Re: [Ccrtp-devel] More about ccRTP on ARM

[Federico Montesino Pouzols]

        Yes, it helps everyone, since this is a bug in ccRTP :). I
have committed your fix to CVS, as well as other fixes for a bug in
the RTCP interval randomization code that I discovered while reviewing
this method.

        However, I had not seen this code failing before. bwfract may
be zero (so it must be checked before the division) but it is unusual
that it is zero. I suggest you to check that the RTCP interval is
rightly computed on ARM (it should be a randomized value around 5
seconds which oscillates between 2.5 and 7.5 seconds, if you do not
change the minimum interval). 

On Mon, Oct 20, 2003 at 04:11:06PM +0200, Jeremy wrote:
> Examining the variables revealed that bwfract is zero.  I changed the
> offending code to:
> 
>     microtimeout_t interval=0;
>     if (bwfract != 0) {
>             interval = static_cast<microtimeout_t>
>                     ((participants * rtcpAvgSize / bwfract) * 1000000);
>     }
>     if ( interval < rtcpMinInterval )
>             interval = rtcpMinInterval;
> 
> and everything seems to working for now.  Now I just have to get broadcast
> done.
> 
> Hope this helps someone :)
> 
> --J
> 
> 
> 
> _______________________________________________
> Ccrtp-devel mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/ccrtp-devel