From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Sullivan Subject: Re: [Patch v4] x86: irq_comm: Add check for RH bit in kvm_set_msi_irq Date: Fri, 13 Mar 2015 09:12:57 -0600 Message-ID: <5502FE79.2040306@gmail.com> References: <55022423.7050101@gmail.com> <5502506C.6000005@gmail.com> <550254C6.6010103@gmail.com> <20150313143957.GA17349@potion.brq.redhat.com> <5502F885.7080508@gmail.com> <20150313150829.GB1711@potion.brq.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: kvm@vger.kernel.org, pbonzini@redhat.com, gleb@kernel.org To: =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= Return-path: Received: from mail-pa0-f52.google.com ([209.85.220.52]:44577 "EHLO mail-pa0-f52.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751986AbbCMPPI (ORCPT ); Fri, 13 Mar 2015 11:15:08 -0400 Received: by padet14 with SMTP id et14so30153451pad.11 for ; Fri, 13 Mar 2015 08:15:07 -0700 (PDT) In-Reply-To: <20150313150829.GB1711@potion.brq.redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: Perfect, thanks for the feedback. I'll get v5 out shortly. On 03/13/2015 09:08 AM, Radim Kr=C4=8Dm=C3=A1=C5=99 wrote: > 2015-03-13 08:47-0600, James Sullivan: >> On 03/13/2015 08:39 AM, Radim Kr=C4=8Dm=C3=A1=C5=99 wrote: >> ... >>> The warning message is very clever: >>> - it contains the magical "may" qualifier and being protected only = by >>> RH=3D1 creates weird-looking code structure, but it is technicall= y right >>> 1) lowest-priority delivery may be set in msi.data, which avoids = our >>> otherwise incorrect behavior with RH=3D1/DM=3D1 >>> 2) RH=3D1/DM=3D0 can't deliver to multiple APICs (broadcast is fo= rbidden), >>> but real hardware may overwrite delivery mode from msi.data >>> - being two lines apart adds to suspicion, yet it can be hint to th= ose >>> possible problems >>> >>> I only fear it is too clever :) >>> >> >> For the error message, how does: >> >> kvm: MSI RH=3D1 unsupported, use low-priority delivery mode >> >> Sit with you? >=20 > I actually liked the former. >=20 > New one doesn't say what is the impact of the error and the advice is > not easy follow -- people usually have no idea what low-priority > delivery mode is and nothing can be done outside of the guest. >=20 > (I put the rant mainly for future reviewers; the alternative I had i= n > was to warn only when DM=3D1.) >=20