From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kevin Hilman Subject: Re: OMAP1 realtime patch Date: Mon, 12 Feb 2007 11:24:30 -0800 Message-ID: <1171308270.20717.32.camel@vence.hilman.org> References: <45C785A4.6080007@googlemail.com> <1170882015.11368.25.camel@vence.hilman.org> <45CB6775.7020500@googlemail.com> <45CF4BD1.6010109@googlemail.com> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <45CF4BD1.6010109@googlemail.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-omap-open-source-bounces@linux.omap.com Errors-To: linux-omap-open-source-bounces@linux.omap.com To: Dirk Behme Cc: linux-omap-open-source@linux.omap.com List-Id: linux-omap@vger.kernel.org On Sun, 2007-02-11 at 18:01 +0100, Dirk Behme wrote: > Dirk Behme wrote: > > Kevin Hilman wrote: > > > >> Here is a patch which leaves out the raw_spinlock conversions. It goes > >> on the OMAP git tree on top of -rt5 (ignoring conflicts.) I've left out > >> all the raw_spinlock conversions. This has been tested on an > >> OMAP1623/H2 platform. > > > > Update of Kevins update. Tested on OSK. > > > > Changes: > > > > - Remove kernel/printk.c "+ zap_rt_locks();". It's already in -rt5. > > > > - Re-add HRTIMER conversion to ads7846.c > > > > With this patch I get Oops below after playing mp3 via NFS. Add > > raw_spinlock_t to sound/arm/omap/omap-alsa-dma.c like in my original > > patch doesn't seem to help. > > Re-adding raw_spinlock_t to omap-alsa.h helps against this Oops. My guess the oops you sent is that you saw this crash when doing a Ctrl-C or otherwise stopping the playback, correct? This oops is triggered by the same thread trying to take the same spinlock (now an rt_mutex under PREEMPT_RT.) The following patch fixes the nested spinlocks, and removes the raw_spinlock. It applies on top of your latest patch. On my OSK, I do notice however that using -rt (with and without this fix) mp3 playback has some audible noise under PREEMPT_RT. Do you hear this too? Kevin