[FFmpeg-devel] [PATCH] avfilter: add limiter filter
James Almer
jamrial at gmail.com
Tue Jul 4 00:43:55 EEST 2017
On 7/3/2017 4:45 PM, Paul B Mahol wrote:
> Signed-off-by: Paul B Mahol <onemda at gmail.com>
> ---
> doc/filters.texi | 17 +++
> libavfilter/Makefile | 1 +
> libavfilter/allfilters.c | 1 +
> libavfilter/limiter.h | 33 ++++++
> libavfilter/vf_limiter.c | 228 ++++++++++++++++++++++++++++++++++++++
> libavfilter/x86/Makefile | 2 +
> libavfilter/x86/vf_limiter.asm | 80 +++++++++++++
> libavfilter/x86/vf_limiter_init.c | 41 +++++++
> 8 files changed, 403 insertions(+)
> create mode 100644 libavfilter/limiter.h
> create mode 100644 libavfilter/vf_limiter.c
> create mode 100644 libavfilter/x86/vf_limiter.asm
> create mode 100644 libavfilter/x86/vf_limiter_init.c
>
> diff --git a/doc/filters.texi b/doc/filters.texi
> index 930ca4c..dbf0fc1 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -9639,6 +9639,23 @@ The formula that generates the correction is:
> where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
> distances from the focal point in the source and target images, respectively.
>
> + at section limiter
> +
> +Limits the pixel components values to the specified range [min, max].
> +
> +The filter accepts the following options:
> +
> + at table @option
> + at item min
> +Lower bound. Defaults to the lowest allowed value for the input.
> +
> + at item max
> +Upper bound. Defaults to the highest allowed value for the input.
> +
> + at item planes
> +Specify which planes will be processed.
> + at end table
> +
> @section loop
>
> Loop video frames.
> diff --git a/libavfilter/Makefile b/libavfilter/Makefile
> index f023a0d..4d85f65 100644
> --- a/libavfilter/Makefile
> +++ b/libavfilter/Makefile
> @@ -216,6 +216,7 @@ OBJS-$(CONFIG_INTERLACE_FILTER) += vf_interlace.o
> OBJS-$(CONFIG_INTERLEAVE_FILTER) += f_interleave.o
> OBJS-$(CONFIG_KERNDEINT_FILTER) += vf_kerndeint.o
> OBJS-$(CONFIG_LENSCORRECTION_FILTER) += vf_lenscorrection.o
> +OBJS-$(CONFIG_LIMITER_FILTER) += vf_limiter.o
> OBJS-$(CONFIG_LOOP_FILTER) += f_loop.o
> OBJS-$(CONFIG_LUMAKEY_FILTER) += vf_lumakey.o
> OBJS-$(CONFIG_LUT_FILTER) += vf_lut.o
> diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
> index c1c5233..0a990ca 100644
> --- a/libavfilter/allfilters.c
> +++ b/libavfilter/allfilters.c
> @@ -228,6 +228,7 @@ static void register_all(void)
> REGISTER_FILTER(INTERLEAVE, interleave, vf);
> REGISTER_FILTER(KERNDEINT, kerndeint, vf);
> REGISTER_FILTER(LENSCORRECTION, lenscorrection, vf);
> + REGISTER_FILTER(LIMITER, limiter, vf);
> REGISTER_FILTER(LOOP, loop, vf);
> REGISTER_FILTER(LUMAKEY, lumakey, vf);
> REGISTER_FILTER(LUT, lut, vf);
> diff --git a/libavfilter/limiter.h b/libavfilter/limiter.h
> new file mode 100644
> index 0000000..54d423d
> --- /dev/null
> +++ b/libavfilter/limiter.h
> @@ -0,0 +1,33 @@
> +/*
> + * 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 AVFILTER_LIMITER_H
> +#define AVFILTER_LIMITER_H
> +
> +#include <stddef.h>
> +#include <stdint.h>
> +
> +typedef struct LimiterDSPContext {
> + void (*limiter)(const uint8_t *src, uint8_t *dst,
> + ptrdiff_t slinesize, ptrdiff_t dlinesize,
> + int w, int h, int min, int max);
> +} LimiterDSPContext;
> +
> +void ff_limiter_init_x86(LimiterDSPContext *dsp, int bpp);
> +
> +#endif /* AVFILTER_LIMITER_H */
> diff --git a/libavfilter/vf_limiter.c b/libavfilter/vf_limiter.c
> new file mode 100644
> index 0000000..d6f5745
> --- /dev/null
> +++ b/libavfilter/vf_limiter.c
> @@ -0,0 +1,228 @@
> +/*
> + * 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/attributes.h"
> +#include "libavutil/common.h"
> +#include "libavutil/eval.h"
> +#include "libavutil/imgutils.h"
> +#include "libavutil/opt.h"
> +#include "libavutil/pixdesc.h"
> +#include "avfilter.h"
> +#include "formats.h"
> +#include "internal.h"
> +#include "limiter.h"
> +#include "video.h"
> +
> +typedef struct LimiterContext {
> + const AVClass *class;
> + int min;
> + int max;
> + int planes;
> + int nb_planes;
> + int linesize[4];
> + int width[4];
> + int height[4];
> +
> + LimiterDSPContext dsp;
> +} LimiterContext;
> +
> +#define OFFSET(x) offsetof(LimiterContext, x)
> +#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_VIDEO_PARAM
> +
> +static const AVOption limiter_options[] = {
> + { "min", "set min value", OFFSET(min), AV_OPT_TYPE_INT, {.i64=0}, 0, 65535, .flags = FLAGS },
> + { "max", "set max value", OFFSET(max), AV_OPT_TYPE_INT, {.i64=65535}, 0, 65535, .flags = FLAGS },
> + { "planes", "set planes", OFFSET(planes), AV_OPT_TYPE_INT, {.i64=15}, 0, 65535, .flags = FLAGS },
> + { NULL }
> +};
> +
> +AVFILTER_DEFINE_CLASS(limiter);
> +
> +static av_cold int init(AVFilterContext *ctx)
> +{
> + LimiterContext *s = ctx->priv;
> +
> + if (s->min > s->max)
> + return AVERROR(EINVAL);
> + return 0;
> +}
> +
> +static int query_formats(AVFilterContext *ctx)
> +{
> + static const enum AVPixelFormat pix_fmts[] = {
> + AV_PIX_FMT_YUVA444P, AV_PIX_FMT_YUV444P, AV_PIX_FMT_YUV440P,
> + AV_PIX_FMT_YUVJ444P, AV_PIX_FMT_YUVJ440P,
> + AV_PIX_FMT_YUVA422P, AV_PIX_FMT_YUV422P, AV_PIX_FMT_YUVA420P, AV_PIX_FMT_YUV420P,
> + AV_PIX_FMT_YUVJ422P, AV_PIX_FMT_YUVJ420P,
> + AV_PIX_FMT_YUVJ411P, AV_PIX_FMT_YUV411P, AV_PIX_FMT_YUV410P,
> + AV_PIX_FMT_YUV420P9, AV_PIX_FMT_YUV422P9, AV_PIX_FMT_YUV444P9,
> + AV_PIX_FMT_YUV420P10, AV_PIX_FMT_YUV422P10, AV_PIX_FMT_YUV444P10,
> + AV_PIX_FMT_YUV420P12, AV_PIX_FMT_YUV422P12, AV_PIX_FMT_YUV444P12, AV_PIX_FMT_YUV440P12,
> + AV_PIX_FMT_YUV420P14, AV_PIX_FMT_YUV422P14, AV_PIX_FMT_YUV444P14,
> + AV_PIX_FMT_YUV420P16, AV_PIX_FMT_YUV422P16, AV_PIX_FMT_YUV444P16,
> + AV_PIX_FMT_YUVA420P9, AV_PIX_FMT_YUVA422P9, AV_PIX_FMT_YUVA444P9,
> + AV_PIX_FMT_YUVA420P10, AV_PIX_FMT_YUVA422P10, AV_PIX_FMT_YUVA444P10,
> + AV_PIX_FMT_YUVA420P16, AV_PIX_FMT_YUVA422P16, AV_PIX_FMT_YUVA444P16,
> + AV_PIX_FMT_GBRP, AV_PIX_FMT_GBRP9, AV_PIX_FMT_GBRP10,
> + AV_PIX_FMT_GBRP12, AV_PIX_FMT_GBRP14, AV_PIX_FMT_GBRP16,
> + AV_PIX_FMT_GBRAP, AV_PIX_FMT_GBRAP12, AV_PIX_FMT_GBRAP16,
> + AV_PIX_FMT_GRAY8, AV_PIX_FMT_GRAY10, AV_PIX_FMT_GRAY12, AV_PIX_FMT_GRAY16,
> + AV_PIX_FMT_NONE
> + };
> +
> + AVFilterFormats *fmts_list = ff_make_format_list(pix_fmts);
> + if (!fmts_list)
> + return AVERROR(ENOMEM);
> + return ff_set_common_formats(ctx, fmts_list);
> +}
> +
> +static void limiter8(const uint8_t *src, uint8_t *dst,
> + ptrdiff_t slinesize, ptrdiff_t dlinesize,
> + int w, int h, int min, int max)
> +{
> + int x, y;
> +
> + for (y = 0; y < h; y++) {
> + for (x = 0; x < w; x++) {
> + dst[x] = av_clip(src[x], min, max);
> + }
> +
> + dst += dlinesize;
> + src += slinesize;
> + }
> +}
> +
> +static void limiter16(const uint8_t *ssrc, uint8_t *ddst,
> + ptrdiff_t slinesize, ptrdiff_t dlinesize,
> + int w, int h, int min, int max)
> +{
> + const uint16_t *src = (const uint16_t *)ssrc;
> + uint16_t *dst = (uint16_t *)ddst;
> + int x, y;
> +
> + dlinesize /= 2;
> + slinesize /= 2;
> +
> + for (y = 0; y < h; y++) {
> + for (x = 0; x < w; x++) {
> + dst[x] = av_clip(src[x], min, max);
> + }
> +
> + dst += dlinesize;
> + src += slinesize;
> + }
> +}
> +
> +static int config_props(AVFilterLink *inlink)
> +{
> + AVFilterContext *ctx = inlink->dst;
> + LimiterContext *s = ctx->priv;
> + const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(inlink->format);
> + int vsub, hsub, ret;
> +
> + s->nb_planes = av_pix_fmt_count_planes(inlink->format);
> +
> + if ((ret = av_image_fill_linesizes(s->linesize, inlink->format, inlink->w)) < 0)
> + return ret;
> +
> + hsub = desc->log2_chroma_w;
> + vsub = desc->log2_chroma_h;
> + s->height[1] = s->height[2] = AV_CEIL_RSHIFT(inlink->h, vsub);
> + s->height[0] = s->height[3] = inlink->h;
> + s->width[1] = s->width[2] = AV_CEIL_RSHIFT(inlink->w, hsub);
> + s->width[0] = s->width[3] = inlink->w;
> +
> + if (desc->comp[0].depth == 8)
> + s->dsp.limiter = limiter8;
> + else
> + s->dsp.limiter = limiter16;
> +
> + if (ARCH_X86)
> + ff_limiter_init_x86(&s->dsp, desc->comp[0].depth);
> +
> + return 0;
> +}
> +
> +static int filter_frame(AVFilterLink *inlink, AVFrame *in)
> +{
> + AVFilterContext *ctx = inlink->dst;
> + LimiterContext *s = ctx->priv;
> + AVFilterLink *outlink = ctx->outputs[0];
> + AVFrame *out;
> + int p;
> +
> + if (av_frame_is_writable(in)) {
> + out = in;
> + } else {
> + out = ff_get_video_buffer(outlink, outlink->w, outlink->h);
> + if (!out) {
> + av_frame_free(&in);
> + return AVERROR(ENOMEM);
> + }
> + av_frame_copy_props(out, in);
> + }
> +
> + for (p = 0; p < s->nb_planes; p++) {
> + if (!((1 << p) & s->planes)) {
> + if (out != in)
> + av_image_copy_plane(out->data[p], out->linesize[p], in->data[p], in->linesize[p],
> + s->linesize[p], s->height[p]);
> + continue;
> + }
> +
> + s->dsp.limiter(in->data[p], out->data[p],
> + in->linesize[p], out->linesize[p],
> + s->width[p], s->height[p],
> + s->min, s->max);
> + }
> +
> + if (out != in)
> + av_frame_free(&in);
> +
> + return ff_filter_frame(outlink, out);
> +}
> +
> +static const AVFilterPad inputs[] = {
> + {
> + .name = "default",
> + .type = AVMEDIA_TYPE_VIDEO,
> + .filter_frame = filter_frame,
> + .config_props = config_props,
> + },
> + { NULL }
> +};
> +
> +static const AVFilterPad outputs[] = {
> + {
> + .name = "default",
> + .type = AVMEDIA_TYPE_VIDEO,
> + },
> + { NULL }
> +};
> +
> +AVFilter ff_vf_limiter = {
> + .name = "limiter",
> + .description = NULL_IF_CONFIG_SMALL("Limit pixels components to the specified range."),
> + .priv_size = sizeof(LimiterContext),
> + .priv_class = &limiter_class,
> + .init = init,
> + .query_formats = query_formats,
> + .inputs = inputs,
> + .outputs = outputs,
> + .flags = AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC,
> +};
> diff --git a/libavfilter/x86/Makefile b/libavfilter/x86/Makefile
> index f598ceb..3431625 100644
> --- a/libavfilter/x86/Makefile
> +++ b/libavfilter/x86/Makefile
> @@ -8,6 +8,7 @@ OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun_init.o
> OBJS-$(CONFIG_HQDN3D_FILTER) += x86/vf_hqdn3d_init.o
> OBJS-$(CONFIG_IDET_FILTER) += x86/vf_idet_init.o
> OBJS-$(CONFIG_INTERLACE_FILTER) += x86/vf_interlace_init.o
> +OBJS-$(CONFIG_LIMITER_FILTER) += x86/vf_limiter_init.o
> OBJS-$(CONFIG_MASKEDMERGE_FILTER) += x86/vf_maskedmerge_init.o
> OBJS-$(CONFIG_NOISE_FILTER) += x86/vf_noise.o
> OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7_init.o
> @@ -33,6 +34,7 @@ X86ASM-OBJS-$(CONFIG_GRADFUN_FILTER) += x86/vf_gradfun.o
> X86ASM-OBJS-$(CONFIG_HQDN3D_FILTER) += x86/vf_hqdn3d.o
> X86ASM-OBJS-$(CONFIG_IDET_FILTER) += x86/vf_idet.o
> X86ASM-OBJS-$(CONFIG_INTERLACE_FILTER) += x86/vf_interlace.o
> +X86ASM-OBJS-$(CONFIG_LIMITER_FILTER) += x86/vf_limiter.o
> X86ASM-OBJS-$(CONFIG_MASKEDMERGE_FILTER) += x86/vf_maskedmerge.o
> X86ASM-OBJS-$(CONFIG_PP7_FILTER) += x86/vf_pp7.o
> X86ASM-OBJS-$(CONFIG_PSNR_FILTER) += x86/vf_psnr.o
> diff --git a/libavfilter/x86/vf_limiter.asm b/libavfilter/x86/vf_limiter.asm
> new file mode 100644
> index 0000000..7d61c26
> --- /dev/null
> +++ b/libavfilter/x86/vf_limiter.asm
> @@ -0,0 +1,80 @@
> +;*****************************************************************************
> +;* x86-optimized functions for limiter filter
> +;*
> +;* 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/x86/x86util.asm"
> +
> +SECTION_RODATA
> +
> +pb_0: times 16 db 0
> +
> +SECTION .text
> +
> +INIT_XMM sse2
> +
> +cglobal limiter_8bit, 8, 9, 3, src, dst, slinesize, dlinesize, w, h, min, max, x
More than 7 gprs is x86_64 only.
> + movsxdifnidn wq, wd
> + add srcq, wq
> + add dstq, wq
> + neg wq
> + SPLATB_REG m1, min, [pb_0]
> + SPLATB_REG m2, max, [pb_0]
> +.nextrow:
> + mov xq, wq
> +
> + .loop:
> + movu m0, [srcq + xq]
> + CLIPUB m0, m1, m2
> + mova [dstq+xq], m0
> + add xq, mmsize
> + jl .loop
> +
> + add srcq, slinesizeq
> + add dstq, dlinesizeq
> + sub hd, 1
> + jg .nextrow
> + ret
> +
> +cglobal limiter_16bit, 8, 9, 3, src, dst, slinesize, dlinesize, w, h, min, max, x
> + shl wd, 1
> + add srcq, wq
> + add dstq, wq
> + neg wq
> + pinsrw m1, minw, 0
movd m1, mind
> + SPLATW m1, m1
> + pinsrw m2, maxw, 0
movd m2, maxd
You also could, as part of making this work on x86_32, load these two
straight from stack with using r6m/r7m. That would leave you with
exactly seven gprs to store both pointers, both strides, both dimensions
and the temp value x.
> + SPLATW m2, m2
> +.nextrow:
> + mov xq, wq
> +
> + .loop:
> + movu m0, [srcq + xq]
> + pmaxuw m0, m1
> + pminuw m0, m2
These two are SSE4 instructions. You should be getting the relevant
warnings during assembly...
See libavcodec/x86/hevc_sao_10bit.asm for an example of SSE2 emulation
of pminuw, which you can move to x86util so it can be used here.
> + mova [dstq+xq], m0
> + add xq, mmsize
> + jl .loop
> +
> + add srcq, slinesizeq
> + add dstq, dlinesizeq
> + sub hd, 1
> + jg .nextrow
> + ret
> +
> + ret
> diff --git a/libavfilter/x86/vf_limiter_init.c b/libavfilter/x86/vf_limiter_init.c
> new file mode 100644
> index 0000000..348bb23
> --- /dev/null
> +++ b/libavfilter/x86/vf_limiter_init.c
> @@ -0,0 +1,41 @@
> +/*
> + * 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/x86/cpu.h"
> +
> +#include "libavfilter/limiter.h"
> +
> +void ff_limiter_8bit_sse2(const uint8_t *src, uint8_t *dst,
> + ptrdiff_t slinesize, ptrdiff_t dlinesize,
> + int w, int h, int min, int max);
> +void ff_limiter_16bit_sse2(const uint8_t *src, uint8_t *dst,
> + ptrdiff_t slinesize, ptrdiff_t dlinesize,
> + int w, int h, int min, int max);
> +
> +void ff_limiter_init_x86(LimiterDSPContext *dsp, int bpp)
> +{
> + int cpu_flags = av_get_cpu_flags();
> +
> + if (EXTERNAL_SSE2(cpu_flags)) {
> + if (bpp <= 8) {
> + dsp->limiter = ff_limiter_8bit_sse2;
> + } else if (bpp <= 16) {
> + dsp->limiter = ff_limiter_16bit_sse2;
> + }
> + }
> +}
>
More information about the ffmpeg-devel
mailing list