<div dir="ltr">Merged, sent implementation.<div><br></div><div>Regards,</div><div>Praveen Tiwari<br><div><br></div><div><br></div></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Wed, Nov 20, 2013 at 6:08 PM, chen <span dir="ltr"><<a href="mailto:chenm003@163.com" target="_blank">chenm003@163.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div style="line-height:1.7;font-size:14px;font-family:arial"><pre>At 2013-11-20 19:45:24,<a href="mailto:praveen@multicorewareinc.com" target="_blank">praveen@multicorewareinc.com</a> wrote:
># HG changeset patch
># User Praveen Tiwari
># Date 1384947915 -19800
># Node ID c1e556f54d61422d153ff67f4830dc62dd1111d9
># Parent a7fb47a7eddf18634449a5ac898f7c2d029048e9
>asm code for pixeladd_ps_4x4 and testbench integration
>
>diff -r a7fb47a7eddf -r c1e556f54d61 source/common/CMakeLists.txt
>--- a/source/common/CMakeLists.txt Wed Nov 20 12:57:57 2013 +0530
>+++ b/source/common/CMakeLists.txt Wed Nov 20 17:15:15 2013 +0530
>@@ -113,7 +113,7 @@
>
> if(ENABLE_PRIMITIVES_ASM)
> set(C_SRCS asm-primitives.cpp pixel.h mc.h ipfilter8.h blockcopy8.h)
>- set(A_SRCS pixel-a.asm const-a.asm cpu-a.asm sad-a.asm mc-a.asm mc-a2.asm ipfilter8.asm pixel-util.asm blockcopy8.asm)
>+ set(A_SRCS pixel-a.asm const-a.asm cpu-a.asm sad-a.asm mc-a.asm mc-a2.asm ipfilter8.asm pixel-util.asm blockcopy8.asm pixeladd8.asm)
> if (NOT X64)
> set(A_SRCS ${A_SRCS} pixel-32.asm)
> endif()
>diff -r a7fb47a7eddf -r c1e556f54d61 source/common/x86/asm-primitives.cpp
>--- a/source/common/x86/asm-primitives.cpp Wed Nov 20 12:57:57 2013 +0530
>+++ b/source/common/x86/asm-primitives.cpp Wed Nov 20 17:15:15 2013 +0530
>@@ -633,6 +633,13 @@
> p.calcrecon[BLOCK_32x32] = x265_calcRecons32_sse4;
> p.calcresidual[BLOCK_16x16] = x265_getResidual16_sse4;
> p.calcresidual[BLOCK_32x32] = x265_getResidual32_sse4;
>+
>+ // This function pointer initialization is temporary will be removed
>+ // later with macro definitions. It is used to avoid linker errors
>+ // until all partitions are coded and commit smaller patches, easier to
>+ // review.
>+
>+ p.chroma_add_ps[X265_CSP_I420][CHROMA_4x4] = x265_pixel_add_ps_4x4_sse4;
> }
> if (cpuMask & X265_CPU_AVX)
> {
>diff -r a7fb47a7eddf -r c1e556f54d61 source/common/x86/pixel.h
>--- a/source/common/x86/pixel.h Wed Nov 20 12:57:57 2013 +0530
>+++ b/source/common/x86/pixel.h Wed Nov 20 17:15:15 2013 +0530
>@@ -313,7 +313,8 @@
> SETUP_CHROMA_PIXELSUB_PS_FUNC(8, 32, cpu);
>
> #define SETUP_LUMA_PIXELSUB_PS_FUNC(W, H, cpu) \
>- void x265_pixel_sub_ps_ ## W ## x ## H ## cpu(int16_t *dest, intptr_t destride, pixel *src0, pixel *src1, intptr_t srcstride0, intptr_t srcstride1);
>+ void x265_pixel_sub_ps_ ## W ## x ## H ## cpu(int16_t *dest, intptr_t destride, pixel *src0, pixel *src1, intptr_t srcstride0, intptr_t srcstride1);\
>+ void x265_pixel_add_ps_ ## W ## x ## H ## cpu(pixel *dest, int destride, pixel *src0, int16_t *scr1, int srcStride0, int srcStride1);
>
> #define LUMA_PIXELSUB_DEF(cpu) \
> SETUP_LUMA_PIXELSUB_PS_FUNC(4, 4, cpu); \
>@@ -342,6 +343,8 @@
> SETUP_LUMA_PIXELSUB_PS_FUNC(64, 16, cpu); \
> SETUP_LUMA_PIXELSUB_PS_FUNC(16, 64, cpu);
>
>+// void x265_pixeladd_ps_4x4_sse4(pixel *dest, int destride, pixel *src0, int16_t *scr1, int srcStride0, int srcStride1);
>+
