Subject: Re: High-resolution clcoks (time vultures on pmaxes)?
To: Jonathan Stone <firstname.lastname@example.org>
From: Chris G Demetriou <Chris_G_Demetriou@LAGAVULIN.PDL.CS.CMU.EDU>
Date: 01/28/1995 21:17:05
> The Dave Mills paper in SIGCOMM 94 describes the 5k/240 as having a
> high-resolution 25 MHz clock as part of the TurboChannel
> interface on one of the ASICs.
Is this paper anywhere around, online?
Looking at the TC docs i have around (most of which apply to later
DEC hardware, anyway), i can't seem to find a reference to this. Do
you know what ASIC it's on?
> It seems a reasonable guess that
> this is the same as the ``free-running counter'' on the XINE;
> except it should run at 25 MHz, not 12.5.
Where's the "free-running counter" live?
> Has anyone redone the NetBSD system clock code to use this time source
> for the system clock?
if it's a reasonably univeral register then the thing to do would be
to use it for microtime() only -- at each clock interrupt, get its
value, then when microtime is called, use the offset from that value,