[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: crypto loopback fs stuff
- To: "Brian S. Julin" <bri@tull.umassp.edu>
- Subject: Re: crypto loopback fs stuff
- From: Dale Amon <amon@vnl.com>
- Date: Wed, 21 Feb 2001 17:01:25 +0000
- Cc: linux-crypto@nl.linux.org
- In-Reply-To: <Pine.LNX.4.21.0102211137290.18236-100000@tull.umassp.edu>; from bri@tull.umassp.edu on Wed, Feb 21, 2001 at 11:52:13AM -0500
- References: <Pine.LNX.4.21.0102211137290.18236-100000@tull.umassp.edu>
- Sender: owner-linux-crypto@nl.linux.org
- User-Agent: Mutt/1.3.12i
I'm forwarding this on as it seems highly relevant to the
problems a number of us are seeing:
On Wed, Feb 21, 2001 at 11:52:13AM -0500, Brian S. Julin wrote:
>
> Hi, I'm not on the crypto list but I have some
> info that probably should be posted there, so if
> you'd do me the favor, thanks.
>
> I was having similar problems to the one you described
> on the linux-crypto list, except I was using small
> partitions/files and when the mke2fs hung, it would only
> sit in the D state, it would not crash the kernel and
> did not generate a segfault. This was with both
> SCSI and IDE drives, so it's definitely not driver
> problems.
>
> So I got the lastest greatest bleeding edge kernel
> and applied Jens Axboe's loop-4 patch to it, and I
> had the same result. I mailed Jens, and he said
> to wait and give the upcoming loop-5 a try, which I did.
> loop-5 (and loop-5B) got rid of the mke2fs problem, but
> the machine now hangs mount in the D state when using
> crypt (nonencrypted loopback seems to work fine.)
>
> Jens says he thinks he's fixed what he can of the
> loop driver, and that he may look at it further, but
> this particular problem is probably in the crypto
> stuff. It will probably require someone deep into
> the loop_gen crypto driver to fix it; until then,
> assume things are just totally broken.
>
> If anyone wants their problem reports to be "meaningful",
> I would recommend patching Linus's 2.4.X-preY releases
> with Jen's latest loop patch (go to where you get your
> kernel and cd ../people/axboe/patches).
>
> As things now stand on my system:
>
> AES -- refuses to mount
> Blowfish -- segfaults losetup, div0
> serpent, rc5, twofish -- /bin/mount hands in D state
> rc6 -- segfaults losetup, EIP=0
>
> Note also if you are using funny block sizes, Alan Cox
> mentioned something about fixing ext2fs block size
> detection in 2.4.1-ac20
>
> --
> Brian
>
--
------------------------------------------------------
Use Linux: A computer Dale Amon, CEO/MD
is a terrible thing Village Networking Ltd
to waste. Belfast, Northern Ireland
------------------------------------------------------
Linux-crypto: cryptography in and on the Linux system
Archive: http://mail.nl.linux.org/linux-crypto/