dp tests

[Raphael Collet]

Denys Duchier wrote:
> I created a debug build of mozart and I discovered a problem that's even
> more troubling than the previous one.  dp_cell_stationary fails because
> the engine encounters a bytecode that's clearly not a bytecode.
> 
> This happens both with gcc 4.3.1 and gcc 3.4.6, so it doesn't seem to be
> a compiler problem.
> 
> Can anyone else confirm?

I reran the test suite on my machine (MacOSX with gcc 4.0.1), and the 
test passes.  I already saw that kind of message before (a bytecode that 
is not a bytecode).  The bug was in the builder, but it has been fixed. 
  (Well, at least I did something that seemed to work.)

Mmmm, it's difficult to tell more than that right now... :-/

Note, however, that the test dp_lock_replicated always fails.  That's a 
known problem, which hasn't found a solution yet :-(


Cheers,
raph