[PATCH] os/access: fix regression in server interpreted auth

[Dave Airlie]

On 18 February 2015 at 07:24, Keith Packard <keithp at keithp.com> wrote:
> Dave Airlie <airlied at gmail.com> writes:
>
>> From: Dave Airlie <airlied at redhat.com>
>>
>> This was reported on irc on Fedora when rawhide went to 1.17.1.
>>
>> regression occured in: 2566835b4374edb3e5a8353d4f7c9e7ec4851c57
>>  os: Eliminate uninitialized value warnings from access.c
>>
>> siAddrMatch doesn't need addr to be a useful value, it checks
>> some things like localuser without having an address at all.
>>
>> Signed-off-by: Dave Airlie <airlied at redhat.com>
>
> Ok, auditing this took a bit of time, but I think things are reasonably
> clear to me. The various siAddrMatch functions will all work fine with a
> NULL addr as long as len is zero. And, the code paths to get here don't
> obviously make sure this is always true, so I'd be happier if we just
> made sure len and addr agreed about empty addresses:

That seems like it should work fine,

its easy to test yourself, create a testuser, xhost +SI:localuser:testuser
and su to them and see can they use the display.

but r-b by me for whatever works.
Dave.