[GHC] #15799: GHC panic (and warnings)
GHC
ghc-devs at haskell.org
Mon Nov 12 19:35:36 UTC 2018
#15799: GHC panic (and warnings)
-------------------------------------+-------------------------------------
Reporter: Iceland_jack | Owner: (none)
Type: bug | Status: new
Priority: normal | Milestone:
Component: Compiler | Version: 8.6.1
Resolution: | Keywords:
Operating System: Unknown/Multiple | Architecture:
| Unknown/Multiple
Type of failure: None/Unknown | Test Case:
Blocked By: | Blocking:
Related Tickets: #12045 | Differential Rev(s):
Wiki Page: |
-------------------------------------+-------------------------------------
Comment (by goldfire):
Discussion this morning essentially agreed with my "first instinct" above.
But with a new spin on things: The `tcm_smart` setting on `TyCoMapper` is
really about what to do about `Refl` casts. Do we preserve them? Discard
them? Or, in our new mode, discard some of them? Previously, the "dumb"
mode meant to preserve all casts. This is important for `Note [The well-
kinded type invariant]` (where we sometimes use reflexive casts to make a
type's kind looked zonked even when the type is not yet zonked). The
"smart" mode dropped all reflexive casts. Now, we want a third mode that
drops all casts that look reflexive to the type checker.
By elucidating what we really mean by "smart", having a third setting does
not look so unusual, after all.
--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15799#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
More information about the ghc-tickets
mailing list