Hi folks
I was asked recently if one could optimize runtime performance of the Linux-Kernel by compiling it with an optimized compiler.
The answer: No
Even trying to use the armv7l hardfloat gcc causes the kernel build to fail.