[Bug 664274] [API] codecparsers: add gst_mpeg_video_parse_slice() and gst_mpeg_video_parse_sequence_scalable_extension()

GStreamer (bugzilla.gnome.org) bugzilla at gnome.org
Wed Jan 30 14:13:13 PST 2013


https://bugzilla.gnome.org/show_bug.cgi?id=664274
  GStreamer | gst-plugins-bad | git

--- Comment #14 from Gwenole Beauchesne <gb.devel at gmail.com> 2013-01-30 22:13:09 UTC ---
Review of attachment 234829:
 --> (https://bugzilla.gnome.org/review?bug=664274&attachment=234829)

Hi, on second thought, the slice API is not fully consistent with the rest
(seq_hdr, pic_hdr, et al.). In particular, the offset for slice() is relative
to the start-code prefix, whereas it is relative to the actual data (after the
start code + type) for others. I suggest we globally change the MPEG-2 video
parser API to take a const GstMpegVideoPacket * instead of multiple arguments
like data, size, offset. That way, the slice API would also know about the type
to calculate slice_vertical_position without parsing it again.

Besides, this would also simplify a bit client applications since they usually
know about the GstMpegVideoPacket to use and would just need to pass it along
to the parser, instead of propagating all its fields as multiple arguments.

That's on Tim's call now. :)

BTW, you don't need to check for height > 2800 again if slice_vertical_position
is already initialized to zero, before hand. i.e. shift + add is probably
cheaper than a mispredicted branch in the common case. Not that this is really
important, just something that came to my mind. :)

-- 
Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.
You are the assignee for the bug.


More information about the gstreamer-bugs mailing list