← Back to team overview

rohc team mailing list archive

Re: TS_Stride when generating dynamic part RTP header

 

Josephine,

> Thanks Didier. I've posted the bugs.

Thank you.

Didier Barvaux
Viveris Technologies
http://www.tech.viveris.com/opensource/



> On Mon, Feb 8, 2010 at 1:11 PM, Didier Barvaux <
> didier.barvaux@xxxxxxxxxxxxxxxxxxxx> wrote:
> 
> > Hi Josephine,
> >
> > > Looking at rtp_code_dynamic_rtp_part() in c_rtp.c, it appears that
> > > there's a small error in line 568:
> > >
> > >             rohc_debugf(3, "send ts_stride = %04x on 4 bytes\n",
> > > ts_stride);
> > >
> > > Shouldn't it be
> > >
> > >           "  ... send ts_stride = %08x .. " ?
> >
> > Yes, that's a mistake. There should even be a '0x' prefix in front
> > of the '%08x' format string so that we know that it is a
> > hexadecimal value.
> >
> > Moreover reading the source code around the debug trace you mention,
> > make me discover another problem. Thanks for that Josephine.
> >
> > The TS_STRIDE field must be encoded with SDVL according to §4.5.6 of
> > RFC3095, but the ROHC library does not. With SDVL encoding, the
> > length of the TS_STRIDE field may vary from 1 byte to 4 bytes.
> > That's why one reads "1-4 octets, if TSS = 1" along the TS_Stride
> > field in §5.7.7.6 of RFC3095. The ROHC decompressor is also
> > affected by the problem.
> >
> > Josephine, could you please open a bug in the bugtracker of the ROHC
> > library for both problems ? Go to
> > https://bugs.launchpad.net/rohc/+filebug and answer the different
> > questions.
> >
> > Regards,
> >
> > Didier Barvaux
> > Viveris Technologies
> > http://www.tech.viveris.com/opensource/
> >
> >
> > _______________________________________________
> > Mailing list: https://launchpad.net/~rohc
> > <https://launchpad.net/%7Erohc> Post to     :
> > rohc@xxxxxxxxxxxxxxxxxxx Unsubscribe : https://launchpad.net/~rohc
> > <https://launchpad.net/%7Erohc> More help   :
> > https://help.launchpad.net/ListHelp
> >



References