Thanks. When I followed the ticket, it gave me the impression failure for dynamic GHCi seems relatively reasonable, so I assumed this was acceptable. Should've just asked! On Thu, Aug 22, 2013 at 2:16 PM, Simon Marlow wrote:

On 22/08/13 19:24, git@git.haskell.org wrote:

...

Repository : ssh://git@git.haskell.org/**testsuitehttp://git@git.haskell.org/testsuite

On branch : master Link : http://ghc.haskell.org/trac/**ghc/changeset/** a5875c4e4fb83993142c8e8294b475**16adf6a77f/testsuitehttp://ghc.haskell.org/trac/ghc/changeset/a5875c4e4fb83993142c8e8294b47516ad...

------------------------------**------------------------------**---

...

commit a5875c4e4fb83993142c8e8294b475**16adf6a77f Author: Austin Seipp Date: Thu Aug 22 13:22:22 2013 -0500

linker_unload fails when using dynamic GHCi.

This is because dynamic GHCi can't load archive files, and linker_unload explicitly does this via loadArchive().

Sigh. I've no idea why the linker has an explicit check for this, there's no reason why it can't work. I'll fix it.

Cheers, Simon

______________________________**_________________ ghc-commits mailing list ghc-commits@haskell.org http://www.haskell.org/**mailman/listinfo/ghc-commitshttp://www.haskell.org/mailman/listinfo/ghc-commits

-- Regards, Austin - PGP: 4096R/0x91384671