Showing the changes to libavfilter (4 of 4 files). Show all
Unified · Split
| @@ -6,7 +6,7 @@ OBJS-$(CONFIG_BLEND_FILTER) += x86/vf_blend_init.o · 5 unmodified lines | |||
| 6 | OBJS-$(CONFIG_BWDIF_FILTER) += x86/vf_bwdif_init.o | 6 | OBJS-$(CONFIG_BWDIF_FILTER) += x86/vf_bwdif_init.o |
| 7 | OBJS-$(CONFIG_COLORSPACE_FILTER) += x86/colorspacedsp_init.o | 7 | OBJS-$(CONFIG_COLORSPACE_FILTER) += x86/colorspacedsp_init.o |
| 8 | OBJS-$(CONFIG_CONVOLUTION_FILTER) += x86/vf_convolution_init.o | 8 | OBJS-$(CONFIG_CONVOLUTION_FILTER) += x86/vf_convolution_init.o |
| 9 | OBJS-$(CONFIG_EQ_FILTER) += x86/vf_eq.o | 9 | OBJS-$(CONFIG_EQ_FILTER) += x86/vf_eq_init.o |
| 10 | OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp_init.o | 10 | OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp_init.o |
| 11 | OBJS-$(CONFIG_GBLUR_FILTER) += x86/vf_gblur_init.o | 11 | OBJS-$(CONFIG_GBLUR_FILTER) += x86/vf_gblur_init.o |
| 12 | OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun_init.o | 12 | OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun_init.o |
| @@ -43,6 +43,7 @@ X86ASM-OBJS-$(CONFIG_BLEND_FILTER) += x86/vf_blend.o · 30 unmodified lines | |||
| 43 | X86ASM-OBJS-$(CONFIG_BWDIF_FILTER) += x86/vf_bwdif.o | 43 | X86ASM-OBJS-$(CONFIG_BWDIF_FILTER) += x86/vf_bwdif.o |
| 44 | X86ASM-OBJS-$(CONFIG_COLORSPACE_FILTER) += x86/colorspacedsp.o | 44 | X86ASM-OBJS-$(CONFIG_COLORSPACE_FILTER) += x86/colorspacedsp.o |
| 45 | X86ASM-OBJS-$(CONFIG_CONVOLUTION_FILTER) += x86/vf_convolution.o | 45 | X86ASM-OBJS-$(CONFIG_CONVOLUTION_FILTER) += x86/vf_convolution.o |
| 46 | X86ASM-OBJS-$(CONFIG_CONVOLUTION_FILTER) += x86/vf_eq.o | ||
| 46 | X86ASM-OBJS-$(CONFIG_FRAMERATE_FILTER) += x86/vf_framerate.o | 47 | X86ASM-OBJS-$(CONFIG_FRAMERATE_FILTER) += x86/vf_framerate.o |
| 47 | X86ASM-OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp.o | 48 | X86ASM-OBJS-$(CONFIG_FSPP_FILTER) += x86/vf_fspp.o |
| 48 | X86ASM-OBJS-$(CONFIG_GBLUR_FILTER) += x86/vf_gblur.o | 49 | X86ASM-OBJS-$(CONFIG_GBLUR_FILTER) += x86/vf_gblur.o |
| @@ -0,0 +1,82 @@ | |||
| 1 | ;***************************************************************************** | ||
| 2 | ;* x86-optimized functions for eq filter | ||
| 3 | ;* | ||
| 4 | ;* Original MPlayer filters by Richard Felker. | ||
| 5 | ;* | ||
| 6 | ;* This file is part of FFmpeg. | ||
| 7 | ;* | ||
| 8 | ;* FFmpeg is free software; you can redistribute it and/or modify | ||
| 9 | ;* it under the terms of the GNU General Public License as published by | ||
| 10 | ;* the Free Software Foundation; either version 2 of the License, or | ||
| 11 | ;* (at your option) any later version. | ||
| 12 | ;* | ||
| 13 | ;* FFmpeg is distributed in the hope that it will be useful, | ||
| 14 | ;* but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 15 | ;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | ||
| 16 | ;* GNU General Public License for more details. | ||
| 17 | ;* | ||
| 18 | ;* You should have received a copy of the GNU General Public License along | ||
| 19 | ;* with FFmpeg; if not, write to the Free Software Foundation, Inc., | ||
| 20 | ;* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. | ||
| 21 | ;***************************************************************************** | ||
| 22 | |||
| 23 | %include "libavutil/x86/x86util.asm" | ||
| 24 | |||
| 25 | SECTION .text | ||
| 26 | |||
| 27 | INIT_MMX mmxext | ||
| 28 | cglobal process_one_line, 5, 7, 5, src, dst, contrast, brightness, w | ||
| 29 | movd m3, contrastd | ||
| 30 | movd m4, brightnessd | ||
| 31 | movsx r5d, contrastw | ||
| 32 | movsx r6d, brightnessw | ||
| 33 | SPLATW m3, m3, 0 | ||
| 34 | SPLATW m4, m4, 0 | ||
| 35 | |||
| 36 | DEFINE_ARGS src, dst, tmp, scalar, w | ||
| 37 | xor tmpd, tmpd | ||
| 38 | pxor m0, m0 | ||
| 39 | pxor m1, m1 | ||
| 40 | mov scalard, wd | ||
| 41 | and scalard, mmsize-1 | ||
| 42 | sar wd, 3 | ||
| 43 | cmp wd, 1 | ||
| 44 | jl .loop1 | ||
| 45 | |||
| 46 | .loop0: | ||
| 47 | movu m1, [srcq] | ||
| 48 | mova m2, m1 | ||
| 49 | punpcklbw m1, m0 | ||
| 50 | punpckhbw m2, m0 | ||
| 51 | psllw m1, 4 | ||
| 52 | psllw m2, 4 | ||
| 53 | pmulhw m1, m3 | ||
| 54 | pmulhw m2, m3 | ||
| 55 | paddw m1, m4 | ||
| 56 | paddw m2, m4 | ||
| 57 | packuswb m1, m2 | ||
| 58 | movu [dstq], m1 | ||
| 59 | add srcq, mmsize | ||
| 60 | add dstq, mmsize | ||
| 61 | sub wd, 1 | ||
| 62 | cmp wd, 0 | ||
| 63 | jne .loop0 | ||
| 64 | |||
| 65 | .loop1: | ||
| 66 | cmp scalard, 0 | ||
| 67 | je .end | ||
| 68 | movzx tmpd, byte [srcq] | ||
| 69 | imul tmpd, r5d | ||
| 70 | sar tmpd, 12 | ||
| 71 | add tmpd, r6d | ||
| 72 | movd m1, tmpd | ||
| 73 | packuswb m1, m0 | ||
| 74 | movd tmpd, m1 | ||
| 75 | mov [dstq], tmpb | ||
| 76 | inc srcq | ||
| 77 | inc dstq | ||
| 78 | dec scalard | ||
| 79 | jmp .loop1 | ||
| 80 | |||
| 81 | .end: | ||
| 82 | RET | ||
| @@ -1,96 +0,0 @@ | |||
| 1 | /* | ||
| 2 | * | ||
| 3 | * Original MPlayer filters by Richard Felker. | ||
| 4 | * | ||
| 5 | * This file is part of FFmpeg. | ||
| 6 | * | ||
| 7 | * FFmpeg is free software; you can redistribute it and/or modify | ||
| 8 | * it under the terms of the GNU General Public License as published by | ||
| 9 | * the Free Software Foundation; either version 2 of the License, or | ||
| 10 | * (at your option) any later version. | ||
| 11 | * | ||
| 12 | * FFmpeg is distributed in the hope that it will be useful, | ||
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | ||
| 15 | * GNU General Public License for more details. | ||
| 16 | * | ||
| 17 | * You should have received a copy of the GNU General Public License along | ||
| 18 | * with FFmpeg; if not, write to the Free Software Foundation, Inc., | ||
| 19 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. | ||
| 20 | */ | ||
| 21 | |||
| 22 | #include "libavutil/attributes.h" | ||
| 23 | #include "libavutil/cpu.h" | ||
| 24 | #include "libavutil/mem.h" | ||
| 25 | #include "libavutil/x86/asm.h" | ||
| 26 | #include "libavfilter/vf_eq.h" | ||
| 27 | |||
| 28 | #if HAVE_MMX_INLINE && HAVE_6REGS | ||
| 29 | static void process_MMX(EQParameters *param, uint8_t *dst, int dst_stride, | ||
| 30 | const uint8_t *src, int src_stride, int w, int h) | ||
| 31 | { | ||
| 32 | int i; | ||
| 33 | int pel; | ||
| 34 | int dstep = dst_stride - w; | ||
| 35 | int sstep = src_stride - w; | ||
| 36 | short brvec[4]; | ||
| 37 | short contvec[4]; | ||
| 38 | int contrast = (int) (param->contrast * 256 * 16); | ||
| 39 | int brightness = ((int) (100.0 * param->brightness + 100.0) * 511) / 200 - 128 - contrast / 32; | ||
| 40 | |||
| 41 | brvec[0] = brvec[1] = brvec[2] = brvec[3] = brightness; | ||
| 42 | contvec[0] = contvec[1] = contvec[2] = contvec[3] = contrast; | ||
| 43 | |||
| 44 | while (h--) { | ||
| 45 | __asm__ volatile ( | ||
| 46 | "movq (%5), %%mm3 \n\t" | ||
| 47 | "movq (%6), %%mm4 \n\t" | ||
| 48 | "pxor %%mm0, %%mm0 \n\t" | ||
| 49 | "movl %4, %%eax \n\t" | ||
| 50 | ".p2align 4 \n\t" | ||
| 51 | "1: \n\t" | ||
| 52 | "movq (%0), %%mm1 \n\t" | ||
| 53 | "movq (%0), %%mm2 \n\t" | ||
| 54 | "punpcklbw %%mm0, %%mm1\n\t" | ||
| 55 | "punpckhbw %%mm0, %%mm2\n\t" | ||
| 56 | "psllw $4, %%mm1 \n\t" | ||
| 57 | "psllw $4, %%mm2 \n\t" | ||
| 58 | "pmulhw %%mm4, %%mm1 \n\t" | ||
| 59 | "pmulhw %%mm4, %%mm2 \n\t" | ||
| 60 | "paddw %%mm3, %%mm1 \n\t" | ||
| 61 | "paddw %%mm3, %%mm2 \n\t" | ||
| 62 | "packuswb %%mm2, %%mm1 \n\t" | ||
| 63 | "add $8, %0 \n\t" | ||
| 64 | "movq %%mm1, (%1) \n\t" | ||
| 65 | "add $8, %1 \n\t" | ||
| 66 | "decl %%eax \n\t" | ||
| 67 | "jnz 1b \n\t" | ||
| 68 | : "=r" (src), "=r" (dst) | ||
| 69 | : "0" (src), "1" (dst), "r" (w>>3), "r" (brvec), "r" (contvec) | ||
| 70 | : "%eax" | ||
| 71 | ); | ||
| 72 | |||
| 73 | for (i = w&7; i; i--) { | ||
| 74 | pel = ((*src++ * contrast) >> 12) + brightness; | ||
| 75 | if (pel & ~255) | ||
| 76 | pel = (-pel) >> 31; | ||
| 77 | *dst++ = pel; | ||
| 78 | } | ||
| 79 | |||
| 80 | src += sstep; | ||
| 81 | dst += dstep; | ||
| 82 | } | ||
| 83 | __asm__ volatile ( "emms \n\t" ::: "memory" ); | ||
| 84 | } | ||
| 85 | #endif | ||
| 86 | |||
| 87 | av_cold void ff_eq_init_x86(EQContext *eq) | ||
| 88 | { | ||
| 89 | #if HAVE_MMX_INLINE && HAVE_6REGS | ||
| 90 | int cpu_flags = av_get_cpu_flags(); | ||
| 91 | |||
| 92 | if (cpu_flags & AV_CPU_FLAG_MMX) { | ||
| 93 | eq->process = process_MMX; | ||
| 94 | } | ||
| 95 | #endif | ||
| 96 | } | ||
| @@ -0,0 +1,54 @@ | |||
| 1 | /* | ||
| 2 | * | ||
| 3 | * Original MPlayer filters by Richard Felker. | ||
| 4 | * | ||
| 5 | * This file is part of FFmpeg. | ||
| 6 | * | ||
| 7 | * FFmpeg is free software; you can redistribute it and/or modify | ||
| 8 | * it under the terms of the GNU General Public License as published by | ||
| 9 | * the Free Software Foundation; either version 2 of the License, or | ||
| 10 | * (at your option) any later version. | ||
| 11 | * | ||
| 12 | * FFmpeg is distributed in the hope that it will be useful, | ||
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | ||
| 15 | * GNU General Public License for more details. | ||
| 16 | * | ||
| 17 | * You should have received a copy of the GNU General Public License along | ||
| 18 | * with FFmpeg; if not, write to the Free Software Foundation, Inc., | ||
| 19 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. | ||
| 20 | */ | ||
| 21 | |||
| 22 | #include "libavutil/attributes.h" | ||
| 23 | #include "libavutil/cpu.h" | ||
| 24 | #include "libavutil/mem.h" | ||
| 25 | #include "libavutil/x86/cpu.h" | ||
| 26 | #include "libavutil/x86/asm.h" | ||
| 27 | #include "libavfilter/vf_eq.h" | ||
| 28 | |||
| 29 | extern void ff_process_one_line_mmxext(const uint8_t *src, uint8_t *dst, short contrast, | ||
| 30 | short brightness, int w); | ||
| 31 | |||
| 32 | static void process_mmxext(EQParameters *param, uint8_t *dst, int dst_stride, | ||
| 33 | const uint8_t *src, int src_stride, int w, int h) | ||
| 34 | { | ||
| 35 | short contrast = (short) (param->contrast * 256 * 16); | ||
| 36 | short brightness = ((short) (100.0 * param->brightness + 100.0) * 511) | ||
| 37 | / 200 - 128 - contrast / 32; | ||
| 38 | |||
| 39 | while (h--) { | ||
| 40 | ff_process_one_line_mmxext(src, dst, contrast, brightness, w); | ||
| 41 | src += src_stride; | ||
| 42 | dst += dst_stride; | ||
| 43 | } | ||
| 44 | emms_c(); | ||
| 45 | } | ||
| 46 | |||
| 47 | av_cold void ff_eq_init_x86(EQContext *eq) | ||
| 48 | { | ||
| 49 | int cpu_flags = av_get_cpu_flags(); | ||
| 50 | |||
| 51 | if (cpu_flags & AV_CPU_FLAG_MMXEXT) { | ||
| 52 | eq->process = process_mmxext; | ||
| 53 | } | ||
| 54 | } | ||