> There's a comment at the beginning of ghc/driver/mangler/ghc-asm.prl
> which seems to relate to this.  You might want to take a look.

That explains quite a lot... in particular why it works with 2.7.2. Looks
like the mangler understands most of the issues already, it just needs
modifying to understand the modified assembly layout of gcc 2.95.2. I would
let it discard the P% selectors, but live with the extra indirection for
now. If gcc ever switches to use ADDIL like HP's compiler the linker
fastaccess optimization will be able to take effect and I think the result
will be faster overall.

> 3.  write a native code generator back-end for HP-PA.
> 
> :-)

In an ideal world absolutey, and for Itanium too while we're about it.


I think the end is in sight with this problem...

Regards,
Andy.

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

Reply via email to