c - What's the difference between hard and soft floating point numbers? -
when compile c code cross toolchain, linker prints pages of warnings saying executable uses hard floats libc uses soft floats. what's difference?
hard floats use on-chip floating point unit. soft floats emulate 1 in software. difference speed. it's strange see both used on same target architecture, since chip either has fpu or doesn't. can enable soft floating point in gcc -msoft-float. may want recompile libc use hardware floating point if use it.
Comments
Post a Comment