It's a typo. Fortunately it doesn't make everything bad. Because I
doesn't use max_port_events at all. I will fix it later.
Thanks,
lin
> Hmmm,
>
> is this a finished patch? A cursory look turns up this:
>
> 1642 + if (getrctl ("process.max-port-events", NULL, rblk, RCTL_FIRST)
> == -1) {
> 1643 + perror ("getrctl");
> 1644 + exit (1);
> 1645 + } else {
> 1646 + if (max_port_evnets > rctlblk_get_value(rblk))
> 1647 + max_port_evnets = rctlblk_get_value(rblk);
> 1648 + GAM_DEBUG(DEBUG_INFO, "FENKERNEL : max event of a port:
> %u\n",
> max_port_evnets);
> 1649 + }
>
> Reading the code, 'max_port_evnets' is obviously a typo-ed version of
> 'max_port_events'. However, rather then correcting the typo, a
> max_port_evnets variable was added, directly above (!) the spot where
> the correctly spelled variable is declared:
>
> 1309 +
> 1310 +static ulong max_port_evnets = 512;
> 1311 +static ulong max_port_events = 256;
> 1312 +static GList *pn_vq; /* the queue of ports which don't have the
> max objs */
>
>
> Cheers,
>
> Menno
>
>
--
x82120 / +86 10 82618200
