Floating point library for embedded application

Question

I am developing a program for Cortex-M3. It does not have a floating point coprocessor. The Standard C library can emulate floating point operations, but I do not use it because of its size.

Is there a good and free c library that can emulate floating point arithmetic aimed at ARM processors? Currently, when I use floating point operators, I have such communication errors:

undefined reference to `__adddf3'
undefined reference to `__subdf3'
undefined reference to `__divdf3'
undefined reference to `__extendsfdf2'
undefined reference to `__muldf3'

Perhaps such a library should implement them.

Answer 1

( ) ? , . ++, ( - C).

(, ..), ​​ .

Answer 2

, , , , , , .

, , , , , , , . , - , , .

, , , . , , , .

Answer 3

, . libgcc ( gcc), . , .

Answer 4

, newlib ? , gcc .

Answer 5

, , . .

Answer 6

gcc , gcc build voodoo (multilib, thumb, thumb2 soft float), , . gcc, . , , newlib gcc, , , ​​ . , .

Answer 7

Suppose you are using gcc try compilation with the -static-libgcc flag and checking the final binary size. I do not know if the linker will optimize unused calls, but I think it will.