> On Mar 27, 2014, at 13:14, "Alain O'Dea" <[email protected]> wrote: > >> On 26 March 2014 16:58, a b <[email protected]> wrote: >> > I'll give that a shot tomorrow. This is made more challenging due to the >> > code that calls collect2 is itself already compiled rather than being a >> > script or Makefile. >> >> collect2 should not be referenced in the Makefile directly, as it is a >> compiler backend. If it is, I would say that Makefile needs revising. >> >> It looks like you have the backend stepping in messing up the linking step. >> I came to this conclusion by reading the following documentation: > > ghc is a build-chain abstraction so it hides everything that would normally > be done in a Makefile for a native build. The idea being that you can run > ghc the same way regardless of the underlying platform even in the absence of > the GCC toolchain. It hides the details at that level entirely. It's > intended to work like javac in that you run ghc and it does everything > necessary to get to a built product. Normally that's an advantage because I > don't have to care, right now it's an annoyance since it's broken :)
Patch submitted to GHC: https://ghc.haskell.org/trac/ghc/ticket/8921#comment:4 ------------------------------------------- smartos-discuss Archives: https://www.listbox.com/member/archive/184463/=now RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00 Modify Your Subscription: https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb Powered by Listbox: http://www.listbox.com
