[Applied Oneiric] Re: [PATCH 2/2] [oneiric CVE 2/2] Change check_ruid flag to a more reasonable type

John Johansen john.johansen at canonical.com
Thu Aug 11 18:07:06 UTC 2011

On 08/11/2011 10:42 AM, Leann Ogasawara wrote:
> On Thu, 2011-08-11 at 11:28 -0600, Tim Gardner wrote:
>> I don't think this one is strictly necessary since it doesn't change
>> code behavior, and needlessly diverges from upstream. A uid_t is an
>> 'unsigned int' which is as good as a boolean in this case.
> John, before I go yanking this from Oneiric master-next, what's the
> status of this patch landing upstream?  I'd assumed since it was
> associated with the CVE, that is was making it's way up and likely to
> hit stable (ie, we'd be able to drop it in favor of the upstream patch
> later).  As Tim has pointed out, minimal divergence from upstream is
>   the ideal scenario.

I sent it upstream but haven't heard back from tyler on it yet.  I think
where this is going will depend on upstream.

There was a pull request for the original patch sent to linus and CC'd
to stable but last I checked it wasn't sucked in yet.

   If the original patch gets pulled in then I see this patch probably
just going to current and not stable.  However if the first patch gets
NAK'd I see the two patches combining.

I agree with tim that it isn't necessary, it is really only syntactic
sugar.  I would wait on this one, if upstream takes it we can pull it
in so we match, otherwise just ignore it.

I included mostly to document that the uid_t mistake was noticed
and the follow up patch sent.

More information about the kernel-team mailing list