RFA: fix address in call to val_print

[Joel Brobecker]

> My reasoning was that I do not know whether it is easy to make a test
> case or not.  I would have to look into it.

Aha, I see.

What I understood from your original patch was that Paul encountered
an actual problem in the Archer GDB, which prompted the patch. The
scenario in which Archer GDB fails before the patch and passes after
was all that I was wondering about. If we can put that example and
stuff it in a testcase now, then let's do it if you have the time.
If it depends on some functionality that's still only in the Archer
repo, then let's defer that until that piece of functionaly is in
the FSF repo.

In the meantime, the code patch is OK to go in (unless you'd prefer
to hold off on it, but I would imagine that the fewer local changes
in the Archer repo, the better).

-- 
Joel