[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20090722213942.GA4337@josh-work.beaverton.ibm.com>
Date: Wed, 22 Jul 2009 14:39:46 -0700
From: Josh Triplett <josh@...htriplett.org>
To: Peter Zijlstra <peterz@...radead.org>
Cc: Arjan van de Ven <arjan@...radead.org>,
john stultz <johnstul@...ibm.com>,
Thomas Gleixner <tglx@...utronix.de>,
lkml <linux-kernel@...r.kernel.org>,
Thomas Gleixner <tglx@...x.de>, Ingo Molnar <mingo@...e.hu>,
Andi Kleen <ak@...e.de>, nikolag@...ibm.com,
Darren Hart <dvhltc@...ibm.com>
Subject: Re: [RFC][PATCH] Introduce CLOCK_REALTIME_COARSE
On Mon, Jul 20, 2009 at 03:49:14PM +0200, Peter Zijlstra wrote:
> On Mon, 2009-07-20 at 06:33 -0700, Arjan van de Ven wrote:
> > > Maybe power64, sparc64 and s390x qualify, but certainly nothing on x86
> > > does.
> >
> > the x86 on my desk disagrees.
>
> >From what I know even nehalem doesn't have fully synced tscs when your
> machine is large enough, and the timers are still a tad expensive.
All single-socket Nehalem boards have synced TSCs, and some multi-socket
Nehalem boards have synced TSCs, notably Intel's. I don't know if you
can safely rely on the latter, though.
> Maybe your desktop is next-gen? 't would be nice to finally have an x86
> that has usable clock and timer hardware.
x86 CPUs with the ARAT feature ("always running APIC timer") will let
you use the local APIC even in deep C-states. I don't know about the
overhead of the local APIC.
- Josh Triplett
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists