[GHC] #9131: Experiment with a dedicated solver for Coercible

[GHC]

#9131: Experiment with a dedicated solver for Coercible
-------------------------------------+------------------------------------
        Reporter:  nomeata           |            Owner:
            Type:  task              |           Status:  new
        Priority:  low               |        Milestone:
       Component:  Compiler          |          Version:  7.8.2
      Resolution:                    |         Keywords:
Operating System:  Unknown/Multiple  |     Architecture:  Unknown/Multiple
 Type of failure:  None/Unknown      |       Difficulty:  Unknown
       Test Case:                    |       Blocked By:
        Blocking:                    |  Related Tickets:
-------------------------------------+------------------------------------

Comment (by nomeata):

 > In effect we already have a separate solver, namely special-purpose
 handling of the Coercible constraints, which works interleaved with the
 current one.

 I’d disagree. What we have is a custom lookup routine that creates the
 instances on demand instead of looking them up in the instance environment
 (and chooses one if there are several, i.e. the issue of ordering). How
 the instances are used to solve a constraint is still up to the general
 code, nothing `Coercible`-specific in there.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9131#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler