Subject: Re: Some more 'D' evidence.
To: Paul Kranenburg <pk@cs.few.eur.nl>
From: David Gilbert <dgilbert@jaywon.pci.on.ca>
List: port-sparc
Date: 04/08/1996 21:19:58
>>>>> "Paul" == Paul Kranenburg <pk@cs.few.eur.nl> writes:

Paul> All cases that I know of so far, involve a vnode that's being
Paul> flushed out to disk after going "inactive". The process in state
Paul> `D' is waiting in vflushbuf(). Obviously, there's a race
Paul> somewhere for the buffers that were (once) associated with the
Paul> vnode in question.

	Is there any current workaround?  I'm getting hit by this
about once a week depending on what I'm doing with the machine.  I
don't know if I want to start poking around there as well... I'm
having enough trouble groking what little I've been poking around in
now.

Dave.

-- 
----------------------------------------------------------------------------
|David Gilbert, PCI, Richmond Hill, Ontario.  | Two things can only be     |
|Mail:      dgilbert@jaywon.pci.on.ca         |  equal if and only if they |
|http://www.pci.on.ca/~dgilbert               |   are precisely opposite.  |
---------------------------------------------------------GLO----------------