• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 100 subscribers
  • Views 442 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

OMAP3530 new customer HW , hanging when uncompressing kernel : HELP

Colombo Carlo
Mastermind 24120 points
Other Parts Discussed in Thread: AM3517

Hi Champs ,

please my customer just got latest pcb of his board ( beagleboard like , differnet memories , no POP ) .

now we are string hw debugging . Uboot is up and runningOK, we can also do fatload of big files in RAM and everythings seems OK ( we crossc checked with emulator ) .

Now  when we launch the Kernel uncompressing from Uboot, it starts and the in the middle 9 ( after few ..... )  it hangs ...

we connected the emulator and we can see a Data abort exception for th CORTEXA8 .

please  what do you suggest ?   could it be a wrong kernel compoiling or it must be related to something wrong in HW ? 

I asked them to load ITBOK  and  do checks , but in the meantime  any idea / suggestion ?

thank you very muhc

bye

Carlo

  • 0
    TI__Expert 8180 points

    Hi Carlo,

    I do remember facing similar issue while AM3517 bring-up, and it came out as a hardware DDR interface issue. As you mentioned, u-boot comes up without any issues; but the difference here is, u-boot doesn't enable caches/MMU but when kernel starts first thing it does is enable MMU and Cache.

    Can you disable both instruction and Data caches, please follow below steps -

    De-compressor -

     - Go to the file "arch/arm/boot/compressed/head.S" and search for string "cache_on".

     - Comment this while line, this would enable de-compressing without caches.

     

    Kernel -

    - Open menuconfig ("make menuconfig")

    - Go to the "System Type" and enable both the options

     [ * ] Disable I-Cache (I-bit)                                                     

     [ * ] Disable D-Cache (C-bit)   

     

    If the issue is really related to cache, then definitely this should give us some pointers.

    Thanks,

    Vaibhav

  • 0 in reply to Vaibhav Hiremath
    TI__Mastermind 24120 points

    Hi Vaibhav

    thank you very much for your precious inputs .

    please  back to your hw  what was the issue ?  a missing connection to  RAM ?  bad  routing ?  how did  you find it ?

    as you can imagine we are in deep stress so anything from your experience  is evry evry welcome

    thank you very veyr much

    bye

    carlo

  • 0
    TI__Guru* 86700 points

    Some things to try:

    -When you get the data abort, see if you can read/write DDR memory successfully with the emulator.  If the data is not stable, you probably need to adjust your DDR AC timings for the memory you have chosen.  Can you post the memory you are using, its configuration (x32, size, addr bits, etc), and the SDRC register dump?

    -Check power across the board.  Ensure you have adquate supply, no dropouts, noise, etc on all rails, especially 1.8V

    Regards,

    James

  • 0 in reply to JJD
    TI__Mastermind 24120 points

    Hi James

    thank you very very much

    I ll try and be back

    bye

    Carlo

  • 0 in reply to Vaibhav Hiremath
    TI__Mastermind 24120 points

    Hi  Vaibhav 

    we did the test and  now  the  kernel  can complete uncompressing ( " done "  message  on Hypertermnila  )   but once it starts  it hangs ...

    we  cross checked with emulator  and  memory looks  fine .   We tested  also  filling complete  memory with random data and reading back  with emulator  and everything is fine ...

    please  what is  your feeling ? 

    I am asking also  details  on momory and data as  very precious  post ...

    big trouble is  consultnat who made  their xloder is no more there   so  they are not 100% in control now ...

    what can I test to be sure  HW is OK ,  and  be sure it is "just" a sw issue ?

    do yo have any idea/test  to perform ?

    any idea  is very welcome

    bye

    Carlo

  • 0 in reply to Colombo Carlo
    TI__Expert 8180 points

    Hi Carlo,

    What did you fix to get through Uncompressing step? What was the root-cause?

    Since now the Uncompressing kernel is going through well, I suspect kernel might be crashing in very early stage of booting, you can try following steps -

     

    - Build the kernel without any drivers, try booting.

    - Put a infinite while loop Or breakpoint in your board_init function (should be under arch/arm/mach-omap2/board-<your board file>.c) then do step-by-step execution of code to figure out where exactly it is crashing.

     

    Thanks,

    Vaibhav

  • 0 in reply to Vaibhav Hiremath
    TI__Mastermind 24120 points

    Hi Vaibah ,

    thank you  let me try

    the memory they use is MT46H64M32LF   Micron . 

    any  suggetsion or settings  to copy ?

    thank you

    bye

    Carlo

  • 0 in reply to Colombo Carlo
    Prodigy 140 points

    hello,

    I am facing problem similar to this.

    my board is booting but when I am typing commands to test u-boot it hangs and jumps to 0x0001408C

    I am using cortexA8 am3517 processor and Micron MT29C4G48MAZAPAKQ-5IT chip

    what are the common reasons for this ??

    what test I should do so that I can figure out the problem ??

    any suggestion will be helpful for me.

    thanks