On 07/30/2012 02:56 PM, David Basden wrote:
>>>>> Right now I really don't know where I should look for the solution of
>>>>> the problem. It seems that the tuner reset does not have any effect on the 
>>>>> tuner whatsoever.
> 
> Can I suggest setting GPIO5 to 1, leave it there, and see if it breaks. If it
> doesn't, GPIO5 on the RTL isn't setup correctly somehow.
> 
> At the same time, I was rereading code from:
> 
> http://git.linuxtv.org/anttip/media_tree.git/blob/3efd26330fda97e06279cbca170ae4a0dee53220:/drivers/media/dvb/dvb-usb/rtl28xxu.c#l898
> 
> and at no point is GPIO5 actually set to an output or enabled that I can find.
> rtl2832u_frontend_attach skips doing so. (Actually, I seem to remember running
> into this problem while trying to use some DVB driver code as an example of
> how to setup the RTL to talk to the FC0012)
> 
> Try giving the patch below a go. Sorry, I don't have a build environment to
> hand, so there might be a typo I haven't picked up, but the upshot is that 
> I'm moving the FC0012 detection to the end, setting up GPIO5, resetting the
> tuner and then trying to probe for the FC0012.
> 
> Please let me know if this helps :)
> 
> David
> 
> --- a/rtl28xxu.c	2012-07-30 22:31:53.789638678 +1000
> +++ b/rtl28xxu.c	2012-07-30 22:48:35.774607232 +1000
> @@ -550,15 +550,6 @@
>  
>  	priv->tuner = TUNER_NONE;
>  
> -	/* check FC0012 ID register; reg=00 val=a1 */
> -	ret = rtl28xxu_ctrl_msg(adap->dev, &req_fc0012);
> -	if (ret == 0 && buf[0] == 0xa1) {
> -		priv->tuner = TUNER_RTL2832_FC0012;
> -		rtl2832_config = &rtl28xxu_rtl2832_fc0012_config;
> -		info("%s: FC0012 tuner found", __func__);
> -		goto found;
> -	}
> -
>  	/* check FC0013 ID register; reg=00 val=a3 */
>  	ret = rtl28xxu_ctrl_msg(adap->dev, &req_fc0013);
>  	if (ret == 0 && buf[0] == 0xa3) {
> @@ -640,6 +631,71 @@
>  		goto unsupported;
>  	}
>  
> +	/* If it's a FC0012, we need to bring GPIO5/RESET
> +	   out of floating or it's not going to show up.
> +	   We set GPIO5 to an output, enable the output, then
> +	   reset the tuner by bringing GPIO5 high then low again.
> +
> +	   We're testing this last so that we don't accidentally
> +	   mess with other hardware that wouldn't like us messing
> +	   with whatever is connected to the rtl2832's GPIO5
> +	*/
> +
> +	/* close demod I2C gate */
> +	ret = rtl28xxu_ctrl_msg(adap->dev, &req_gate_close);
> +	if (ret)
> +		goto err;
> +
> +	/* Set GPIO5 to be an output */
> +	ret = rtl28xx_rd_reg(adap->dev, SYS_GPIO_DIR, &val);
> +	if (ret)
> +		goto err;
> +
> +	val &= 0xdf;
> +	ret = rtl28xx_wr_reg(adap->dev, SYS_GPIO_DIR, val);
> +	if (ret)
> +		goto err;
> +
> +	/* enable as output GPIO5 */
> +	ret = rtl28xx_rd_reg(adap->dev, SYS_GPIO_OUT_EN, &val);
> +	if (ret)
> +		goto err;
> +
> +	val |= 0x20;
> +	ret = rtl28xx_wr_reg(adap->dev, SYS_GPIO_OUT_EN, val);
> +	if (ret)
> +		goto err;
> +
> +	/* set GPIO5 high to reset fc0012 (if it exists) */
> +	ret = rtl28xx_rd_reg(adap->dev, SYS_GPIO_OUT_VAL, &val);
> +	if (ret)
> +		goto err;
> +
> +	val |= 0x20; 
> +	ret = rtl28xx_wr_reg(adap->dev, SYS_GPIO_OUT_VAL, val);
> +	if (ret)
> +		goto err;
> +
> +	/* bring GPIO5 low again after reset */
> +	val &= 0xdf;
> +	ret = rtl28xx_wr_reg(adap->dev, SYS_GPIO_OUT_VAL, val);
> +	if (ret)
> +		goto err;
> +
> +	/* re-open demod I2C gate */
> +	ret = rtl28xxu_ctrl_msg(adap->dev, &req_gate_open);
> +	if (ret)
> +		goto err;
> +
> +	/* check FC0012 ID register; reg=00 val=a1 */
> +	ret = rtl28xxu_ctrl_msg(adap->dev, &req_fc0012);
> +	if (ret == 0 && buf[0] == 0xa1) {
> +		priv->tuner = TUNER_RTL2832_FC0012;
> +		rtl2832_config = &rtl28xxu_rtl2832_fc0012_config;
> +		info("%s: FC0012 tuner found", __func__);
> +		goto found;
> +	}
> +
>  unsupported:
>  	/* close demod I2C gate */
>  	ret = rtl28xxu_ctrl_msg(adap->dev, &req_gate_close);
> 

…sorry for the delay,
After applied patch no luck - in attach is dmesg for working original
Realtek driver(dvb_usb_rtl2832), and second one(dvb-usb-rtl28xxu)
rtl2832 part by Thomas with tuner issue, still not working.
Most intriguing is tuner get stucked by tuning(t-zapping)!

Cheers,
poma

ps.
Thank you so far, really thorough explanation!