Re: race between tcp_close and delack?

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Fri, Oct 10, 2008 at 11:51:25AM +0200, Edgar Fuß wrote:
> > Problem: we check if a timer is being invoked with tcp_timers_invoking(),
> > but this forgets the t_delack_ch timer.
> Sorry, I don't understand. I may be stupid, but tcp_timers_invoking() reads
> 
>       for (i = 0; i < TCPT_NTIMERS; i++)
>               if (callout_invoking(&tp->t_timer[i]))
>                       count++;
>       if (callout_invoking(&tp->t_delack_ch))
>               count++;
> 
> so I'm under the impression it does care about t_delack_ch.

gasp, you're right. So this isn't my bug :( Sigh.


-- 
Manuel Bouyer, LIP6, Universite Paris VI.           
Manuel.Bouyer%lip6.fr@localhost
     NetBSD: 26 ans d'experience feront toujours la difference
--