RE: [PATCH 0/6] staging: iio: tsl2x7x: staging cleanups

[Jon Brenner <Jon.Brenner@ams.com>]

Hi Brian,
Not sure about the DTS interrupt setting you mentioned but would like to know how you resolve - as I am planning on getting a Pi0w for some 'home personal projects'.

Anyway (and I am sure you probably already know this - but just in case) WRT the device registers (on the TSL2771 for example);
If you enable interrupts (ie. Reg#0x00= 0x37) and set the persistence register to 0x00 (is. Reg# 0x0C = 0x00), you should see interrupts occurring every integration period.
That way you can see if you're at least getting them.

You probably know this too - on earlier parts such as these, interrupts were cleared via use of the "command register" special function bits.  (see data sheets for details).
Now a day - interrupts are cleared in a more conventional manner 

Jon

-----Original Message-----
From: Brian Masney [mailto:masneyb@onstation.org] 
Sent: Friday, June 30, 2017 10:21 AM
To: jic23@kernel.org; linux-iio@vger.kernel.org
Cc: devel@driverdev.osuosl.org; lars@metafoo.de; gregkh@linuxfoundation.org; linux-kernel@vger.kernel.org; Jon Brenner <Jon.Brenner@ams.com>; pmeerw@pmeerw.net; knaack.h@gmx.de
Subject: Re: [PATCH 0/6] staging: iio: tsl2x7x: staging cleanups

On Thu, Jun 29, 2017 at 01:03:46PM -0400, Brian Masney wrote:
> This begins my work to clean this driver up and eventually move it out 
> of staging. Driver changes were tested using a TSL2771 hooked up to a 
> Raspberry Pi 2.
> 
> Thanks to Jon Brenner at AMS/TAOS for loaning me some hardware samples 
> to test my driver changes.
> 
> Brian Masney (6):
>   staging: iio: tsl2x7x: add of_match table for device tree support
>   staging: iio: tsl2x7x: remove redundant power_state sysfs attribute
>   staging: iio: tsl2x7x: remove tsl2x7x_i2c_read()
>   staging: iio: tsl2x7x: cleaned up i2c calls in tsl2x7x_als_calibrate()
>   staging: iio: tsl2x7x: refactor {read,write}_event_value to allow
>     handling multiple iio_event_infos
>   staging: iio: tsl2x7x: use usleep_range() instead of mdelay()
> 
>  .../devicetree/bindings/trivial-devices.txt        |  10 +
>  drivers/staging/iio/light/tsl2x7x.c                | 316 ++++++++++-----------
>  2 files changed, 153 insertions(+), 173 deletions(-)

Hi Jonathon,

Hold off on applying this series. There are several other formatting warnings from checkpatch with this series that I need to fix. Sorry about the noise. I'd appreciate it if you could at least look at the functionality of my changes and I'll resubmit next week.

I held back several other changes related to the event subsystem until I'm able to properly test my changes. I'm having trouble getting the interrupts to work. I wired the interrupt pin on the sensor to GPIO pin 17 on my Raspberry Pi 2 and added the following section to arch/arm/boot/dts/bcm2836-rpi-2-b.dts for my sensor:

&i2c1 {
        tsl2771@29 {
                compatible = "amstaos,tsl2771";
                reg = <0x39>;
                interrupt-parent = <&gpio>;
                interrupts = <17 2>;
        };
};

I start up iio_event_monitor, and run these commands:

echo 1 > events/in_intensity0_thresh_rising_en
echo 256 > events/in_intensity0_thresh_rising_value

When I shine a light on the ALS sensor, and the reading goes above 256, I do not get any events back from iio_event_monitor.

I'm honestly not sure about the 2 in the 'interrupts = <17 2>;' line.
I looked at how interrupts were setup in device tree overlays in the official Raspberry Pi kernel and tried several variations from there.
I have more reading to do. :)

Thanks,

Brian