2010-04-27 17:07:07

by Justin P. Mattock

[permalink] [raw]
Subject: [refpolicy] Opensuse11.2 wiredness

A few weeks ago, I had no problem building opensuse11.2,
and doing the fixes that went into play prior too
to get the policy up and running on this system.

Now after things have settled down for me I decided to throw
opensuse11.2 back in and look into the init_upstart bug:
https://bugzilla.novell.com/show_bug.cgi?id=582399

but now maybe updates and so forth(or I did something wrong),
setting the init_upstart=1 in /etc/selinux/refpolicy-standard/booleans
has no affect(prior too this was always used), Now in order
to get the policy to load the way that it should
I needed to create /etc/seliux/refpolicy-standard/booleans.local
and add init_upstart=1

what/why would this be?(isn't boolens all that is needed)

keep in mind this is a monolithic policy(not binary).

Justin P. Mattock


2010-04-27 22:18:18

by Justin P. Mattock

[permalink] [raw]
Subject: [refpolicy] Opensuse11.2 wiredness

On Tue, Apr 27, 2010 at 10:07 AM, Justin P. Mattock
<[email protected]> wrote:
> A few weeks ago, I had no problem building opensuse11.2,
> and doing the fixes that went into play prior too
> to get the policy up and running on this system.
>
> Now after things have settled down for me I decided to throw
> opensuse11.2 back in and look into the init_upstart bug:
> https://bugzilla.novell.com/show_bug.cgi?id=582399
>
> but now maybe updates and so forth(or I did something wrong),
> setting the init_upstart=1 in /etc/selinux/refpolicy-standard/booleans
> has no affect(prior too this was always used), Now in order
> to get the policy to load the way that it should
> I needed to create /etc/seliux/refpolicy-standard/booleans.local
> and add init_upstart=1
>
> what/why would this be?(isn't boolens all that is needed)
>
> keep in mind this is a monolithic policy(not binary).
>
> Justin P. Mattock
>

hmm.. I think I need to look more into
this,but from just looking around I'm not
getting into my user context, looking at the
pam logs I see unable to get context,
then looking at /etc/pam.d/* I see there
using pam_unix2 instead of pam_unix
which I think is not set correctly causing
the boolean file to not be properly read
(but could be wrong).

--
Justin P. Mattock