On Tue, 23 Oct 2007 05:29:21 -0400 Daniel Veillard wrote:
On Tue, Oct 23, 2007 at 04:07:34PM +0900, Saori Fukuta wrote:
On Mon, 22 Oct 2007 12:39:50 -0400 Daniel Veillard wrote:
The following patch finishes the cleanup for NUMA parsing code: - the cpuset parsing is moved to xml.c - some comments and cleanups of the include then add the output of a (cpus '...') line based on the /domain/vcpu/@cpuset attributes, this is parsed and reserialized as ranges to avoid any possibility of misinterpretation of say ^ or any special syntax we may want to add in the future. A few things to note: - dependant on the tiny patch I sent earlier today - if we notice that the set covers all CPU maybe we should avoid outputing (cpus '...'), trivial to add - mostly untested yet
I think we also need to fix for "start/define/dumpxml". So, I attached the patch for them. Could you check it ?
Okay, I looked at it, the only worry I have is that it takes the cpus values as emitted in the S-Expr and output it directly in the XML. I guess it really depends how xend exports this information, reading the code it seems they only generate a comma separated CPU list on output (in which case reusing libvirt parsing/serialization to compact it to ranges would be nicer), but I don't have machines with enough CPUs to really see this, could you have a lookm before I check this in ?
Well, it may not be good idea to emit the S-Expr directly to output as you are worried. I confirmed how I can get the cpus values at fedora8(xen-3.1.0-10.fc8), and I expected to be emitted "(cpus '0,1')", when I specified "cpuset='0,1'". But xend shows nothing about cpus. Then, I could not get the cpus values by sexpr_node(root, "domain/cpus"). I guess this is a bug of Xen, and this has not been fixed with latest upstream Xen. So I will work about this for Xen before discussing about how xend exports. Regards, Saori Fukuta