Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 24 Jun 2005 20:36:36 +0100 (BST)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        Peter Wemm <peter@wemm.org>
Cc:        Perforce Change Reviews <perforce@FreeBSD.org>, Peter Wemm <peter@FreeBSD.org>, John Baldwin <jhb@FreeBSD.org>
Subject:   Re: PERFORCE change 78828 for review
Message-ID:  <20050624203602.M42596@fledge.watson.org>
In-Reply-To: <200506231604.23036.peter@wemm.org>
References:  <200506222312.j5MNChMh079831@repoman.freebsd.org> <200506231444.08027.jhb@FreeBSD.org> <200506231604.23036.peter@wemm.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 23 Jun 2005, Peter Wemm wrote:

> On Thursday 23 June 2005 11:44 am, John Baldwin wrote:
>> On Wednesday 22 June 2005 07:12 pm, Peter Wemm wrote:
>>> http://perforce.freebsd.org/chv.cgi?CH=78828
>>>
>>> Change 78828 by peter@peter_melody on 2005/06/22 23:12:33
>>>
>>> 	Freeze things for the dump snapshot.
>>
>> This may be a bad idea since we may panic with sched_lock held and it
>> would be very nice to get a dump there and not just deadlock.
>
> Yeah.  I'm really not sure what to do though because with interrupts
> still happening, it upsets some of the driver dump routines.  eg: when
> you call this from ddb.

Aren't we supposed to be running with interrupts disabled, and having 
IPI'd the other processors to stop (along with a barrier to make sure they 
did), and operating entirely polled?

Robert N M Watson



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050624203602.M42596>