2009-11-12 21:04:34

by Daniel Walsh

[permalink] [raw]
Subject: [refpolicy] kernel_storage.patch

http://people.fedoraproject.org/~dwalsh/SELinux/F12/kernel_storage.patch

add /dev/mtd*

Fix a couple of interfaces


2009-11-19 14:03:57

by cpebenito

[permalink] [raw]
Subject: [refpolicy] kernel_storage.patch

On Thu, 2009-11-12 at 16:04 -0500, Daniel J Walsh wrote:
> http://people.fedoraproject.org/~dwalsh/SELinux/F12/kernel_storage.patch
>
> add /dev/mtd*
>
> Fix a couple of interfaces

I don't understand why there would be a relabelto fixed_disk_device_t.
That implies the program is SELinux-aware, so this sounds like a bug
somewhere else. Otherwise merged.


--
Chris PeBenito
Tresys Technology, LLC
(410) 290-1411 x150

2009-11-19 14:19:30

by Daniel Walsh

[permalink] [raw]
Subject: [refpolicy] kernel_storage.patch

On 11/19/2009 09:03 AM, Christopher J. PeBenito wrote:
> On Thu, 2009-11-12 at 16:04 -0500, Daniel J Walsh wrote:
>> http://people.fedoraproject.org/~dwalsh/SELinux/F12/kernel_storage.patch
>>
>> add /dev/mtd*
>>
>> Fix a couple of interfaces
>
> I don't understand why there would be a relabelto fixed_disk_device_t.
> That implies the program is SELinux-aware, so this sounds like a bug
> somewhere else. Otherwise merged.
>
>
I believe the app is doing a mknod and then running restorecon which would cause this?

2009-11-19 14:51:06

by cpebenito

[permalink] [raw]
Subject: [refpolicy] kernel_storage.patch

On Thu, 2009-11-19 at 09:19 -0500, Daniel J Walsh wrote:
> On 11/19/2009 09:03 AM, Christopher J. PeBenito wrote:
> > On Thu, 2009-11-12 at 16:04 -0500, Daniel J Walsh wrote:
> >> http://people.fedoraproject.org/~dwalsh/SELinux/F12/kernel_storage.patch
> >>
> >> add /dev/mtd*
> >>
> >> Fix a couple of interfaces
> >
> > I don't understand why there would be a relabelto fixed_disk_device_t.
> > That implies the program is SELinux-aware, so this sounds like a bug
> > somewhere else. Otherwise merged.
> >
> >
> I believe the app is doing a mknod and then running restorecon which would cause this?

But mknod would result in a device node, not a symlink, which is what is
being dontaudited for relabelto.

--
Chris PeBenito
Tresys Technology, LLC
(410) 290-1411 x150