From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934032AbdJQH4e (ORCPT ); Tue, 17 Oct 2017 03:56:34 -0400 Received: from Galois.linutronix.de ([146.0.238.70]:54929 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757305AbdJQH43 (ORCPT ); Tue, 17 Oct 2017 03:56:29 -0400 Message-Id: <20171017075456.694047687@linutronix.de> User-Agent: quilt/0.63-1 Date: Tue, 17 Oct 2017 09:54:56 +0200 From: Thomas Gleixner To: LKML Cc: Dexuan Cui , x86@kernel.org, Bjorn Helgaas , Marc Zyngier , KY Srinivasan , Stephen Hemminger , linux-pci@vger.kernel.org, devel@linuxdriverproject.org, Josh Poulson , Haiyang Zhang , Simon Xiao , Saeed Mahameed , Mihai Costache , Jork Loeser Subject: [patch 0/3] x86/PCI/MSI: Make sure that irq reservation mode works everywhere Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Dexuan reported that the recent rework of the vector allocation mode in x86 broke HyperV PCI passtrough because the rework missed to add the MSI_FLAG_MUST_REACTIVATE flag to the HyperV/PCI interrupt domain info. The simple solution would be to set the flag in the HyperV/PCI driver but it's better to make this generic and let the PCI/MSI core code set the flag when reservation mode is enabled. That ensures that future users of this wont trip over the same problem. Thanks, tglx --- arch/x86/Kconfig | 2 +- arch/x86/kernel/apic/msi.c | 5 ++--- drivers/pci/msi.c | 2 ++ kernel/irq/Kconfig | 3 +++ 4 files changed, 8 insertions(+), 4 deletions(-)