On Mon, 16 May 2011, Michal Nazarewicz wrote:
> > Now that accessing current->comm needs to be protected,
> > @@ -2868,6 +2868,10 @@ sub process {
> > WARN("usage of NR_CPUS is often wrong - consider using
> > cpu_possible(), num_possible_cpus(), for_each_possible_cpu(), etc\n" .
> > $herecurr);
> > }
> > +# check for current->comm usage
> > + if ($line =~ /\b(?:current|task|tsk|t)\s*->\s*comm\b/) {
>
> Not a checkpatch.pl expert but as far as I'm concerned, that looks reasonable.
>
> I was sort of worried that t->comm could produce quite a few false positives
> but all its appearances in the kernel (seem to) refer to task.
>
It's guaranteed to generate false positives since perf events uses a field
of the same name to store a thread's comm, so I think the most important
thing is for the checkpatch output to specify that this _may_ be a
dereference of a thread's comm that needs get_task_comm() or %ptc.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>