Ian, Manuel, David:

You three seems to be central to this problem.  Might you be able to resolve it 
together?

David: you're the one who has managed to provoke it, so you are necessarily 
involved.   Ian knows the build system..

Simon

| -----Original Message-----
| From: Manuel M T Chakravarty [mailto:[email protected]]
| Sent: 02 November 2011 23:11
| To: Simon Peyton-Jones
| Cc: Ben Lippmeier; Ian Lynagh; Simon Marlow
| Subject: Re: Safe Haskell haddock segfault
| 
| I had this error as well, but could resolve it by fixing a bit-rotted rule in
| rules/haddock.mk:
| 
|   
https://github.com/mchakravarty/ghc/commit/cc3c2bd9bc2d95e2a3bac70d126c9f2e33f22cf2
| 
| The problem was that haddock ran on the packages dph-seq and dph-par before 
these
| packages had been completely built.
| 
| So, I wonder:
| 
| * if you restart validate with --no-clean (maybe a few times), does your 
build go
| through?
| * is it possible that your 2 line patch affects the order in which components 
are
| built?
| * did the dph-par library already get built in your log before the failure?
| 
| Manuel
| 
| Simon Peyton-Jones:
| > Manuel
| >
| > Any ideas?  This is some interaction with DPH...
| >
| > Simon
| >
| > -----Original Message-----
| > From: David Terei [mailto:[email protected]]
| > Sent: 02 November 2011 06:00
| > To: Simon Marlow; Simon Peyton-Jones; Ian Lynagh
| > Subject: Safe Haskell haddock segfault
| >
| > Hi all,
| >
| > I was hoping one of you could point out what the bug was that I've
| > been running into today. This is the bug that caused Ian to revert
| > some of my patches.
| >
| > So if I apply this patch 2 line patch:
| >
| > 
https://github.com/dterei/ghc/commit/5d0fa2617efdd36c16efc9cb1cd4a2c0a94a05b6
| >
| > To the HEAD of ghc, which is currently working fine, I get the
| > following build problem:
| >
| > libraries/dph/dph-par/../dph-common/Data/Array/Parallel/PArray.hs:217:10:
| >    Warning: orphan instance:
| >      instance (PA a, Show a) => Show (PArray a)
| >  96% ( 25 / 26) in 'Data.Array.Parallel.PArray'
| >  19% ( 11 / 58) in 'Data.Array.Parallel.Lifted'
| >   0% (  0 /  6) in 'Data.Array.Parallel.PArr'  46% ( 21 / 46) in
| > 'Data.Array.Parallel.Prim'GHC error in desugarer lookup in
| > dph-par:Data.Array.Parallel.Prelude.Bool:
| >  Can't find interface-file declaration for type constructor or class
| > Data.Array    Probable cause: bug in .hi-boot file, or inconsistent
| > .hi file
| >    Use -ddump-if-trace to get an idea of which file caused the error
| > haddock: panic! (the 'impossible' happened)
| >  (GHC version 7.3 for x86_64-unknown-linux):
| >        initDs IOEnv failure
| >
| > Please report this as a GHC bug:  http://www.haskell.org/ghc/reportabug
| >
| > I'm a little lost at the moment why such a simple changing is causing
| > a failure in a distant place. Any ideas would be appreciated.
| >
| > Cheers,
| > David
| >
| 


_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc

Reply via email to