[GHC] #13027: Core lint errors compiling containers HEAD with GHC HEAD
GHC
ghc-devs at haskell.org
Wed Jan 11 02:57:23 UTC 2017
#13027: Core lint errors compiling containers HEAD with GHC HEAD
-------------------------------------+-------------------------------------
Reporter: erikd | Owner:
Type: bug | Status: new
Priority: normal | Milestone: 8.2.1
Component: Compiler | Version: 8.1
Resolution: | Keywords:
Operating System: Unknown/Multiple | Architecture:
| Unknown/Multiple
Type of failure: Building GHC | Test Case:
failed | simplCore/should_compile/T13027
Blocked By: | Blocking:
Related Tickets: | Differential Rev(s):
Wiki Page: |
-------------------------------------+-------------------------------------
Comment (by dfeuer):
Is there a reason ''not'' to make `exprOkForSpeculation` more lenient for
`reallyUnsafePtrEquality#`? Having it otherwise seems inconsistent, but
it's also possible that it doesn't enable useful floating or that it's too
much trouble to make it happen. I dug around just a bit, and the only
other lazy primitives I could find that might be safe for speculation were
`unpackClosure#` and (under certain circumstances, maybe) `seq`.
I'm somewhat concerned about the fact that the problem tracking
evaluatedness was only revealed "by chance", because
`exprOkForSpeculation` is (seemingly) too conservative about
`reallyUnsafePtrEquality#`. Is there some way to make this tracking more
robust? I understand that you and Max Bolingbroke worked on some possibly-
related ideas in "Types are Calling Conventions". What's your intuition
about whether the type system can help here without making life horrible?
--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13027#comment:22>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
More information about the ghc-tickets
mailing list