I have discussed with Stefano about current xl/xend situation:
1). current issue seems casued by xl side;
2). we don't need apply Stefano's patch, at least now;
Stefano, you will check xl side, right?
Ian Jackson wrote:
> Liu, Jinsong writes ("RE: [Xen-devel] [PATCH] qemu-xen: fix cpu
>> Stabellini, I read your attached patch, it's OK.
> So we should apply Stefano's patch ? I'm not sure I'm convinced.
>> In fact, we firstly implemented xenstore watch by same scheme of
>> your patch, watching each cpu node status:
>> /local/domain/xx/cpu/yy/availability=offline (online)
>> However, we finally didn't use this scheme. We watch 'common' node
>> instead: /local/domain/xx/cpu in this way, only 1 watch point need.
>> Considering vcpu number may become more and more in the future (say,
>> more than 128), it's more simple and reasonable. (Watches can be
>> set at points in the hierarchy and an individual watch will be
>> triggered when anything at or below that point in the hierachy
> Here you seem to be saying that the scheme that you implemented (in
> xend, I take it) is not the same as the one in Stefano's patch.
>> Yes, I think there must be some confusion.
>> Currently 'xm vcpu-set' command works fine with both PV and HVM vcpu
> Right, good.
>> Stabellini/Jackson, would your please tell me what xl recently
>> happened for vcpu hotplug?
> Well, we accepted patches to try to implement "xl vcpu-set" to make it
> work like "xm vcpu-set" but they are apparently wrong.
>> is there any different ideas xl and xend about xenstore syntax?
> xl should use the same syntax in xenstore as xend currently does. If
> xm (and therefore xend) and qemu currently work properly then changing
> qemu will break xend. Instead, libxl should be changed to use the
> same scheme as xend.
>> (Each time 'xm vcpu-set' executed, xend will write all xenstore cpu
>> node status)
Xen-devel mailing list