Messages in this thread | | | From | "Stephen C. Tweedie" <> | Date | Wed, 15 Sep 1999 00:58:45 +0100 (BST) | Subject | Re: > 15,000 Simultaneous Connections |
| |
Hi,
On Wed, 08 Sep 1999 11:17:39 -0700, Jason Nordwick <nordwick@erdos.askjeeves.com> said:
>> if someone is trying to implement the API described in banga's paper, >> you should send some e-mail to the authors (jeff mogul and gaurav >> banga) and ask them if they have more research on this besides what >> is described in that paper. at usenix, i asked about some details >> during gaurav's presentation, and jeff mogul hinted that there were >> some things they had been working on (new APIs to support sharing the >> event queue over multiple threads?) but that weren't included in the >> paper.
> If people are going to implement this, I think that it would be good > if there was an API writeup first. I would like to implement > something similar in FreeBSD and I think that it would be nice to see > both implementations have the same API and effects.
We already have a queued event model. I have already submitted a fix for 2.3 to add si_band reason codes to the events. The only thing missing is POSIX shared signal queues: any takers?
--Stephen
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |