[FFmpeg-cvslog] r21174 - in trunk/libavcodec: h264.c h264.h h264_direct.c h264_mvpred.h svq3.c

Michael Niedermayer michaelni
Fri Jan 15 04:12:04 CET 2010


On Fri, Jan 15, 2010 at 12:31:31AM +0000, M?ns Rullg?rd wrote:
> michael <subversion at mplayerhq.hu> writes:
> 
> > Author: michael
> > Date: Tue Jan 12 22:36:26 2010
> > New Revision: 21174
> >
> > Log:
> > Split motion vector prediction off h264.c/h.
> >
> > Added:
> >    trunk/libavcodec/h264_mvpred.h
> >       - copied, changed from r21173, trunk/libavcodec/h264.c
> > Modified:
> >    trunk/libavcodec/h264.c
> >    trunk/libavcodec/h264.h
> >    trunk/libavcodec/h264_direct.c
> >    trunk/libavcodec/svq3.c
> 
> h264.h and h264_mvpred.h are interdependent in a way that breaks the
> tms470 compiler (and probably other TI compilers) which chokes on
> forward-declared static inline functions.

Is that a violation of the C spec or compiler bug?
Anyway, the 2 obvious things to try would be
1. make decode_mb_skip inline static as well, maybe this fixes it
2. move decode_mb_skip() to h264.c and rename to ff_h264_... 
either way these should not affect speed meassureably but a benchmark is
probably a good idea.
If its a violation of the C spec ill try to fix it tomorrow when i have time
and noone did before me. If its a compiler issue someone with the compiler
is probably in a better position to commit a workaround.
And yes you have my ok to commit either 1 or 2

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Its not that you shouldnt use gotos but rather that you should write
readable code and code with gotos often but not always is less readable
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-cvslog/attachments/20100115/0e9e66bb/attachment.pgp>



More information about the ffmpeg-cvslog mailing list