#4973: building ghc7.0.1.20110217 under x86 solaris fails
----------------------------------+-----------------------------------------
Reporter: maeder | Owner: igloo
Type: bug | Status: patch
Priority: highest | Milestone: 7.2.1
Component: Compiler | Version: 7.0.2
Resolution: | Keywords:
Testcase: | Blockedby:
Difficulty: | Os: Solaris
Blocking: | Architecture: x86
Failure: Building GHC failed |
----------------------------------+-----------------------------------------
Comment(by kgardas):
Simon, perhaps it's not clear, but we do have a situation where current
HEAD is working well on Solaris 11, but is not working at all due to
relocations errors on Solaris 10. IMHO all the source code of GHC is the
same for both OSes. Yes, I know that such issue might be solved by -fPIC.
In fact I've done this before for Solaris 11. Please see rts/ghc.mk and
search for On Solaris we get errors like:
Unfortunately now on Solaris 10 when everything is working well on Solaris
11 I don't see any -fPIC missing anywhere. I may reinvestigate this issue
again in next two weeks, but IMHO this might as well be some bug either in
gcc or as on Solaris 10 which is already fixed on Solaris 11. If this is
the case, then we do have two possibilities: either (a) disable shared
libs support on Solaris 10 and leave it enabled on Solaris 11 or (b)
workaround issue by using -mimpure-text -Wl,-ztextwarn on both platforms.
Currently I'm slightly inclined to (a) since (b) is still not the best on
Solaris 10. The issue is: I'm able to build whole GHC, but I'm not able to
build executable which would use shared libs due to conflicting -dn and -P
options going to linker from GCC. So far I've not been able to find the
right workaround for this. Perhaps Christian might try this? Of course the
advantage of (b) is that you see all the warnings which would be error w/o
using -mimpure-text and you can fix the code yourself if this is GHC
issue...
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/4973#comment:39>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
