From: Keir Fraser Subject: Re: [Patch] continue_hypercall_on_cpu rework using tasklets Date: Wed, 14 Apr 2010 07:46:56 +0100 Message-ID: References: <4BC543F3.6080500@ts.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <4BC543F3.6080500@ts.fujitsu.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Juergen Gross Cc: "xen-devel@lists.xensource.com" List-Id: xen-devel@lists.xenproject.org On 14/04/2010 05:26, "Juergen Gross" wrote: >> Why do you change the interface of continue_hypercall_on_cpu()? What's a >> 'hdl' anyway? > > I need a way to find the migrate_info structure in case of nested calls of > continue_hypercall_on_cpu(). Originally this was done by storing it in the > vcpu structure, but this can't be done any more using tasklets. In my first > attempt I saved it in the per-cpu area, but this approach isn't working if > continue_hypercall_on_cpu() is called concurrently. So the cleanest way is > to pass it via a parameter. The per-cpu area method should work fine, since Xen is non-preemptive? I don't think the concurrency you are concerned about can happen. -- Keir