Re: [PATCH v1] watchdog: Use a reference cycle counter to avoid scaling issues

[Thomas Gleixner]

On Fri, 24 Apr 2015, Andi Kleen wrote:
> > There are better ways to do that than using heuristics. We have to
> > deal with 3 variants of the reference counter:
> > 
> > 1) Core and Atom: counts bus cycles and we know that frequency already
> >    	    	  from the local apic calibration
> > 
> > 2) Nehalem, Westmere: Same as TSC
> > 
> > 3) Sandybridge and later:  XCLK which is 100MHz
> > 
> > No magic calibration, just use the information which we have on our
> > hands already.
> 
> This is a really bad idea. We basically would need to maintain a big
> switch with model numbers, with new cases added for every new CPU.
> Would be a maintenance nightmare.

Nonsense. We have already enough family specific switch cases which
need to be maintained for every new cpu anyway. So they can simply
provide that extra bit of information. It's neither rocket science nor
a nightmare.

We don't need to calibrate stuff which we can access by simpler means.

Thanks,

	tglx


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/