Re: [GHC] #487: powerpc/linux segfaulting binaries

#487: powerpc/linux segfaulting binaries
----------------------+-----------------------------------------------------
 Reporter:  dons      |          Owner:  nobody  
     Type:  bug       |         Status:  assigned
 Priority:  normal    |      Milestone:          
Component:  Compiler  |        Version:  6.4.1   
 Severity:  normal    |     Resolution:  None    
 Keywords:            |     Difficulty:  Unknown 
 Testcase:  N/A       |   Architecture:  powerpc 
       Os:  Linux     |  
----------------------+-----------------------------------------------------
Changes (by igloo):

  * testcase:  => N/A

Old description:

> {{{
> (Just so we don't forget) On powerpc/linux both yi and
> hmp3 segfault immediately when built the -fvia-C or
> -fasm ways, with (in the case of hmp3):
>
> (gdb) where
> #0  0x100529d4 in __stginit_Binary_ ()
> #1  0x1046ed78 in StgRun ()
> #2  0x104671e0 in hs_add_root ()
> #3  0x10467128 in startupHaskell ()
> #4  0x10464f18 in main ()
>
> and in __stginit_Main_() with yi.
>
> They die before any code in Main.main is executed.
> Not much other info seems available. Both these
> applications are built with the ncurses binding, and
> this appears to be the same problem repored to the
> lists back in June 05 by J. Bobbio:
>
> http://www.mail-archive.com/glasgow-haskell-
> bugs@xxxxxxxxxxx/msg07478.html
>
> A test case is attached to that mail.
> To reproduce the hmp3 crash:
>   $ darcs get --partial
> http://www.cse.unsw.edu.au/~dons/code/hmp3
>
> And follow the cabalised build process. Both
> applications work on apple/powerpc, interestingly enough.
>
> -- Don
> }}}

New description:

 (Just so we don't forget) On powerpc/linux both yi and
 hmp3 segfault immediately when built the -fvia-C or
 -fasm ways, with (in the case of hmp3):

 {{{
 (gdb) where
 #0  0x100529d4 in __stginit_Binary_ ()
 #1  0x1046ed78 in StgRun ()
 #2  0x104671e0 in hs_add_root ()
 #3  0x10467128 in startupHaskell ()
 #4  0x10464f18 in main ()
 }}}

 and in __stginit_Main_() with yi.

 They die before any code in Main.main is executed.
 Not much other info seems available. Both these
 applications are built with the ncurses binding, and
 this appears to be the same problem repored to the
 lists back in June 05 by J. Bobbio:

 http://www.mail-archive.com/glasgow-haskell-bugs@xxxxxxxxxxx/msg07478.html

 A test case is attached to that mail.
 To reproduce the hmp3 crash:
   $ darcs get --partial
 http://www.cse.unsw.edu.au/~dons/code/hmp3

 And follow the cabalised build process. Both
 applications work on apple/powerpc, interestingly enough.

 -- Don

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/487>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler_______________________________________________
Glasgow-haskell-bugs mailing list
Glasgow-haskell-bugs@xxxxxxxxxxx
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs