Browse thread
[Caml-list] ocamlc -linkall: problems with external functions?
- Ranjan Bagchi
[
Home
]
[ Index:
by date
|
by threads
]
[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
| Date: | -- (:) |
| From: | Ranjan Bagchi <ranjan@f...> |
| Subject: | [Caml-list] ocamlc -linkall: problems with external functions? |
Hi --
I'm experimenting with mod_caml 1.0.6 (which is very cool, thanks Rich),
and I've found something interesting (this may be an ocaml thing, a
mod_caml thing, or an apache thing since it intersects):
The following line in the Makefile:
mod_caml_executable: $(ALL_CMOS) apache_c.o wrappers.o dummy.o
$(OCAMLC) -custom $(OCAMLCFLAGS) $(ALL_CMAS) $^ -o $@
if changed to
mod_caml_executable: $(ALL_CMOS) apache_c.o wrappers.o dummy.o
$(OCAMLC) -linkall -custom $(OCAMLCFLAGS) $(ALL_CMAS) $^ -o $@
results in the very strange startup error (when doing
/usr/sbin/apachectl start):
Fatal error: unknown C primitive `pcre_get_match_limit_wrapper'
However, in both mod_caml_executables, nm | grep
pcre_get_match_limit_wrapper results in
080571a0 T pcre_get_match_limit_wrapper
Is this a bug? What's happening? Is there a workaround?
Thanks,
Ranjan
(PS -- why I'm doing this. I'm trying to load into the executable the
rpc library. The reason why is that Dynlink seems to really dislike
loading -- at run time -- the rpc.cma and equeue.cma libraries.
However, I've found that building an executable with --linkall produces
a system that Dynlink can use to load in code building **using** rpc:
stub modules, etc)
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners