<div dir="ltr"><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)"><br></div><div class="gmail_quote">---------- Forwarded message ----------<br>From: <b class="gmail_sendername"></b> <span dir="ltr"><<a href="mailto:vignesh@multicorewareinc.com">vignesh@multicorewareinc.com</a>></span><br>Date: Tue, May 2, 2017 at 3:16 PM<br>Subject: [x265] [PATCH 1 of 3] SEA motion search:Setup asm primitives for integral calculation<br>To: <a href="mailto:x265-devel@videolan.org">x265-devel@videolan.org</a><br><br><br># HG changeset patch<br>
# User Vignesh Vijayakumar<br>
# Date 1493699293 -19800<br>
# Tue May 02 09:58:13 2017 +0530<br>
# Node ID cb67dffd0e2a596c8d3c6d042b8e6c<wbr>532487d427<br>
# Parent 5bc5e73760cdb61d2674e74cc52149<wbr>fa0603af8a<br>
SEA motion search:Setup asm primitives for integral calculation<br>
<br>
diff -r 5bc5e73760cd -r cb67dffd0e2a source/common/primitives.cpp<br>
--- a/source/common/primitives.cpp Sat Apr 22 17:00:28 2017 -0700<br>
+++ b/source/common/primitives.cpp Tue May 02 09:58:13 2017 +0530<br>
@@ -57,6 +57,7 @@<br>
void setupIntraPrimitives_c(<wbr>EncoderPrimitives &p);<br>
void setupLoopFilterPrimitives_c(<wbr>EncoderPrimitives &p);<br>
void setupSaoPrimitives_c(<wbr>EncoderPrimitives &p);<br>
+void setupSeaIntegralPrimitives_c(<wbr>EncoderPrimitives &p);<br>
<br>
void setupCPrimitives(<wbr>EncoderPrimitives &p)<br>
{<br>
@@ -66,6 +67,7 @@<br>
setupIntraPrimitives_c(p); // intrapred.cpp<br>
setupLoopFilterPrimitives_c(p)<wbr>; // loopfilter.cpp<br>
setupSaoPrimitives_c(p); // sao.cpp<br>
+ setupSeaIntegralPrimitives_c(<wbr>p); // framefilter.cpp<br>
}<br>
<br>
void setupAliasPrimitives(<wbr>EncoderPrimitives &p)<br>
diff -r 5bc5e73760cd -r cb67dffd0e2a source/common/primitives.h<br>
--- a/source/common/primitives.h Sat Apr 22 17:00:28 2017 -0700<br>
+++ b/source/common/primitives.h Tue May 02 09:58:13 2017 +0530<br>
@@ -202,6 +202,18 @@<br>
<br>
typedef void (*pelFilterLumaStrong_t)(<wbr>pixel* src, intptr_t srcStep, intptr_t offset, int32_t tcP, int32_t tcQ);<br>
typedef void (*pelFilterChroma_t)(pixel* src, intptr_t srcStep, intptr_t offset, int32_t tc, int32_t maskP, int32_t maskQ);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline"></div>typedef void(*integral4h_t)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral8h_t)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral12h_t)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral16h_t)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral24h_t)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral32h_t)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline"></div><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline"></div>typedef void(*integral4v_t)(uint32_t *sum, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral8v_t)(uint32_t *sum, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral12v_t)(uint32_t *sum, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral16v_t)(uint32_t *sum, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral24v_t)(uint32_t *sum, intptr_t stride);<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+typedef void(*integral32v_t)(uint32_t *sum, intptr_t stride);<br>
<br><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)">Just needed two typedef here, one for horitontal and one for vertical rest of the typedef are redudent here. </div><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)"><br></div>
/* Function pointers to optimized encoder primitives. Each pointer can reference<br>
* either an assembly routine, a SIMD intrinsic primitive, or a C function */<br>
@@ -342,6 +354,19 @@<br>
pelFilterLumaStrong_t pelFilterLumaStrong[2]; // EDGE_VER = 0, EDGE_HOR = 1<br>
pelFilterChroma_t pelFilterChroma[2]; // EDGE_VER = 0, EDGE_HOR = 1<br>
<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral4h_t integral_init4h;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral8h_t integral_init8h;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral12h_t integral_init12h;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral16h_t integral_init16h;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral24h_t integral_init24h;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral32h_t integral_init32h;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral4v_t integral_init4v;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral8v_t integral_init8v;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral12v_t integral_init12v;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral16v_t integral_init16v;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral24v_t integral_init24v;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+ integral32v_t integral_init32v;<br>
<div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0);display:inline">>></div>+</div><div class="gmail_quote"><br></div><div class="gmail_quote"><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)">An array of appropiate size for horizontal and another for vertical.</div><br></div><div class="gmail_quote"><br>
/* There is one set of chroma primitives per color space. An encoder will<br>
* have just a single color space and thus it will only ever use one entry<br>
* in this array. However we always fill all entries in the array in case<br>
diff -r 5bc5e73760cd -r cb67dffd0e2a source/common/x86/asm-<wbr>primitives.cpp<br>
--- a/source/common/x86/asm-<wbr>primitives.cpp Sat Apr 22 17:00:28 2017 -0700<br>
+++ b/source/common/x86/asm-<wbr>primitives.cpp Tue May 02 09:58:13 2017 +0530<br>
@@ -114,6 +114,7 @@<br>
#include "blockcopy8.h"<br>
#include "intrapred.h"<br>
#include "dct8.h"<br>
+#include "seaintegral.h"<br>
}<br>
<br>
#define ALL_LUMA_CU_TYPED(prim, fncdef, fname, cpu) \<br>
diff -r 5bc5e73760cd -r cb67dffd0e2a source/common/x86/seaintegral.<wbr>h<br>
--- /dev/null Thu Jan 01 00:00:00 1970 +0000<br>
+++ b/source/common/x86/<wbr>seaintegral.h Tue May 02 09:58:13 2017 +0530<br>
@@ -0,0 +1,41 @@<br>
+/****************************<wbr>******************************<wbr>*******************<br>
+* Copyright (C) 2013-2017 MulticoreWare, Inc<br>
+*<br>
+* Authors: Vignesh V Menon <<a href="mailto:vignesh@multicorewareinc.com">vignesh@multicorewareinc.com</a>><br>
+* Jayashri Murugan <<a href="mailto:jayashri@multicorewareinc.com">jayashri@multicorewareinc.com</a><wbr>><br>
+*<br>
+* This program is free software; you can redistribute it and/or modify<br>
+* it under the terms of the GNU General Public License as published by<br>
+* the Free Software Foundation; either version 2 of the License, or<br>
+* (at your option) any later version.<br>
+*<br>
+* This program is distributed in the hope that it will be useful,<br>
+* but WITHOUT ANY WARRANTY; without even the implied warranty of<br>
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the<br>
+* GNU General Public License for more details.<br>
+*<br>
+* You should have received a copy of the GNU General Public License<br>
+* along with this program; if not, write to the Free Software<br>
+* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02111, USA.<br>
+*<br>
+* This program is also available under a commercial proprietary license.<br>
+* For more information, contact us at license @ <a href="http://x265.com" rel="noreferrer" target="_blank">x265.com</a>.<br>
+*****************************<wbr>******************************<wbr>******************/<br>
+<br>
+#ifndef X265_SEAINTEGRAL_H<br>
+#define X265_SEAINTEGRAL_H<br>
+<br>
+void PFX(integral4h_avx2)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
+void PFX(integral8h_avx2)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
+void PFX(integral12h_avx2)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
+void PFX(integral16h_avx2)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
+void PFX(integral24h_avx2)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
+void PFX(integral32h_avx2)(uint32_t *sum, pixel *pix, intptr_t stride);<br>
+void PFX(integral4v_avx2)(uint32_t *sum, intptr_t stride);<br>
+void PFX(integral8v_avx2)(uint32_t *sum, intptr_t stride);<br>
+void PFX(integral12v_avx2)(uint32_t *sum, intptr_t stride);<br>
+void PFX(integral16v_avx2)(uint32_t *sum, intptr_t stride);<br>
+void PFX(integral24v_avx2)(uint32_t *sum, intptr_t stride);<br>
+void PFX(integral32v_avx2)(uint32_t *sum, intptr_t stride);<br>
+<br>
+#endif //X265_SEAINTEGRAL_H<br>
diff -r 5bc5e73760cd -r cb67dffd0e2a source/encoder/framefilter.cpp<br>
--- a/source/encoder/framefilter.<wbr>cpp Sat Apr 22 17:00:28 2017 -0700<br>
+++ b/source/encoder/framefilter.<wbr>cpp Tue May 02 09:58:13 2017 +0530<br>
@@ -34,108 +34,126 @@<br>
<br>
static uint64_t computeSSD(pixel *fenc, pixel *rec, intptr_t stride, uint32_t width, uint32_t height);<br>
static float calculateSSIM(pixel *pix1, intptr_t stride1, pixel *pix2, intptr_t stride2, uint32_t width, uint32_t height, void *buf, uint32_t& cnt);<br>
+namespace X265_NS<br>
+{<br>
+ void integral_init4h_c(uint32_t *sum, pixel *pix, intptr_t stride)<br>
+ {<br>
+ int32_t v = pix[0] + pix[1] + pix[2] + pix[3];<br>
+ for (int16_t x = 0; x < stride - 4; x++)<br>
+ {<br>
+ sum[x] = v + sum[x - stride];<br>
+ v += pix[x + 4] - pix[x];<br>
+ }<br>
+ }<br>
<br><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)">Why removed "ststic" from C primitives?</div><br><br>
-static void integral_init4h(uint32_t *sum, pixel *pix, intptr_t stride)<br>
-{<br>
- int32_t v = pix[0] + pix[1] + pix[2] + pix[3];<br>
- for (int16_t x = 0; x < stride - 4; x++)<br>
+ void integral_init8h_c(uint32_t *sum, pixel *pix, intptr_t stride)<br>
{<br>
- sum[x] = v + sum[x - stride];<br>
- v += pix[x + 4] - pix[x];<br>
+ int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7];<br>
+ for (int16_t x = 0; x < stride - 8; x++)<br>
+ {<br>
+ sum[x] = v + sum[x - stride];<br>
+ v += pix[x + 8] - pix[x];<br>
+ }<br>
}<br>
-}<br>
<br>
-static void integral_init8h(uint32_t *sum, pixel *pix, intptr_t stride)<br>
-{<br>
- int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7];<br>
- for (int16_t x = 0; x < stride - 8; x++)<br>
+ void integral_init12h_c(uint32_t *sum, pixel *pix, intptr_t stride)<br>
{<br>
- sum[x] = v + sum[x - stride];<br>
- v += pix[x + 8] - pix[x];<br>
+ int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
+ pix[8] + pix[9] + pix[10] + pix[11];<br>
+ for (int16_t x = 0; x < stride - 12; x++)<br>
+ {<br>
+ sum[x] = v + sum[x - stride];<br>
+ v += pix[x + 12] - pix[x];<br>
+ }<br>
}<br>
-}<br>
<br>
-static void integral_init12h(uint32_t *sum, pixel *pix, intptr_t stride)<br>
-{<br>
- int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
- pix[8] + pix[9] + pix[10] + pix[11];<br>
- for (int16_t x = 0; x < stride - 12; x++)<br>
+ void integral_init16h_c(uint32_t *sum, pixel *pix, intptr_t stride)<br>
{<br>
- sum[x] = v + sum[x - stride];<br>
- v += pix[x + 12] - pix[x];<br>
+ int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
+ pix[8] + pix[9] + pix[10] + pix[11] + pix[12] + pix[13] + pix[14] + pix[15];<br>
+ for (int16_t x = 0; x < stride - 16; x++)<br>
+ {<br>
+ sum[x] = v + sum[x - stride];<br>
+ v += pix[x + 16] - pix[x];<br>
+ }<br>
}<br>
-}<br>
<br>
-static void integral_init16h(uint32_t *sum, pixel *pix, intptr_t stride)<br>
-{<br>
- int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
- pix[8] + pix[9] + pix[10] + pix[11] + pix[12] + pix[13] + pix[14] + pix[15];<br>
- for (int16_t x = 0; x < stride - 16; x++)<br>
+ void integral_init24h_c(uint32_t *sum, pixel *pix, intptr_t stride)<br>
{<br>
- sum[x] = v + sum[x - stride];<br>
- v += pix[x + 16] - pix[x];<br>
+ int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
+ pix[8] + pix[9] + pix[10] + pix[11] + pix[12] + pix[13] + pix[14] + pix[15] +<br>
+ pix[16] + pix[17] + pix[18] + pix[19] + pix[20] + pix[21] + pix[22] + pix[23];<br>
+ for (int16_t x = 0; x < stride - 24; x++)<br>
+ {<br>
+ sum[x] = v + sum[x - stride];<br>
+ v += pix[x + 24] - pix[x];<br>
+ }<br>
}<br>
-}<br>
<br>
-static void integral_init24h(uint32_t *sum, pixel *pix, intptr_t stride)<br>
-{<br>
- int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
- pix[8] + pix[9] + pix[10] + pix[11] + pix[12] + pix[13] + pix[14] + pix[15] +<br>
- pix[16] + pix[17] + pix[18] + pix[19] + pix[20] + pix[21] + pix[22] + pix[23];<br>
- for (int16_t x = 0; x < stride - 24; x++)<br>
+ void integral_init32h_c(uint32_t *sum, pixel *pix, intptr_t stride)<br>
{<br>
- sum[x] = v + sum[x - stride];<br>
- v += pix[x + 24] - pix[x];<br>
+ int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
+ pix[8] + pix[9] + pix[10] + pix[11] + pix[12] + pix[13] + pix[14] + pix[15] +<br>
+ pix[16] + pix[17] + pix[18] + pix[19] + pix[20] + pix[21] + pix[22] + pix[23] +<br>
+ pix[24] + pix[25] + pix[26] + pix[27] + pix[28] + pix[29] + pix[30] + pix[31];<br>
+ for (int16_t x = 0; x < stride - 32; x++)<br>
+ {<br>
+ sum[x] = v + sum[x - stride];<br>
+ v += pix[x + 32] - pix[x];<br>
+ }<br>
}<br>
-}<br>
<br>
-static void integral_init32h(uint32_t *sum, pixel *pix, intptr_t stride)<br>
-{<br>
- int32_t v = pix[0] + pix[1] + pix[2] + pix[3] + pix[4] + pix[5] + pix[6] + pix[7] +<br>
- pix[8] + pix[9] + pix[10] + pix[11] + pix[12] + pix[13] + pix[14] + pix[15] +<br>
- pix[16] + pix[17] + pix[18] + pix[19] + pix[20] + pix[21] + pix[22] + pix[23] +<br>
- pix[24] + pix[25] + pix[26] + pix[27] + pix[28] + pix[29] + pix[30] + pix[31];<br>
- for (int16_t x = 0; x < stride - 32; x++)<br>
+ void integral_init4v_c(uint32_t *sum4, intptr_t stride)<br>
{<br>
- sum[x] = v + sum[x - stride];<br>
- v += pix[x + 32] - pix[x];<br>
+ for (int x = 0; x < stride; x++)<br>
+ sum4[x] = sum4[x + 4 * stride] - sum4[x];<br>
}<br>
-}<br>
<br>
-static void integral_init4v(uint32_t *sum4, intptr_t stride)<br>
-{<br>
- for (int x = 0; x < stride; x++)<br>
- sum4[x] = sum4[x + 4 * stride] - sum4[x];<br>
-}<br>
+ void integral_init8v_c(uint32_t *sum8, intptr_t stride)<br>
+ {<br>
+ for (int x = 0; x < stride; x++)<br>
+ sum8[x] = sum8[x + 8 * stride] - sum8[x];<br>
+ }<br>
<br>
-static void integral_init8v(uint32_t *sum8, intptr_t stride)<br>
-{<br>
- for (int x = 0; x < stride; x++)<br>
- sum8[x] = sum8[x + 8 * stride] - sum8[x];<br>
-}<br>
+ void integral_init12v_c(uint32_t *sum12, intptr_t stride)<br>
+ {<br>
+ for (int x = 0; x < stride; x++)<br>
+ sum12[x] = sum12[x + 12 * stride] - sum12[x];<br>
+ }<br>
<br>
-static void integral_init12v(uint32_t *sum12, intptr_t stride)<br>
-{<br>
- for (int x = 0; x < stride; x++)<br>
- sum12[x] = sum12[x + 12 * stride] - sum12[x];<br>
-}<br>
+ void integral_init16v_c(uint32_t *sum16, intptr_t stride)<br>
+ {<br>
+ for (int x = 0; x < stride; x++)<br>
+ sum16[x] = sum16[x + 16 * stride] - sum16[x];<br>
+ }<br>
<br>
-static void integral_init16v(uint32_t *sum16, intptr_t stride)<br>
-{<br>
- for (int x = 0; x < stride; x++)<br>
- sum16[x] = sum16[x + 16 * stride] - sum16[x];<br>
-}<br>
+ void integral_init24v_c(uint32_t *sum24, intptr_t stride)<br>
+ {<br>
+ for (int x = 0; x < stride; x++)<br>
+ sum24[x] = sum24[x + 24 * stride] - sum24[x];<br>
+ }<br>
<br>
-static void integral_init24v(uint32_t *sum24, intptr_t stride)<br>
-{<br>
- for (int x = 0; x < stride; x++)<br>
- sum24[x] = sum24[x + 24 * stride] - sum24[x];<br>
-}<br>
+ void integral_init32v_c(uint32_t *sum32, intptr_t stride)<br>
+ {<br>
+ for (int x = 0; x < stride; x++)<br>
+ sum32[x] = sum32[x + 32 * stride] - sum32[x];<br>
+ }<br>
<br>
-static void integral_init32v(uint32_t *sum32, intptr_t stride)<br>
-{<br>
- for (int x = 0; x < stride; x++)<br>
- sum32[x] = sum32[x + 32 * stride] - sum32[x];<br>
+ void setupSeaIntegralPrimitives_c(<wbr>EncoderPrimitives &p)<br>
+ {<br>
+ p.integral_init4h = integral_init4h_c;<br>
+ p.integral_init8h = integral_init8h_c;<br>
+ p.integral_init12h = integral_init12h_c;<br>
+ p.integral_init16h = integral_init16h_c;<br>
+ p.integral_init24h = integral_init24h_c;<br>
+ p.integral_init32h = integral_init32h_c;<br>
+ p.integral_init4v = integral_init4v_c;<br>
+ p.integral_init8v = integral_init8v_c;<br>
+ p.integral_init12v = integral_init12v_c;<br>
+ p.integral_init16v = integral_init16v_c;<br>
+ p.integral_init24v = integral_init24v_c;<br>
+ p.integral_init32v = integral_init32v_c;<br>
+ }<br>
}<br>
<br>
void FrameFilter::destroy()<br>
@@ -833,47 +851,47 @@<br>
uint32_t *sum4x4 = m_frame->m_encData->m_<wbr>meIntegral[11] + (y + 1) * stride - padX;<br>
<br>
/*For width = 32 */<br>
- integral_init32h(sum32x32, pix, stride);<br>
+ primitives.integral_init32h(<wbr>sum32x32, pix, stride);<br>
if (y >= 32 - padY)<br>
- integral_init32v(sum32x32 - 32 * stride, stride);<br>
- integral_init32h(sum32x24, pix, stride);<br>
+ primitives.integral_init32v(<wbr>sum32x32 - 32 * stride, stride);<br>
+ primitives.integral_init32h(<wbr>sum32x24, pix, stride);<br>
if (y >= 24 - padY)<br>
- integral_init24v(sum32x24 - 24 * stride, stride);<br>
- integral_init32h(sum32x8, pix, stride);<br>
+ primitives.integral_init24v(<wbr>sum32x24 - 24 * stride, stride);<br>
+ primitives.integral_init32h(<wbr>sum32x8, pix, stride);<br>
if (y >= 8 - padY)<br>
- integral_init8v(sum32x8 - 8 * stride, stride);<br>
