Other Parts Discussed in Thread: HALCOGEN, , LAUNCHXL2-570LC43
My main symptom is in the CCS console any printf statement results in the following response:
"Invalid CIO command (0). Memory map prevented reading 0x0800970C"
This memory location corresponds to the CIOBUF section that I believe is temporary data used by printf that the debugger reads to display the printf statement. (See Below)
During my debugging efforts, I've noticed that depending on the HalCoGen-configured User-Stack size and binary size that this issue presents whenever this CIOBUF section falls above 0x0800_8000. My theory is that this may be a configuration setting on the Debugger (perhaps a GEL file?) and not the processor/HAL code.
Miscellany
- Disabling the MPU any various nMPU regions has no impact on this issue
- This issue appears intermittent based on application binary (Perhaps smaller stack allocations happen to result in better chance of the CIOBUF symbols from falling under 0x0800_8000)
- No HalCoGen MPU or RAM settings seems to correspond to the observed 0x0800_8000 boundary.
- I'm building my binary outside of CCS but am using CCS to flash the binary and use the debugger - perhaps a setting is getting lost between the build step and debugger configuration used by CCS?
If this is a debugger setting any instructions on parameters or files to investigate would be helpful. Thanks!
my HL_sys_link.cmd file:
