[Bug 675772] byte_reader + adapter: _masked_scan_uin32() could be optimized

[GStreamer (bugzilla.gnome.org)]

https://bugzilla.gnome.org/show_bug.cgi?id=675772
  GStreamer | gstreamer (core) | 1.x

Nicolas Dufresne <nicolas.dufresne> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED

--- Comment #6 from Nicolas Dufresne <nicolas.dufresne at collabora.co.uk> 2013-08-10 23:11:36 UTC ---
I have experimented with ageneric solution:

see http://cgit.collabora.com/git/user/nicolas/gstreamer.git/log/?h=boyer-moore

But this endup being slower. Instead it was decided to optimize certain know
case. The H264/MPEG/VC1 case has been fixed:

commit de75bca9b387c404a29cb057046d0db1d601794c
Author: Nicolas Dufresne <nicolas.dufresne at collabora.com>
Date:   Thu Aug 8 12:08:31 2013 +0200

    bytereader: Accelerate MPEG/H264 start code scanning

    Accelerate MPEG/H264 start code scanning using Boyer-Moor bad character
    heuristic.

    https://bugzilla.gnome.org/show_bug.cgi?id=702357

I think we can also optimizing the cases with mask 0xFFFFFFFF and 0xFFFFFF00
(found in Matroska), where generating table of offset would be possible since
we don't need to mask each bytes. Thus I'll keep this open until I'm done
experimenting.

p.s. the Adapter version has not been optimized since it's not used for
0xffffff00/0x00000100 start codes.

-- 
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.