I'm using MCUXpresso IDE v11.2.1 and made a project for LPC_11xx with lpcopen and CMSIS-DSP.
when I change the optimization to size i get a hardfault when it calls the CMSIS-DSP arm_fir_fast_q15 function.
Seems to work at the -O3 optimization.
Solved! Go to Solution.
I found out that the fir data has to be aligned with 4 bytes:
static __attribute__((aligned(4))) q15_t firStateQ15[BLOCK_SIZE + NUM_TAPS - 1];
Hello jespermadsen,
You can run into that function, and check the disassemble to confirm which code
made hardfault.
Regards,
Alice
I found out that the fir data has to be aligned with 4 bytes:
static __attribute__((aligned(4))) q15_t firStateQ15[BLOCK_SIZE + NUM_TAPS - 1];