Re: Meson I2C driver working

[Robert Swindells <rjs%fdy2.co.uk@localhost>]

Vincent DEFERT <vincent.defert%posteo.net@localhost> wrote:
> I have updated https://www.defert.com/meson-i2c.tgz accordingly.
>
> I have used an external 24C256 for my tests.
> I tested both 100kHz and 400kHz and verified the clock frequency with an 
> oscilloscope.
> I also tested both i2c_A and i2c_B, available on the ODROID C2's GPIO 
> connector.
> I used my own test program to make sure my code works with large blocks 
> of data (seeprom(4) works only one byte at a time).

I spotted a few more things, diff at:

<https://ftp.NetBSD.org/pub/NetBSD/misc/rjs/i2c.diff.gz>

Prints this at startup:

mesoniic0 at simplebus4: Meson I2C (100000 Hz)
mesoniic0: interrupting on GIC irq 49
iic0 at mesoniic0: I2C bus

Earlier in the email thread you identified that the interrupt number
is wrong in the Linux device tree sources, your "override file" in the
latest tarball doesn't modify this but does change some other nodes.

We have used device tree overlays to make the device tree match hardware
changes to the default configuration of a board, there are some in
sys/arch/arm/dts/overlays.

I think I would just expect to have the interrupt change in the main
"override file".

Have been using the file here:

<https://ftp.NetBSD.org/pub/NetBSD/misc/rjs/meson-gxbb-odroidc2.dts>

If you don't object to any changes that I have made then I will commit
them.

Robert Swindells