On 19 September 2017 15:28:42 BST, Patrick Venture <venture@xxxxxxxxxx> wrote: >+linux-iio > >On Mon, Sep 18, 2017 at 10:45 PM, Joel Stanley <joel@xxxxxxxxx> wrote: >> >> Hey Patrick, >> >> On Tue, Sep 19, 2017 at 2:55 AM, Patrick Venture <venture@xxxxxxxxxx> >wrote: >> > I have a system where there is a GPIO gating whether a line to the >ADC is >> > active. I'm routing the ADC channels through iio-hwmon. >> > >> > I'm considering that it might be prudent to tell iio-hwmon that it >needs to >> > set the GPIO and wait some configurable amount of time before >trying to read >> > that channel. >> > >> > >https://github.com/torvalds/linux/blob/master/drivers/hwmon/iio_hwmon.c#L43 >> > >> > I'm curious if there are strong thoughts about whether that kind of >thing >> > belongs in that level or whether it should be pushed to the >aspeed_adc >> > driver. >> >> That's a tricky one. >> >> I'd suggest the code to flip the GPIO should go in the aspeed_adc.c, >> and failing that in the iio subsystem, not in the hwmon bridge. > >Ok, so I can extend the properties of the aspeed_adc to support >specifying a GPIO per channel, and then add the logic to handle this >possible per channel configuration. I had to ask, but I assume >there's some standard to how these things are specifying in the >device-tree. For instance, each fan has its own configuration under >the aspeed-pwm-tacho driver, so arguably it could be channel@{index} >to specify for whichever have such a requirement. Wait up. No this doesn't belong in the ADC driver unless it is a feature of the ADC itself. You need to represent the hardware doing the gating explicitly. > >> >> I assume this GPIO needs to be switched at runtime (ie, it's not >> something that is muxed one way or the other at boottime and left >> static)? > >Correct. > >> >> Is it a case that the GPIO should be asserted while the ADC is >> sampling, and then deasserted afterwards? > >That is also correct. There is the mux framework (just merged recently) for doing exactly this. Can send links later but on phone only today. J > >> >> > Also, please let me know if I'm not reaching the correct people for >this >> > particular discussion. >> >> You could instead go to the iio list. That would be better as others such as Peter Rosin who wrote the mux framework may reply. Jonathan >> >> Cheers, >> >> Joel -- Sent from my Android device with K-9 Mail. Please excuse my brevity. -- To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
