On Friday, November 14, 2014 10:32:51 PM Richard Guy Briggs wrote:
On 14/11/13, Steve Grubb wrote:
> On Thursday, November 13, 2014 08:08:52 PM Richard Guy Briggs wrote:
> > > So what terrible things happen to userspace if
> > > AUDIT_VERSION_BACKLOG_WAIT_TIME becomes 0x03 instead of 0x02?
> >
> > But it won't. It gets the value of
> > AUDIT_FEATURE_BITMAP_BACKLOG_WAIT_TIME, which is 0x00000002.
> >
> > I think you meant to ask about AUDIT_VERSION_LATEST, which would become
> > 3.
> >
> > You *did* already ask that question in a previous thread, and there
> > didn't seem to be a concern. Steve Grubb could likely answer this
> > question better than me.
>
> The audit 2.4.1 package has been pushed to everything from F20 -> rawhide.
> If you don't see any problems, then its safe. But check carefully around
> the things that you did change. Right now, we only are caring about only
> one kernel feature, --loginuid-immutable. Check that it still works,
> auditctl -s.
Here's my output, which I assume looks sane:
[root@f20 ~]# rpm -q audit
audit-2.4.1-1.fc20.x86_64
[root@f20 ~]# auditctl -s
enabled 1
flag 1
pid 307
rate_limit 0
backlog_limit 320
lost 0
backlog 0
backlog_wait_time 60000
loginuid_immutable 0 unlocked
Looks like good output to me, Steve?
--
paul moore
security and virtualization @ redhat