Merge lp://qastaging/~sdeziel/apparmor-profiles/thunderbird-enigmail-1.9 into lp://qastaging/apparmor-profiles

Heh, I was going to complain about the /usr/bin/locale Uxr, rule but there's at least those three other Uxr rules right next to it.

I'm surprised about the silenced denials -- those seem wide-ranging and potentially problematic. I might have even thought that thunderbird should have ~/.thunderbird/** rwlk, access.

The static names in /tmp/ are interesting. Those may need more research to see if those need a CVE. (It's possible to use static names in /tmp safely, but the [0-9]* regex there gives me a bad feeling.)

Thanks

On 2016-04-18 04:36 PM, Seth Arnold wrote:
> I'm surprised about the silenced denials -- those seem wide-ranging
> and potentially problematic. I might have even thought that
> thunderbird should have ~/.thunderbird/** rwlk, access.

The web view doesn't make it very easy to spot but those rules apply
only to the _subprofile_ gpg2.

> The static names in /tmp/ are interesting. Those may need more
> research to see if those need a CVE. (It's possible to use static
> names in /tmp safely, but the [0-9]* regex there gives me a bad
> feeling.)

When the base file already exists, a number is appended, that's only how
far I checked this.

On Mon, Apr 18, 2016 at 09:57:24PM -0000, Simon Déziel wrote:
> On 2016-04-18 04:36 PM, Seth Arnold wrote:
> > I'm surprised about the silenced denials -- those seem wide-ranging
> > and potentially problematic. I might have even thought that
> > thunderbird should have ~/.thunderbird/** rwlk, access.
>
> The web view doesn't make it very easy to spot but those rules apply
> only to the _subprofile_ gpg2.

Thanks for highlighting that.

> > The static names in /tmp/ are interesting. Those may need more
> > research to see if those need a CVE. (It's possible to use static
> > names in /tmp safely, but the [0-9]* regex there gives me a bad
> > feeling.)
>
> When the base file already exists, a number is appended, that's only how
> far I checked this.

It's a bit dubious, but looking at the gpg2 subprofile, there's other
similar dubious /tmp/ usage already.

I've merged this branch after applying the changes to the 16.10 tree as
well.

--
Steve Beattie
<email address hidden>
http://NxNW.org/~steve/

On 2016-04-30 12:45 PM, Steve Beattie wrote:
> I've merged this branch after applying the changes to the 16.10 tree as
> well.

Thanks!