Re: [RFC PATCH V2 8/8] x86/hvm: factor out vm_event related functions into separate file

From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Tamas K Lengyel <tamas.lengyel@zentific.com>, xen-devel@lists.xen.org
Cc: kevin.tian@intel.com, wei.liu2@citrix.com,
	ian.campbell@citrix.com, stefano.stabellini@eu.citrix.com,
	ian.jackson@eu.citrix.com, tim@xen.org, jbeulich@suse.com,
	eddie.dong@intel.com, andres@lagarcavilla.org,
	jun.nakajima@intel.com, rshriram@cs.ubc.ca, keir@xen.org,
	dgdegra@tycho.nsa.gov, yanghy@cn.fujitsu.com
Subject: Re: [RFC PATCH V2 8/8] x86/hvm: factor out vm_event related functions into separate file
Date: Thu, 22 Jan 2015 16:32:09 +0000	[thread overview]
Message-ID: <54C12609.8010408@citrix.com> (raw)
In-Reply-To: <1421594281-27658-9-git-send-email-tamas.lengyel@zentific.com>

On 18/01/15 15:18, Tamas K Lengyel wrote:
> -void hvm_event_cr0(unsigned long value, unsigned long old)
> -{
> -    vm_event_request_t req = {
> -        .reason = VM_EVENT_REASON_CR0,
> -        .vcpu_id = current->vcpu_id,
> -        .cr_event.new_value = value,
> -        .cr_event.old_value = old
> -    };
> -
> -    long parameters = current->domain->arch.hvm_domain
> -                        .params[HVM_PARAM_MEMORY_EVENT_CR0];

(I realise this is probably not the best patch, but) As we are redoing
the API with a hope of including PV and ARM guests, can we remove this
use of hvm params?  This would probably involve a new setup hypercall
subop to set up reporting preferences.

~Andrew