From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail-ew0-f228.google.com ([209.85.219.228]:62053 "EHLO mail-ew0-f228.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753147Ab0BBRQO (ORCPT ); Tue, 2 Feb 2010 12:16:14 -0500 Received: by ewy28 with SMTP id 28so131280ewy.28 for ; Tue, 02 Feb 2010 09:16:13 -0800 (PST) Message-ID: <4B685DD5.3080508@gmail.com> Date: Tue, 02 Feb 2010 17:16:05 +0000 From: Nameer Kazzaz MIME-Version: 1.0 To: "Igor M. Liplianin" CC: paul10@planar.id.au, linux-media , =?UTF-8?B?Q2hyaXN0aWFuIEjDvHBwZQ==?= Subject: Re: DM1105: could not attach frontend 195d:1105 References: <3bf14d196e3bc8717d910d09a623f98e@mail.velocitynet.com.au> <4B66C649.7010500@gmail.com> <4B68430A.7010500@gmail.com> <201002021857.17353.liplianin@me.by> In-Reply-To: <201002021857.17353.liplianin@me.by> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Sender: linux-media-owner@vger.kernel.org List-ID: Igor M. Liplianin wrote: > On 2 февраля 2010 17:21:46 Nameer Kazzaz wrote: > >> Hi Igor, >> What do you think ? if I can help you solve this, let me know what I >> can do. >> >> Thanks >> Nameer >> >> Nameer Kazzaz wrote: >> >>> Hi there, >>> dmesg output with patched dm1105.c against current v4l-dvb 'modprob >>> dm1105 card=4' >>> >>> dm1105 0000:05:0f.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 >>> DVB: registering new adapter (dm1105) >>> dm1105 0000:05:0f.0: MAC dd49b0dc >>> dm1105 0000:05:0f.0: could not attach frontend >>> dm1105 0000:05:0f.0: PCI INT A disabled >>> >>> Thanks >>> Nameer Kazzaz >>> >>> Igor M. Liplianin wrote: >>> >>>> On 20 ?????? 2010 23:20:20 paul10@planar.id.au wrote: >>>> >>>>> Igor wrote: >>>>> >>>>>> Oh, that is wrong. It is registers addresses, Never touch this. >>>>>> >>>>>> Let's look on that part of code: >>>>>> >>>>>> /* GPIO's for LNB power control */ >>>>>> #define DM1105_LNB_MASK 0x00000000 // later in >>>>>> >>>>> code write it to >>>>> >>>>> >>>>>> DM1105_GPIOCTR, all GPIO's as OUT >>>>>> #define DM1105_LNB_OFF 0x00020000 // later in >>>>>> >>>>> code write it to >>>>> >>>>> >>>>>> DM1105_GPIOVAL, set GPIO17 to HIGH >>>>>> >>>>>> But you have not to change this. >>>>>> Right way is to write another entry in cards structure and so on. >>>>>> Better leave it to me. >>>>>> >>>>>> Regards >>>>>> Igor >>>>>> >>>>> Thanks for all your help, I understand better now. I have moved to >>>>> code >>>>> like that at the bottom. It still doesn't work, but feels a lot >>>>> closer. >>>>> >>>>> Before I keep playing with values, I want to check I'm on the right >>>>> track. >>>>> Does it look right? Specific questions: >>>>> 1. I see there is a hw_init function. Should I be using that? I >>>>> put the >>>>> logic into fe_attach because there was already card-specific logic in >>>>> there. But this feels like hw initialisation. >>>>> >>>>> 2. Should I set the control to input or output? I'm assuming input >>>>> = 1. >>>>> >>>>> 3. Would pin 15 be numbered from the left or right - is it 0x4, or >>>>> 0x2000? >>>>> >>>>> Thanks, >>>>> >>>>> Paul >>>>> >>>>> *** dm1105.c.old 2010-01-13 16:15:00.000000000 +1100 >>>>> --- dm1105.c 2010-01-21 08:13:14.000000000 +1100 >>>>> *************** >>>>> *** 51,56 **** >>>>> --- 51,57 ---- >>>>> #define DM1105_BOARD_DVBWORLD_2002 1 >>>>> #define DM1105_BOARD_DVBWORLD_2004 2 >>>>> #define DM1105_BOARD_AXESS_DM05 3 >>>>> + #define DM1105_BOARD_UNBRANDED 4 >>>>> >>>>> /* ----------------------------------------------- */ >>>>> /* >>>>> *************** >>>>> *** 171,176 **** >>>>> --- 172,181 ---- >>>>> #define DM05_LNB_13V 0x00020000 >>>>> #define DM05_LNB_18V 0x00030000 >>>>> >>>>> + /* GPIO's for demod reset for unbranded 195d:1105 */ >>>>> + #define UNBRANDED_DEMOD_MASK 0x00008000 >>>>> + #define UNBRANDED_DEMOD_RESET 0x00008000 >>>>> + >>>>> static unsigned int card[] = {[0 ... 3] = UNSET }; >>>>> module_param_array(card, int, NULL, 0444); >>>>> MODULE_PARM_DESC(card, "card type"); >>>>> *************** >>>>> *** 206,211 **** >>>>> --- 211,219 ---- >>>>> [DM1105_BOARD_AXESS_DM05] = { >>>>> .name = "Axess/EasyTv DM05", >>>>> }, >>>>> + [DM1105_BOARD_UNBRANDED] = { >>>>> + .name = "Unbranded 195d:1105", >>>>> + }, >>>>> }; >>>>> >>>>> static const struct dm1105_subid dm1105_subids[] = { >>>>> *************** >>>>> *** 229,234 **** >>>>> --- 237,246 ---- >>>>> .subvendor = 0x195d, >>>>> .subdevice = 0x1105, >>>>> .card = DM1105_BOARD_AXESS_DM05, >>>>> + }, { >>>>> + .subvendor = 0x195d, >>>>> + .subdevice = 0x1105, >>>>> + .card = DM1105_BOARD_UNBRANDED, >>>>> }, >>>>> }; >>>>> >>>>> *************** >>>>> *** 698,703 **** >>>>> --- 710,727 ---- >>>>> dm1105dvb->fe->ops.set_voltage = >>>>> dm1105dvb_set_voltage; >>>>> >>>>> break; >>>>> + case DM1105_BOARD_UNBRANDED: >>>>> + printk(KERN_ERR "Attaching as board_unbranded\n"); >>>>> + outl(UNBRANDED_DEMOD_MASK, dm_io_mem(DM1105_GPIOCTR)); >>>>> + outl(UNBRANDED_DEMOD_RESET , >>>>> dm_io_mem(DM1105_GPIOVAL)); >>>>> + dm1105dvb->fe = dvb_attach( >>>>> + si21xx_attach, &serit_config, >>>>> + &dm1105dvb->i2c_adap); >>>>> + if (dm1105dvb->fe) >>>>> + dm1105dvb->fe->ops.set_voltage = >>>>> + dm1105dvb_set_voltage; >>>>> + >>>>> + break; >>>>> case DM1105_BOARD_DVBWORLD_2002: >>>>> case DM1105_BOARD_AXESS_DM05: >>>>> default: >>>>> >>>> Some things are missed, like keep GPIO15 high in set_voltage function. >>>> Try attached patch against current v4l-dvb tree with modprobe option >>>> card=4 >>>> modprobe dm1105 card=4 >>>> > Hi Nameer, > You can modify sended by me patch to guess GPIO pin. Simply try all of them, it is only 17. > Just replace all appearances DM1105_GPIO(15) with number you want. > Ok I will do that. Thanks Nameer