Hello Alan,
thanks much for testing.
I have rebased RPi related changes to the actual master
head and rebuild toolchain to Newlib 2.4.0.20160527 and
GCC 6-20160526. The test on RPi1 from U-boot goes
OK so I have pushed changes to the master.
On Monday 30 of May 2016 23:04:36 Alan Cudmore wrote:
> Here are the tests I was able to run:
> In general, the HDMI console works on the Model A+ ( should also work on
> the Model B/B+ ) and the Pi Zero.
>
> I still have firmware issues with the Pi 2, where it only seems to work
> with my older RTEMS SD boot disk. To make matters worse, I’m not sure what
> version of the firmware that is.
This would worth to add debug prints and find where boot stucks.
I can prepare that on my branch or GitHub but I have this week
so busy that I do not get to such reparation now.
> With the old firmware, the Pi2 image boots fine with the UART, and it
> almost works with the HDMI console.. The console output seems to get
> corrupted ( see attached picture )
This looks like some memory corruption or missing flush
of data from the cache. May be some attributes for framebuffer
mapping by MMU should be changed. I try to think about that.
> I have also attached a PDF that hopefully describes the tests I did.
>
> I’m confident that things are working well with the single core Pi models
> with the latest firmware and console HDMI output. I think it’s worth
> getting the code in and trying to solve the Pi2 issues.
Thanks again and best wishes,
Pavel
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
