Using FPU of nrf 52

RE: Using FPU of nrf 52

jcormier — Tue, 18 Apr 2017 15:16:35 GMT

Thanks, will check it out.

RE: Using FPU of nrf 52

Sigurd — Tue, 18 Apr 2017 14:42:05 GMT

Yes, you have the linker flag called LDFLAGS += -mfloat-abi=hard -mfpu=fpv4-sp-d16, and CFLAG CFLAGS += -mfloat-abi=hard -mfpu=fpv4-sp-d16 and also might need to add these ARM_MATH_CM4 flags:

CFLAGS += -DARM_MATH_CM4
ASMFLAGS += -DARM_MATH_CM4

Take look at e.g. the makefile for the example, located in the folder SDK_folder\examples\peripheral\fpu_fft\pca10040\blank\armgcc\

RE: Using FPU of nrf 52

jcormier — Tue, 18 Apr 2017 14:35:27 GMT

Sigurd, Does this "Use single precision" option have a similar option for building with GCC?

RE: Using FPU of nrf 52

Sigurd — Tue, 18 Apr 2017 14:31:08 GMT

Hi,

There is a FPU example in the SDK. It's located in the folder

<SDK_InstallFolder>\examples\peripheral\fpu_fft

In Keil the Floating Point Hardware option must be set to "Use Single Precision":