On 2012-01-31 11:40:15 -0500, Justin Ruggles wrote: > From: Reimar Döffinger <[email protected]> > > Ensures alignment and avoids using uninitialized data. > > Signed-off-by: Justin Ruggles <[email protected]> > --- > libavcodec/fraps.c | 11 +++++++++-- > 1 files changed, 9 insertions(+), 2 deletions(-) > > diff --git a/libavcodec/fraps.c b/libavcodec/fraps.c > index 1444eda..5f5e55e 100644 > --- a/libavcodec/fraps.c > +++ b/libavcodec/fraps.c > @@ -46,6 +46,7 @@ typedef struct FrapsContext{ > AVCodecContext *avctx; > AVFrame frame; > uint8_t *tmpbuf; > + int tmpbuf_size; > DSPContext dsp; > } FrapsContext; > > @@ -276,7 +277,10 @@ static int decode_frame(AVCodecContext *avctx, > offs[planes] = buf_size; > for(i = 0; i < planes; i++){ > is_chroma = !!i; > - s->tmpbuf = av_realloc(s->tmpbuf, offs[i + 1] - offs[i] - 1024 + > FF_INPUT_BUFFER_PADDING_SIZE); > + av_fast_padded_malloc(&s->tmpbuf, &s->tmpbuf_size, > + offs[i + 1] - offs[i] - 1024); > + if (!s->tmpbuf) > + return AVERROR(ENOMEM); > if(fraps2_decode_plane(s, f->data[i], f->linesize[i], > avctx->width >> is_chroma, > avctx->height >> is_chroma, buf + offs[i], offs[i + 1] - > offs[i], is_chroma, 1) < 0) { > av_log(avctx, AV_LOG_ERROR, "Error decoding plane %i\n", i); > @@ -318,7 +322,10 @@ static int decode_frame(AVCodecContext *avctx, > } > offs[planes] = buf_size; > for(i = 0; i < planes; i++){ > - s->tmpbuf = av_realloc(s->tmpbuf, offs[i + 1] - offs[i] - 1024 + > FF_INPUT_BUFFER_PADDING_SIZE); > + av_fast_padded_malloc(&s->tmpbuf, &s->tmpbuf_size, > + offs[i + 1] - offs[i] - 1024); > + if (!s->tmpbuf) > + return AVERROR(ENOMEM); > if(fraps2_decode_plane(s, f->data[0] + i + (f->linesize[0] * > (avctx->height - 1)), -f->linesize[0], > avctx->width, avctx->height, buf + offs[i], offs[i + 1] > - offs[i], 0, 3) < 0) { > av_log(avctx, AV_LOG_ERROR, "Error decoding plane %i\n", i);
probably ok if the uninitialized data just involves the FF_INPUT_BUFFER_PADDING. Janne _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
