Re: Unicode mangling (was Re: [PATCH] Java: New C++ ABI compatibility changes.)

[Gavin Romig-Koch <gavin at redhat dot com>]

Tom Tromey writes:
 > Gavin> Also, while I'm talking, I'll venture into an area I'm less
 > Gavin> confident about.  It seems to me we can't use UTF8 as an
 > Gavin> encoding from gcc to the assembler unless the assembler allows
 > Gavin> any 8bit character in identifiers (which seems unlikely).
 > 
 > We actually only need a weaker constraint, namely that the assembler
 > will allow any byte with the 8th bit set.  Maybe that is what you
 > meant -- it is unclear to me.

No, I ment the stronger constraint - but your right, we only need the
weaker one (though it took me writing another note trying to explain
why you were wrong for me to actually see this :-).

Still, it doesn't seem to me that many assemblers would allow even
the weaker constraint - do you think I'm wrong?

                                              -gavin...