GHC panic: Can't serialise IfaceTcTyVar
This bug currently prevents vector from being built with GHC HEAD. Here is a minimized test case:
module Bug where
data Foo a x = Foo x
refoo :: Foo a x -> Foo b x
refoo (Foo x) = Foo x
{-# RULES
"refoo/refoo" forall s.
refoo (refoo s) = s #-}You'll need to compile with optimizations on to trigger this:
$ /opt/ghc/head/bin/ghc -O1 Bug.hs
[1 of 1] Compiling Bug (.hs -> .o)
Bug.hs:10:1: warning: [-Winline-rule-shadowing]
Rule "refoo/refoo" may never fire
because ‘refoo’ might inline first
Probable fix: add an INLINE[n] or NOINLINE[n] pragma for ‘refoo’
Bug.hs:10:1: warning: [-Winline-rule-shadowing]
Rule "refoo/refoo" may never fire
because ‘refoo’ might inline first
Probable fix: add an INLINE[n] or NOINLINE[n] pragma for ‘refoo’
ghc: panic! (the 'impossible' happened)
(GHC version 8.1.20161125 for x86_64-unknown-linux):
Can't serialise IfaceTcTyVar
a_ary[sk:1]
Call stack:
CallStack (from HasCallStack):
prettyCurrentCallStack, called at compiler/utils/Outputable.hs:1076:58 in ghc:Outputable
callStackDoc, called at compiler/utils/Outputable.hs:1080:37 in ghc:Outputable
pprPanic, called at compiler/iface/IfaceType.hs:1331:10 in ghc:IfaceType
Please report this as a GHC bug: http://www.haskell.org/ghc/reportabugIf I had to guess, this commit ( http://git.haskell.org/ghc.git/commit/5f349fe24066e7b0af85934664e27636d2e84fe5 ) is responsible. Simon, do you know what's going on here?
Trac metadata
| Trac field | Value |
|---|---|
| Version | 8.1 |
| Type | Bug |
| TypeOfFailure | OtherFailure |
| Priority | high |
| Resolution | Unresolved |
| Component | Compiler |
| Test case | |
| Differential revisions | |
| BlockedBy | |
| Related | |
| Blocking | |
| CC | simonpj |
| Operating system | |
| Architecture |