diff options
Diffstat (limited to 'libavcodec/x86/hpeldsp_init.c')
| -rw-r--r-- | libavcodec/x86/hpeldsp_init.c | 80 |
1 files changed, 65 insertions, 15 deletions
diff --git a/libavcodec/x86/hpeldsp_init.c b/libavcodec/x86/hpeldsp_init.c index 1cc3bacd15..bcae22fa44 100644 --- a/libavcodec/x86/hpeldsp_init.c +++ b/libavcodec/x86/hpeldsp_init.c @@ -3,20 +3,20 @@ * Copyright (c) 2000, 2001 Fabrice Bellard * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at> * - * This file is part of Libav. + * This file is part of FFmpeg. * - * Libav is free software; you can redistribute it and/or + * 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. * - * Libav is distributed in the hope that it will be useful, + * 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 Libav; if not, write to the Free Software + * License along with FFmpeg; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * * MMX optimization by Nick Kurshev <nickols_k@mail.ru> @@ -40,6 +40,14 @@ void ff_put_pixels16_x2_mmxext(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h); void ff_put_pixels16_x2_3dnow(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h); +void ff_put_pixels16_x2_sse2(uint8_t *block, const uint8_t *pixels, + ptrdiff_t line_size, int h); +void ff_avg_pixels16_x2_sse2(uint8_t *block, const uint8_t *pixels, + ptrdiff_t line_size, int h); +void ff_put_pixels16_y2_sse2(uint8_t *block, const uint8_t *pixels, + ptrdiff_t line_size, int h); +void ff_avg_pixels16_y2_sse2(uint8_t *block, const uint8_t *pixels, + ptrdiff_t line_size, int h); void ff_put_no_rnd_pixels8_x2_mmxext(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h); void ff_put_no_rnd_pixels8_x2_3dnow(uint8_t *block, const uint8_t *pixels, @@ -74,10 +82,12 @@ void ff_avg_pixels8_y2_mmxext(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h); void ff_avg_pixels8_y2_3dnow(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h); -void ff_avg_pixels8_xy2_mmxext(uint8_t *block, const uint8_t *pixels, - ptrdiff_t line_size, int h); void ff_avg_pixels8_xy2_3dnow(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h); +void ff_avg_approx_pixels8_xy2_mmxext(uint8_t *block, const uint8_t *pixels, + ptrdiff_t line_size, int h); +void ff_avg_approx_pixels8_xy2_3dnow(uint8_t *block, const uint8_t *pixels, + ptrdiff_t line_size, int h); #define avg_pixels8_mmx ff_avg_pixels8_mmx #define avg_pixels8_x2_mmx ff_avg_pixels8_x2_mmx @@ -156,32 +166,49 @@ CALL_2X_PIXELS_EXPORT(ff_put_pixels16_xy2_mmx, ff_put_pixels8_xy2_mmx, 8) CALL_2X_PIXELS(avg_pixels16 ## CPUEXT, ff_avg_pixels8 ## CPUEXT, 8) \ CALL_2X_PIXELS(avg_pixels16_x2 ## CPUEXT, ff_avg_pixels8_x2 ## CPUEXT, 8) \ CALL_2X_PIXELS(avg_pixels16_y2 ## CPUEXT, ff_avg_pixels8_y2 ## CPUEXT, 8) \ - CALL_2X_PIXELS(avg_pixels16_xy2 ## CPUEXT, ff_avg_pixels8_xy2 ## CPUEXT, 8) + CALL_2X_PIXELS(avg_pixels16_xy2 ## CPUEXT, ff_avg_pixels8_xy2 ## CPUEXT, 8) \ + CALL_2X_PIXELS(avg_approx_pixels16_xy2## CPUEXT, ff_avg_approx_pixels8_xy2## CPUEXT, 8) HPELDSP_AVG_PIXELS16(_3dnow) HPELDSP_AVG_PIXELS16(_mmxext) #endif /* HAVE_YASM */ +#define SET_HPEL_FUNCS_EXT(PFX, IDX, SIZE, CPU) \ + if (HAVE_MMX_EXTERNAL) \ + c->PFX ## _pixels_tab IDX [0] = PFX ## _pixels ## SIZE ## _ ## CPU; + +#if HAVE_MMX_INLINE #define SET_HPEL_FUNCS(PFX, IDX, SIZE, CPU) \ do { \ - c->PFX ## _pixels_tab IDX [0] = PFX ## _pixels ## SIZE ## _ ## CPU; \ + SET_HPEL_FUNCS_EXT(PFX, IDX, SIZE, CPU) \ c->PFX ## _pixels_tab IDX [1] = PFX ## _pixels ## SIZE ## _x2_ ## CPU; \ c->PFX ## _pixels_tab IDX [2] = PFX ## _pixels ## SIZE ## _y2_ ## CPU; \ c->PFX ## _pixels_tab IDX [3] = PFX ## _pixels ## SIZE ## _xy2_ ## CPU; \ } while (0) +#else +#define SET_HPEL_FUNCS(PFX, IDX, SIZE, CPU) \ + do { \ + SET_HPEL_FUNCS_EXT(PFX, IDX, SIZE, CPU) \ + } while (0) +#endif static void hpeldsp_init_mmx(HpelDSPContext *c, int flags, int cpu_flags) { -#if HAVE_MMX_INLINE SET_HPEL_FUNCS(put, [0], 16, mmx); SET_HPEL_FUNCS(put_no_rnd, [0], 16, mmx); SET_HPEL_FUNCS(avg, [0], 16, mmx); SET_HPEL_FUNCS(avg_no_rnd, , 16, mmx); SET_HPEL_FUNCS(put, [1], 8, mmx); SET_HPEL_FUNCS(put_no_rnd, [1], 8, mmx); - SET_HPEL_FUNCS(avg, [1], 8, mmx); -#endif /* HAVE_MMX_INLINE */ + if (HAVE_MMX_EXTERNAL) { + c->avg_pixels_tab[1][0] = ff_avg_pixels8_mmx; + c->avg_pixels_tab[1][1] = ff_avg_pixels8_x2_mmx; + } +#if HAVE_MMX_INLINE + c->avg_pixels_tab[1][2] = avg_pixels8_y2_mmx; + c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_mmx; +#endif } static void hpeldsp_init_mmxext(HpelDSPContext *c, int flags, int cpu_flags) @@ -193,6 +220,7 @@ static void hpeldsp_init_mmxext(HpelDSPContext *c, int flags, int cpu_flags) c->avg_pixels_tab[0][0] = avg_pixels16_mmxext; c->avg_pixels_tab[0][1] = avg_pixels16_x2_mmxext; c->avg_pixels_tab[0][2] = avg_pixels16_y2_mmxext; + c->avg_pixels_tab[0][3] = avg_pixels16_xy2_mmxext; c->put_pixels_tab[1][1] = ff_put_pixels8_x2_mmxext; c->put_pixels_tab[1][2] = ff_put_pixels8_y2_mmxext; @@ -200,6 +228,7 @@ static void hpeldsp_init_mmxext(HpelDSPContext *c, int flags, int cpu_flags) c->avg_pixels_tab[1][0] = ff_avg_pixels8_mmxext; c->avg_pixels_tab[1][1] = ff_avg_pixels8_x2_mmxext; c->avg_pixels_tab[1][2] = ff_avg_pixels8_y2_mmxext; + c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_mmxext; if (!(flags & CODEC_FLAG_BITEXACT)) { c->put_no_rnd_pixels_tab[0][1] = put_no_rnd_pixels16_x2_mmxext; @@ -207,8 +236,8 @@ static void hpeldsp_init_mmxext(HpelDSPContext *c, int flags, int cpu_flags) c->put_no_rnd_pixels_tab[1][1] = ff_put_no_rnd_pixels8_x2_mmxext; c->put_no_rnd_pixels_tab[1][2] = ff_put_no_rnd_pixels8_y2_mmxext; - c->avg_pixels_tab[0][3] = avg_pixels16_xy2_mmxext; - c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_mmxext; + c->avg_pixels_tab[0][3] = avg_approx_pixels16_xy2_mmxext; + c->avg_pixels_tab[1][3] = ff_avg_approx_pixels8_xy2_mmxext; } if (flags & CODEC_FLAG_BITEXACT && CONFIG_VP3_DECODER) { @@ -227,6 +256,7 @@ static void hpeldsp_init_3dnow(HpelDSPContext *c, int flags, int cpu_flags) c->avg_pixels_tab[0][0] = avg_pixels16_3dnow; c->avg_pixels_tab[0][1] = avg_pixels16_x2_3dnow; c->avg_pixels_tab[0][2] = avg_pixels16_y2_3dnow; + c->avg_pixels_tab[0][3] = avg_pixels16_xy2_3dnow; c->put_pixels_tab[1][1] = ff_put_pixels8_x2_3dnow; c->put_pixels_tab[1][2] = ff_put_pixels8_y2_3dnow; @@ -234,6 +264,7 @@ static void hpeldsp_init_3dnow(HpelDSPContext *c, int flags, int cpu_flags) c->avg_pixels_tab[1][0] = ff_avg_pixels8_3dnow; c->avg_pixels_tab[1][1] = ff_avg_pixels8_x2_3dnow; c->avg_pixels_tab[1][2] = ff_avg_pixels8_y2_3dnow; + c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_3dnow; if (!(flags & CODEC_FLAG_BITEXACT)){ c->put_no_rnd_pixels_tab[0][1] = put_no_rnd_pixels16_x2_3dnow; @@ -241,8 +272,8 @@ static void hpeldsp_init_3dnow(HpelDSPContext *c, int flags, int cpu_flags) c->put_no_rnd_pixels_tab[1][1] = ff_put_no_rnd_pixels8_x2_3dnow; c->put_no_rnd_pixels_tab[1][2] = ff_put_no_rnd_pixels8_y2_3dnow; - c->avg_pixels_tab[0][3] = avg_pixels16_xy2_3dnow; - c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_3dnow; + c->avg_pixels_tab[0][3] = avg_approx_pixels16_xy2_3dnow; + c->avg_pixels_tab[1][3] = ff_avg_approx_pixels8_xy2_3dnow; } if (flags & CODEC_FLAG_BITEXACT && CONFIG_VP3_DECODER) { @@ -259,11 +290,27 @@ static void hpeldsp_init_sse2(HpelDSPContext *c, int flags, int cpu_flags) // these functions are slower than mmx on AMD, but faster on Intel c->put_pixels_tab[0][0] = ff_put_pixels16_sse2; c->put_no_rnd_pixels_tab[0][0] = ff_put_pixels16_sse2; + c->put_pixels_tab[0][1] = ff_put_pixels16_x2_sse2; + c->put_pixels_tab[0][2] = ff_put_pixels16_y2_sse2; + c->put_pixels_tab[0][3] = ff_put_pixels16_xy2_sse2; c->avg_pixels_tab[0][0] = ff_avg_pixels16_sse2; + c->avg_pixels_tab[0][1] = ff_avg_pixels16_x2_sse2; + c->avg_pixels_tab[0][2] = ff_avg_pixels16_y2_sse2; + c->avg_pixels_tab[0][3] = ff_avg_pixels16_xy2_sse2; } #endif /* HAVE_SSE2_EXTERNAL */ } +static void hpeldsp_init_ssse3(HpelDSPContext *c, int flags, int cpu_flags) +{ +#if HAVE_SSSE3_EXTERNAL + c->put_pixels_tab[0][3] = ff_put_pixels16_xy2_ssse3; + c->avg_pixels_tab[0][3] = ff_avg_pixels16_xy2_ssse3; + c->put_pixels_tab[1][3] = ff_put_pixels8_xy2_ssse3; + c->avg_pixels_tab[1][3] = ff_avg_pixels8_xy2_ssse3; +#endif +} + av_cold void ff_hpeldsp_init_x86(HpelDSPContext *c, int flags) { int cpu_flags = av_get_cpu_flags(); @@ -279,4 +326,7 @@ av_cold void ff_hpeldsp_init_x86(HpelDSPContext *c, int flags) if (EXTERNAL_SSE2(cpu_flags)) hpeldsp_init_sse2(c, flags, cpu_flags); + + if (EXTERNAL_SSSE3(cpu_flags)) + hpeldsp_init_ssse3(c, flags, cpu_flags); } |