Re: ...unofficial 2.4.3 int patch...

[Jon Masters <jonathan@jonmasters.org>]

Hi,
	I have applied the kerneli 2.4.3.1 patch to a clean 2.4.3 source
without trouble and have built a patched util-linux however attempting
to run losetup as:

losetup-crypto -e rc6 /dev/loop0 /dev/hd<drive number>

results only in either a seg fault or non fatal oops but the loop device
used cannot be used again until a reboot.

Here's part of a trace:

(Line numbers added obviously)

---begin---
0	open("/dev/hdg6", O_RDWR|0x8000)        = 3
1	open("/dev/loop1", O_RDWR|0x8000)       = 4
2	mlockall(0x3, 0x4)                      = 0
3	ioctl(4, LOOP_SET_FD, 0x3)              = -1 ENOSYS (Function
	not implemented)
4	ioctl(4, LOOP_SET_FD, 0x3)              = 0
5	ioctl(4, LOOP_SET_STATUS <unfinished ...>
6	+++ killed by SIGSEGV +++
---end---

Line 3 would seem to indicate that I am doing something wrong.

Has anyone else had problems? If not can people please suggest things
for me to try in order to help sort this out...

This patched losetup also seems to work on 2.2.18 with serpent. I wish
to use rc6 however no option appears in that box's kernel config for
this - although losetup doesn't complain about a lack of rc6 until after
it has requested a password.

I'm afraid that I was using a much older kernel for a lot of crypto
until recently and then was unable to use crypto at all for a while on
my main desktop box so a lot of things have changed since I last looked
at this - I apologise for not being up to speed on what's happening.

--jcm

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