1d186e9e12
* qatar/master: Revert "swscale: use 15-bit intermediates for 9/10-bit scaling." swscale: use 15-bit intermediates for 9/10-bit scaling. dct32: Add SSE2 ASM optimizations Correct chroma vector calculation for RealVideo 3. lavf: Add an option to discard corrupted frames mpegts: Mark wrongly-sized packets as corrupted mpegts: Move scan test to handle_packets mpegts: Mark corrupted packets mpegts: Reset continuity counter on seek mpegts: Fix for continuity counter mpegts: Silence "can't seek" warning on unseekable apichange: add an entry for AV_PKT_FLAG_CORRUPT avpacket: signal possibly corrupted packets mpeg4videodec: remove dead code that would have detected erroneous encoding aac: Remove some suspicious illegal memcpy()s from LTP. bink: Eliminate unnecessary shadow declaration. Conflicts: doc/APIchanges libavcodec/version.h libavformat/avformat.h libavformat/options.c libavformat/utils.c Merged-by: Michael Niedermayer <michaelni@gmx.at>
71 lines
2.5 KiB
C
71 lines
2.5 KiB
C
/*
|
|
* 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 "libavutil/cpu.h"
|
|
#include "libavcodec/dsputil.h"
|
|
#include "libavcodec/dct.h"
|
|
#include "fft.h"
|
|
|
|
av_cold void ff_fft_init_mmx(FFTContext *s)
|
|
{
|
|
#if HAVE_YASM
|
|
int has_vectors = av_get_cpu_flags();
|
|
if (has_vectors & AV_CPU_FLAG_AVX && HAVE_AVX && s->nbits >= 5) {
|
|
/* AVX for SB */
|
|
s->imdct_calc = ff_imdct_calc_sse;
|
|
s->imdct_half = ff_imdct_half_avx;
|
|
s->fft_permute = ff_fft_permute_sse;
|
|
s->fft_calc = ff_fft_calc_avx;
|
|
s->fft_permutation = FF_FFT_PERM_AVX;
|
|
} else if (has_vectors & AV_CPU_FLAG_SSE && HAVE_SSE) {
|
|
/* SSE for P3/P4/K8 */
|
|
s->imdct_calc = ff_imdct_calc_sse;
|
|
s->imdct_half = ff_imdct_half_sse;
|
|
s->fft_permute = ff_fft_permute_sse;
|
|
s->fft_calc = ff_fft_calc_sse;
|
|
s->fft_permutation = FF_FFT_PERM_SWAP_LSBS;
|
|
} else if (has_vectors & AV_CPU_FLAG_3DNOWEXT && HAVE_AMD3DNOWEXT) {
|
|
/* 3DNowEx for K7 */
|
|
s->imdct_calc = ff_imdct_calc_3dn2;
|
|
s->imdct_half = ff_imdct_half_3dn2;
|
|
s->fft_calc = ff_fft_calc_3dn2;
|
|
} else if (has_vectors & AV_CPU_FLAG_3DNOW && HAVE_AMD3DNOW) {
|
|
/* 3DNow! for K6-2/3 */
|
|
s->imdct_calc = ff_imdct_calc_3dn;
|
|
s->imdct_half = ff_imdct_half_3dn;
|
|
s->fft_calc = ff_fft_calc_3dn;
|
|
}
|
|
#endif
|
|
}
|
|
|
|
#if CONFIG_DCT
|
|
av_cold void ff_dct_init_mmx(DCTContext *s)
|
|
{
|
|
#if HAVE_YASM
|
|
int has_vectors = av_get_cpu_flags();
|
|
if (has_vectors & AV_CPU_FLAG_AVX && HAVE_AVX)
|
|
s->dct32 = ff_dct32_float_avx;
|
|
else if (has_vectors & AV_CPU_FLAG_SSE2 && HAVE_SSE)
|
|
s->dct32 = ff_dct32_float_sse2;
|
|
else if (has_vectors & AV_CPU_FLAG_SSE && HAVE_SSE)
|
|
s->dct32 = ff_dct32_float_sse;
|
|
#endif
|
|
}
|
|
#endif
|
|
|