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

RE: [Xen-ia64-devel] Behaviour of DOM0_GETMEMLIST ?


  • To: "Tristan Gingold" <Tristan.Gingold@xxxxxxxx>, <xen-ia64-devel@xxxxxxxxxxxxxxxxxxx>
  • From: "Magenheimer, Dan (HP Labs Fort Collins)" <dan.magenheimer@xxxxxx>
  • Date: Fri, 12 May 2006 10:02:29 -0700
  • Delivery-date: Fri, 12 May 2006 10:02:46 -0700
  • List-id: Discussion of the ia64 port of Xen <xen-ia64-devel.lists.xensource.com>
  • Thread-index: AcZ11kK52UjmKNaqQ8m+mk5nxrCFdQADr4xg
  • Thread-topic: [Xen-ia64-devel] Behaviour of DOM0_GETMEMLIST ?

> is this comment correct (dom0_ops.c).  I don't see where 
> pages are allocated.
> 
> 
>    /*
>      * NOTE: DOM0_GETMEMLIST has somewhat different semantics 
> on IA64 -
>      * it actually allocates and maps pages.
>      */
>     case DOM0_GETMEMLIST:
>     {

I think it is no longer correct.  I think it was written
by Matt Chapman last summer.  At that time (until Kevin
changed it recently), physical pages were allocated for
pseudo-physical pages the first time a domain accessed a
psuedo-physical address.  If a domain had use of a page
but never accessed it, Xen never assigned it to the domain.
This was more flexible and better for certain future
implementations (such as memory oversubscription and
pull-based migration), but was different than the Xen/x86
implementation.  So rather than make core Xen tools more
flexible, it was changed to be more like Xen/x86.

Dan


_______________________________________________
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®.