[Letux-kernel] [PATCH 00/20] A bunch of JZ4730 fixups for letux-kernel

[Paul Boddie]

On Wednesday, 13 January 2021 14:46:35 CET H. Nikolaus Schaller wrote:
> 
> Am 12.01.2021 um 22:32 schrieb Paul Boddie <paul at boddie.org.uk>:
> > 
> > I have attached the driver with my own modifications for comparison with
> > previous versions. I cannot promise that it will work, but you should be
> > able to see what I have been trying to do.
> 
> Yes, the idea is good!
> 
> What I have done is to take the latest letux-5.11-rc3 and replace the i2c
> driver with yours (it seems to be based on an older version of what I had
> tried last, but should not matter since interfaces are the same).
> 
> Unfortunately the only effect I can see is the kernel being stuck after
> a timeout:

I actually forgot to include some code that tests the bus conditions after an 
address is sent. So, a slightly updated version of the code is included.

What I left out was a test for the mysterious STX flag which needs to be clear 
before an attempt is made to read data, at least in my experience, and one of 
the ACKF and DRF flags must also be set.

[...]

> What I assume from earlier experiments is that not processing an IRQ doesn't
> work in this case. This makes the interrupt handler being called
> immediately again.

The IRQ should still be handled since all returning from jz4730_i2c_irq 
appears to involve IRQ_HANDLED, but that might not stop the IRQ from occurring 
again immediately, I suppose.

> AFAIR, the kernel should throttle IRQs that are not processed too often
> but the broken sched-timer may stop this function.
> 
> So unfortunately that does not work.
> 
> I think I'll try to make the LCDC work first and then I can focus again on
> the I2C driver. Debugging both in parallel doesn't work well...

Agreed. The I2C functionality isn't particularly important compared to the 
LCD.

Paul