> > Sure! I don't know what this didn't cross my mind earlier, I guess > > I was in a bit of a crunch... Will do that tomorrow (I'm also planning > > on submitting a bunch of patches tomorrow as well, if things go as > > planned - so that'll be a good time for that). > > OK, thanks! OK, I ended up checking in both patches (there was simply too much confusion on the second one, so it was simpler if I just change the use of builtin_type_int into builtin_type_int32). Hopefully, this should clear the way for you to commit your series of patches! 2008-09-10 Joel Brobecker * ada-lang.c (ada_evaluate_subexp) [BINOP_ADD, BINOP_SUB]: Add special handling for pointer types. 2008-09-10 Joel Brobecker * ada-lang.c (ada_evaluate_subexp) [OP_ATR_SIZE]: Use archecture-neutral builtin_type_int32 instead of builtin_type_int. As I said, using int32 is not completely ideal, but should be good enough in practice. If someone ever hits that limitation, I'll think about writing a function that returns a type that's big enough to hold a given LONGEST/ULONGEST... Actual patches attached. Tested on x86-linux by running the gdb.ada testcases. Cheers, -- Joel