You could write a memory compare check in MLO to see that the file is indeed loaded correctly to help debug the problem. I wrote a routine to write the memory address to each cell in memory and then compare. Not totally comprehensive for stuck bits, but gave a warm fuzzy that memory was somewhat ok. Writing all 1's and then all 0's and compare would be another memory test.
On Wednesday, November 29, 2017 at 2:28:03 PM UTC-7, [email protected] wrote: > > Hi, > > I'm currently trying to modify the (custom) MLO of my bare metal > application in order to work with the PocketBeagle. At the moment this MLO > works fine with BBB. > > According to the (incredibly small) manual of the Octavo SoC I changed the > initialisation of the DDR and EMIF registers - but this does not seem to be > everything which is necessary. > > Now with these register settings my MLO successfully loads the main > application into the DDR RAM but when the SoC jumps to the start adress of > it, it stalls. So it seems there is still a problem with the RAM > initialisation. Either writing the application to the RAM does not work or > execution of the loaded application is not possible. > > What I can see at the moment: the BBB initialisation of the RAM contains a > lot more than shown in the Octavo manual, there are a lot of idle and > shut-down states set which are not mentioned in Octavo manual. > > So...any idea what else is different in initialisation of BBB and > PocketBeagle? Are these additional idle/shut-down register settings needed > too? > > Thanks! > > -- For more options, visit http://beagleboard.org/discuss --- You received this message because you are subscribed to the Google Groups "BeagleBoard" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/beagleboard/89126aef-1c13-4a93-9607-10d8cedb7739%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
