some progress

Ed Sutter edsutterjr at
Wed Jul 1 10:44:12 UTC 2015

That's perfect, cause something in that code isn't complete.
My guess is that it has to do with initialization more fundamental
than the UART itself, like clock management or power control.
When I get this stuff cleaned up I'll send it over.
> Ed,
> This is great progress!  Since the time you gave me the diff that
> updates the files up until now, I was mostly walking through what you
> had updated and was working on mostly UART-specific stuff in cpuio.c.
> On Tue, Jun 30, 2015 at 8:55 PM, Ed Sutter <edsutterjr at> wrote:
>> Jarielle,
>> I made some progress regarding boot from sd...
>> I took a step back and decided to get the LEDs initialized.
>> Doing that allows me to work with a simpler device to verify startup.
>> Anyway, I managed to get the GPIO pins initialized correctly so
>> that I could control the LEDs.  Then I built a SD card with the
>> proper prefixes etc.. in there and was able to boot up the led
>> pattern as expected.
>> That tells me that it boots correctly, but is not initializing the UART
>> as it should. On the other hand, since our UART-boot does work, we can
>> at least say that the code in cpuio.c currently is NOT breaking the
>> uart initialization that the rom-bootloader does for us.  So, my hunch
> I would say so as well.
>> is that some clock that drives the UART is not being turned on in our
>> code (but it is turned on when the bootloader is forced to boot through
>> the UART).
>> Anyway, that's my state at the moment.  Done for the night.
>> The one tricky part that I'm glad to have out of the way is the prefix
>> needed at the base of the image.  That wasn't obvious to me.
>> Ed
> Now I am going to try to replicate your approach when trying to get
> UART to work.
> I am the one who needs to catch up to you now. :-)
>> _______________________________________________
>> umon-devel mailing list
>> umon-devel at

More information about the umon-devel mailing list