Date: Sun, 07 Mar 2010 16:15:23 +0200 From: Alexander Motin <mav@FreeBSD.org> To: Kevin Day <toasty@dragondata.com> Cc: freebsd-hackers@freebsd.org Subject: Re: ACPI/power implementation causing performance loss with i7/Nehalem turbo boost Message-ID: <4B93B4FB.2090408@FreeBSD.org> In-Reply-To: <1267870984.00226502.1267859402@10.7.7.3> References: <1267863782.00226478.1267851002@10.7.7.3> <1267867387.00226487.1267855802@10.7.7.3> <1267870984.00226502.1267859402@10.7.7.3>
next in thread | previous in thread | raw e-mail | index | archive | help
Kevin Day wrote: > On Mar 6, 2010, at 12:05 AM, Daniel O'Connor wrote: > Yeah, sorry for not mentioning that I had tried this and didn't see any change, so I thought I was on the wrong track. > > # sysctl hw.acpi.cpu.cx_lowest=C3 > hw.acpi.cpu.cx_lowest: C1 -> C3 > > but it doesn't look like it's ever leaving C1: > > hw.acpi.cpu.cx_lowest: C3 > dev.cpu.0.cx_supported: C1/3 C2/205 C3/245 > dev.cpu.0.cx_lowest: C3 > dev.cpu.0.cx_usage: 100.00% 0.00% 0.00% last 500us You have too high interrupt rate for the C-states entering latencies reported by your ACPI. You should significantly reduce number of interrupts per core. For example, by setting: kern.hz=100 hint.apic.0.clock=0 hint.atrtc.0.clock=0 -- Alexander Motin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4B93B4FB.2090408>