Hi all,
on our new Controller module where we use the QorIQ LS1 device we had a strange behavior!
After Power ON the BL starts normally and shows CPU revision LS1021A (DCFG_CCSR_SVR register is 0x87081120). But after booting our OS and then rebooting from OS/Linux the bootloader seems to be running on different CPU???... The BL shows that CPU is LS1020 and DCFG_CCSR_SVR register read is 0x87081020. We also stopped in BL CLI after reboot and manually read the register as a memory – it shows different value!
All that is OK according to documentation, but why on Earth the hard-coded CPU revision and personality is changing? Could that be, that the POR processor IO state is wrong and this is causing this weird behavior? Or simply a new silicon bug?
We realy appreciate your response.
Best regards and thank you in advance,
Christian
Hello Ufedor,
thank you very much for your help and the rapid answer ... ;-)!
Best regards and thank you in advance
Christian S.
Please note that IFC_A16 must be pulled down and IFC_A17 must be pulled up during power-on reset for the LS1021A.
It is required to check levels on these signals during POR.
If further issue investigation will be needed please consider to create a technical case:
https://community.freescale.com/thread/381898
and provide the processor connection schematics for inspection.
Hello Ufedor,
thank you very much for your help and the rapid answer!
The IFC_A16 Pin has an pull down but there isn't a pull up at IFC_A17! Where I can find in the documents this requirement?
Best regards and thank you in advance
Christian S.
http://cache.nxp.com/files/32bit/doc/app_note/AN4878.pdf
page 24