[libre-riscv-dev] [isa-dev] Re: FP reciprocal sqrt extension proposal

[lkcl]

https://libre-riscv.org/rv_major_opcode_1010011/

okaaay, so that's just done, what i did was to split out the use of funct5 
(when either rs2 or funct3 are "specifiers") and you can see that, 
actually, rs2=00000 is used for FSQRT.X, and it might be better, instead of 
using a new funct5, to just use rs2=00001 for FRSQRT.X

funct3 in both cases is the "rounding mode".

l.