From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ryan Harper Subject: Re: [PATCH] add dom0 vcpu hotplug control Date: Tue, 7 Jun 2005 17:35:14 -0500 Message-ID: <20050607223514.GH28421@us.ibm.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Ian Pratt Cc: Ryan Harper , xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org * Ian Pratt [2005-06-07 17:27]: > > > > -----Original Message----- > > From: xen-devel-bounces@lists.xensource.com > > [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of > > Ryan Harper > > Sent: 07 June 2005 23:03 > > To: xen-devel@lists.xensource.com > > Subject: [Xen-devel] [PATCH] add dom0 vcpu hotplug control > > > > This patch adds new control messages for vcpu hotplug events. > > Via the xm vcpu_hotplug sub-program, VCPUS in domains can be > > enabled/disabled when CONFIG_HOTPLUG_CPU is enabled in the > > target domain's kernel. > > > > Currently there is nothing that tracks whether a VCPU is up or down. > > My previous [1]patch added a new per-VCPU flag (VCPUF_down) > > which could be used to keep track of which VCPUS are up and > > down. Right now, there isn't a hypercall that returns the > > per-VCPU flag status (something equivalent to the per-domain > > flags in get_dominfo ). Have we thought about a > > get_vcpu_info hypercall? I'd like to get that implemented so > > I can report VCPU state in the xm list --vcpus output. That > > would also make it easier to skip sending control messages > > that don't change the VCPU's state (e.g. sending a down > > message to a vcpu that is already down). > > I haven't looked at the code, but can you not use DOM0_GETVCPUCONTEXT to > iterate over all the VCPUs for a domain? (e.g. if you ask for VCPU N and > it doesn't exist (isn't up) you'll get the state for the next highest > numbered one, or ESRCH if there isn't one) That's the behaviour I'd > expect. That would work. Moving forward, there may be more per-VCPU info that we want to collect and that it would make sense to factor out per-VCPU info from DOM0_GETDOMINFO and create a DOM0_GETVCPUINFO call. If that is not desirable at this point, I can work with DOM0_GETVCPUCONTEXT. -- Ryan Harper Software Engineer; Linux Technology Center IBM Corp., Austin, Tx (512) 838-9253 T/L: 678-9253 ryanh@us.ibm.com