Re: [Haskell-cafe] Calling Haskell from C

Apologies. I put the output of “-l” (not “-L”) here: https://gist.github.com/idontgetoutmuch/654007ac57e863a293b76a0cf98250c1 https://gist.github.com/idontgetoutmuch/654007ac57e863a293b76a0cf98250c1 I couldn’t immediately see anything that tells me how @rpath is resolved.

On 13 Jan 2017, at 17:46, Brandon Allbery wrote:

On Fri, Jan 13, 2017 at 12:35 PM, mailto:dominic@steinitz.org> wrote:

...

$ otool -L a.out

I said "-l". "-L" shows how it resolves shared objects; "-l" dumps the load commands table, which includes RPATH entries that tell it how to resolve @rpath.

...

DYLD_FALLBACK_LIBRARY_PATH=/Users/dom/Library/Haskell/ghc-8.0.1/lib/test-via-c-0.1.0.0/bin:/Library/Frameworks/GHC.framework/Versions/8.0.1-x86_64/usr/lib/ghc-8.0.1/rts ~/Dropbox/Private/TestU01-1.2.3/examples/src $ ./a.out

This is odd, unless Apple changed things yet again (I should not be surprised, the loader stuff is about as stable as water /o\ )

It’s not clear to me why it’s looking for `gcc/5` which is not installed when it looks as though it should be looking for `gcc/6`.

That might be indicated by the RPATH entries, since DYLD_FALLBACK_LIBRARY_PATH changes how lookups work (so does DYLD_LIBRARY_PATH, in a different way, which is why it's dangerous).

-- brandon s allbery kf8nh sine nomine associates allbery.b@gmail.com mailto:allbery.b@gmail.com ballbery@sinenomine.net mailto:ballbery@sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net http://sinenomine.net/ Dominic Steinitz dominic@steinitz.org http://idontgetoutmuch.wordpress.com