< Previous by Date Date Index Next by Date >
< Previous in Thread Thread Index  

RE: [reSIProcate] resip::AbstractFifo::getNext(int ms)


Yes the issue is with either negative or 0 time.  For windows (at least) 0
time = infinite wait.

" If that's the case, then this  
routine has been very broken for a while."

Well - this function was completely broken, up until a few months ago.  The
condition wasn't being rechecked.  In fact, this function wasn't even used
anywhere in the stack until repro came along.  Currently repro is the only
user of getNext with a timeout.

Scott

-----Original Message-----
From: Alan Hawrylyshen [mailto:alan@xxxxxxxxxx] 
Sent: Tuesday, August 09, 2005 7:38 PM
To: Scott Godin
Cc: stensil; resiprocate-devel@xxxxxxxxxxxxxxxxxxx
Subject: Re: [reSIProcate] resip::AbstractFifo::getNext(int ms)


On Aug 9, 2005, at 13.07, Scott Godin wrote:

> Hi Stensil,
>
> Yes, I think you are right.  I've committed a fix for this to the SVN
> head.  Thanks!
>

I had a quick conversation with Jason about this and we were  
unconvinced that this was a problem. Can someone quickly just write  
out in plain english how the deadlock manifests itself? Is is because  
we would wait for negative time?  If that's the case, then this  
routine has been very broken for a while.

Thanks
Alan