Re: netbsd-5 NFS(?) lock up

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Tue, Mar 31, 2009 at 07:43:29AM +0000, Andrew Doran wrote:
> > so we'll end up calling biodone() with a buffer that was not the one
> > given to sdstrategy(), in the same thread context. Most drivers don't do
> > this (maybe some advanced hardware raid drivers can do this too).
> > 
> > Maybe some upper level I/O subsystems are not prepared to deal with this.
> 
> I think it's unlikely. Note that biodone2() can block. Can it cause a
> problem for scsipi or the driver?

It shouldn't be a problem for the driver. As for scsipi, it prevent new
I/O from making progress while the scsipi channel thread is blocked in
biodone2. But this wouln't prevent already-running processes from continue
running. I'll have to double-check but I think the server was not even
pingable.

I'm trying to reproduce it on a test box, but so far I've only seen it
on the live server :( I may be able to make more experiments tonight;
anything special I should check ?

-- 
Manuel Bouyer, LIP6, Universite Paris VI.           
Manuel.Bouyer%lip6.fr@localhost
     NetBSD: 26 ans d'experience feront toujours la difference
--