[libre-riscv-dev] fp special functions

[Luke Kenneth Casson Leighton]

http://www.andraka.com/files/crdcsrvy.pdf

do you think we can justify adding the transcendental functions,
perhaps using a pipelined version of the cordic algorithm?
http://www.myhdl.org/docs/examples/sinecomp/

natural-logarithm version of CORDIC - *sigh* very obtuse paper,
there's probably better:
https://www.researchgate.net/profile/Daniel_Llamocca/publication/230668515_A_fixed-point_implementation_of_the_natural_logarithm_based_on_a_expanded_hyperbolic_CORDIC_algorithm/links/0912f50660aa5e6682000000/A-fixed-point-implementation-of-the-natural-logarithm-based-on-a-expanded-hyperbolic-CORDIC-algorithm.pdf?origin=publication_detail

l.