We have developed a custom target using a MCF8548E. Unfortunately, I cannot seem to get floating point arithmetic to work on it right. It does it, but hiccups (I get a signal) and I get a break in 2_math_emu_stub() when adding two floats togther.
I used the system builder to put together the image and tried to specify ‘-fe’ as an argument to procnto, but it made no difference. It is clear to me that the FPU emulator is being used. If I delete fpemu.so.2, I get errors running the program as it cannot find the FPU emulation library.
The binary program that adds the two floats together works fine on a commercial eval system with a MCF8548E. So, I know the compilation/linking process is ok.
Has anyone seen this or have any ideas? I’ve been scratching my head so much I dug another hole in it.