Re: encrypted ext2 on cd?

Andrew E. Mileski (aem@netcom.ca)
Tue, 10 Jun 1997 21:54:12 -0400 (EDT)


I'm currently working on a mjaor overhaul of the loop driver.

IMHO, your problem _may_ have something to do with block size.
I've recently had another report about using a loop device with
a CD-ROM. This seems to be a _recent_ problem AFAIK, and may have
something to do with changes elsewhere in the kernel - the loop
driver is simple, but breaks some assumptions about driver
behaviour.

Dou you have similar problems using a loop _without_ encryption
on a plain CD-ROM drive (not image)?

Anyways, I don't have a solution for you right now, but I will look
into it.

P.S. If you feel like trying out my development patch, just holler.
It doesn't include encryption though.

--
Andrew E. Mileski   mailto:aem@netcom.ca

> Greetings, > > Recently I tried to write an encrypted ext2fs to a CD-R. I basically > dd'ed a 650M file and created the encrypted fs in that file, then wrote > that file to the CD-R using cdwrite. I can mount the image via loopback > and all is well, but when I tried to do the same with the CD, the mount > fails with: > > VFS: Can't find an ext2 filesystem on dev 07:00. > EXT2-fs error (device 07:00): ext2_check_descriptors: Block bitmap for group 0 not in group (block 365000324)! > EXT2-fs: group descriptors corrupted ! > > When I use losetup and attach the CD to a loop device, then run e2fsck > on it, it produces lots of block bitmap not in group errors. When I > dd the CD image to a file, the file shows no errors and mounts fine as > an encrypted ext2fs. BTW, a standard ext2fs works just fine written to > a CD-R. > > I'm not familiar enough with ext2fs/encryption patch/CD drivers to know > where to look. My system is: > > - Linux 2.0.29+"old" encryption patch (running on p133) > - mount-2.5k > - e2fsprogs-1.09 > - Toshiba 3401B CD-ROM > - Phillips 2000 CD-R > - Adaptec 1542-CF > > I would appreciate any help in getting this working. > > Thanks, > > James. > > -- > jtsiao@madoka.jpl.nasa.gov || SCSI is *NOT* magic. There are *fundamental > http://www.roadkill.com/thoth || technical reasons* why it is necessary to > || sacrifice a young goat to your SCSI chain > #include <disclaimer.h> || now and then. -- jfw@proteon.com > >