From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756118AbcGZNvR (ORCPT ); Tue, 26 Jul 2016 09:51:17 -0400 Received: from Galois.linutronix.de ([146.0.238.70]:43510 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751261AbcGZNvO (ORCPT ); Tue, 26 Jul 2016 09:51:14 -0400 Date: Tue, 26 Jul 2016 15:49:02 +0200 (CEST) From: Thomas Gleixner To: Foster Snowhill cc: LKML , Matthias Prager , Jason Taylor , Marc Zyngier Subject: Re: PROBLEM: Intel I210AT NIC resets while using PCI passthrough on ESXi (regression) In-Reply-To: Message-ID: References: User-Agent: Alpine 2.11 (DEB 23 2013-08-11) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 25 Jul 2016, Foster Snowhill wrote: > On 25.07.16 13:56, Thomas Gleixner wrote: > > On Sat, 23 Jul 2016, Foster Snowhill wrote: > >> [1.] One line summary of the problem: > >> > >> Intel I210AT NIC resets while using PCI passthrough on ESXi (regression) > > > > That has been reported before and so far this has been believed to be a vmware > > issue. Native hardware and KVM do not expose that. See: > > > > http://marc.info/?l=linux-kernel&m=145280623530135&w=2 > > http://marc.info/?l=linux-kernel&m=145879968005421&w=2 > > This is indeed the same symptoms I encountered in my case. More than that, I have an LSI 2008 adapter which is also being passed through in the same manner as the NIC, absolutely the same problem there too. Before doing the bisect I considered it a separate issue and planned to work on it later. > > > Could you please give the patch below a try? It might be related, but I'm not > > sure whether it will cure that particular vmware oddity. > > Patch fixed the issue for me. Did 4-5 reboots just to be sure, all working properly, no adapter resets. Just for the sake of experiment built a clean 4.7 kernel without the patch, issue still present there as expected. > Can you please test something else with that patch? After starting the NIC, can you please shutdown the interface, so the interrupt gets shutdown as well. (It should vanish from /proc/interrupts). Does that work correctly? Thanks, tglx