[FFmpeg-devel] [PATCH 04/10] avformat/matroskadec: Use generic size check for signed integers

[James Almer]

On 3/8/2019 6:25 AM, Andreas Rheinhardt wrote:
> and drop the redundant checks contained in ebml_read_uint and
> ebml_read_sint.
> 
> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt at googlemail.com>
> ---
>  libavformat/matroskadec.c | 7 +------
>  1 file changed, 1 insertion(+), 6 deletions(-)
> 
> diff --git a/libavformat/matroskadec.c b/libavformat/matroskadec.c
> index 88e80b2fda..55a153d982 100644
> --- a/libavformat/matroskadec.c
> +++ b/libavformat/matroskadec.c
> @@ -882,9 +882,6 @@ static int ebml_read_uint(AVIOContext *pb, int size, uint64_t *num)
>  {
>      int n = 0;
>  
> -    if (size > 8)
> -        return AVERROR_INVALIDDATA;
> -
>      /* big-endian ordering; build up number */
>      *num = 0;
>      while (n++ < size)
> @@ -901,9 +898,6 @@ static int ebml_read_sint(AVIOContext *pb, int size, int64_t *num)
>  {
>      int n = 1;
>  
> -    if (size > 8)
> -        return AVERROR_INVALIDDATA;
> -
>      if (size == 0) {
>          *num = 0;
>      } else {
> @@ -1161,6 +1155,7 @@ static int ebml_parse_elem(MatroskaDemuxContext *matroska,
>  {
>      static const uint64_t max_lengths[EBML_TYPE_COUNT] = {
>          [EBML_UINT]  = 8,
> +        [EBML_SINT]  = 8,
>          [EBML_FLOAT] = 8,
>          // max. 16 MB for strings
>          [EBML_STR]   = 0x1000000,

Should be ok.