[GHC] #9813: Error when reifying type constructor
GHC
ghc-devs at haskell.org
Mon Feb 15 18:36:15 UTC 2016
#9813: Error when reifying type constructor
-------------------------------------+-------------------------------------
Reporter: owst | Owner:
Type: bug | Status: new
Priority: normal | Milestone:
Component: Template Haskell | Version: 7.8.3
Resolution: | Keywords:
Operating System: Unknown/Multiple | Architecture:
| Unknown/Multiple
Type of failure: None/Unknown | Test Case:
Blocked By: | Blocking:
Related Tickets: | Differential Rev(s): Phab:D1899
Wiki Page: |
-------------------------------------+-------------------------------------
Comment (by owst):
Replying to [comment:27 rwbarton]:
> This all sounds great except for one thing. Do we have a plan for
actually loosening the stage restriction? The stage restriction exists for
technical reasons, not conceptual reasons.
>
> The stage restriction currently applies to the module as a whole. It
says you cannot use an identifier that was defined in a module within a
splice in the same module. Clearly if the use is in a later declaration
group than the definition, there is no problem in principle with lifting
the restriction. But in order to do so, we would have to run the
declaration group containing the definition through the entire compiler
pipeline, right up to linking, so that we can load it to run the splice in
the latter declaration group to begin type checking it. GHC just isn't
organized to work that way currently.
An incredibly good point. To answer the first question simply: I don't,
no.
I haven't hacked on GHC before, so was attempting to flesh out my original
bug report to find a change that I might get my teeth into. I have to
admit I hadn't thought about the technical barriers and thus rather
haphazardly threw about the "let's loosen the stage restriction" part!
> The user's guide has this to say on the subject: ...
Indeed - I pasted that very snippet in comment:14, but I hadn't
appreciated what the technical barriers were.
What do you think is a sensible step forward? Change 2.b. in my previous
comment to "Process NS first, then S. Splices within S can refer (via
reify but ''not'' evaluation) to all names declared in NS or previous
DGs), but no names within S"? That way, my original program in this ticket
would work, but I wouldn't need any of the complexity of changing the
stage restriction too.
--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9813#comment:28>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
More information about the ghc-tickets
mailing list