Implementation of strchr for AArch64. Speedups taken from micro-bench show the improvements relative to the standard C code. The use of LD1 means we have identical code for both big- and little-endian systems.
