1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-26 19:01:44 +02:00
FFmpeg/libavcodec/loongarch/h264qpel_lsx.c
yuanhecai f6077cc666
avcodec/la: Add LSX optimization for h264 qpel.
./configure --disable-lasx
ffmpeg -i 1_h264_1080p_30fps_3Mbps.mp4 -f rawvideo -y /dev/null -an
before: 214fps
after:  274fps

Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2023-05-25 21:05:01 +02:00

488 lines
16 KiB
C

/*
* Loongson LSX optimized h264qpel
*
* Copyright (c) 2023 Loongson Technology Corporation Limited
* Contributed by Hecai Yuan <yuanhecai@loongson.cn>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "h264qpel_loongarch.h"
#include "libavutil/loongarch/loongson_intrinsics.h"
#include "libavutil/attributes.h"
static void put_h264_qpel16_hv_lowpass_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t dstStride, ptrdiff_t srcStride)
{
put_h264_qpel8_hv_lowpass_lsx(dst, src, dstStride, srcStride);
put_h264_qpel8_hv_lowpass_lsx(dst + 8, src + 8, dstStride, srcStride);
src += srcStride << 3;
dst += dstStride << 3;
put_h264_qpel8_hv_lowpass_lsx(dst, src, dstStride, srcStride);
put_h264_qpel8_hv_lowpass_lsx(dst + 8, src + 8, dstStride, srcStride);
}
void ff_put_h264_qpel16_mc22_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
put_h264_qpel16_hv_lowpass_lsx(dst, src, stride, stride);
}
static void put_h264_qpel16_h_lowpass_lsx(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride)
{
put_h264_qpel8_h_lowpass_lsx(dst, src, dstStride, srcStride);
put_h264_qpel8_h_lowpass_lsx(dst+8, src+8, dstStride, srcStride);
src += srcStride << 3;
dst += dstStride << 3;
put_h264_qpel8_h_lowpass_lsx(dst, src, dstStride, srcStride);
put_h264_qpel8_h_lowpass_lsx(dst+8, src+8, dstStride, srcStride);
}
static void put_h264_qpel16_v_lowpass_lsx(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride)
{
put_h264_qpel8_v_lowpass_lsx(dst, (uint8_t*)src, dstStride, srcStride);
put_h264_qpel8_v_lowpass_lsx(dst+8, (uint8_t*)src+8, dstStride, srcStride);
src += 8*srcStride;
dst += 8*dstStride;
put_h264_qpel8_v_lowpass_lsx(dst, (uint8_t*)src, dstStride, srcStride);
put_h264_qpel8_v_lowpass_lsx(dst+8, (uint8_t*)src+8, dstStride, srcStride);
}
void ff_put_h264_qpel16_mc21_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 256;
put_h264_qpel16_h_lowpass_lsx(halfH, src, 16, stride);
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
put_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
void ff_put_h264_qpel16_mc12_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 256;
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
put_h264_qpel16_v_lowpass_lsx(halfH, src, 16, stride);
put_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
void ff_put_h264_qpel16_mc32_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 256;
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
put_h264_qpel16_v_lowpass_lsx(halfH, src + 1, 16, stride);
put_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
void ff_put_h264_qpel16_mc23_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 256;
put_h264_qpel16_h_lowpass_lsx(halfH, src + stride, 16, stride);
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
put_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
static void avg_h264_qpel16_v_lowpass_lsx(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride)
{
avg_h264_qpel8_v_lowpass_lsx(dst, (uint8_t*)src, dstStride, srcStride);
avg_h264_qpel8_v_lowpass_lsx(dst+8, (uint8_t*)src+8, dstStride, srcStride);
src += 8*srcStride;
dst += 8*dstStride;
avg_h264_qpel8_v_lowpass_lsx(dst, (uint8_t*)src, dstStride, srcStride);
avg_h264_qpel8_v_lowpass_lsx(dst+8, (uint8_t*)src+8, dstStride, srcStride);
}
void ff_avg_h264_qpel16_mc02_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
avg_h264_qpel16_v_lowpass_lsx(dst, src, stride, stride);
}
void ff_avg_h264_qpel16_mc03_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[256];
put_h264_qpel16_v_lowpass_lsx(half, src, 16, stride);
avg_pixels16_l2_8_lsx(dst, src + stride, half, stride, stride);
}
void ff_avg_h264_qpel16_mc23_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 256;
put_h264_qpel16_h_lowpass_lsx(halfH, src + stride, 16, stride);
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
avg_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
void ff_avg_h264_qpel16_mc21_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 256;
put_h264_qpel16_h_lowpass_lsx(halfH, src, 16, stride);
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
avg_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
void ff_avg_h264_qpel16_mc01_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[256];
put_h264_qpel16_v_lowpass_lsx(half, src, 16, stride);
avg_pixels16_l2_8_lsx(dst, src, half, stride, stride);
}
void ff_avg_h264_qpel16_mc32_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 256;
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
put_h264_qpel16_v_lowpass_lsx(halfH, src + 1, 16, stride);
avg_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
void ff_avg_h264_qpel16_mc12_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[512];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 256;
put_h264_qpel16_hv_lowpass_lsx(halfHV, src, 16, stride);
put_h264_qpel16_v_lowpass_lsx(halfH, src, 16, stride);
avg_pixels16_l2_8_lsx(dst, halfH, halfHV, stride, 16);
}
static void avg_h264_qpel16_hv_lowpass_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t dstStride, ptrdiff_t srcStride)
{
avg_h264_qpel8_hv_lowpass_lsx(dst, src, dstStride, srcStride);