remove unused line</pre><pre> </pre><pre><div><div class="h5">> CHROMA_PIXELSUB_DEF(_sse4);
> LUMA_PIXELSUB_DEF(_sse4);
>
>diff -r a7fb47a7eddf -r c1e556f54d61 source/common/x86/pixeladd8.asm
>--- /dev/null Thu Jan 01 00:00:00 1970 +0000
>+++ b/source/common/x86/pixeladd8.asm Wed Nov 20 17:15:15 2013 +0530
>@@ -0,0 +1,79 @@
>+;*****************************************************************************
>+;* Copyright (C) 2013 x265 project
>+;*
>+;* Authors: Praveen Kumar Tiwari <<a href="mailto:praveen@multicorewareinc.com" target="_blank">praveen@multicorewareinc.com</a>>
>+;*
>+;* This program is free software; you can redistribute it and/or modify
>+;* it under the terms of the GNU General Public License as published by
>+;* the Free Software Foundation; either version 2 of the License, or
>+;* (at your option) any later version.
>+;*
>+;* This program is distributed in the hope that it will be useful,
>+;* but WITHOUT ANY WARRANTY; without even the implied warranty of
>+;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>+;* GNU General Public License for more details.
>+;*
>+;* You should have received a copy of the GNU General Public License
>+;* along with this program; if not, write to the Free Software
>+;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02111, USA.
>+;*
>+;* This program is also available under a commercial proprietary license.
>+;* For more information, contact us at <a href="mailto:licensing@multicorewareinc.com" target="_blank">licensing@multicorewareinc.com</a>.
>+;*****************************************************************************/
>+
>+%include "x86inc.asm"
>+%include "x86util.asm"
>+
>+SECTION_RODATA 32
>+
>+SECTION .text
>+
>+;-----------------------------------------------------------------------------
>+; void pixel_add_ps_4x4(pixel *dest, int destride, pixel *src0, int16_t *scr1, int srcStride0, int srcStride1)
>+;-----------------------------------------------------------------------------
>+INIT_XMM sse4
>+cglobal pixel_add_ps_4x4, 6, 6, 2, dest, destride, src0, scr1, srcStride0, srcStride1
>+
>+add r5, r5
>+
>+movd m0, [r2]
>+pmovzxbw m0, m0
>+movh m1, [</div></div>r3]
we can merge movd and pmovzxbw, in Intel documents, this instruction is not need alignment to 16-bytes bound</pre><pre><div><div class="h5">>+
>+paddw m0, m1
>+packuswb m0, m0
>+
>+movd [r0], m0
>+
>+movd m0, [r2 + r4]
>+pmovzxbw m0, m0
>+movh m1, [r3 + r5]
>+
>+paddw m0, m1
>+packuswb m0, m0
>+
>+movd [r0 + r1], m0
>+
>+movd m0, [r2 + 2 * r4]
>+pmovzxbw m0, m0
>+movh m1, [r3 + 2 * r5]
>+
>+paddw m0, m1
>+packuswb m0, m0
>+
>+movd [r0 + 2 * r1], m0
>+
>+lea r0, [r0 + 2 * r1]
>+lea r2, [r2 + 2 * r4]
>+lea r3, [r3 + 2 * r5]
>+
>+movd m0, [r2 + r4]
>+pmovzxbw m0, m0
>+movh m1, [r3 + r5]
>+
>+paddw m0, m1
>+packuswb m0, m0
>+
>+movd [r0 + r1], </div></div>m0
>+
>+RET
>_______________________________________________
>x265-devel mailing list
><a href="mailto:x265-devel@videolan.org" target="_blank">x265-devel@videolan.org</a>
><a href="https://mailman.videolan.org/listinfo/x265-devel" target="_blank">https://mailman.videolan.org/listinfo/x265-devel</a>
</pre></div><br>_______________________________________________<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" target="_blank">https://mailman.videolan.org/listinfo/x265-devel</a><br>
<br></blockquote></div><br></div>