Date: Wed, 26 Jul 2000 15:06:28 -0400 (EDT) From: Brian Fundakowski Feldman <green@FreeBSD.org> To: Sheldon Hearn <sheldonh@uunet.co.za> Cc: Kirk McKusick <mckusick@mckusick.com>, current@FreeBSD.org Subject: Re: Panic: lockmgr: pid 5, not exclusive lock holder 0 unlocking Message-ID: <Pine.BSF.4.21.0007261458260.18690-100000@green.dyndns.org> In-Reply-To: <408.964610749@axl.ops.uunet.co.za>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 26 Jul 2000, Sheldon Hearn wrote: > > Hi Kirk, > > On Tue, 25 Jul 2000 11:28:47 MST, Kirk McKusick wrote: > > > Modified files: > > sys/kern vfs_bio.c > > Log: > > Now that buffer locks can be recursive, we need to delete the panics > > that complain about them. > > This looks related. I get it pretty consistently on shutdown, provided > the machine did some work subsequent to going multi-user. [Script cut] Yeah, several people are noticing that one. Robert Watson and I were trying to debug it, but it was really late at night so personally I gave up on it until I could reproduce it here. I got a different one. I'd like to track it down but it hasn't repeated itself (yet?). My call stack was (disregarding ATA's call stack mostly): panic() bufdone() checking that B_DONE's not set (which it is) cluster_callback() calling bp->b_iodone() biodone() calling bufdone() ad_interrupt() calling biodone() I implemented a little circular buffer which holds (calling function, bp) pairs for bufdone(), but so far I have had no crash yet (yay? :-/). When I do, the circular buffer (thanks, Greg!) will help me out to actually debug this instead of just wondering why B_DONE was set. -- Brian Fundakowski Feldman \ FreeBSD: The Power to Serve! / green@FreeBSD.org `------------------------------' To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0007261458260.18690-100000>