• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 1428 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.

Undefined symbols when including SafeTI library

AG
Prodigy 40 points

Other Parts Discussed in Thread: HALCOGEN

I'm trying to include the SafeTI library in a project but I get an undefined symbols error when including sl_types.h, sl_api.h, sl_priv.h:

Error:

undefined first referenced
symbol in file
--------- ----------------
StackModeABT ./source/sl_asm_api.obj
StackModeFIQ ./source/sl_asm_api.obj
StackModeIRQ ./source/sl_asm_api.obj
StackModeSVC ./source/sl_asm_api.obj
StackModeSYS ./source/sl_asm_api.obj
StackModeUND ./source/sl_asm_api.obj

I tried following the instruction from an earlier posting which advised to add a call to SL_Init_StackPointers(); in the HL_sys_startup.c function but this didn't really solve the problem.

Any help would be appreciated.

  • 0
    TI__Mastermind 49120 points
    Hello,

    These are global variables defined in the sl_asm_api.asm file and need to be defined by the application as the base addresses for the different stacks. The default stack sizes are defined in the sl_config.h header file.

    Regards,
    Sunil
  • 0 in reply to Sunil Oak
    Prodigy 40 points

    Hi Sunil,

    Thank you for your reply. How and where in the code would I implement the definition of those global variables? If I call SL_Init_StackPointers in the main() function the linker still won't see it and fail linkage due to undefined symbols. I'm using Halcogen and CCS.

    Thanks for your help.

  • 0 in reply to AG
    TI__Mastermind 49120 points
    There are multiple ways you can do this. You can do simple #define to define each of these symbols. This has the limitation of the stack base addresses being hardcoded inside a source or header file. A better way to define these symbols is to use the linker command file.

    You need to define a STACKS region:
    STACKS (RW) : origin=0x08000000 length=0x00001800

    Then you can assign the stacks to this region as shown below (note the stack pointer assignments):

    .intvecs : {} > VECTORS
    .text align(8) : {} > FLASH0 | FLASH1
    .const align(8) : {} > FLASH0 | FLASH1
    .cinit align(8) : {} > FLASH0 | FLASH1
    .pinit align(8) : {} > FLASH0 | FLASH1
    .STACK_DATA_svc : {. += 1024;} > STACKS, RUN_START(StackModeSVC)
    .STACK_DATA_fiq : {. += 1024;} > STACKS, RUN_START(StackModeFIQ)
    .STACK_DATA_irq : {. += 1024;} > STACKS, RUN_START(StackModeIRQ)
    .STACK_DATA_abt : {. += 1024;} > STACKS, RUN_START(StackModeABT)
    .STACK_DATA_und : {. += 1024;} > STACKS, RUN_START(StackModeUND)
    .STACK_DATA_sys : {. += 1024;} > STACKS, RUN_START(StackModeSYS)
    .bss : {} > RAM
    .data : {} > RAM
    .sysmem : {} > RAM

    Regards,
    Sunil
  • 0 in reply to Sunil Oak
    Prodigy 40 points
    Thanks Sunil, that worked. I have also tried to use the SafetyLib example that came with Halcogen. The code compiles and runs alright but no matter which test I turn on (e.g. SL_SelfTest_SRAM) it returns 0 which forces the code into a while(1) loop.
    I assume that is not the expected behavior. Is there a configuration I'm missing?
  • 0 in reply to AG
    TI__Mastermind 49120 points
    How is the SRAM region configured in the MPU? Can you configure this to be a write-through region? I think it is a write-back cacheable region by default (as configured by HALCoGen).
  • 0 in reply to Sunil Oak
    Prodigy 40 points
    Region 3 seems to be the SRAM region as it extends from 0x08000000 to 0x0807FFFF. This region is of type: NORMAL_OIWBWA_NONSHARED
    and permission: PRIV_RW_USER_RW_NOEXEC. None of the sub-regions are disabled. I changed that to type NORMAL_OIWTNOWA_NONSHARED but the problem persists.
    I debugged this problem a little more and SL_SelfTest_SRAM() seems to return false because nERROR is active (SL_ESM_nERROR_Active() returns true). I could reset it by adding SL_Clear_nERROR() calls. What surprised me was that the esmCallBackFunction in the example never gets called, I had expected to see it executed during each test, is it normal that it doesn't?