Should vfslide1* do nanbox checking?

[avakar]

I'm guessing it should, based on the spike implementation and the text in extension D:

Apart from transfer operations described in the previous paragraph, all other floating-point operations on narrower n-bit operations, n<FLEN, check if the input operands are correctly NaN-boxed

However, the text on vslide1down.vx states:

If XLEN > SEW, the least-significant bits are copied over and the high SEW-XLEN bits are ignored.

and then the text on vfslide1down.vf states, somewhat vaguely:

The vfslide1down instruction is defined analogously to vslide1down, but sources its scalar argument from an f register.

This has already led to some confusion, I suggest at least adding a note. Same issue for vfslide1up.vf of course.

[aswaterman]

Since this is not a transfer operation as described in the previous paragraph, it seems like this is nailed down pretty clearly.