Question about bound checking in rows display

[Jose E. Marchesi]

Hello!

> I am working on implementing the bound checking for row displays, and I
> think I understand the overall algorithm. The loop in
> a68_lower_collateral_clause iterates over each sub multiple, and the // XXX
> comment is where the bounds checks are intended to be added.
>
> I have one question about the interpretation of the row display. In your
> previous email you mentioned that it depends on whether the target is a row
> row or a row of rows. By the time execution reaches the // XXX block,
> should that distinction already have been resolved by the frontend (i.e.
> this code is only handling one of those cases), or is this where the
> lowering code is expected to distinguish between them?

That must have been already been determined by the parser.
You can check the mode of p,  MOID (p).