCoeffVer_32]
+    add             r5, r4
+%else
+    lea             r5, [tab_ChromaCoeffVer_32 + r4]
+%endif
+
+    lea             r4, [r1 * 3]
+    sub             r0, r1
+%ifidn %1,pp
+    mova            m7, [pw_512]
+%else
+    add             r3d, r3d
+    mova            m7, [pw_2000]
+%endif
+    lea             r6, [r3 * 3]
+%rep 2
+    PROCESS_CHROMA_AVX2_W8_16R %1
+    lea             r2, [r2 + r3 * 4]
+%endrep
+    RET
+%endmacro
+
+FILTER_VER_CHROMA_AVX2_8x32 pp
+FILTER_VER_CHROMA_AVX2_8x32 ps
+
 %macro PROCESS_CHROMA_AVX2_W8_4R 0
     movq            xm1, [r0]                       ; m1 = row 0
     movq            xm2, [r0 + r1]                  ; m2 = row 1