mirror of
https://github.com/ultravideo/uvg266.git
synced 2024-11-25 02:44:07 +00:00
1de102be61
- Enforces a little bit more hierarchy. Compilation units are in strategies and whatever inline includes they have are in a folder with the same name as the strategy.
59 lines
2 KiB
C
59 lines
2 KiB
C
/*****************************************************************************
|
|
* This file is part of Kvazaar HEVC encoder.
|
|
*
|
|
* Copyright (C) 2013-2014 Tampere University of Technology and others (see
|
|
* COPYING file).
|
|
*
|
|
* Kvazaar is free software: you can redistribute it and/or modify
|
|
* it under the terms of the GNU General Public License version 2 as published
|
|
* by the Free Software Foundation.
|
|
*
|
|
* Kvazaar 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 Kvazaar. If not, see <http://www.gnu.org/licenses/>.
|
|
****************************************************************************/
|
|
|
|
/*
|
|
* \file
|
|
*/
|
|
#include "strategyselector.h"
|
|
#include "image.h"
|
|
#include <immintrin.h>
|
|
#include <assert.h>
|
|
|
|
#ifdef __GNUC__
|
|
__attribute__ ((__target__ ("sse2")))
|
|
#endif
|
|
static unsigned reg_sad_sse2(const pixel * const data1, const pixel * const data2,
|
|
const int width, const int height, const unsigned stride1, const unsigned stride2)
|
|
{
|
|
int y, x;
|
|
unsigned sad = 0;
|
|
__m128i sse_inc = _mm_setzero_si128 ();
|
|
long long int sse_inc_array[2];
|
|
|
|
for (y = 0; y < height; ++y) {
|
|
for (x = 0; x <= width-16; x+=16) {
|
|
const __m128i a = _mm_loadu_si128((__m128i const*) &data1[y * stride1 + x]);
|
|
const __m128i b = _mm_loadu_si128((__m128i const*) &data2[y * stride2 + x]);
|
|
sse_inc = _mm_add_epi32(sse_inc, _mm_sad_epu8(a,b));
|
|
}
|
|
|
|
for (; x < width; ++x) {
|
|
sad += abs(data1[y * stride1 + x] - data2[y * stride2 + x]);
|
|
}
|
|
}
|
|
_mm_storeu_si128((__m128i*) sse_inc_array, sse_inc);
|
|
sad += sse_inc_array[0] + sse_inc_array[1];
|
|
|
|
return sad;
|
|
}
|
|
|
|
static int strategy_register_picture_sse2(void* opaque) {
|
|
return strategyselector_register(opaque, "reg_sad", "sse2", 10, ®_sad_sse2);
|
|
}
|