From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Kiszka Subject: Re: [PATCH 4/5] msix: Don't process table changes while disabled Date: Tue, 18 Oct 2011 13:35:37 +0200 Message-ID: <4E9D6489.7000608@siemens.com> References: <16be9aa8fe4b5ebef73192c487b3cdf61c7a14a9.1318924251.git.jan.kiszka@siemens.com> <20111018111809.GC28776@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Avi Kivity , Marcelo Tosatti , "kvm@vger.kernel.org" To: "Michael S. Tsirkin" Return-path: Received: from goliath.siemens.de ([192.35.17.28]:20696 "EHLO goliath.siemens.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754706Ab1JRLfl (ORCPT ); Tue, 18 Oct 2011 07:35:41 -0400 In-Reply-To: <20111018111809.GC28776@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 2011-10-18 13:18, Michael S. Tsirkin wrote: > On Tue, Oct 18, 2011 at 09:50:53AM +0200, Jan Kiszka wrote: >> As long as MSI-X is disabled, it's incorrect to invoke >> msix_handle_mask_update on per-vector mask changes. That may misguide >> the config notifier callback or spuriously trigger an MSI event. >> >> Signed-off-by: Jan Kiszka > > Same question as on the previous patch. At the moment, the only user > (virtio) seems to check msix_enabled before registering > notifiers. See virtio_pci_query_guest_notifiers. > Is there some way you see to trigger incorrect behaviour? Then it's not an urgent fix, just a conceptual one (the user should not have to worry about enable/disabled, we want the core to do the boring jobs). Jan -- Siemens AG, Corporate Technology, CT T DE IT 1 Corporate Competence Center Embedded Linux