PCH vs. type canonicalization

[David Edelsohn]

>>>>> Jason Merrill writes:

Jason> I suppose the right fix is to stop using the address in calculating the
Jason> hash of a type.  Alternately, we could do something along the lines of the
Jason> reorder functions for sorted field structures.

	Avoiding addresses in hash calculations is a long-standing
enhancement request:

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=4520

David