Subject: Re: 1.0A Repeated "panic: ffs_alloccg" Any ideas?
To: Michael L. VanLoon -- HeadCandy.com <michaelv@HeadCandy.com>
From: Mark Delany <markd@bushwire.apana.org.au>
List: current-users
Date: 10/17/1994 06:41:13
> >> >Yes, pine 3.90 is often hanging around at or about the time of the
> 
> >> It is early days yet, but the tentative result is that pine is the
> >> culprit possibly due to ftruncate() as you suggest. Perhaps that fix
> >> you alluded to was compromised with the integration of the 4.4
> >> sources?

> >Yup, I have 3.90 installed, and according to the accounting, sendmail is alway
> s the 
> >last process run before the crash. I'll disable the use of pine for awhile and
>  monitor
> >the results.
> >
> >bry
> 
> I'll just throw in that I use MH for all my mail handling.  I receive
> the current-users list, plus send and receive private email on my home
> machine through a PPP link.  With my setup, I haven't experienced any
> of these crashes, and sendmail is quite busy here sometimes.  Just
> another point of data. :-)

I've now switched to elm and the panics have stopped. I think Bryan's
point is that pine is in the process of sending a mail, having just
forked sendmail, when he gets his panics.

It's almost certainly not sendmail as my system relays mail for about
40 uucp sites as well as acting as a secondary MX for a large
pub.access site here and it never crashes with that traffic (so long
as I don't touch pine that is).

If Bryan's panics likewise stop subsequent to disabling pine, I think
the correlation is starting to look significant.


M..