[GHC] #9370: unfolding info as seen when building a module depends on flags in a previously-compiled module

[GHC]

#9370: unfolding info as seen when building a module depends on flags in a
previously-compiled module
-------------------------------------+-------------------------------------
        Reporter:  carter            |                Owner:  richardfung
            Type:  bug               |               Status:  new
        Priority:  high              |            Milestone:  8.2.1
       Component:  Compiler          |              Version:  7.8.3
      Resolution:                    |             Keywords:  newcomer,
                                     |  Inlining
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Compile-time      |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #8635             |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by bgamari):

 > I would rather do things the right way and thus try my hand at the first
 solution. Is this something that is still suitable for a newcomer?

 I'm happy to hear that you are willing to give it another go.

 Regarding Simon's suggestion, I suspect you could pull it off, although it
 will naturally require some learning. You'll likely want to start reading
 `compiler/iface/LoadIface.hs`. In particular pay attention to the
 `ignore_prags` argument to `loadDecl`. This is how we currently tell the
 typechecker not to bother typechecking unfoldings. You will likely want to
 remove this; instead you want to fork off typechecking as an interleaved
 computation with `TcRnMonad.forkM`. Then just make sure that there are no
 unconditional strictness demands on the pragma field and add a condition
 on `-fignore-interface-pragmas` to ignore the presence of the pragmas when
 necessary.

 As always, let me know if I can be of help, especially if any of the above
 seems incorrect; it is largely the result of a cursory glance over the
 relevant code and I could have easily missed something.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9370#comment:41>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler