comparison utils.c @ 11311:ee2e050815be libavcodec

Fix avcodec_align_dimensions to return values suitably aligned for FLV decoding with SSE and add a avcodec_align_dimensions2 taht returns the stride alignment requirements independently from doing the width/height padding.

11310:1ff8ae765206

#define _XOPEN_SOURCE 600

#include "libavutil/avstring.h"
#include "libavutil/integer.h"
#include "libavutil/crc.h"
#include "avcodec.h"
#include "dsputil.h"
#include "opt.h"
#include "imgconvert.h"
#include "audioconvert.h"

    enum PixelFormat pix_fmt;
}InternalBuffer;

#define INTERNAL_BUFFER_SIZE 32

void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height){
    int w_align= 1;
    int h_align= 1;

    switch(s->pix_fmt){
    case PIX_FMT_YUV420P:

    *width = FFALIGN(*width , w_align);
    *height= FFALIGN(*height, h_align);
    if(s->codec_id == CODEC_ID_H264)
        *height+=2; // some of the optimized chroma MC reads one line too much
}

int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h){
    if((int)w>0 && (int)h>0 && (w+128)*(uint64_t)(h+128) < INT_MAX/8)
        return 0;

        AVPicture picture;
        int stride_align[4];

        avcodec_get_chroma_sub_sample(s->pix_fmt, &h_chroma_shift, &v_chroma_shift);

        avcodec_align_dimensions(s, &w, &h);

        if(!(s->flags&CODEC_FLAG_EMU_EDGE)){
            w+= EDGE_WIDTH*2;
            h+= EDGE_WIDTH*2;
        }

            // increase alignment of w for next try (rhs gives the lowest bit set in w)
            w += w & ~(w-1);

            unaligned = 0;
            for (i=0; i<4; i++){
//STRIDE_ALIGN is 8 for SSE* but this does not work for SVQ1 chroma planes
//we could change STRIDE_ALIGN to 16 for x86/sse but it would increase the
//picture size unneccessarily in some cases. The solution here is not
//pretty and better ideas are welcome!
#if HAVE_MMX
                if(s->codec_id == CODEC_ID_SVQ1 || s->codec_id == CODEC_ID_VP5 ||
                   s->codec_id == CODEC_ID_VP6 || s->codec_id == CODEC_ID_VP6F ||
                   s->codec_id == CODEC_ID_VP6A)
                    stride_align[i]= 16;
                else
#endif
                stride_align[i] = STRIDE_ALIGN;
                unaligned |= picture.linesize[i] % stride_align[i];
            }
        } while (unaligned);

        tmpsize = ff_fill_pointer(&picture, NULL, s->pix_fmt, h);

11311:ee2e050815be

#define _XOPEN_SOURCE 600

#include "libavutil/avstring.h"
#include "libavutil/integer.h"
#include "libavutil/crc.h"
#include "libavutil/pixdesc.h"
#include "avcodec.h"
#include "dsputil.h"
#include "opt.h"
#include "imgconvert.h"
#include "audioconvert.h"

    enum PixelFormat pix_fmt;
}InternalBuffer;

#define INTERNAL_BUFFER_SIZE 32

void avcodec_align_dimensions2(AVCodecContext *s, int *width, int *height, int linesize_align[4]){
    int w_align= 1;
    int h_align= 1;

    switch(s->pix_fmt){
    case PIX_FMT_YUV420P:

    *width = FFALIGN(*width , w_align);
    *height= FFALIGN(*height, h_align);
    if(s->codec_id == CODEC_ID_H264)
        *height+=2; // some of the optimized chroma MC reads one line too much

    linesize_align[0] =
    linesize_align[1] =
    linesize_align[2] =
    linesize_align[3] = STRIDE_ALIGN;
//STRIDE_ALIGN is 8 for SSE* but this does not work for SVQ1 chroma planes
//we could change STRIDE_ALIGN to 16 for x86/sse but it would increase the
//picture size unneccessarily in some cases. The solution here is not
//pretty and better ideas are welcome!
#if HAVE_MMX
    if(s->codec_id == CODEC_ID_SVQ1 || s->codec_id == CODEC_ID_VP5 ||
       s->codec_id == CODEC_ID_VP6 || s->codec_id == CODEC_ID_VP6F ||
       s->codec_id == CODEC_ID_VP6A) {
        linesize_align[0] =
        linesize_align[1] =
        linesize_align[2] = 16;
    }
#endif
}

void avcodec_align_dimensions(AVCodecContext *s, int *width, int *height){
    int chroma_shift = av_pix_fmt_descriptors[s->pix_fmt].log2_chroma_w;
    int linesize_align[4];
    int align;
    avcodec_align_dimensions2(s, width, height, linesize_align);
    align = FFMAX(linesize_align[0], linesize_align[3]);
    linesize_align[1] <<= chroma_shift;
    linesize_align[2] <<= chroma_shift;
    align = FFMAX3(align, linesize_align[1], linesize_align[2]);
    *width=FFALIGN(*width, align);
}

int avcodec_check_dimensions(void *av_log_ctx, unsigned int w, unsigned int h){
    if((int)w>0 && (int)h>0 && (w+128)*(uint64_t)(h+128) < INT_MAX/8)
        return 0;

        AVPicture picture;
        int stride_align[4];

        avcodec_get_chroma_sub_sample(s->pix_fmt, &h_chroma_shift, &v_chroma_shift);

        avcodec_align_dimensions2(s, &w, &h, stride_align);

        if(!(s->flags&CODEC_FLAG_EMU_EDGE)){
            w+= EDGE_WIDTH*2;
            h+= EDGE_WIDTH*2;
        }

            // increase alignment of w for next try (rhs gives the lowest bit set in w)
            w += w & ~(w-1);

            unaligned = 0;
            for (i=0; i<4; i++){
                unaligned |= picture.linesize[i] % stride_align[i];
            }
        } while (unaligned);

        tmpsize = ff_fill_pointer(&picture, NULL, s->pix_fmt, h);