[opam-devel] conf- libraries and stub cmas: time for ocaml-discover?
Sylvain Le Gall
gildor at ocamlcore.org
Mon Aug 26 21:40:37 BST 2013
I was thinking to add build information to the variable you can
substitute in .ab files. These files already allow the embed
$(version) but we can add $(build_information) as well. We just need
to define the right format.
We can also add all this information directly in the META (findlib)
file with a specific key to make it easily accessible through findlib
query API.
Also, on a more general topics, you know that most of these
information are already available in the .cma file ?
$ ocamlobjinfo /usr/lib/ocaml/pcre/pcre.cma
Force custom: no
Extra C object files: -lpcre_stubs -L/usr/lib/x86_64-linux-gnu -lpcre
Extra C options:
Extra dynamically-loaded libraries: dllpcre_stubs.so
Unit name: Pcre
Force link: no
2013/8/26 Christophe TROESTLER <Christophe.Troestler at umons.ac.be>:
> On Mon, 26 Aug 2013 11:57:40 +0100, Anil Madhavapeddy wrote:
>>
>> [...] Ideally, we would be able to query three types of information
>> from a conf package installation:
>>
>> - an ML file that has values exposing cflags/lflags as OCaml values.
>> This can be used by discover.ml scripts (as used in Core or Lwt)
>>
>> - a CMA file that embeds the dynamic linking flags via cclib, including
>> adding -Wl,--no-as-needed on newer systems like Ubuntu. This lets a
>> Ctypes using dynamic linking (Ctypes.Foreign) just add a CMA to the
>> link line and have the correct gcc/clang libraries work.
>>
>> - The final thing is some way to pass the right CFLAGS/LDFLAGS to
>> compile up generated C stubs with the right include/library paths to
>> gcc (either from Ctypes, or via the traditional camlidl/swig outputs).
>> Presumably this would be a raw shell fragment in a file, as usually
>> output by autoconf.
>
> Sounds good to me. For the last point, it is also important that
> these flags are available in a way that makes adding ocamlbuild rules
> using these flags easy. Maybe, the C compiler and CFLAGS/LDFLAGS can
> be duplicated in the first ML file (and then the configure phase — say
> in oasis — may generate the proper rules).
>
> Best,
> C.
More information about the opam-devel
mailing list