[patch] Re: Another weird sparc solaris 2.8 compiler error

Fri Feb 2 01:10:00 GMT 2001

Alexandre Petit-Bianco writes:
> I'll put that on my todo list. 

It's harder than I thought, as the new mangling code requires support
from other parts of the compiler that are asking for resources usually
provided by a front-end implementation -- we would end up linking a
huge amount of stubs, 38 as I now try to link with
 toplev.o libbackend.a obstack.o ../libiberty/libiberty.a 
I'm going to try an other way. Put unicode_mangling_length,
append_unicode_mangled_name and append_gpp_mangled_name in a separate
file, which will only require obstack.o and would jcf.h (for
UTF8_GET.) It should suffice as the mangling we have to perform is
fairly simple (for example, it doesn't require compression.)
./A