[rsbac] Newbie: 2.4.18+1.2.1 doesn't start, complains of invalid device

Tomasz Korycki rsbac@rsbac.org
Mon Oct 7 20:01:03 2002

On Mon, 2002-10-07 at 03:34, Amon Ott wrote:
> On Saturday, 5. October 2002 06:16, Tomasz Korycki wrote:
> > A newbie here, but I'll try to make it brief: RSBAC kernel never starts, 
> > complains about invalid device until syslog fills /var. Here's how I got 
> there:
> That looks fine - except the version: I never made a 1.2.1-final patch for 
> 2.4.18...

My bad: it was 2.4.18 + 1.2.0 (from the /kernels/)

>>>>>>>>>>>>>>>>> snip! <<<<<<<<<<<<<<<<<<<
> Somehow the RSBAC initialization either failed to add the root device to its 
> internal device list, 

When the kernel gets built, the root device it recognizes is the correct

> or some strange umount stuff happened.

Can You elucidate a bit? What should I watch for? When the system comes
down it does not complain when umounting stuff. Besides, that's the
value of journalling...

> It might be related to reiserfs, which is weird in many respects.

Well, at least in softmode, it *does* get mounted (since the rc scripts
get read), also when I just "off" RSBAC in config. So there does not
seem to be a problem with reiserfs per se, or with a kernel.

>>>>>>>>>>>>>>>>>>> snip! <<<<<<<<<<<<<<<<<
> When a module cannot get its attributes, access is denied and notification 
> fails. This is the expected behaviour.

Yes, I thought so, that's why I didn't (yet) worry about it, I just
mentioned it for completeness. Besides, those messages start *much*
later, so by themselves they can't be the cause.

>>>>>>>>>>>>>>>>>>> snip! <<<<<<<<<<<<<<<<<
> Without softmode enabled, please do the following:

Just to confirm: *without* softmode and *not* maintenance, right? Just
the regular, fully-armed kernel?

> 1. check, if your root partition gets added in rsbac_do_init() (see boot 
> messages)

I found no "rsbac_do_init" anywhere in copies of boot.log, syslog or
messages that I saved outside of /var/log

> 2. add the kernel parameter rsbac_debug_adf and check, whether the root fs 
> gets umounted at some place.

That's with that new kernel, right? And You mean root fs being
unmounted? Frankly, I don't think I need to bother, as the kernel can't
even find a device, let alone a filesystem on it, so it is not likely to
unmount it...

> 3. If not, please also add rsbac_debug_stack, take one boot log until the 
> first INVALIDDEV messages appear and send it to me. I need the init messages 
> up to the part you included.

That's what I think I'll do right now. What facility and priority does
RSBAC use? I'd like to make it go "somewhere else", to get a clearer

> To get you up and running, another root fs, e.g. ext3, might help.

Well, I do not think reiserfs is a problem here: the kernel works fine
with it, as long as RSBAC is not enabled. So I'll persist a while longer
with it.

Thanks, and I'll let You know as soon as I know!