Messages in this thread | | | Date | Sun, 6 Jan 2002 05:01:12 +0100 (CET) | From | Ingo Molnar <> | Subject | Re: [announce] [patch] ultra-scalable O(1) SMP and UP scheduler |
| |
On Sat, 5 Jan 2002, Davide Libenzi wrote:
> Ingo, you don't need that many queues, 32 are more than sufficent. If > you look at the distribution you'll see that it matters ( for > interactive feel ) only the very first ( top ) queues, while lower > ones can very easily tollerate a FIFO pickup w/out bad feelings.
I have no problem with using 32 queues as long as we keep the code flexible enough to increase the queue length if needed. I think we should make it flexible and not restrict ourselves to something like word size. (with this i'm not suggesting that you meant this, i'm just trying to make sure.) I saw really good (behavioral, latency, not performance) effects of the longer queue under high load, but this must be weighed against the cache footprint of the queues.
Ingo
- 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/
| |