Fix implementation of FMIN/FMAX NaN case

[riscv-isa-sim.git] / riscv / insns / fsqrt_q.h

require_extension('Q');
require_fp;
softfloat_roundingMode = RM;
WRITE_FRD(f128_sqrt(f128(FRS1)));
set_fp_exceptions;