Re: Our SMP implementation scalability

2007-01-18 Thread Thomas E. Spanjaard
Dmitri Nikulin wrote: I really really hope NetBSD doesn't end up with similarly complex SMP, because that could really kill it, with a minor fraction of FreeBSD's developer resources and an onus to continue supporting many architectures. Well, it appears that at least for userland they're movin

Re: Our SMP implementation scalability

2007-01-17 Thread Dmitri Nikulin
On 1/18/07, Kris Kennaway <[EMAIL PROTECTED]> wrote: I disagree with you about FreeBSD's SMP architecture being difficult to develop and optimize (there are several really good tools for that, like witness, lock profiling, pmc, and dtrace, etc), but at least there are half a dozen or more people

Re: Our SMP implementation scalability

2007-01-17 Thread hui
On Thu, Jan 18, 2007 at 09:26:32AM +1100, Dmitri Nikulin wrote: > To memory-quote Matt, it should scale really well in theory, because > most things are naturally lockless and so aren't co-dependent. FreeBSD > scales badly (though improving) because there are complex locks > everywhere, and even in

Re: Our SMP implementation scalability

2007-01-17 Thread Dmitri Nikulin
On 1/18/07, Erik Wikström <[EMAIL PROTECTED]> wrote: On 2007-01-17 16:31, Petr Janda wrote: Even more interesting would be a guess of how good it might scale when you finally get rid of BGL. To memory-quote Matt, it should scale really well in theory, because most things are naturally lockless

Re: Our SMP implementation scalability

2007-01-17 Thread Erik Wikström
On 2007-01-17 16:31, Petr Janda wrote: Simon 'corecode' Schubert wrote: Petr Janda wrote: I was reading an older news article comparing FreeBSD's SMPng and Linux's SMP saying that Linux scales nicely on 10+ cpu's but FreeBSD has problem with 6 or more. So my question is, would DragonFly's SMP

Re: Our SMP implementation scalability

2007-01-17 Thread Petr Janda
Could you take a guess (just for being informed) how much of the system is under the BGL and how much is MPSAFE? Petr Simon 'corecode' Schubert wrote: Petr Janda wrote: I was reading an older news article comparing FreeBSD's SMPng and Linux's SMP saying that Linux scales nicely on 10+ cpu's b

Re: Our SMP implementation scalability

2007-01-17 Thread Simon 'corecode' Schubert
Petr Janda wrote: I was reading an older news article comparing FreeBSD's SMPng and Linux's SMP saying that Linux scales nicely on 10+ cpu's but FreeBSD has problem with 6 or more. So my question is, would DragonFly's SMP scale as good as Linux? not even close. we're still under the giant lo

Our SMP implementation scalability

2007-01-17 Thread Petr Janda
I was reading an older news article comparing FreeBSD's SMPng and Linux's SMP saying that Linux scales nicely on 10+ cpu's but FreeBSD has problem with 6 or more. So my question is, would DragonFly's SMP scale as good as Linux?