miri::intrinsics

Function apply_random_float_error_to_imm

Source 
fn apply_random_float_error_to_imm<'tcx>(
    ecx: &mut MiriInterpCx<'tcx>,
    val: ImmTy<'tcx>,
    ulp_exponent: u32,
) -> InterpResult<'tcx, ImmTy<'tcx>>
Expand description

Applies a random ULP floating point error to val and returns the new value. So if you want an X ULP error, ulp_exponent should be log2(X).

Will fail if val is not a floating point number.