On Tue, Apr 25, 2006 at 08:05:25PM +0400, Dmitry Morozovsky wrote:
> On Tue, 25 Apr 2006, Kostik Belousov wrote:
> 
> KB> > I just made lab machine with serial console, compile minimal kernel 
> with quotas 
> KB> > and KDB+WITNESS, and immediately after ``quotacheck /var && quotaon 
> /var'' got
> KB> > 
> KB> > kdb_backtrace(d663aba0,c051f402,c05f7da3,c05fe731,c32cb414) at 
> KB> > kdb_backtrace+0x29
> KB> > vfs_badlock(c05f7da3,c05fe731,c32cb414) at vfs_badlock+0x11
> KB> > assert_vop_locked(c32cb414,c05fe731,c30ed400,c04eac2c,c05f6cf4) at 
> 
> KB> Obviously, revisions 1.78, 1.79 of the sys/ufs/ufs/ufs_quota.c
> KB> shall be MFCed. Try this patch (note, I does not tested it):
> 
> Yup, at least immediate drop to debugger disappeared. I'll check now with our 
> situation: quotas with at least one user over quota *and* makinfg snapshot on 
> this filesystem.

Your problem shall persist, this patch has nothing to do with it.
I am looking for debugging information on deadlock.

Attachment: pgps6AW6sYbCg.pgp
Description: PGP signature

Reply via email to