Subject: Re: ffs_alloccg
To: Ken Hornstein <email@example.com>
From: None <Postmaster@spinne.web.net>
Date: 10/19/1994 11:17:53
> [David Maxwell wrote:]
> >We've also been experiencing crashes which can be induced at will with pine.
> >(This is an old copy compiled before upgrading to 1.0)
> Just curious ... has anyone tried running ktrace on the process to figure out
> exactly _which_ system call is causing the crash?
At your suggestion I tried this. (No one using the machine... may as well let
it crash) I hadn't yet due to previous problems with getting a stable kernel
Needless to say pine works fine now. Ktraced or not, even run multiple times.
It no longer exhibits after-run on the drive either. The HD used to jump around
for a while after quitting pine.
Greg tells me the only thing he has finished was the make world that was on
hold until the kernel was stable enough. Checking the .so libs shows that
they haven't changed at all either.
In the meantime though I have rebuilt httpd (NCSA, was CERN) lynx, elm,
No more ffs_alloccgs at the moment, but now we get an occasional
panic: nfsreq nogrps
of which we had 2-3 yesterday but we're up 23:21 right now...