From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752310Ab2LTBc2 (ORCPT ); Wed, 19 Dec 2012 20:32:28 -0500 Received: from mail-we0-f172.google.com ([74.125.82.172]:56332 "EHLO mail-we0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751397Ab2LTBcX (ORCPT ); Wed, 19 Dec 2012 20:32:23 -0500 From: Grant Likely Subject: Re: [PATCH] mfd: db8500-prcmu: fix irqdomain usage To: Linus Walleij , Samuel Ortiz , linux-kernel@vger.kernel.org Cc: Anmar Oueja , Linus Walleij , stable@kernel.org, Lee Jones In-Reply-To: <1355928161-8082-1-git-send-email-linus.walleij@stericsson.com> References: <1355928161-8082-1-git-send-email-linus.walleij@stericsson.com> Date: Thu, 20 Dec 2012 01:32:18 +0000 Message-Id: <20121220013218.A55E93E0AD7@localhost> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 19 Dec 2012 15:42:41 +0100, Linus Walleij wrote: > From: Linus Walleij > > This fixes two issues with the DB8500 PRCMU irqdomain: > - You have to state the irq base 0 to get a linear domain > for the DT case from irq_domain_add_simple() > - The irqdomain was not used to translate the initial irq > request using irq_create_mapping() making the linear > case fail as it was lacking a proper descriptor. > > I took this opportunity to fix two lines of whitespace > errors in related code as I was anyway messing around with > it. > > Cc: stable@kernel.org > Cc: Lee Jones > Signed-off-by: Linus Walleij > --- > Hi Sam, this is a regression and as such should go into the > -rc series. As you can see I also added the Cc: stable tag. > > Yours, > Linus Walleij > --- > drivers/mfd/db8500-prcmu.c | 8 ++++---- > 1 file changed, 4 insertions(+), 4 deletions(-) > > diff --git a/drivers/mfd/db8500-prcmu.c b/drivers/mfd/db8500-prcmu.c > index dc8826d..fcac8e0 100644 > --- a/drivers/mfd/db8500-prcmu.c > +++ b/drivers/mfd/db8500-prcmu.c > @@ -2524,7 +2524,7 @@ static bool read_mailbox_0(void) > > for (n = 0; n < NUM_PRCMU_WAKEUPS; n++) { > if (ev & prcmu_irq_bit[n]) > - generic_handle_irq(IRQ_PRCMU_BASE + n); > + generic_handle_irq(irq_create_mapping(db8500_irq_domain,n)); This looks wrong. It probably works at the moment, but calling irq_create_mapping when trying to process an IRQ is backwards. There certainly shouldn't be any processing of an irq that isn't already mapped. irq_find_mapping() should be used instead. g.