Floating point arithmetic in the kernel

In kernel space, I need to perform matrix multiplication whose elements are floats (say 0.8666, etc.). Is it possible to do floating point arithmetic in kernel space. I am using 3.12 cores on an ARM based board.

+3


source to share


1 answer


I understand that the kernel is always related to FPU emulation, so this does work, but it is almost always wiser to use scaled integers.



+1


source







All Articles