Today I realized that I can't run an application based on floating point. int main(void) { double x = 3.1415; printf("%lf",x); return 0; } How can I make working floating point ? I should to specify some arguments in command line at compiling ? EABI ? Regards !
Floagin Point - Illegal Instruction
Seeing as no one else responded I can only suggest a few things to investigate. I have seen some comments about floating point libraries for this platform. -msoft-float -D__GCC_FLOAT_NOT_NEEDED -march=armv4 -mtune=arm920t Not sure what this is for, but it looks like it leaves out the floating point library. Maybe, floating point is in by default. I know, say for AVR 8bit chips and GCC, you have an option of bringing in or leaving out the floating point library (libm, I think). Also, printf . . . what are you printing to? Thought printk was the call for output to the console. HTH
Hello, compiling attached program with arm-linux-gcc test_float.c -o test_float works fine on my mini2440 ( I am running 2.6.32.2 kernel )
Hi azzido, maybe its help: Any weeks ago, I built kernel 2.6.29 from sources of these site. cp config_mini2440_t35 .config done and before make I walked thru make menuconfig. I found there was no floting point emulation activated but a hint I have to activate one. I have never tested the result leaving floting point emulation unchecked. skip