From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH] [RFC] Activate VMX on demand Date: Tue, 04 Nov 2008 15:37:38 +0200 Message-ID: <49105022.6050207@redhat.com> References: <1225729181-24431-1-git-send-email-agraf@suse.de> <491025AE.1030601@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org, Sander.Vanleeuwen@sun.com, kraxel@redhat.com, anthony@codemonkey.ws, zach@vmware.com To: Alexander Graf Return-path: Received: from mx2.redhat.com ([66.187.237.31]:44001 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752055AbYKDNhr (ORCPT ); Tue, 4 Nov 2008 08:37:47 -0500 In-Reply-To: Sender: kvm-owner@vger.kernel.org List-ID: Alexander Graf wrote: >> >> This should be done in x86.c and made to apply to svm as well. Other >> VMMs would use efer.svme as an indication that svm is in use, similar >> to cr4.vmxe today. > > Right, that sounds good. So just track the usage counter in x86.c and > then split the hardware_{en,dis}able functions? What's there to split in hardware_enable()? > Or do you want to actually call hardware_{en,dis}able on the usage > counter events? Yes. > There is no real need to enable/disable everything just because of the > usage counter. All it does is enable the feature, enable vmxe, and vmxon. I don't think we gain anything by splitting it. -- error compiling committee.c: too many arguments to function