[Sumover-dev] 64bit rat and common patches

Douglas Kosovic douglask at itee.uq.edu.au
Fri Jul 20 09:10:40 BST 2007


Hi Piers,

Actually the following patch for rat/ui_send_rtp.c that I sent as part of
rat-64bit.patch wasn't necessary :

----
void ui_send_rtp_cname(session_t *sp, char *addr, uint32_t ssrc)
{
- ui_info_update_sdes(sp, addr, "rtp.source.cname", 
rtp_get_sdes(sp->rtp_session[0], ssrc, RTCP_SDES_CNAME), ssrc);
+ ui_info_update_sdes(sp, addr, "rtp.source.cname", 
rtp_get_sdes(sp->rtp_session[0], (unsigned long)ssrc, RTCP_SDES_CNAME), 
(unsigned long)ssrc);
}
---

if you could undo that one, that'll be good.

Thanks,
Doug

----- Original Message ----- 
From: "Douglas Kosovic" <douglask at itee.uq.edu.au>
To: "'sumover-dev'" <sumover-dev at cs.ucl.ac.uk>
Sent: Friday, July 20, 2007 12:31 PM
Subject: [Sumover-dev] 64bit rat and common patches


> Hi Piers,
>
> Please find attached a couple of 64bit patch files for rat and libcommon.
>
> With common, sizeof was being performed on a pointer rather than a struct.
> The struct in question happens to be 4 bytes in size which is coincidently
> the size of pointers on 32bit systems.
>
> With 64bit systems, the RAT UI would show ghost ssrc entries as a 64bit
> value, the upper 32bits were random values, but the lower 32bits
> corresponded to existing entries.
>
>
> Thanks,
> Doug 





More information about the Sumover-dev mailing list