Subject: Re: crashing sharks (was Re: More ELF stuff)
To: None <Richard.Earnshaw@arm.com>
From: Neil Hoggarth <njh@kernighan.demon.co.uk>
List: port-arm32
Date: 02/15/2001 20:38:20
On Thu, 15 Feb 2001, Richard Earnshaw wrote:

> Symptoms are that the machine locks solid (I think it's disk IO related,
> but it could be soft-dep or any of a number of other things).  I can't get
> any backtrace out of the machine since the console is dead.  Certainly the
> crash seems to happen most when there is heavy disk io involved.  I've
> seen reports that this might pre-date 1.5 (see PR port-arm32/11921 and
> port-arm32/9859).

I get something very like this on a CATS running 1.5BETA if, and only if,
I use a kernel which *lacks* the following from its configuration:

file-system NFS
file-system MSDOSFS
options NFSSEVER
options NFS_BOOT_BOOTP
options NFS_BOOT_DHCP

Those lines basically taken from a diff of a "it works" configuration
against a "it doesn't work" configuration. I don't actually use any NFS on
this machine, and my prefered configuration had all the options commented
out, but this was what triggered lockups on heavy disk activity for me.
I've always intended to do more testing and find out which of these (or
which combination) actually makes the difference and file a PR.
Unfortunately I've had zero free time for NetBSD hacking recently.

Regards,

Neil.