Transformation of control flow into monads: support continue, break and return

[W95Psp]

No description provided.

[W95Psp]

This is related to #28: causes an issue with tls_cryptolib

[lassebramer]

Ran into this error:

error[CE0001]: (Diagnostics.Context.Phase CfIntoMonads): something is not implemented yet. This is discussed in issue 96: please upvote or comment this issue if you see this error message.
               TODO: Monad for loop-related control flow
   --> pasta/src/pasta.rs:783:5
    |
783 | /     for _ in 0..loop_upper_bound {
784 | |         if check_not_zero(r.clone()) && degree_polyx(r.clone()) >= degree_polyx(d.clone()) {
785 | |             let t: Polyx = divide_leading_terms(r.clone(), d.clone());
786 | |
...   |
790 | |         }
791 | |     }
    | |_____^

Full function here:

pub fn divide_poly(n: Polyx, d: Polyx) -> (Polyx, Polyx) {
    let mut q: Polyx = Seq::<FpVesta>::create(n.len());
    let mut r: Polyx = n.clone();

    let mut loop_upper_bound = d.len();
    if q.len() > d.len() {
        loop_upper_bound = q.len();
    }
    for _ in 0..loop_upper_bound {
        if check_not_zero(r.clone()) && degree_polyx(r.clone()) >= degree_polyx(d.clone()) {
            let t: Polyx = divide_leading_terms(r.clone(), d.clone());

            q = add_polyx(q, t.clone());
            let aux_prod: Polyx = mul_polyx(d.clone(), t.clone());
            r = sub_polyx(r, aux_prod);
        }
    }

    (trim_poly(q), trim_poly(r))
}

[W95Psp]

Thanks for your bug report! This for loop contains no break or continue; thus I suspect hax is not reconstructing properly a for loop. I opened #105 for tracking that specifically.

[W95Psp]

Closing in favor of #15