2022-10-25
12:25 PM
- last edited on
2024-02-15
09:44 PM
by
Lina_DABASINSKA
version 4.19.1
platform: CORTEX_M7 + IAR 8.X
file name: touchgfx_core_release.a
file location: \4.19.1\touchgfx\lib\core\cortex_m7\IAR8.x
can we build the .a file? if not, have you disable asset() in release version? or if you still prefer to keep assert() in release version, please add --no_path_in_file_macros in IAR option, because not any end user will interested the path of c:\jenkins\workspace\... which increase code size of application.
Thanks
Richard L.
2022-10-25 01:43 PM
Is there a question? Are you asking how to get in CubeIDE (gcc) the same effect as with the IAR compiler option?
If so - gcc has a similar option since v.8.
2022-10-25 02:04 PM
sorry the question may not clear,
Thank you for the quick response.
Richard L.
2023-11-22 02:05 PM
@RichardLC - Hi Richard, did you find a resolution to this. I'm struggling with the same issue as I would like to generate the same binary file on different machines for QA purposes. We're using IAR as well, but the touchgfx_core_release.a file makes this impossible.
Thanks
2023-11-22 02:55 PM
Isn't this like a debug record in the ELF object/library, so it can associate a break point or trace back to a specific file / function / line ?
I can see leaking the data into a final binary image is problematic*, but is that what we're talking about, or assert()s with __FILE__ designations? If it's static data, then you need to get the library rebuilt cleanly, from whomever owns it.
*Been doing forensics for 30 years. I clean and strip a lot of my stuff. Lot of *** removed from .ELF's I post, makes them a lot smaller. https://github.com/cturvey/stm32extldr/blob/main/u5_w25q128/CLIVEONE-W25Q128_STM32U545-PB10-PA4-PB1-PB0-PA7-PA6.stldr