| Message ID | 20181016210128.7652-3-sw@jkqxz.net |
|---|---|
| State | Accepted |
| Commit | da6db843b80e6140dbf28f05195b6a17018fd68c |
| Headers | show |
On Tue, Oct 16, 2018 at 2:08 PM Mark Thompson <sw@jkqxz.net> wrote: > Any of these buffers (for both H.264 and H.265) might reasonably be > parsed using the bitstream reader, so include padding on all of them. > --- > This seems simpler than adding it to specific cases as problems are found. > > Entire patchset looks great to me. Thanks for cleaning up my mess. Aman > > libavcodec/cbs_h2645.c | 3 ++- > libavcodec/cbs_h264_syntax_template.c | 2 +- > 2 files changed, 3 insertions(+), 2 deletions(-) > > diff --git a/libavcodec/cbs_h2645.c b/libavcodec/cbs_h2645.c > index a1b92c87ce..e55bd00183 100644 > --- a/libavcodec/cbs_h2645.c > +++ b/libavcodec/cbs_h2645.c > @@ -319,7 +319,8 @@ static int cbs_h2645_read_more_rbsp_data(GetBitContext > *gbc) > #define byte_alignment(rw) (get_bits_count(rw) % 8) > > #define allocate(name, size) do { \ > - name ## _ref = av_buffer_allocz(size); \ > + name ## _ref = av_buffer_allocz(size + \ > + AV_INPUT_BUFFER_PADDING_SIZE); \ > if (!name ## _ref) \ > return AVERROR(ENOMEM); \ > name = name ## _ref->data; \ > diff --git a/libavcodec/cbs_h264_syntax_template.c > b/libavcodec/cbs_h264_syntax_template.c > index 7d0ce5918a..9e29132fec 100644 > --- a/libavcodec/cbs_h264_syntax_template.c > +++ b/libavcodec/cbs_h264_syntax_template.c > @@ -718,7 +718,7 @@ static int > FUNC(sei_user_data_registered)(CodedBitstreamContext *ctx, RWContext > *payload_size = i + current->data_length; > #endif > > - allocate(current->data, current->data_length + > AV_INPUT_BUFFER_PADDING_SIZE); > + allocate(current->data, current->data_length); > for (j = 0; j < current->data_length; j++) > xu(8, itu_t_t35_payload_byte[i], current->data[j], 0x00, 0xff, 1, > i + j); > > -- > 2.19.1 > > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-devel >
On 17/10/18 01:40, Aman Gupta wrote: > On Tue, Oct 16, 2018 at 2:08 PM Mark Thompson <sw@jkqxz.net> wrote: > >> Any of these buffers (for both H.264 and H.265) might reasonably be >> parsed using the bitstream reader, so include padding on all of them. >> --- >> This seems simpler than adding it to specific cases as problems are found. >> >> > Entire patchset looks great to me. Applied. Thank you! - Mark
diff --git a/libavcodec/cbs_h2645.c b/libavcodec/cbs_h2645.c index a1b92c87ce..e55bd00183 100644 --- a/libavcodec/cbs_h2645.c +++ b/libavcodec/cbs_h2645.c @@ -319,7 +319,8 @@ static int cbs_h2645_read_more_rbsp_data(GetBitContext *gbc) #define byte_alignment(rw) (get_bits_count(rw) % 8) #define allocate(name, size) do { \ - name ## _ref = av_buffer_allocz(size); \ + name ## _ref = av_buffer_allocz(size + \ + AV_INPUT_BUFFER_PADDING_SIZE); \ if (!name ## _ref) \ return AVERROR(ENOMEM); \ name = name ## _ref->data; \ diff --git a/libavcodec/cbs_h264_syntax_template.c b/libavcodec/cbs_h264_syntax_template.c index 7d0ce5918a..9e29132fec 100644 --- a/libavcodec/cbs_h264_syntax_template.c +++ b/libavcodec/cbs_h264_syntax_template.c @@ -718,7 +718,7 @@ static int FUNC(sei_user_data_registered)(CodedBitstreamContext *ctx, RWContext *payload_size = i + current->data_length; #endif - allocate(current->data, current->data_length + AV_INPUT_BUFFER_PADDING_SIZE); + allocate(current->data, current->data_length); for (j = 0; j < current->data_length; j++) xu(8, itu_t_t35_payload_byte[i], current->data[j], 0x00, 0xff, 1, i + j);