Messages in this thread | | | Date | Mon, 22 May 2006 13:30:39 +1000 | From | Peter Williams <> | Subject | Re: 2.6.17-rc2+ regression -- audio skipping |
| |
Mike Galbraith wrote: > On Mon, 2006-05-22 at 12:43 +1000, Con Kolivas wrote: >> On Monday 22 May 2006 12:14, Mike Galbraith wrote: >>> In my tree, I don't use the expired array for anything except batch >>> tasks any more for this very reason. The latency just hurts too bad. >> So it's turning your tree into a single priority array design effectively just >> like staircase ;) ? > > Similar I suppose. I still do have dual arrays though, because it's an > almost free way to deal with batch tasks (and is now named accordingly). > > I also still use sleep_avg, but I keep it sane, and with the full > dynamic spread instead of gravitating to either full or empty, and I > monitor for cpu hungry tasks, and let them climb the ladder to where the > action is. That way, it retains the pleasant interactivity of the > current design, yet is absolutely starvation proof.
What about the batch tasks? How do you ensure that they don't get starved? Remember they're "batch" tasks not "background" tasks.
Do you force an array switch after a while and if so how do you handle the adverse effects that will have on any non batch tasks on the other queue?
Peter -- Peter Williams pwil3058@bigpond.net.au
"Learning, n. The kind of ignorance distinguishing the studious." -- Ambrose Bierce - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |