On Tue Apr 03 2007 at 14:18:19 +0000, Antti Kantee wrote:
> 
> Module Name:	src
> Committed By:	pooka
> Date:		Tue Apr  3 14:18:19 UTC 2007
> 
> Modified Files:
> 	src/sys/fs/smbfs: smbfs_vfsops.c
> 
> Log Message:
> In sync, skip vnodes which have a null v_data pointer.  This is a
> workaround for the problem analyzed more deeply in kern/30831.  In
> short, the problem is keeping the vnode on the mount point vnode
> list during reclaim.  If reclaim happens to sleep (as is a possibility
> with smbfs due to calling vrele() and therefore possibly VOP_INACTIVE),
> code going through the entire mountpoint vnode list will hit
> half-reclaimed vnodes.

And the mandatory "remember something when pressing enter":

Thanks to Patrick Welche for supplying debugging data and testing the
workaround.

-- 
Antti Kantee <pooka@iki.fi>                     Of course he runs NetBSD
http://www.iki.fi/pooka/                          http://www.NetBSD.org/
    "la qualité la plus indispensable du cuisinier est l'exactitude"