Adaptable, Fast, Area-Efficient Architecture for Logarithm Approximation with Arbitrary Accuracy on FPGA