Coda File System

Re: venus dying with a SIGBUS

From: Brett Lymn <blymn_at_baesystems.com.au>
Date: Wed, 27 Jun 2007 09:41:43 +0930
On Tue, Jun 26, 2007 at 03:51:29PM -0400, Jan Harkes wrote:
> 
> I'm interested to know which tweaks were needed.
>

I am not sure if it is a NetBSD specific, the tweaks consisted of
changing some LWP thread functions from (void)(...) to (int)(...) and
cast the LWP argument to a (char *) when creating the LWP thread.
Hopefully that is clear :) I think this affects NetBSD-current, there
were some LWP related changes that went in recently.  I will shoot you
the patches tonight.
 
> 
> I think SIGBUS is typically an indication when we are trying to
> dereference a pointer which is not correctly aligned (multiple of 8
> bytes in the case of a 64-bit architecture).
> 

Ah yes, I forgot about an unaligned access.

> If you can start the client and then attach gdb, gdb will trap the bus
> error before it is passed on to the application. I wonder if the problem
> is with the call to MarinerLog, or with a call made from that function.
> 

Yes, will give that a go tonight for you.

-- 
Brett Lymn
Received on 2007-06-26 20:13:31