From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from viefep27-int.chello.at ([62.179.121.47]) by www.linuxtv.org with esmtp (Exim 4.63) (envelope-from ) id 1Jt1HE-0005kT-RR for linux-dvb@linuxtv.org; Mon, 05 May 2008 16:02:49 +0200 Message-ID: <481F1365.50808@hispeed.ch> Date: Mon, 05 May 2008 16:02:13 +0200 From: Roland Scheidegger MIME-Version: 1.0 To: Ruediger Dohmhardt References: <481B7D43.2050200@hispeed.ch> <481CD27A.9080300@freenet.de> In-Reply-To: <481CD27A.9080300@freenet.de> Cc: linux-dvb@linuxtv.org Subject: Re: [linux-dvb] mantis crash... List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: linux-dvb-bounces@linuxtv.org Errors-To: linux-dvb-bounces+mchehab=infradead.org@linuxtv.org List-ID: Ruediger Dohmhardt wrote: > Roland Scheidegger schrieb: >> This was reported before, the current mantis driver will cause a >> page fault in mantis_dvb_start_feed. >> > Dear Roland, > > I applied your patch to the code I got from > > http://jusst.de/hg/mantis > > Now the mantis driver (for 2033) works fine for me. > > Setup: 2.6.22.19 vdr-1.5.18/xineliboutput > > Unfortunately, as soon as I insert the CAM module, the driver > crashes. According to the logfile, it looks that "vdr" still gets the > information about the CAM and tries to access it. Then the driver > crashes. Maybe someone still has an idea how to dig into this. Yes, the crash is due to the same (or similar) bogus hif reads/writes as that fixed in the patch obviously - that code gets only used with a CAM. There are mmreads/mmwrites to MANTIS_GPIF_HIFADDR, MANTIS_GPIF_HIFDIN and MANTIS_GPIF_HIFDOUT - I'm sure all of them will blow up (they are all the exact same value anyway). But I wouldn't know how to fix this - I suspect those values are actually masks which you'd or with some other values for accessing values not directly accessible in registers. That's just speculation however, I don't even have a clue what "HIF" stands for (well I guess something with interface?). Fixing it would require a chip datasheet at least (or (working) example code), or reverse engeneering or whatever. Even if this particular hif read/write issue was fixed I'd highly doubt though the cam code would actually work in its current state - apparently it was never tested nor working. So Manu Abraham needs to fix it I suspect - unfortunately there were no updates for over two weeks now. Roland _______________________________________________ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb