Re: kern/44897: bge initialization/run failures

[David Holland <dholland-bugs%netbsd.org@localhost>]

The following reply was made to PR kern/44897; it has been noted by GNATS.

From: David Holland <dholland-bugs%netbsd.org@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/44897: bge initialization/run failures
Date: Sun, 17 Jul 2011 03:03:01 +0000

 Not sent to gnats.
 
    ------
 
 From: Jukka Ruohonen <jruohonen%iki.fi@localhost>
 To: netbsd-bugs%NetBSD.org@localhost
 Cc: Erik Berls <cyber%NetBSD.org@localhost>
 Subject: Re: kern/44897: bge initialization/run failures
 Date: Mon, 4 Jul 2011 23:12:18 +0300
 Mail-Followup-To: netbsd-bugs%NetBSD.org@localhost, Erik Berls 
<cyber%NetBSD.org@localhost>
 
 On Mon, Jul 04, 2011 at 08:10:08PM +0000, Erik Berls wrote:
 >  With -current (6/28, from nyftp) I get a hang right after acpicpu1
 >  (it's a dual core system) regardless of azalia state.
 
 Can you verify this with more recent -curent (from July)? And if the problem
 is still present, please file a separate problem report. Thanks.
 
 
 From: Erik Berls <cyber%netbsd.org@localhost>
 To: jruohonen%iki.fi@localhost, netbsd-bugs%netbsd.org@localhost, Erik Berls 
<cyber%netbsd.org@localhost>
 Subject: Re: kern/44897: bge initialization/run failures
 Date: Mon, 4 Jul 2011 16:35:50 -0700
 
 Submitted as port-amd64/45118
 
 Summary:
 GENERIC, Source update as of July 4, 2011 14:30 PST:
 - booted normally has the same issue (hang).
 - booted with -12c & 'disable azalia' gets to multi user, but then
 suspicious issues with arp
 - booted with -1 gets to multi user
 - booted -with -2 hangs after audio0 at hdafg4
 
 
 Running non-smp under some nfs load managed to get -current to panic, as well.
 *sigh*
 kern/45119
 
 
 From: Erik Berls <cyber%netbsd.org@localhost>
 To: jruohonen%iki.fi@localhost, netbsd-bugs%netbsd.org@localhost, Erik Berls 
<cyber%netbsd.org@localhost>
 Subject: Re: kern/44897: bge initialization/run failures
 Date: Mon, 4 Jul 2011 17:08:33 -0700
 
 I'm now able to replicate this on 5.0.2 by pushing NFS traffic (remote
 builds) to the system.
 
 I get the set of 3 messages accompanied by a network hang.  after a
 minute or two it recovers. (usually about 4 sets worth)
 
 
 On a whim from the other tests, i tried running 5.0.2 with -1:
 longer times to any recovery.
 
 I was able to make -current panic in the whole mess (kern/45119)