Part Number: MSP430FR2355
Good morning,
I encountered a strange behaviour during the debugging process of my firmware application on a MSP430FR2355 using Code Composer Studio 20.3.1.5.
That's the code (note that I placed those ASM comments just as a marker in order to find it easily in the assembly code):
__asm(";------------- BEGIN list_for_each_entry_safe"); // marker
list_for_each_entry_safe(event, tmp, &gpt_internal.used_events_queue, head) { // <=== I have the issue here
__asm(";------------- END list_for_each_entry_safe"); // marker
// rest of the code...
}
The list_for_each_entry_safe is the same as linux/list.h (v2.6.32).
When debugging the list_for_each_entry_safe body with TI v21.6.1.LTS I saw an odd behaviour situation:
the event variable, at the first iteration, contained invalid data, causing the firmware to crash (PC=0x0004).
Needles to say: all the structures were correctly initialized.
The same code, with TI v16.9.4.LTS, produced the expected results.
After some debugging I could have find out the source of the issue: the increment part of the for loop is being evaluated BEFORE each execution of the loop body and not after. That could explain the issue encountered.
If needed I can provide the generated ASM source files of that particular module (with TI16 and TI21 both).
Here are the compiler flags:
-vmspx --use_hw_mpy=F5 --include_path="${CCS_BASE_ROOT}/msp430/include" --
include_path="${CG_TOOL_ROOT}/include" <project include dirs> --advice:power=all --advice:hw_config=all -
-define=__MSP430FR2355__ --define=_FRWP_ENABLE --define=_INFO_FRWP_ENABLE -g --
printf_support=minimal --diag_warning=225 --diag_wrap=off --display_error_number --
silicon_errata=CPU21 --silicon_errata=CPU22 --silicon_errata=CPU40 --src_interlist --
gen_preprocessor_listing.
The optimization flags were the same for both compilations, I was only changing the compiler version in the project settings.
Did I miss something? Or did I do something wrong?
Thank you in advance.
Regards,
