Yes I agree that since increases the delay fixes this issue the problem is 
caused by:

"some sort of startup delay".

However you are pointing to a grander problem:

As the DE0_Nano_SoC is flaged EOL with the DE10_Nano (with hdmi) set as 
replacement
framebuffersupport would be something hard to keep out and.

by the time the UIO devices show up.  One thing that could be causing 
> problems is if the FPGA is programmed by U-Boot and then reprogrammed 
> by the Linux kernel.  If anything was talking to the FPGA fabric (like 
> the HDMI interface or any of the other "extra" logic in the example 
> designs and not just the HM2 logic) that could trigger BadThings up to 
> and including a kernel panic or system halt. 


REmoving the devicetree requirement cleared being able to start machinekit 
without having to reconfigure the fpga on MK startup,,, fine.

But I have tested what happens when you have the display running (linux 
desktop) and then try to reconfigure the fpga with the same bitfile as 
loaded by uboot.

Result is a blank screen....
THis illustrates that loading/running a different bitfile / machinekit 
configuration will allways require a reboot when using the local soc 
display.

1 --> It may be possible to remedy this by reloading the alt_vip_fb driver 
(loaded as module), and resetting / restarting the x-server. ?

2 --> I have spent some time studying the partial-fpga-reconfiguration docs 
and it looks doable for me to create a partitioned design with.
the main partition containing the display and other qsys cores + - the uio. 
and a freezable bridge 
+ 1 new partition for each hm2 config.



On Thursday, 31 August 2017 23:54:04 UTC+2, Charles Steinkuehler wrote:
>
> On 8/31/2017 4:01 PM, Michael Brown wrote: 
> > THat said .. if the signature or board name check fails this should not 
> > cause any kernel panics requiring a reboot. 
>
> Agreed! 
>
> I haven't replicated your results here so I'm not 100% sure what's 
> happening, but from my experience the FPGA is programmed and running 
> by the time the UIO devices show up.  One thing that could be causing 
> problems is if the FPGA is programmed by U-Boot and then reprogrammed 
> by the Linux kernel.  If anything was talking to the FPGA fabric (like 
> the HDMI interface or any of the other "extra" logic in the example 
> designs and not just the HM2 logic) that could trigger BadThings up to 
> and including a kernel panic or system halt. 
>
> It's also possible there's a startup delay in the FPGA reset logic 
> that's causing the problem.  The FPGA may be "running", but still 
> being held in an internal reset condition. 
>
> -- 
> Charles Steinkuehler 
> [email protected] <javascript:> 
>

-- 
website: http://www.machinekit.io blog: http://blog.machinekit.io github: 
https://github.com/machinekit
--- 
You received this message because you are subscribed to the Google Groups 
"Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
Visit this group at https://groups.google.com/group/machinekit.
For more options, visit https://groups.google.com/d/optout.

Reply via email to