[MLton] Experimental 64Bit

[Matthew Fluet]

Ray wrote:
> I see the same floating point exception on a FC 5, dual AMD64.
> 
> [raymond at corwin ~]$ which mlton
> /usr/local/bin/mlton
> [raymond at corwin ~]$ mlton
> Floating point exception
> [raymond at corwin ~]$ gdb /usr/local/lib/mlton/mlton-compile
> GNU gdb Red Hat Linux (6.3.0.0-1.134.fc5rh)
> 
> (gdb) run
> Starting program: /usr/local/lib/mlton/mlton-compile
> (no debugging symbols found)
> 
> Program received signal SIGFPE, Arithmetic exception.
> 0x0000003282608adb in do_lookup_x () from /lib64/ld-linux-x86-64.so.2
> (gdb) bt
> #0  0x0000003282608adb in do_lookup_x ()
> from /lib64/ld-linux-x86-64.so.2

Actually, this (along with the previous information about .gnu.hash) 
provided enough info.  Google for "SIGFPE do_lookup_x" and I pretty 
quickly found:
   http://www.redhat.com/archives/fedora-list/2006-November/msg04167.html
which seems to sufficiently explain the observed behavior.

In any case, early next week I'll try building new binaries, either with 
--hash-style=sysv or by running FC5 (and/or Ubuntu Edgy) under VMWare 
and building from source.