On Tue, Jun 18, 2013 at 03:45:15PM +0530, Viresh Kumar wrote:
On 18 June 2013 15:33, Peter Zijlstra peterz@infradead.org wrote:
On Tue, Jun 11, 2013 at 04:32:44PM +0530, Viresh Kumar wrote:
In build_sched_groups() we don't need to call get_group() for cpus which are already covered in previous iterations. So, call get_group() after checking if cpu is covered or not.
Aside from not needing it; doing it would mark the group used and eventually leak it since we wouldn't connect it and not find it again to free it, right?
So effectively this fixes a memory leak?
Exactly. To be precise: In cases where sg->cpumask contained more than one cpu (For any topology level), this patch helps freeing sg's memory for all cpus leaving the group leader.
May I update the logs again?
Sure. Just send a new patch and I'll substitute.