Re: kern/37933: very frequent crashes after suspend/resume

["Jared D. McNeill" <jmcneill%invisible.ca@localhost>]

The following reply was made to PR kern/37933; it has been noted by GNATS.

From: "Jared D. McNeill" <jmcneill%invisible.ca@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost, 
 netbsd-bugs%netbsd.org@localhost, smb%cs.columbia.edu@localhost
Subject: Re: kern/37933: very frequent crashes after suspend/resume
Date: Sat, 09 Feb 2008 21:16:56 -0500

 Steven M. Bellovin wrote:
 > The following reply was made to PR kern/37933; it has been noted by GNATS.
 > 
 > From: "Steven M. Bellovin" <smb%cs.columbia.edu@localhost>
 > To: gnats-bugs%NetBSD.org@localhost
 > Cc: gnats-admin%NetBSD.org@localhost
 > Subject: Re: kern/37933: very frequent crashes after suspend/resume
 > Date: Sun, 10 Feb 2008 00:15:15 +0000
 > 
 >  I managed to get a core dump.  I booted single-user, set ddb.onpanic=0,
 >  zeroed most of main memory via a program that malloced a lot of
 >  memory with _malloc_options=Z to make the core dump smaller, and did two
 >  sysctl machdep.sleep_state=3.  On the second resume, it panicked.  The
 >  resulting dump and kernel file (with symbols) are at 
 >  
 >  http://www.machshav.com/~smb/netbsd.9.core.gz
 >  http://www.machshav.com/~smb/netbsd.9.gz
 >  
 >  The traceback in the dump (and for that matter the panic string printed
 >  at crash time) are different than in the original PR, but I suspect
 >  that that's more a helpful clue than a hindrance -- the crash is 100%
 >  reproducible, which suggests that the different backtraces are symptoms
 >  of the same underlying issue.
 >  
 >  This dump, btw, is from 4.99.53, from sources retrieved today.
 
 Are you using cpuctl to offline all APs before you suspend?