On Sun, Mar 11, 2018 at 7:55 PM, James Almer <jamr...@gmail.com> wrote:
> On 3/11/2018 3:30 PM, Mark Thompson wrote:
>> The SEI NAL is unused since 69062d0f9b6aef5d9d9b8c9c9b5cfb23037caddb,
>> while the AUD NAL is small and would more sensibly be on the stack.
>> ---
>> libavcodec/h264_metadata_bsf.c | 12 +++++-------
>> 1 file changed, 5 insertions(+), 7 deletions(-)
>>
>> diff --git a/libavcodec/h264_metadata_bsf.c b/libavcodec/h264_metadata_bsf.c
>> index d340c55990..760fe99c41 100644
>> --- a/libavcodec/h264_metadata_bsf.c
>> +++ b/libavcodec/h264_metadata_bsf.c
>> @@ -47,9 +47,6 @@ typedef struct H264MetadataContext {
>>
>> int done_first_au;
>>
>> - H264RawAUD aud_nal;
>> - H264RawSEI sei_nal;
>> -
>> int aud;
>>
>> AVRational sample_aspect_ratio;
>> @@ -263,7 +260,9 @@ static int h264_metadata_filter(AVBSFContext *bsf,
>> AVPacket *out)
>> 0x3ff, // 0, 1, 2, 3, 4, 5, 6, 7, 8, 9
>> };
>> int primary_pic_type_mask = 0xff;
>> - H264RawAUD *aud = &ctx->aud_nal;
>> + H264RawAUD aud = {
>> + .nal_unit_header.nal_unit_type = H264_NAL_AUD,
>> + };
>
> Afaik every other field is not zero initialized if you do this, unlike
> if you keep it in H264MetadataContext.
> Not sure if that may have some consequences or not here.
>
All other members are initialized with zero if you use any sort of
initializer syntax.
- Hendrik
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
