RE: "Hello World" program core dumps with GHC 4.08.1 on HPUX 10.2 0

[BENNETT,ANDY (HP-Unitedkingdom,ex1)]

Hi Simon,

Thanks for the information about the JMP_ calls usage.

> However, I'll glady commit your change to JMP_ on HPUX if it 
> fixes your problems - actually we're very grateful to have someone 
> working on this!
> 
> Have you tried building the CVS sources, BTW?  It'd be nice 
> if we had a working HPUX port for the forthcoming release.

I'm trying at the moment. I'm getting multiple errors during the 'hsc' link,
"Target of unconditional branch is out of range" for parser/Parser.o and
another module happy generated too. The errors are associated with long
calls (-mlong-calls used to fix this, it is not a gcc option anymore). The
idea now being that a compiler will generate a BRANCH AND LINK ("bl")
instruction that has a range of +/- 0x40000 bytes relative to the
instruction's position, and it also tags it as such for fixup during
linking. If the target is out range (beyond the 0x40000 limit), the linker
is supposed to substitute a long call (2 instructions) for the "bl". With
gcc 2.95.2, it doesn't look like this is happening in this case.

It was interesting; I searched on the internet to see if I could find
something helpful on the problem. I found on a GCC info site someone had had
the exact same problem with building GHC on GCC 2.95.?, but not with GCC
2.7.?. There was no good solution, other than break up the source file into
pieces. I'm not totally 100% certain about the fixup process and how it
should operate, so I'll keep playing to see if I figure why it doesn't work.

If I successfully build, I'll let you know.

Regards,
Andy.

_______________________________________________
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs