[Ffmpeg-devel] H.264 encoder

Christian Iversen chrivers
Wed Oct 4 20:21:28 CEST 2006


On Wednesday 04 October 2006 19:58, Rich Felker wrote:
> On Wed, Oct 04, 2006 at 07:11:13PM +0200, Christian Iversen wrote:
> > On Wednesday 04 October 2006 19:06, Panagiotis Issaris wrote:
> > > Hi,
> > >
> > > On wo, 2006-10-04 at 19:02 +0200, Panagiotis Issaris wrote:
> > > >[...]
> > > >
> > > > > Is this expression equiv. to val--; ?
> > > >
> > > > Yes.
> > >
> > > Well. yes, in case val>0.
> >
> > Haha, yeah ;)
> >
> > Ok, I thought C used logical SHR, so it all makes sense now. Thanks!
>
> It depends on whether the type is signed or unsigned. Actually it's
> unspecified for shifts to signed values which are negative, but any
> real-world implementation will have correct signed shifts. This is one
> of the (fewer, these days, at least) remaining idiocies in the C
> standard, akin to the stupidity of allowing a floating-point
> implementation where every float expression evaluates to zero..

That, you have to explain :)

Do you mean for machines with no FPU, or?

-- 
Regards,
Christian Iversen




More information about the ffmpeg-devel mailing list