[x265] [PATCH 4 of 6] asm-intra_pred_ang32_26: improve performance by ~65% over SSE4

[praveen at multicorewareinc.com]

# HG changeset patch
# User Praveen Tiwari <praveen at multicorewareinc.com>
# Date 1426674481 -19800
# Node ID e92e3e58a30c0bc51c3e4bc7f69cdb1308c34a41
# Parent  125516e35a73d951a3f49053536b3a3fecd93e08
asm-intra_pred_ang32_26: improve performance by ~65% over SSE4

AVX2:
intra_ang_32x32[26]     31.05x   367.98          11426.67

SSE4:
intra_ang_32x32[26]     10.78x   1052.93         11354.30

diff -r 125516e35a73 -r e92e3e58a30c source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Wed Mar 18 15:08:26 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp	Wed Mar 18 15:58:01 2015 +0530
@@ -1532,6 +1532,7 @@
         p.cu[BLOCK_16x16].intra_pred[22] = x265_intra_pred_ang16_22_avx2;
         p.cu[BLOCK_32x32].intra_pred[34] = x265_intra_pred_ang32_34_avx2;
         p.cu[BLOCK_32x32].intra_pred[2] = x265_intra_pred_ang32_2_avx2;
+        p.cu[BLOCK_32x32].intra_pred[26] = x265_intra_pred_ang32_26_avx2;
 
         // copy_sp primitives
         p.cu[BLOCK_16x16].copy_sp = x265_blockcopy_sp_16x16_avx2;
diff -r 125516e35a73 -r e92e3e58a30c source/common/x86/intrapred.h
--- a/source/common/x86/intrapred.h	Wed Mar 18 15:08:26 2015 +0530
+++ b/source/common/x86/intrapred.h	Wed Mar 18 15:58:01 2015 +0530
@@ -196,6 +196,7 @@
 void x265_intra_pred_ang16_22_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
 void x265_intra_pred_ang32_34_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
 void x265_intra_pred_ang32_2_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
+void x265_intra_pred_ang32_26_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
 void x265_all_angs_pred_4x4_sse4(pixel *dest, pixel *refPix, pixel *filtPix, int bLuma);
 void x265_all_angs_pred_8x8_sse4(pixel *dest, pixel *refPix, pixel *filtPix, int bLuma);
 void x265_all_angs_pred_16x16_sse4(pixel *dest, pixel *refPix, pixel *filtPix, int bLuma);
diff -r 125516e35a73 -r e92e3e58a30c source/common/x86/intrapred8.asm
--- a/source/common/x86/intrapred8.asm	Wed Mar 18 15:08:26 2015 +0530
+++ b/source/common/x86/intrapred8.asm	Wed Mar 18 15:58:01 2015 +0530
@@ -11381,3 +11381,58 @@
     INTRA_PRED_ANG32_ALIGNR_STORE  7
     INTRA_PRED_ANG32_ALIGNR_STORE 11
     RET
+
+INIT_YMM avx2
+cglobal intra_pred_ang32_26, 3, 4, 1
+    lea    r3, [3 * r1]
+
+    movu    m0,           [r2 + 1]
+
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+
+    lea    r0, [r0 + 4 * r1]
+    movu   [r0],          m0
+    movu   [r0 + r1],     m0
+    movu   [r0 + r1 * 2], m0
+    movu   [r0 + r3],     m0
+    RET
+