[ocaml-ctypes] Spinning process on osx

[David Sheets]

On Wed, Jan 29, 2014 at 10:38 AM, Daniel Bünzli
<daniel.buenzli at erratique.ch> wrote:
> Le samedi, 28 décembre 2013 à 21:23, Jeremy Yallop a écrit :
>> There do seem to have been some issues with -no_compact_unwind and libffi:
>>
>> https://ghc.haskell.org/trac/ghc/ticket/5019
>
> This is becoming slightly annoying for me as I would like to use a 4.01.0 feature only (Printexc.get_callstack).
>
> I have to admit I have a very poor understanding of the issue, -no_compact_unwind is not documented in man ld, but the option -warn_compact_unwind says this [1]. Here are the options I'm contemplating, I'll gladly take comments before I spend time on this.
>
> * Patch the compiler to remove the option, what will it entail ?
> * Try with libffi HEAD to see if the issue was maybe fixed. To brew experts, is there an quick easy way/pointer to instructions of doing that with brew ? I'm spoilt by opam pin.

We did this for libsodium a few months ago. See
<https://github.com/amirmc/homebrew/commit/f1ae9826cd61cb0ae3b2ea0271dee66699c53734>.
To understand the magic, I used the wiki enterable via
<https://github.com/Homebrew/homebrew/wiki/Formula-Cookbook>. Good
luck, it's Ruby.

David

> Thanks,
>
> Daniel
>
> [1]
> -warn_compact_unwind
> When producing a final linked image, the linker processes the
> __eh_frame section and produces an __unwind_info section.
> Most FDE entries in the __eh_frame can be represented by a
> 32-bit value in the __unwind_info section. The option issues
> a warning for any function whose FDE cannot be expressed in
> the compact unwind format.
>
>
>
> _______________________________________________
> Ctypes mailing list
> Ctypes at lists.ocaml.org
> http://lists.ocaml.org/listinfo/ctypes