Messages in this thread | | | Date | Wed, 14 Aug 2013 12:57:23 -0400 | From | Tejun Heo <> | Subject | Re: [PATCH v7 2/2] mm: make lru_add_drain_all() selective |
| |
Hello, Chris.
On Wed, Aug 14, 2013 at 12:03:39PM -0400, Chris Metcalf wrote: > Tejun, I don't know if you have a better idea for how to mark a > work_struct as being "not used" so we can set and test it here. > Is setting entry.next to NULL good? Should we offer it as an API > in the workqueue header?
Maybe simply defining a static cpumask would be cleaner?
> We could wrap the whole thing in a new workqueue API too, of course > (schedule_on_each_cpu_cond_sequential??) but it seems better at this > point to wait until we find another caller with similar needs, and only > then factor the code into a new workqueue API.
We can have e.g. __schedule_on_cpu(fn, pcpu_works) but yeah it seems a bit excessive at this point.
Thanks.
-- tejun
| |