From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752922Ab2ITJcH (ORCPT ); Thu, 20 Sep 2012 05:32:07 -0400 Received: from mx1.redhat.com ([209.132.183.28]:56535 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752119Ab2ITJcF (ORCPT ); Thu, 20 Sep 2012 05:32:05 -0400 Message-ID: <505AE281.1030603@redhat.com> Date: Thu, 20 Sep 2012 12:31:45 +0300 From: Avi Kivity User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:15.0) Gecko/20120828 Thunderbird/15.0 MIME-Version: 1.0 To: Suresh Siddha CC: hpa@zytor.com, mingo@kernel.org, torvalds@linux-foundation.org, andreas.herrmann3@amd.com, bp@amd64.org, robert.richter@amd.com, linux-kernel@vger.kernel.org Subject: Re: [PATCH 3/6] x86, kvm: use kernel_fpu_begin/end() in kvm_load/put_guest_fpu() References: <1345842782-24175-1-git-send-email-suresh.b.siddha@intel.com> <1345842782-24175-4-git-send-email-suresh.b.siddha@intel.com> <50599AE2.80601@redhat.com> <1348075103.26695.301.camel@sbsiddha-desk.sc.intel.com> <5059FF67.6050005@redhat.com> <1348075534.26695.303.camel@sbsiddha-desk.sc.intel.com> In-Reply-To: <1348075534.26695.303.camel@sbsiddha-desk.sc.intel.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 09/19/2012 08:25 PM, Suresh Siddha wrote: > On Wed, 2012-09-19 at 20:22 +0300, Avi Kivity wrote: >> On 09/19/2012 08:18 PM, Suresh Siddha wrote: >> >> > These routines (kvm_load/put_guest_fpu()) are already called with >> > preemption disabled but as you mentioned, we don't want the preemption >> > to be disabled completely between the kvm_load_guest_fpu() and >> > kvm_put_guest_fpu(). >> > >> > Also KVM already has the preempt notifier which is doing the >> > kvm_put_guest_fpu(), so something like the appended should address this. >> > I will test this shortly. >> > >> >> Note, we could also go in a different direction and make >> kernel_fpu_begin() use preempt notifiers and thus make its users >> preemptible. But that's for a separate patchset. > > yep, but we need the fpu buffer to save/restore the kernel fpu state. > > KVM already has those buffers allocated in the guest cpu state and hence > it all works out ok. But yes, we can revisit this in future. kernel_fpu_begin() can allocate it. It means changing the APIs, but changing the behaviour to be preemptible is a bigger change anyway. -- error compiling committee.c: too many arguments to function