On Thu, Nov 22, 2018 at 5:44 PM Richard Guy Briggs <rgb(a)redhat.com> wrote:
On 2018-11-21 18:24, Paul Moore wrote:
> There are many places, notably audit_log_task_info() and
> audit_log_exit(), that take task_struct pointers but in reality they
> are always working on the current task. This patch eliminates the
> task_struct arguments and uses current directly which allows a number
> of cleanups as well.
>
> Signed-off-by: Paul Moore <paul(a)paul-moore.com>
Ack. This looks like a welcome simplification. This will cause a merge
conflict with my ghak59 patch 4&5 (but Jan's patch didn't) which is
fine. I'd like to rebase ghak59 on this once this is merged to get the
EOE record in the right place after kill_trees...
Thanks for the review, I just merged it into audit/next so you should
be able to rebase your patches now.
One minor comment in __audit_free() below once we've established
tsk ==
current...
...
Once we've established that tsk == current, I think it would read
better
if we used current in the two filter calls and in fact simplify it even
more and just eliminate the task_struct parameter from both audit_filter
calls since it will only ever be current.
Yeah, I went back and forth on this too, but eventually decided to
leave this as-is for the moment - another one of those judgement
calls. In the case of the filter functions, both were pretty small
and I could *maybe* see them being useful for tasks other than current
... maybe not. We can always convert them in the future too.
--
paul moore
www.paul-moore.com