Linking a dynamic library (libjvm.dylib) in Mac OS X (rpath issue)

Question

I do have an application that requires linkage with libjvm (a library from the JDK needed to do JNI bindings). When I tell the location of libjvm.dylib using -L it successfully compiles and links. However when I run the binary I get:

dyld: Library not loaded: @rpath/libjvm.dylib
  Referenced from: <my home directory>/./mybinary
  Reason: image not found

So far I found out that I can run my binary specifying LD_LIBRARY_PATH like so:

LD_LIBRARY_PATH=<path to libfolder installation> ./mybinary

But of course I do not want that. Why should I specify the exact location anyway if I have to give it again and again each time I start the application?!

I also learned that dynamic libraries on mac os x do get a kind of stamp which tells there location. However I don't know what rpath is (seems like a variable to me, but how can I set it during linking?).

The application is built using haskell, but I can equally well link the object files manually using ld. However, I'm stuck on that rpath thing - is it maybe special to the JDK libraries?

Here is what I do in order to build:

ghc --make Main.hs mycbinding.o -ljvm -L<javahome>/jre/lib/server -o mybinary

Answer 1

From Apple's dyld man page:

@rpath/

  Dyld maintains a current stack of paths called the run path list.
  When @rpath is encountered it is substituted with each path in the
  run path list until a loadable dylib if found. The run path stack
  is built from the LC_RPATH load commands in the depencency chain
  that lead to the current dylib load. You can add an LC_RPATH load
  command to an image with the -rpath option to ld(1). You can even add
  a LC_RPATH load command path that starts with @loader_path/, and it
  will push a path on the run path stack that relative to the image
  containing the LC_RPATH. The use of @rpath is most useful when you
  have a complex directory structure of programs and dylibs which can be
  installed anywhere, but keep their relative positions. This scenario
  could be implemented using @loader_path, but every client of a dylib
  could need a different load path because its relative position in the
  file system is different. The use of @rpath introduces a level of
  indirection that simplies things. You pick a location in your directory
  structure as an anchor point. Each dylib then gets an install path that
  starts with @rpath and is the path to the dylib relative to the anchor
  point. Each main executable is linked with -rpath @loader_path/zzz,
  where zzz is the path from the executable to the anchor point. At runtime
  dyld sets it run path to be the anchor point, then each dylib is found
  relative to the anchor point.

You need to pass -rpath path/containing/the/library to ld when linking your binary to tell it where to search when expanding the @rpath/ prefix in the shared library load command. With GHC you can use the -optl-Wl argument to have it pass flags through to ld, so you'll want to invoke GHC like so:

ghc --make Main.hs mycbinding.o -ljvm -L<javahome>/jre/lib/server -optl-Wl,-rpath,<javahome>/jre/lib/server -o mybinary