virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-12 19:18:44 +02:00
Code Issues Releases Activity
e82472e379
FFmpeg/libavcodec/mips/vc1dsp_mips.h

218 lines
12 KiB
C
Raw Normal View History

avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
/*
* Copyright (c) 2016 Zhou Xiaoyong <zhouxiaoyong@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
*/
#ifndef AVCODEC_MIPS_VC1DSP_MIPS_H
#define AVCODEC_MIPS_VC1DSP_MIPS_H
#include "libavcodec/vc1dsp.h"
void ff_put_vc1_mspel_mc00_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc01_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc02_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc03_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc10_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc11_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc12_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc13_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc20_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc21_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc22_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc23_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc30_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc31_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc32_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc33_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc00_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc01_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc02_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc03_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc10_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc11_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc12_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc13_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc20_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc21_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc22_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc23_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc30_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc31_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc32_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc33_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc00_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc01_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc02_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc03_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc10_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc11_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc12_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc13_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc20_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc21_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc22_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc23_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc30_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc31_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc32_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_put_vc1_mspel_mc33_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc00_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc01_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc02_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc03_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc10_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc11_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc12_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc13_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc20_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc21_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc22_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc23_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc30_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc31_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc32_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_avg_vc1_mspel_mc33_16_mmi(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride, int rnd);
void ff_vc1_inv_trans_8x8_mmi(int16_t block[64]);
Merge commit '2ec9fa5ec60dcd10e1cb10d8b4e4437e634ea428' * commit '2ec9fa5ec60dcd10e1cb10d8b4e4437e634ea428': idct: Change type of array stride parameters to ptrdiff_t Merged-by: James Almer <jamrial@gmail.com>
2017-03-21 19:29:52 +02:00
void ff_vc1_inv_trans_8x4_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
void ff_vc1_inv_trans_4x8_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
void ff_vc1_inv_trans_4x4_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
Merge commit '2ec9fa5ec60dcd10e1cb10d8b4e4437e634ea428' * commit '2ec9fa5ec60dcd10e1cb10d8b4e4437e634ea428': idct: Change type of array stride parameters to ptrdiff_t Merged-by: James Almer <jamrial@gmail.com>
2017-03-21 19:29:52 +02:00
void ff_vc1_inv_trans_4x4_dc_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
void ff_vc1_inv_trans_4x8_dc_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
void ff_vc1_inv_trans_8x4_dc_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
void ff_vc1_inv_trans_8x8_dc_mmi(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
vc1dsp: Change remaining stride parameters to ptrdiff_t The existing x86 assembly for loop filters uses the stride as a full register without clearing/sign extending the upper half of the registers on x86_64. This avoids crashes if the caller would have passed nonzero bits in the previously undefined upper 32 bits of the parameters. Signed-off-by: Martin Storsjö <martin@martin.st>
2022-03-29 14:34:19 +02:00
void ff_vc1_v_overlap_mmi(uint8_t *src, ptrdiff_t stride);
void ff_vc1_h_overlap_mmi(uint8_t *src, ptrdiff_t stride);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
void ff_vc1_v_s_overlap_mmi(int16_t *top, int16_t *bottom);
vc1dsp: Change remaining stride parameters to ptrdiff_t The existing x86 assembly for loop filters uses the stride as a full register without clearing/sign extending the upper half of the registers on x86_64. This avoids crashes if the caller would have passed nonzero bits in the previously undefined upper 32 bits of the parameters. Signed-off-by: Martin Storsjö <martin@martin.st>
2022-03-29 14:34:19 +02:00
void ff_vc1_h_s_overlap_mmi(int16_t *left, int16_t *right, ptrdiff_t left_stride, ptrdiff_t right_stride, int flags);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
vc1dsp: Change remaining stride parameters to ptrdiff_t The existing x86 assembly for loop filters uses the stride as a full register without clearing/sign extending the upper half of the registers on x86_64. This avoids crashes if the caller would have passed nonzero bits in the previously undefined upper 32 bits of the parameters. Signed-off-by: Martin Storsjö <martin@martin.st>
2022-03-29 14:34:19 +02:00
void ff_vc1_v_loop_filter4_mmi(uint8_t *src, ptrdiff_t stride, int pq);
void ff_vc1_h_loop_filter4_mmi(uint8_t *src, ptrdiff_t stride, int pq);
void ff_vc1_v_loop_filter8_mmi(uint8_t *src, ptrdiff_t stride, int pq);
void ff_vc1_h_loop_filter8_mmi(uint8_t *src, ptrdiff_t stride, int pq);
void ff_vc1_v_loop_filter16_mmi(uint8_t *src, ptrdiff_t stride, int pq);
void ff_vc1_h_loop_filter16_mmi(uint8_t *src, ptrdiff_t stride, int pq);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
void ff_put_no_rnd_vc1_chroma_mc8_mmi(uint8_t *dst /* align 8 */,
uint8_t *src /* align 1 */,
avcodec/mips: Fixed four warnings in vc1dsp Change the stride argument to ptrdiff_t in the following functions: ff_put_no_rnd_vc1_chroma_mc8_mmi, ff_put_no_rnd_vc1_chroma_mc4_mmi, ff_avg_no_rnd_vc1_chroma_mc8_mmi, ff_avg_no_rnd_vc1_chroma_mc4_mmi. Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2019-10-12 04:48:19 +02:00
ptrdiff_t stride, int h, int x, int y);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
void ff_put_no_rnd_vc1_chroma_mc4_mmi(uint8_t *dst /* align 8 */,
uint8_t *src /* align 1 */,
avcodec/mips: Fixed four warnings in vc1dsp Change the stride argument to ptrdiff_t in the following functions: ff_put_no_rnd_vc1_chroma_mc8_mmi, ff_put_no_rnd_vc1_chroma_mc4_mmi, ff_avg_no_rnd_vc1_chroma_mc8_mmi, ff_avg_no_rnd_vc1_chroma_mc4_mmi. Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2019-10-12 04:48:19 +02:00
ptrdiff_t stride, int h, int x, int y);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
void ff_avg_no_rnd_vc1_chroma_mc8_mmi(uint8_t *dst /* align 8 */,
uint8_t *src /* align 1 */,
avcodec/mips: Fixed four warnings in vc1dsp Change the stride argument to ptrdiff_t in the following functions: ff_put_no_rnd_vc1_chroma_mc8_mmi, ff_put_no_rnd_vc1_chroma_mc4_mmi, ff_avg_no_rnd_vc1_chroma_mc8_mmi, ff_avg_no_rnd_vc1_chroma_mc4_mmi. Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2019-10-12 04:48:19 +02:00
ptrdiff_t stride, int h, int x, int y);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
void ff_avg_no_rnd_vc1_chroma_mc4_mmi(uint8_t *dst /* align 8 */,
uint8_t *src /* align 1 */,
avcodec/mips: Fixed four warnings in vc1dsp Change the stride argument to ptrdiff_t in the following functions: ff_put_no_rnd_vc1_chroma_mc8_mmi, ff_put_no_rnd_vc1_chroma_mc4_mmi, ff_avg_no_rnd_vc1_chroma_mc8_mmi, ff_avg_no_rnd_vc1_chroma_mc4_mmi. Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2019-10-12 04:48:19 +02:00
ptrdiff_t stride, int h, int x, int y);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
avcodec/mips: msa optimizations for vc1dsp Performance of WMV3 decoding has speed up from 3.66x to 5.23x tested on 3A4000. Reviewed-by: Shiyou Yin <yinshiyou-hf@loongson.cn> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2019-10-21 09:56:47 +02:00
void ff_vc1_inv_trans_8x8_msa(int16_t block[64]);
void ff_vc1_inv_trans_8x4_msa(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
void ff_vc1_inv_trans_4x8_msa(uint8_t *dest, ptrdiff_t linesize, int16_t *block);
#define FF_PUT_VC1_MSPEL_MC_MSA(hmode, vmode) \
void ff_put_vc1_mspel_mc ## hmode ## vmode ## _msa(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride, int rnd); \
void ff_put_vc1_mspel_mc ## hmode ## vmode ## _16_msa(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride, int rnd);
FF_PUT_VC1_MSPEL_MC_MSA(1, 1);
FF_PUT_VC1_MSPEL_MC_MSA(1, 2);
FF_PUT_VC1_MSPEL_MC_MSA(1, 3);
FF_PUT_VC1_MSPEL_MC_MSA(2, 1);
FF_PUT_VC1_MSPEL_MC_MSA(2, 2);
FF_PUT_VC1_MSPEL_MC_MSA(2, 3);
FF_PUT_VC1_MSPEL_MC_MSA(3, 1);
FF_PUT_VC1_MSPEL_MC_MSA(3, 2);
FF_PUT_VC1_MSPEL_MC_MSA(3, 3);
avcodec/mips: version 1 of vc1dsp optimizations for loongson mmi Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2016-10-10 10:11:37 +02:00
#endif /* AVCODEC_MIPS_VC1DSP_MIPS_H */
Reference in New Issue Copy Permalink