Re: data corruption with dmcrypt/LUKS

[Christoph Anton Mitterer <calestyo@scientia.net>]

Hi Milan Broz

On Sun, 2008-02-03 at 23:06 +0100, Milan Broz wrote:
> Are you sure, that your USB-stick is not faulty ?
I actually tested the stick, too. But I consider problems in the stick
(you mean the key-holding stick, do you?) as highly unlikely. 
If the key would be wrong a good crypto system should give me completely
different data and not just these "minor" faults.


> Could you reproduce it with different piece of hw ?
> (Several strange reports for dm-crypt over USB were identified to be 
> USB hw faults.)
I'll test it tomorrow.



> > 2) The second bug happens only rarely and leads to a panic.
> > Unfortunately it's difficult to reproduce, but it always happened when I
> > mkfs.ext3 on the /dev/mapper/sda2.
> > There's a stack-trace printed which clearly involves some dmcrypt
> > lines...
> But no stack trace attached here... please attach it.
Unfortunately I don't have one,... nothing was written to the logs and I
forgot to write it up :-/



> It can be known bug which was fixed in stable version some time ago
> see http://lkml.org/lkml/2007/7/20/211
Uhm but that patch should be part of 2.6.24, shouldn't it?


> No known bugs causing data corruption, no such reports so far
> for stable kernel.
Uhm ok,.. well as told above I'll make some other tests (without the
USB-sticks) but it would be great some people here could try this, too.

Best wishes,
Chris.

btw: What's about the dmcrypt mailing list,.. I've tried to subscribe
but no answers, and I get not posts (not even my owns).