AnsweredAssumed Answered

LS1046A broken RCW JTAG recovery issue

Question asked by Robert Thorhuus on Nov 29, 2018
Latest reply on Dec 3, 2018 by ufedor



We have a custom board LS1046A design and I broke the RCW which we have in QSPI flash.

I read the RM stating to use the RCW SRC 0x9F and then "just" re-flash the RCW. It says nothing about where the RCW could be, just do not start cores or DDR. Just do it. Reading that you can get the impression that nothing else is needed. But I would be surprised since reading the hard coded RCW values QSPI is not setup.


Then going on the the Target manual for Code Warrior it points out to use the OVERRIDE feature in the JTAG init script following a correct target connection setup. But this does not work. I get the processor to halt but when I then start the flash programmer it gives error when trying the execute the "cw_reset" command in the python scripts.


Just to be clear on one thing: Even though our RCW is totally different from the hard coded 0x9F version I should not need to change the RCW and still be able to flash the QSPI?


There are a lot of open questions not answered in the document. Here is two:

* If you choose a hardcoded RCW, 0x9F for instance, it sounds like you get a certain "profile" of RCW values in all 512 bits. It also sounds that you cannot change a single one of them. But then in the scripts you DO change them. That is contradictory??!!

* In the RM it is stated that you should not run cores or init DDR if using the hardcoded RCW. But in the scripts you DO!!??


If you have any light to shed here I would be very glad since it would benefit our board bring up turn around times greatly