From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758198AbXJCVMl (ORCPT ); Wed, 3 Oct 2007 17:12:41 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752747AbXJCVMd (ORCPT ); Wed, 3 Oct 2007 17:12:33 -0400 Received: from mailbox2.myri.com ([64.172.73.26]:1935 "EHLO myri.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1752694AbXJCVMc (ORCPT ); Wed, 3 Oct 2007 17:12:32 -0400 Message-ID: <470405BB.2000208@myri.com> Date: Wed, 03 Oct 2007 17:12:27 -0400 From: Loic Prylli User-Agent: Thunderbird/2.0.0.4 Mnenhy/0.7.5.0 MIME-Version: 1.0 To: linux-pci@atrey.karlin.mff.cuni.cz, Linux Kernel Mailing List , "Eric W. Biederman" Subject: MSI problem since 2.6.21 for devices not providing a mask in their MSI capability X-Enigmail-Version: 0.95.3 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, We observe a problem with MSI since kernel 2.6.21 where interrupts would randomly stop working. We have tracked it down to the new msi_set_mask_bit definition in 2.6.21. In the MSI case with a device not providing a "native" MSI mask, it was a no-op before, and now it disables MSI in the MSI-ctl register which according to the PCI spec is interpreted as reverting the device to legacy interrupts. If such a device try to generate a new interrupt during the "masked" window, the device will try a legacy interrupt which is generally ignored/never-acked and cause interrupts to no longer work for the device/driver combination (even after the enable bit is restored). Is there anything apart from irq migration that strongly requires masking? Is is possible to do the irq migration without masking? Loic