On Fri, 2009-05-08 at 12:40 -0500, Serge E. Hallyn wrote:
> Quoting Stephen Smalley ([email protected]):
> > On Thu, 2009-05-07 at 21:31 -0500, Serge E. Hallyn wrote:
> > > Hi Stephen,
> > > 
> > > I'm trying to test all of your patches.  But I'm having the (lately
> > > usual) static libs problem.  tests/inherit/Makefile specifies that
> > > selinux_inherit_* should be compiled -static and -lselinux, but
> > > libselinux.a is not installed.  What did you do about this - did
> > > you compile libselinux.a by hand, or did you find a rpm that
> > > installs it?  For now I just removed -static from LDFLAGS.
> > > I don't recall why they were
> > > -static originally.
> > 
> > yum install libselinux-static
> 
> I swear I'd done a yum search libselinux | grep static...
> 
> Thanks, that did it.  Any objection to adding that to the README?

It can go into the SELinux Policy and Userland section added by my
earlier patch to the README; likely should mention that you need the
libselinux-devel package as well.  However, it is somewhat distro- and
release- specific; I think originally everything was provided by the
libselinux package, then they did an initial split into libselinux vs
libselinux-devel, and then they further moved the static lib from
libselinux-devel to its own libselinux-static.  RHEL 5 didn't have a
separate libselinux-static.  And Debian has slightly different package
names (libselinux1, libselinux1-dev) and I think they still keep the
static lib in the -dev package.

> > IIRC, you can't cleanly test the fd:use inheritance check from
> > parent->child if the child is dynamically linked (inheritance of the
> > descriptor to the dynamic linker opened while still in parent context).
> > 
> > > My run, with all of your patches applied, on just-updated f11, gave me
> > > the following failures:
> > > 
> > > SELinux10 - selinux_file
> > >   test14
> > >   probably explained by the fact that selinux_wait_io is labeled
> > >   system_u:object_r:unlabeled_t:s0 ?
> > 
> > Well, it would be left in unlabeled_t after the test policy gets
> > removed.
> 
> Oh I was thinking I'd have to unload manually, but that's bc
> I'd glanced at a line out of your README patch about
> running individual testcases.
> 
> > > SELinux36 - selinux_wait
> > >   test02
> > 
> > The two failures you listed are actual kernel regression, the first
> > fixed by my recent patch (selinux:  Fix send_sigiotask hook) on selinux
> > list and the second fixed by a recent patch by Oleg Nesterov on lkml
> > (do_wait:  do take security_task_wait() into account).  So the above
> > just reflects that the ltp selinux testsuite is operating correctly and
> > detecting improper kernel behavior in your kernel.
> 
> Excellent.
> 
> thanks,
> -serge
-- 
Stephen Smalley
National Security Agency


------------------------------------------------------------------------------
The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
production scanning environment may not be a perfect world - but thanks to
Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
Series Scanner you'll get full speed at 300 dpi even with all image 
processing features enabled. http://p.sf.net/sfu/kodak-com
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to