From mboxrd@z Thu Jan 1 00:00:00 1970 From: valentin.longchamp@epfl.ch (Valentin Longchamp) Date: Wed, 09 Dec 2009 13:44:50 +0100 Subject: [PATCH] mx31moboard: fix typo In-Reply-To: <20091204174623.GW14091@buzzloop.caiaq.de> References: <20091203193034.GD22533@pengutronix.de> <20091203193321.GE22533@pengutronix.de> <4B19473D.7010801@epfl.ch> <20091204174623.GW14091@buzzloop.caiaq.de> Message-ID: <4B1F9BC2.6010205@epfl.ch> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Daniel Mack wrote: > On Fri, Dec 04, 2009 at 06:30:37PM +0100, Valentin Longchamp wrote: >> Guennadi Liakhovetski wrote: >>> ...to second that, yesterday's next also locks up hard on boot on >>> pcm037. .config available on request. >>> >> Same problem here. As a matter of fact, it comes from the mxc-master >> branch. I have bisected out the culprit: >> >> http://git.pengutronix.de/?p=imx/linux-2.6.git;a=commit;h=52939c03e53b151848da9e83fd839bddfda29e78 >> >> It seems conform with the lock syndrom: very early at boot right >> after Uncompessing Linux and before any other message. >> >> The proc stalls on this read: >> srev = __raw_readl(IO_ADDRESS(IIM_BASE_ADDR) + MXC_IIMSREV); >> >> Maybe Guennadi and I have silicon revisions that do not work well >> with this patch (because I assume it worked well for Daniel). > > Grmpf. Yes, it does work right on my LiteKit board, and the reference > manual reads like the register is there for a very long time (as it has > possible values for very old silicon revisions). Well, there are not a lot of information about the IIM in the reference manual (only Tables 13-1 for the different fuses and 13-2 for the silicon revisions). I wanted to check for the registers to find maybe a problem, but found nothing more > > Could you double check whether the clock is running? The only thing I can > think of is differences in the bootloaders. I still have the proprietary > 'losh' running here, which I have no sources for. > The clock is running (well, at least the parent ipg_clk and the gating is configured for the clock running, thanks to clk_enable(&iim_clk); About the bootloader, on our boards we have redboot, adapted from the release Freescale did for the ADS. So it may be a bootloader issue. What could be missing in my bootloader that makes this read fail ? I would say that maybe someting the EMI or SPBA configuration is missing so that I can access the addresses in this range. Any ideas ? Val -- Valentin Longchamp, PhD Student, EPFL-STI-LSRO1 valentin.longchamp at epfl.ch, Phone: +41216937827 http://people.epfl.ch/valentin.longchamp MEA3485, Station 9, CH-1015 Lausanne