Re: Problems with geom_bde
In message [EMAIL PROTECTED], Thorsten Greiner writes : Hi, inspired by the recent posting of the gbde tutorial I went out and set up an encrypted partition using geom_bde on my laptop. When trying to umount this partition the umount fails with: # umount /crypto umount: unmount of /crypto failed: Resource temporarily unavailable Did you call gbde detach before the umount ? You need to unmount first. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Problems with geom_bde
* Poul-Henning Kamp [EMAIL PROTECTED]: Did you call gbde detach before the umount ? No. You need to unmount first. Yes, that's what I tried. I then get the fsync: giving up on dirty message. When I shut down the box, it tries to sync the dirty buffers and finally says giving up on 2 buffers, leaving all my filesystems dirty, and corrupted data on the gbde device (lots of zero bytes in files). Regards -Thorsten __ Bestes Testergebnis: Stiftung Warentest Doppelsieg fur WEB.DE FreeMail und WEB.DE Club. Nur fuer unsere Nutzer! http://f.web.de/?mc=021182 ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Problems with geom_bde
* Poul-Henning Kamp [EMAIL PROTECTED] [2003-09-30 10:54]: Did you call gbde detach before the umount ? You need to unmount first. After investigating further I think I have found the reason for the problem. When I created the disklabel for the partition I was lazy and entered a size of 200 blocks for the slice. I just tried to reproduce the problem and re-created the slice with a size of 1024M (which is 2097152 blocks) and the problem seems to have disappeared - at least I cannot reproduce it anymore. So now I wonder if there is any magic constant which the slice size needs to be a multiple of...? Regards -Thorsten -- There are 10 kinds of people in the world, those who understand binary and those who don't. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Problems with geom_bde
In message [EMAIL PROTECTED], Thorsten Greiner writes: * Poul-Henning Kamp [EMAIL PROTECTED] [2003-09-30 10:54]: Did you call gbde detach before the umount ? You need to unmount first. After investigating further I think I have found the reason for the problem. When I created the disklabel for the partition I was lazy and entered a size of 200 blocks for the slice. I just tried to reproduce the problem and re-created the slice with a size of 1024M (which is 2097152 blocks) and the problem seems to have disappeared - at least I cannot reproduce it anymore. So now I wonder if there is any magic constant which the slice size needs to be a multiple of...? Not that I know off. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Problems with geom_bde
In message: [EMAIL PROTECTED] Poul-Henning Kamp [EMAIL PROTECTED] writes: : So now I wonder if there is any magic constant which the slice size : needs to be a multiple of...? : : Not that I know off. Actually, there is, but it isn't constant. slices need to be an even number of 'cylendars' in the fdisk label. Some BIOSes use this to determine the geometry of the disk. Some BIOSes have issues booting when these numbers are correct. By making the the partition size different, the BIOS was able to properly grok things and booting worked. Warner ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to [EMAIL PROTECTED]
Problems with geom_bde
Hi, inspired by the recent posting of the gbde tutorial I went out and set up an encrypted partition using geom_bde on my laptop. When trying to umount this partition the umount fails with: # umount /crypto umount: unmount of /crypto failed: Resource temporarily unavailable and the following message gets logged: fsync: giving up on dirty: 0xc456ec8c: tag devfs, type VCHR, usecount 5140, writecount 0, refcount 93, flags (VV_OBJBUF), lock type devfs: EXCL (count 1) by thread 0xc4240130 dev ad0s3a.bde This is with a fairly recent kernel, cvsupped a few hours ago: # uname -a FreeBSD tybalt 5.1-CURRENT FreeBSD 5.1-CURRENT #26: Mon Sep 29 22:18:26 CEST 2003 [EMAIL PROTECTED]:/usr/src/sys/i386/compile/TYBALT i386 The kernel config is a standard GENERIC with some device drivers removed. geom_bde was loaded as a module. Please let me know if I can provide further information. Regards -Thorsten -- You know you've landed gear-up when it takes full power to taxi. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to [EMAIL PROTECTED]
