[Letux-kernel] JZ4730 kernel support

[H. Nikolaus Schaller]

Hi Paul,

> Am 15.10.2019 um 18:13 schrieb Paul Boddie <paul at boddie.org.uk>:
> 
> On Tuesday 15. October 2019 09.57.05 H. Nikolaus Schaller wrote:
>> 
>> Thanks! I have applied it and it compiles with the letux_defconfig. So I
>> will include it in the next -rc build.
>> 
>> One tiny note: would be good if there is a signed-off, which is a
>> requirement for upstreaming.
> 
> I'll specify the --signoff option to git format-patch this time.

Or you can use git commit -sm 'message'

> 
>> I have also scanned the source tree for '#if 0' and there is one in
>> arch/mips/jz4740/time.c
> 
> So, I had to take a look at the JZ4740 documentation and my modifications to 
> the Fiasco microkernel to remind myself of how the JZ4730 does things. There 
> are a few differences that require the code to be changed rather than 
> parameterised, and I suspected that there would be other places where similar 
> corresponding changes need to be made.
> 
> Sure enough, I found the drivers/clk/ingenic/tcu.c file containing the timer 
> control functions. Due to the differences in the JZ4730, I ended up having to 
> add separate register definitions in a new file:
> 
> include/linux/mfd/ingenic-tcu-jz4730.h

Good!

> The old jz4730_timer.h header file might conceivably be eliminated at some 
> point given that the above clock driver code and definitions do largely the 
> same thing. This would affect the arch/mips/jz4740/timer.c file, of course.
> 
> There should still be places where the timer interrupts are handled, but I 
> didn't find these just now when quickly searching for possible locations.
> 
> [...]

Ok, that needs more work.

> 
>> Well, I should definitively find time to set up a testing environment...
>> Now as I have achieved a breakthrough with OMAP-PVR/SGX, I hope I can
>> reprioritize some things.
> 
> I read the remarks about PowerVR and saw the article you referenced where 
> someone had managed to configure their OMAP-based system to use the drivers 
> again, so I guess there is hope for this.

I now have it running on beaglebone black (am335x) and pyra (omap5) and
almost on GTA04 (omap3). The latter successfully loads the firmware to
the SGX but fails with a memory mapping problem.

> Not that this affects the Minibook 
> directly, though. ;-)

Well, it negatively affects this project by taking away time :)

> 
> P.S. The cumulative patch (including the previous patches) has been attached, 
> with signed-off, to this message.<patch-48beba89-2.diff>

Thanks! unfortunately I can't apply it... On which commit hash did you base it on?

BR,
Nikolaus