Date: Thu, 24 Nov 2005 08:10:40 +0800 From: David Xu <davidxu@freebsd.org> To: Kris Kennaway <kris@obsecurity.org> Cc: current@FreeBSD.org Subject: Re: 4BSD process starvation during I/O Message-ID: <43850500.5090908@freebsd.org> In-Reply-To: <20051123235741.GA10825@xor.obsecurity.org> References: <20051123201837.GA4822@xor.obsecurity.org> <438500BE.3020507@freebsd.org> <20051123235741.GA10825@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Kris Kennaway wrote: >On Thu, Nov 24, 2005 at 07:52:30AM +0800, David Xu wrote: > > >>Kris Kennaway wrote: >> >> >> >>>Perhaps this can be tweaked. >>> >>>Kris >>> >>>P.S. Please, no responses about how maybe someone could write a new >>>scheduler that doesn't have this property. >>> >>> >>> >>> >>> >>> >>Can you try it again with FULL_PREEMPTION is turned on ? >> >> > >OK. Is this option believed to be "safe" (i.e. largely free from >bugs), and would it be useful to test more widely? > >Kris > > > I didn't test it on big machine. if I remembered it correctly, we only do preemption at user boundary if the FULL_PREEMPTION is turned off, you know system thread won't go to user boundary. :-) David Xu
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?43850500.5090908>