From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH] use KVMState, as upstream do Date: Sun, 07 Jun 2009 13:10:18 +0300 Message-ID: <4A2B920A.8050701@redhat.com> References: <1244139783-15787-1-git-send-email-glommer@redhat.com> <4A2B86B1.2040502@redhat.com> <4A2B9185.9090601@web.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Glauber Costa , kvm@vger.kernel.org To: Jan Kiszka Return-path: Received: from mx2.redhat.com ([66.187.237.31]:50490 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753882AbZFGKKU (ORCPT ); Sun, 7 Jun 2009 06:10:20 -0400 In-Reply-To: <4A2B9185.9090601@web.de> Sender: kvm-owner@vger.kernel.org List-ID: Jan Kiszka wrote: >> What we could do is embed a kvm_context_t variable in the real KVMState, >> and start using that. Then, we could, function by function, use the >> upstream KVMState functionality and remove the corresponding >> kvm_context_t functionality. >> > > That would be great and was also what I had vaguely in mind. Likely > there are already fields that can be used from KVMState (fd and vmfd?). > Obvious candidates are to start using kvm_*_ioctl() everywhere. -- Do not meddle in the internals of kernels, for they are subtle and quick to panic.