[FFmpeg-devel] libossupport status

Michael Niedermayer michaelni
Fri Dec 28 04:25:03 CET 2007


On Thu, Dec 27, 2007 at 06:59:19PM -0500, Rich Felker wrote:
> On Thu, Dec 27, 2007 at 06:33:05PM +0000, M?ns Rullg?rd wrote:
> > Rich Felker <dalias at aerifal.cx> writes:
> > 
> > > Anyway, prototyping usleep is useless. It will work fine without a
> > > prototype.
> > 
> > It will only work until someone uses it with a float or wrong-sized
> > int as argument.  Prototypes exist for a reason.  Do not ignore it.
> 
> Again, the code in question is support for a broken platform. On any
> nonbroken platform, the prototypes will exist. If the argument does
> happen to have the wrong size I suppose strange windows-only bugs
> could result, but very little code in ffmpeg has legitimate use for
> usleep anyway... Actually I wonder why it's used at all.

Theres one on ffmpeg.c rate_emu code, this is valid, its argument is int64_t

Theres one in http.c since
r464 | philipjsg | 2002-05-09 03:19:15 +0200 (Thu, 09 May 2002) | 3 lines

* Add a sleep when doing the post to ffserver. Yes, this is the wrong
  solution.
----
This should be reverted at once!

Theres a large number of invalid looking uses in v4l.c. Theres also a
invalid nanosleep in there. Luckily v4l2.c doesnt contain any.
Ill take the liberty to remove the usleeps() together with the whole AIW
support code. Whoever wants it has to clean it up. The code as it is, is
not fit for ffmpeg svn (doing _uneeded_ colorspace convertion in v4l.c).

Theres also one in bktr.c


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

When you are offended at any man's fault, turn to yourself and study your
own failings. Then you will forget your anger. -- Epictetus
-------------- 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-devel/attachments/20071228/bd6cfee6/attachment.pgp>



More information about the ffmpeg-devel mailing list