Date: Fri, 28 May 1999 03:34:40 -0700 From: Don Lewis <Don.Lewis@tsc.tdk.com> To: Bruce Evans <bde@zeta.org.au>, gibbs@plutotech.com Cc: mauri@aripaev.ee, scsi@FreeBSD.ORG Subject: Re: FW: -STABLE scsi problem (bug?) Message-ID: <199905281034.DAA13244@salsa.gv.tsc.tdk.com> In-Reply-To: Bruce Evans <bde@zeta.org.au> "Re: FW: -STABLE scsi problem (bug?)" (May 28, 2:17am)
next in thread | previous in thread | raw e-mail | index | archive | help
On May 28, 2:17am, Bruce Evans wrote: } Subject: Re: FW: -STABLE scsi problem (bug?) } >>vfs doesn't really know which filesystems have been synced properly, } >>and doesn't call unmount() for any filesystem unless all buffers were } >>synced. The clean bit is set by unmount() so it is left unset for } >>all rw-mounted filesystems if the sync failed. } } >Can't we just walk the buffer list, note the FSes the buffers reference, } >and unmount any FSes that aren't referenced? } } We could, but syncing can only fail if there is an unexpected problem, so } we should be conservative if we don't understand the problem. I think this is overly conservative. The unwritten buffers could be destined for a dead NFS server, and we could be rebooting to clear the stuck mount point. It could be a definite inconvenience if this forces a bunch of huge local filesystems to be fsck'ed. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199905281034.DAA13244>