Open chuckb opened 4 years ago
hi @chuckb , change your Floating Point setting to "Software Implementation" could solve your problem for now.
@mpaland thanks for your great library, please fix the bug.
Is it possible to get context of crash?
gdb
backtrace + other printed info could help ...
This image shows Cortex-M4 processor status after vsnprintf function executed.
Program walk-through functions: vsnprintf_() -> _vsnprintf() -> signal handler called at 0xfffffff9 -> stuck at WWDG_IRQHandler()
This error may contain useful information: Develop Log System needs property 'FPU' fixed. Fixing from 'com.atollic.truestudio.common_options.target.fpucore.DUMMY' to 'com.atollic.truestudio.common_options.target.fpucore.fpv4-sp-d16'.
This is a bit strange - Window Watchdog IRQ is raised, either in place unrelated to FPU or frame for _ftoa is missing.
Cay you try compiling with -O0
to disable optimizations?
Also stepping through _ftoa
may help.
Or you can try returning from IRQ (breaking the infinite loop) - use instruction stepping, change PC to same 'BX LR' instruction and step through it (I hope this works on ARM A processor)
I am compiling with gcc for ARM target...raspberry Pi zero. When I attempt to pass a float, I get a crash. CPU start up code enables the VFP with:
My compile statement is like the following:
My link statement is:
Can you help me with the magic incantation required to get this to work?