Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 20 Apr 2008 22:12:31 -0700
From:      Julian Elischer <julian@elischer.org>
To:        Daniel O'Connor <doconnor@gsoft.com.au>
Cc:        gnn@freebsd.org, freebsd-current@freebsd.org, Andrew Gallatin <gallatin@cs.duke.edu>
Subject:   Re: TSC Timecounter and multi-core/SMP
Message-ID:  <480C223F.2040303@elischer.org>
In-Reply-To: <200804181235.29530.doconnor@gsoft.com.au>
References:  <m2d4oy8n30.wl%gnn@neville-neil.com> <20080418003203.GB11705@dragon.NUXI.org> <48080276.3040203@elischer.org> <200804181235.29530.doconnor@gsoft.com.au>

next in thread | previous in thread | raw e-mail | index | archive | help
Daniel O'Connor wrote:
> On Fri, 18 Apr 2008, Julian Elischer wrote:
>> David O'Brien wrote:
>>> The TSC on K8 is not invariant - its rate of change is affected by
>>> P-state changes.
>>>
>>> The TSC on Greyhound (Family 10h) is invariant.
>>> [but as stated above, is not synced with other cores]
>> You'd think that an invariant sync'd clock (fast to read) of some
>> type would have been done by someone by now.. The software people
>> have been asking for this for the last decade at least.
> 
> Probably because it's very very difficult to get right :)	

and they've had over 20 years to think of something.

> 




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?480C223F.2040303>