Damien Doligez

Results 67 comments of Damien Doligez

In finalise.c : - [x] Please add a comment to `caml_final_merge_finalisable` explaining why the young of the source are added to the old of the target. I think I know...

- [x] in runtime/interp.c: https://github.com/ocaml-multicore/ocaml-multicore/blob/865aabdacb64db24753b9c9268426a8f34751a26/runtime/interp.c#L557 Resetting the `sp` field of a stack block that you're going to free right away.

- [ ] `stack_size_bucket` will only return something if `wosize` is exactly a power of two times `caml_fiber_wsz` but `alloc_stack_noexc` seems to call it with a weaker assumption. https://github.com/ocaml-multicore/ocaml-multicore/blob/865aabdacb64db24753b9c9268426a8f34751a26/runtime/fiber.c#L87 https://github.com/ocaml-multicore/ocaml-multicore/blob/865aabdacb64db24753b9c9268426a8f34751a26/runtime/fiber.c#L142...

So we need to change the grammar in the documentation because the current version has this ambiguity.

One thing has changed: Bigarray is now part of stdlib, so Gerd's solution has become rather easy to implement.

Note: if you update `pos_bol` you will also have to update `pos_cnum`.

>Most likely, the correct fix is to modify the lexer so as to correctly update the pos_bol field after reading a line directive. I don't think this is the right...

I note that you didn't carry over the compiler change from #1642, which would silence the alert for the built-in operations whenever they are specialized (i.e. replaced by non-generic operations...