avg_h264_qpel8_hv_lowpass_lsx(dst + 8, src + 8, dstStride, srcStride);
src += srcStride << 3;
dst += dstStride << 3;
avg_h264_qpel8_hv_lowpass_lsx(dst, src, dstStride, srcStride);
avg_h264_qpel8_hv_lowpass_lsx(dst + 8, src + 8, dstStride, srcStride);
}
void ff_avg_h264_qpel16_mc22_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
avg_h264_qpel16_hv_lowpass_lsx(dst, src, stride, stride);
}
void ff_put_h264_qpel8_mc03_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[64];
put_h264_qpel8_v_lowpass_lsx(half, (uint8_t*)src, 8, stride);
put_pixels8_l2_8_lsx(dst, src + stride, half, stride, stride);
}
void ff_put_h264_qpel8_mc01_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[64];
put_h264_qpel8_v_lowpass_lsx(half, (uint8_t*)src, 8, stride);
put_pixels8_l2_8_lsx(dst, src, half, stride, stride);
}
void ff_put_h264_qpel8_mc30_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[64];
put_h264_qpel8_h_lowpass_lsx(half, src, 8, stride);
put_pixels8_l2_8_lsx(dst, src+1, half, stride, stride);
}
void ff_put_h264_qpel8_mc10_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[64];
put_h264_qpel8_h_lowpass_lsx(half, src, 8, stride);
put_pixels8_l2_8_lsx(dst, src, half, stride, stride);
}
void ff_put_h264_qpel8_mc33_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src + stride, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src + 1, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_put_h264_qpel8_mc13_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src + stride, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_put_h264_qpel8_mc31_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src + 1, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_put_h264_qpel8_mc11_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_put_h264_qpel8_mc32_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 64;
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfH, (uint8_t*)src + 1, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_put_h264_qpel8_mc21_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 64;
put_h264_qpel8_h_lowpass_lsx(halfH, src, 8, stride);
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_put_h264_qpel8_mc23_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 64;
put_h264_qpel8_h_lowpass_lsx(halfH, src + stride, 8, stride);
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_put_h264_qpel8_mc12_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 64;
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfH, (uint8_t*)src, 8, stride);
put_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_put_h264_qpel8_mc02_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
put_h264_qpel8_v_lowpass_lsx(dst, (uint8_t*)src, stride, stride);
}
void ff_put_h264_qpel8_mc22_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
put_h264_qpel8_hv_lowpass_lsx(dst, src, stride, stride);
}
void ff_put_h264_qpel8_mc20_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
put_h264_qpel8_h_lowpass_lsx(dst, src, stride, stride);
}
void ff_avg_h264_qpel8_mc10_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[64];
put_h264_qpel8_h_lowpass_lsx(half, src, 8, stride);
avg_pixels8_l2_8_lsx(dst, src, half, stride, stride);
}
void ff_avg_h264_qpel8_mc20_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
avg_h264_qpel8_h_lowpass_lsx(dst, src, stride, stride);
}
void ff_avg_h264_qpel8_mc30_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t half[64];
put_h264_qpel8_h_lowpass_lsx(half, src, 8, stride);
avg_pixels8_l2_8_lsx(dst, src+1, half, stride, stride);
}
void ff_avg_h264_qpel8_mc11_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_avg_h264_qpel8_mc21_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 64;
put_h264_qpel8_h_lowpass_lsx(halfH, src, 8, stride);
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_avg_h264_qpel8_mc31_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src + 1, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_avg_h264_qpel8_mc02_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
avg_h264_qpel8_v_lowpass_lsx(dst, (uint8_t*)src, stride, stride);
}
void ff_avg_h264_qpel8_mc12_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 64;
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfH, (uint8_t*)src, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_avg_h264_qpel8_mc22_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
avg_h264_qpel8_hv_lowpass_lsx(dst, src, stride, stride);
}
void ff_avg_h264_qpel8_mc32_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfHV = temp;
uint8_t *const halfH = temp + 64;
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfH, (uint8_t*)src + 1, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_avg_h264_qpel8_mc13_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src + stride, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}
void ff_avg_h264_qpel8_mc23_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t temp[128];
uint8_t *const halfH = temp;
uint8_t *const halfHV = temp + 64;
put_h264_qpel8_h_lowpass_lsx(halfH, src + stride, 8, stride);
put_h264_qpel8_hv_lowpass_lsx(halfHV, src, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfHV, stride, 8);
}
void ff_avg_h264_qpel8_mc33_lsx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
uint8_t halfH[64];
uint8_t halfV[64];
put_h264_qpel8_h_lowpass_lsx(halfH, src + stride, 8, stride);
put_h264_qpel8_v_lowpass_lsx(halfV, (uint8_t*)src + 1, 8, stride);
avg_pixels8_l2_8_lsx(dst, halfH, halfV, stride, 8);
}