On Mon, 30 Oct 2000, Jeff V. Merkey wrote: > [...] All protection has to go away in all LAN paths for this to > happen, and user space apps set to ring 0. [...] i found that this is not a requirement for good network scalability. We do not do a syscall for every packet, so the cost evens out. Sure, it does not hurt to not eat ~1 microsecond per system-call, but it causes no overhead or scalability limit otherwise. In the TUX webserver we have user-space modules doing context-switch-less webserving, and it scales quite well, and is generic. Ingo - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
- 2.2.18Pre Lan Performance Rocks! Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance Rocks! Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance Rocks! Andi Kleen
- Re: 2.2.18Pre Lan Performance Rocks! Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance Rocks! Andi Kleen
- Re: 2.2.18Pre Lan Performance Rocks! Ingo Molnar
- Re: 2.2.18Pre Lan Performance Roc... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar
- Re: 2.2.18Pre Lan Performance... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar
- Re: 2.2.18Pre Lan Performance... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar
- Re: 2.2.18Pre Lan Performance... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar
- Re: 2.2.18Pre Lan Performance... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar
- Re: 2.2.18Pre Lan Performance... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar
- Re: 2.2.18Pre Lan Performance... Jeff V. Merkey
- Re: 2.2.18Pre Lan Performance... Ingo Molnar