Messages in this thread | | | Subject | Re: workqueue thing | From | Peter Zijlstra <> | Date | Tue, 22 Dec 2009 12:00:13 +0100 |
| |
On Tue, 2009-12-22 at 08:50 +0900, Tejun Heo wrote: > > > 4) didn't consider the RT needs > > Can you be more specific? What RT needs? It's pretty difficult to > tell when there's no in-kernel user and any shared worker pool would > have pretty similar properties as cmwq.
http://programming.kicks-ass.net/kernel-patches/rt-workqueue-prio/
Most of that passed over lkml too at various times and in various forms, not eve sure that that is the last one, but its the one I found first.
Basically it boils down to letting the worklets run at the priority of the enqueing task and boost current running works with the prio of the highest pending work, including proper barrier support.
| |