Date: Mon, 6 Feb 2006 17:37:18 -0500 (EST) From: Andrew Gallatin <gallatin@cs.duke.edu> To: John Baldwin <jhb@freebsd.org> Cc: freebsd-current@freebsd.org, Andre Oppermann <andre@freebsd.org> Subject: Re: machdep.cpu_idle_hlt and SMP perf? Message-ID: <17383.53150.324978.91528@grasshopper.cs.duke.edu> In-Reply-To: <200602061532.02223.jhb@freebsd.org> References: <17379.56708.421007.613310@grasshopper.cs.duke.edu> <43E74872.7000002@freebsd.org> <17383.42908.349070.31155@grasshopper.cs.duke.edu> <200602061532.02223.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin writes: > On Monday 06 February 2006 14:46, Andrew Gallatin wrote: > > Andre Oppermann writes: > > > Andrew Gallatin wrote: > > > > Why dooes machdep.cpu_idle_hlt=1 drop my 10GbE network rx > > > > performance by a considerable amount (7.5Gbs -> 5.5Gbs)? > > > > <...> > > > > > This may be the same problem OpenBSD has fixed last year in the handling > > > of the idle loop. From the kerneltrap posting: > > > > <....> > > > > > First commit message: > > > http://marc.theaimsgroup.com/?l=openbsd-cvs&m=111692513727274&w=2 > > > > > > The MFC with all changes in one commit message: > > > http://marc.theaimsgroup.com/?l=openbsd-cvs&m=111859519015510&w=2 > > > > The bug they fixes was missing interrupts by both calling APM's idle > > routine, which may hlt, and hlt'ing in the idle loop itself. Since I > > have no idea what acpi is doing, I got excited about this. > > > > Alas, it seems like this isn't it. I pointed cpu_idle_hook back to > > cpu_idle_default and away from acpi_cpu_idle, but that made no > > difference. > > You may be seeing problems because it might simply take a while for the CPU to > wake up from HLT when an interrupt comes in. The 4BSD scheduler tries to do > IPIs to wakeup any sleeping CPUs when it schedules a new thread, but that > would add higher latency for ithreads than just preempting directly to the > ithread. Oh, you have to turn that on, it's off by default > (kern.sched.ipiwakeup.enabled=1). Hmm.. It seems to be on by default. Unfortunately, it does not seem to help. Would you expect ULE to do better? I've noticed that if I screw up the time state of the machine by switching between ACPI-fast and TSC timecounters, performance for TCP ping-pongs goes all over the map... Drew
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?17383.53150.324978.91528>