[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-ia64-devel] RE: Timer merge


  • To: "Magenheimer, Dan \(HP Labs Fort Collins\)" <dan.magenheimer@xxxxxx>
  • From: "Dong, Eddie" <eddie.dong@xxxxxxxxx>
  • Date: Mon, 29 Aug 2005 10:49:37 +0800
  • Cc: "Mallick, Asit K" <asit.k.mallick@xxxxxxxxx>, xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
  • Delivery-date: Mon, 29 Aug 2005 02:47:31 +0000
  • List-id: Discussion of the ia64 port of Xen <xen-ia64-devel.lists.xensource.com>
  • Thread-index: AcWnjSOhSjX2nMKXQI2EGFjNgzyS+AAXrvdAAAAqXDAAAf54IAAFcXoAAATDFSAAADCbAAAOO83gAB29uYAAKFIVoAACuHFQACBS1PAAD2rcAACBlABQAAEE8iAAADG88A==
  • Thread-topic: Timer merge

Sure! 

-----Original Message-----
From: Magenheimer, Dan (HP Labs Fort Collins) [mailto:dan.magenheimer@xxxxxx] 
Sent: 2005年8月29日 10:48
To: Dong, Eddie
Cc: Mallick, Asit K; xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
Subject: RE: Timer merge

OK.  Could you please develop the patch so that it is
entirely controlled by a define/ifdefs?   Since this is a
fundamental change, it would be a good idea to be able
to easily switch back to the old way so that if there
are bugs or performance questions, we can just change
one define and test the old way.  (I think this is
a good development technique... as you saw, I have
several such ifdefs for the various fast paths.)
The define can be enabled/disabled in config.h since
there are likely to be changes in multiple files.

Thanks,
Dan

> -----Original Message-----
> From: Dong, Eddie [mailto:eddie.dong@xxxxxxxxx] 
> Sent: Sunday, August 28, 2005 8:30 PM
> To: Magenheimer, Dan (HP Labs Fort Collins)
> Cc: Mallick, Asit K; xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
> Subject: RE: Timer merge
> 
> Dan:
>       Thanks, we will send out a patch soon for review with 
> 1: guest ITC (delta) /ITM keep in shared page, 2: fast 
> hypercall for TPR,EOI, IVR, 3: ac_timer own machine ITM, 4: 
> short route for guest to set ITM.
>       Eddie.
> 
> 
> > 
> > So Xen/x86 delivers every timer tick to every Linux domain?
> > That seems like a potential performance problem!
> Yes, the overhead spent on IA32 timer is much higher than IA64. 
> Perhaps Linux tickless patch is a better solution.
> 
> 
> 

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.