From owner-freebsd-bugs Sun Mar 17 11:36:23 2002 Delivered-To: freebsd-bugs@freebsd.org Received: from toybox.ca (matchbox.toybox.ca [216.191.236.11]) by hub.freebsd.org (Postfix) with ESMTP id C42DE37B43B; Sun, 17 Mar 2002 11:36:12 -0800 (PST) Received: from cpe001057001952.cpe.net.cable.rogers.com ([24.103.226.50] helo=litebrite) by toybox.ca with asmtp (Exim 3.35 #2) id 16mgSE-000OVh-00; Sun, 17 Mar 2002 14:36:30 -0500 Message-ID: <033201c1cdeb$10697580$1e02070a@litebrite> From: "Roy Hooper" To: "David Malone" , Cc: References: <200203170935.g2H9ZwM95481@freefall.freebsd.org> <20020317133702.GA33770@walton.maths.tcd.ie> Subject: Re: misc/35992: /etc/rc has no way to force fsck -y at next boot, corrupt filesystems with soft updates required manual intervention Date: Sun, 17 Mar 2002 14:36:38 -0500 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2600.0000 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000 Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Sorry to say that it is not. If I encounter it again, i'll dd it up and gzip it (and open a new bug). I didn't think -p isn't supposed to actually examine the filesystem beyond the clean flag. As such the question is really how did the filesystem get corrupt yet keep the clean flag at the next boot? Perhaps a double hard-cold-reset could cause that -- that is exactly what happened at one point, possibly during an fsck. Perhaps when routines like dscheck detect errors, the dirty flag can be forced to be set and cause unmount to refuse to clear it? Roy -- Roy Hooper Project Manager & Senior UNIX Consultant Decisive Technologies Inc. ----- Original Message ----- From: "David Malone" To: Cc: ; Sent: Sunday, March 17, 2002 8:37 AM Subject: Re: misc/35992: /etc/rc has no way to force fsck -y at next boot, corrupt filesystems with soft updates required manual intervention > > Added documentation for the rc.early(8) script to -CURRENT. > > The main bug here is that fsck -p didn't detect the problem. Is a > copy of the currupted filesystem still available? > > David. > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message