From mboxrd@z Thu Jan 1 00:00:00 1970 From: Gleb Natapov Subject: Re: [PATCH] KVM: nVMX: Rework event injection and recovery Date: Thu, 21 Feb 2013 15:45:31 +0200 Message-ID: <20130221134531.GA16817@redhat.com> References: <51250CFA.10600@siemens.com> <20130221092221.GZ3600@redhat.com> <5125EC5D.4050408@siemens.com> <20130221100653.GB3600@redhat.com> <5125F492.8080902@siemens.com> <5125F6C4.7040802@siemens.com> <5125F7FA.8060008@siemens.com> <20130221131332.GA14354@redhat.com> <51261F92.4070705@siemens.com> <20130221133716.GB8847@fermat.math.technion.ac.il> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Jan Kiszka , Marcelo Tosatti , kvm , "Nakajima, Jun" To: "Nadav Har'El" Return-path: Received: from mx1.redhat.com ([209.132.183.28]:31334 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753070Ab3BUNpp (ORCPT ); Thu, 21 Feb 2013 08:45:45 -0500 Content-Disposition: inline In-Reply-To: <20130221133716.GB8847@fermat.math.technion.ac.il> Sender: kvm-owner@vger.kernel.org List-ID: On Thu, Feb 21, 2013 at 03:37:16PM +0200, Nadav Har'El wrote: > On Thu, Feb 21, 2013, Jan Kiszka wrote about "Re: [PATCH] KVM: nVMX: Rework event injection and recovery": > > That generally does not help to inject/report an external IRQ to L1 as > > L1 runs with IRQs disabled around VMLAUNCH/RESUME. > > Good point, I forgot that :( > > So it looks like nested_run_pending was necessary, after all. > Not sure (as in "this is implementation detail that is possible to avoid", not as in "this check here is incorrect!" :)). If interrupts are disabled then vmx_interrupt_allowed() should return false because interrupts are disabled, not because we emulating guest entry. This is easier said that done though, but I'll think about it. Looks like SVM does it this way. -- Gleb.