<br>
<br><tt><font size=2>linuxppc-dev-bounces+lukebrowning=us.ibm.com@ozlabs.org
wrote on 07/16/2006 12:53:54 AM:<br>
<br>
> On Sat, Jul 15, 2006 at 02:00:02AM +0200, Arnd Bergmann wrote:<br>
> <br>
> > What happened to the question whether to use PURR values for
also measuring<br>
> > cycles spent executing the hcall as opposed to cycles that passed
before<br>
> > the hcall returns. Did that turn out not giving extra information
after all<br>
> > or was there a different reason to drop that idea?<br>
> <br>
> Just so people don't forget: this can't be done on all processors.
For<br>
> example, PPC970 and POWER4 don't implement the PURR SPR. And it doesn't<br>
> make sense to use H_PURR to get the software emulated ones there.
Not<br>
> really an issue on POWER4 since they don't do shared processor LPAR,
but<br>
> on JS21 I think they might do?<br>
> <br>
Isn't there someway to do a platform specific processor overlay, where
you </font></tt>
<br><tt><font size=2>get PURR if it exists and TB otherwise.</font></tt>
<br>
<br><tt><font size=2>Luke<br>
</font></tt>