• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 101 subscribers
  • Views 542 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

C6455 EMIF FLASH boot

Bryan Busacco
Expert 1215 points

With respect to the C6455 and the previous question regarding booting from FLASH in 8-bit mode than transitioning to 32-bit modeon the EMIF:

If I use the lower bytes in FLASH at first ( inital boot in 8-bit mode ) and then reconfigure to use 32bits, the byte select lines that are used for the lower address in 8 bit mode,  will still be connected and would always be 0 when I transition to  32bit mode. This means that only every 4th word would be accessible  still making ¾ of the flash useless.  Please verify I was not missing something. How do I get the DSP to first boot in 8-bit mode, then become FLASH efficient ( using all of the FLASH rather than every 4th word ) when I transition to 32-bit mode?

Also, is there a utility for programming the flash from the EMIF?  An application note or example code would be helpful.

 

Thank you,

Bryan Busacco.

  • 0
    TI__Mastermind 30825 points

    Bryan,

    I believe Bernie's response was stating was to program a small portion of the code at the start of the Flash to be 8-bit and 3/4 of this would be wasted.  (i.e. First byte of every word is data, there other 3 bytes 0's) after the portion of the memory that has is for changing the EMIF initialization to 32-bit mode is done, then the rest of the memory can be full 32-bit used.  You'd probably want to treat this as two separate builds.  That said, I'd suggest if this is done, to have the 8-bit code copy to local L2 memory, and execute from the local L2 memory the reconfiguration of the EMIF to 32-bit mode, then jump to the EMIF for the 32-bit execution.  This will ensure no issues of potentially read access of the Flash being corrupted while re-configuring the EMIF (i.e. you don't want to be using it while you're re-configuring it.)

    The better alternative in my opinion, which was also offered by Bernie, of doing this would be to perform the dual boot operation using I2C for the initial boot sequence to configure the Flash, at the end branch to EMIF 0xB000 0000 address for the Flash boot.

     

    I'll check on a Utility/Example Code for programming flash from EMIF.

    Best Regards,

    Chad