Date: Thu, 24 Mar 2005 15:32:36 -0500 From: John Baldwin <jhb@FreeBSD.org> To: freebsd-current@FreeBSD.org Cc: Daniel Eriksson <daniel_k_eriksson@telia.com> Subject: Re: Higher interrupt rate after recent SMP/APIC timer changes? Message-ID: <200503241532.36403.jhb@FreeBSD.org> In-Reply-To: <!~!UENERkVCMDkAAQACAAAAAAAAAAAAAAAAABgAAAAAAAAA0VcX9IoJqUaXPS8MjT1PdsKAAAAQAAAAeMYUhtckfk%2BR4a3yIBKgtAEAAAAA@telia.com> References: <!~!UENERkVCMDkAAQACAAAAAAAAAAAAAAAAABgAAAAAAAAA0VcX9IoJqUaXPS8MjT1PdsKAAAAQAAAAeMYUhtckfk%2BR4a3yIBKgtAEAAAAA@telia.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Friday 11 March 2005 05:51 pm, Daniel Eriksson wrote: > After the recent changes to the use of APIC timers on SMP systems, the > reported interrupt rate has gone up significantly: > > # vmstat -i > interrupt total rate > irq1: atkbd0 3 0 > irq0: clk 55294842 1998 > irq4: sio0 656 0 > irq6: fdc0 13 0 > irq13: npx0 1 0 > irq14: ata0 28299 1 > irq15: ata1 28144 1 > irq17: atapci1+ 60822 2 > irq19: atapci3+ 19848 0 > irq20: ciss0 857173 30 > irq21: em0 1 0 > irq22: em1 1 0 > lapic1: timer 110665583 3998 > lapic0: timer 110649350 3998 > Total 277604736 10031 > > This is an SMP box (dual AMD AthlonMP) running with HZ=2000 and POLLING > enabled. Should I worry about the 4k intr/sec reported for lapic0 and > lapic1, or is this the way things should be? > > I should add that system load has not changed noticeably, leading me to > believe that the lapic* interrupt rate is by design and nothing to worry > about. (But I still wanted to ask.) Yes, the lapic timers run at hz * 2, so that is by design. -- John Baldwin <jhb@FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve" = http://www.FreeBSD.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200503241532.36403.jhb>