Ian Dall wrote:
> I have done this on gcc 2.7.2.1 from FSF. I recall that someone did
> some work for gcc to make it handle 64 bit types better. Has that made
> it into the FSF distribution or should I get it from the netbsd source
> tree?

Matthias Pfaller wrote:
> That was Jon Buller and me. I think Jon sent the code to Richard Kenner
> (is this true Jon?). Should I mail the diffs to the list? I think Jon
> (or I) did this already one or two years ago.

Was it that long ago already?  I did send the code to Richard
Kenner, and I got a response that it would be rolled in sometime.
I don't know if that has happened yet.  (I would hope so...)  The
port-pc532 archives should have the final diffs, and several buggy
versions as well.

As I remember, we were unable to get it to do register to memory,
or memory to register operations.  That would be a win if one of
the int64's was in memory, as it is now it will load the in-memory
argument (and use 2 registers in the process) and then operate.

If you figure that part out, you might want to consider memory to
memory operations too, as in very register tight situations, where
nothing is in the CPU, it would be faster to do a memory to memory
operation than to spill, load, operate, save, etc.  It did not
appear to happen often enough that it was worth the bother (or in
my case, the risk of breaking it) but I didn't run a lot of code
past it either.

Jon Buller