From mboxrd@z Thu Jan 1 00:00:00 1970 From: Heiko =?iso-8859-1?q?St=FCbner?= Subject: Re: Broken IRQ handling for S3C2416 Date: Thu, 27 Jun 2013 16:51:45 +0200 Message-ID: <201306271651.45522.heiko@sntech.de> References: <51CC290B.5080701@inov.pt> <201306271626.42419.heiko@sntech.de> <51CC50B0.2090308@inov.pt> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from gloria.sntech.de ([95.129.55.99]:55131 "EHLO gloria.sntech.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751869Ab3F0Ovt (ORCPT ); Thu, 27 Jun 2013 10:51:49 -0400 In-Reply-To: <51CC50B0.2090308@inov.pt> Sender: linux-samsung-soc-owner@vger.kernel.org List-Id: linux-samsung-soc@vger.kernel.org To: =?iso-8859-1?q?Jos=E9_Miguel_Gon=E7alves?= Cc: linux-samsung-soc@vger.kernel.org, Kukjin Kim Am Donnerstag, 27. Juni 2013, 16:48:16 schrieb Jos=E9 Miguel Gon=E7alve= s: > Hi Heiko, >=20 > On 27-06-2013 15:26, Heiko St=FCbner wrote: > > Hi Jose, > >=20 > > Am Donnerstag, 27. Juni 2013, 13:59:07 schrieb Jos=E9 Miguel Gon=E7= alves: > >> Hi, > >>=20 > >> While migrating from kernel 3.8 to kernel 3.9 I've detected a prob= lem on > >> my S3C2416 based board. I'm using UART 3 as console and with kerne= l 3.9 > >> I'm unable to have a getty working on that port and see a lot of t= hese > >> on dmesg: > >>=20 > >> samsung-uart s3c2440-uart.3: cannot get irq 94 > >>=20 > >> Also WDT is not correctly initialized in 3.8, I see in dmesg: > >>=20 > >> s3c2410-wdt s3c2410-wdt: watchdog inactive, reset disabled, irq di= sabled > >>=20 > >> while in 3.9 I see: > >>=20 > >> s3c2410-wdt s3c2410-wdt: tmr_margin value out of range, default 15= used > >> s3c2410-wdt s3c2410-wdt: failed to install irq (-38) > >> s3c2410-wdt: probe of s3c2410-wdt failed with error -38 > >>=20 > >> So I guess something is broken in 3.9 kernel's IRQ handling for S3= C2416. > >=20 > > when I checked two days ago, the irqs where working fine on my s3c2= 416 > > board. > >=20 > > Basic questions, did you change the init_irq callback to the correc= t new > > function (s3c2416_init_irq) in your board file, can you provide a m= ore > > complete log and of course and is your board-file available somewhe= re? >=20 > My bad! I missed that change on the init_irq callback. Everything is > working fine now. Thank you and sorry for the inconvenience. no problem, glad I could help :-) Heiko