These functions are glue inserted by the compiler for conversions from integer to floating-point types. (floatdisf
converts signed integer to float, floatundisf
converts unsigned integer to float, and aeabi_l2f
is an alias for floatdisf
.)
The implementations of these functions in LLVM can be found at:
- https://llvm.org/svn/llvm-project/compiler-rt/trunk/lib/floatdisf.c
- https://llvm.org/svn/llvm-project/compiler-rt/trunk/lib/floatundisf.c
As the path suggests, they are part of the compiler_rt
library, which is linked in automatically as needed.