Re: loop-aes crashes with any cipher under my standard configuration

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Marcus Roelz wrote:
> I installed loop-aes exactly like described in the readme - but the problem
> is: it crashes when i do mke2fs /dev/loop0
> The first few inodes are written (200 Gig raid5) then mke2fs terminates - the
> same when i try dd if=/dev/zero of=/dev/loop0
> 
> mke2fs /dev/md1 works fine (without loop-aes)
> 
> The crashes occur with serpent128, serpent256, aes (i didn't test the others
> but expect them to fail too)
> 
> My system:
> 
> freshly installed Redhat 7.3 with kernel 2.4.18custom
> loop-AES-v1.6f
> ciphers-v1.0f
> util-linux-2.11u
> gcc 2.96
> 
> Someone on the crypto-list had the same Problem in July 2002 but the thread
> died.

Richard Zidlicky's loop-AES problem was probably a compile problem. He
compiled 3 different kernels with some of the source tree shared and then
compiled 3 copies of loop.o drivers for those 3 kernels. I suggested to him
that he compile 1 kernel and then loop.o for that kernel, and repeat that 3
times.

Last message I receiced from him was on 31 Jul 2002. He was just too busy
with other stuff.

> Any help would be appreciated (spent the last few days messing around with it)

Could you provide exact commands that you used to compile your kernel and
loop.o + loop_serpent.o external modules. And I mean all commands, starting
with "make distclean" or something.

If you compiled your kernel using rpm, please provide your modified .spec
file.

> loopback device without encryption works perfectly....

Was that with loop-AES' loop.o or or RedHat kernel's loop?

> loop-AES seems to work for me with smaller partitions with encryption
> but it crashes on my 200 Gig /dev/md1 Partition - so maybe its a matter of
> size?

Were the smaller partitions raid5 devices or just plain partitions. Is this
problem with loop-AES on top of raid5 device, or problem with loop-AES on
top of big partition?

> what does MAX_DISK_SIZE 1024*1024*1024 in patched-loop.c tell me?
> Bytes ? KBytes? Nothing to do with that at all?

It is in units of 1024 bytes. Meaning that if underlying device has not
exported is size properly, that 1 TB is used as upper limit.

Regards,
Jari Ruusu <jari.ruusu@pp.inet.fi>

-
Linux-crypto:  cryptography in and on the Linux system
Archive:       http://mail.nl.linux.org/linux-crypto/


[Index of Archives]     [Kernel]     [Linux Crypto]     [Gnu Crypto]     [Gnu Classpath]     [Netfilter]     [Bugtraq]
  Powered by Linux