+ primitives.integral_init8v(<wbr>sum32x8 - 8 * stride, stride);<br>
/*For width = 24 */<br>
- integral_init24h(sum24x32, pix, stride);<br>
+ primitives.integral_init24h(<wbr>sum24x32, pix, stride);<br>
if (y >= 32 - padY)<br>
- integral_init32v(sum24x32 - 32 * stride, stride);<br>
+ primitives.integral_init32v(<wbr>sum24x32 - 32 * stride, stride);<br>
/*For width = 16 */<br>
- integral_init16h(sum16x16, pix, stride);<br>
+ primitives.integral_init16h(<wbr>sum16x16, pix, stride);<br>
if (y >= 16 - padY)<br>
- integral_init16v(sum16x16 - 16 * stride, stride);<br>
- integral_init16h(sum16x12, pix, stride);<br>
+ primitives.integral_init16v(<wbr>sum16x16 - 16 * stride, stride);<br>
+ primitives.integral_init16h(<wbr>sum16x12, pix, stride);<br>
if (y >= 12 - padY)<br>
- integral_init12v(sum16x12 - 12 * stride, stride);<br>
- integral_init16h(sum16x4, pix, stride);<br>
+ primitives.integral_init12v(<wbr>sum16x12 - 12 * stride, stride);<br>
+ primitives.integral_init16h(<wbr>sum16x4, pix, stride);<br>
if (y >= 4 - padY)<br>
- integral_init4v(sum16x4 - 4 * stride, stride);<br>
+ primitives.integral_init4v(<wbr>sum16x4 - 4 * stride, stride);<br>
/*For width = 12 */<br>
- integral_init12h(sum12x16, pix, stride);<br>
+ primitives.integral_init12h(<wbr>sum12x16, pix, stride);<br>
if (y >= 16 - padY)<br>
- integral_init16v(sum12x16 - 16 * stride, stride);<br>
+ primitives.integral_init16v(<wbr>sum12x16 - 16 * stride, stride);<br>
/*For width = 8 */<br>
- integral_init8h(sum8x32, pix, stride);<br>
+ primitives.integral_init8h(<wbr>sum8x32, pix, stride);<br>
if (y >= 32 - padY)<br>
- integral_init32v(sum8x32 - 32 * stride, stride);<br>
- integral_init8h(sum8x8, pix, stride);<br>
+ primitives.integral_init32v(<wbr>sum8x32 - 32 * stride, stride);<br>
+ primitives.integral_init8h(<wbr>sum8x8, pix, stride);<br>
if (y >= 8 - padY)<br>
- integral_init8v(sum8x8 - 8 * stride, stride);<br>
+ primitives.integral_init8v(<wbr>sum8x8 - 8 * stride, stride);<br>
/*For width = 4 */<br>
- integral_init4h(sum4x16, pix, stride);<br>
+ primitives.integral_init4h(<wbr>sum4x16, pix, stride);<br>
if (y >= 16 - padY)<br>
- integral_init16v(sum4x16 - 16 * stride, stride);<br>
- integral_init4h(sum4x4, pix, stride);<br>
+ primitives.integral_init16v(<wbr>sum4x16 - 16 * stride, stride);<br>
+ primitives.integral_init4h(<wbr>sum4x4, pix, stride);<br>
if (y >= 4 - padY)<br>
- integral_init4v(sum4x4 - 4 * stride, stride);<br>
+ primitives.integral_init4v(<wbr>sum4x4 - 4 * stride, stride);</div><div class="gmail_quote"><br></div><div class="gmail_quote"><br></div><div class="gmail_quote"><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)"> Use size to calculate index for array of function pointers, once you made the above mentioned changes. </div><br></div><div class="gmail_quote"><br></div><div class="gmail_quote"><br>
}<br>
m_parallelFilter[row].m_<wbr>frameFilter-><wbr>integralCompleted.set(1);<br>
}<br>
<br><br><br><div class="gmail_default" style="font-family:georgia,serif;color:rgb(0,0,0)">This does't apply to latest tip. Please update.</div><br><br>______________________________<wbr>_________________<br>
x265-devel mailing list<br>
<a href="mailto:x265-devel@videolan.org">x265-devel@videolan.org</a><br>
<a href="https://mailman.videolan.org/listinfo/x265-devel" rel="noreferrer" target="_blank">https://mailman.videolan.org/<wbr>listinfo/x265-devel</a><br>
<br></div><br></div>