Ulysse

@gasche @Julow we found one more instance with @Octachron in [3e3ef8f](https://github.com/ocaml/ocaml/pull/13302/commits/3e3ef8fc75f4b7bc7b88db0a1dfda01f56e67eec) following an alternative approach: make compiling fail in translprim / typeopt when some optimization are done on comparison. It...

> * Missing `location` calls in `Ast_iterator`, `Tast_iterator` and `Tast_mapper`. I pushed a commit that update these iterators.

I rebased and cleaned the history a bit. @Octachron the hygiene checks complains that there is no parser changes so we should remove the label. Does it checks that the...

But ideally we should only check that `Longident.t` changed since there are a few utility functions defined this module, it might be tricky to do well.

> This needs a review for correctness to be mergeable. @gasche do you think you could be that reviewer ? I pinged around with not a lot of success...

> The parestree changes issue should be fixed by #13431. I think that even though not all API changes to those modules will be breaking, it's good for ppxlib maintainers...

I rebased the PR after our (offline) discussion, @gasche . Do you think that it needs another independent review ? Could it be yours ?

Thank you for the review, I will squash the two implementation commits together before re-basing on trunk and address your comments.

@Octachron I rebased on trunk once more and addressed your comments in the last commits. I don't know if I should do something about the hygiene of the debugger parser...

> @voodoos , what is the current base of the PR? The failure in the CI looks like #13090 which should have been fixed in the latest trunk. It is...