Re: MPS: hash tables / obarrays

[Gerd Möllmann]

Andrea Corallo <acorallo@gnu.org> writes:

> in case you need you can decouple the optimizations we do un our
> compiler from the GCC ones forcing the number you want in our call to
> 'gcc_jit_context_set_int_option' in comp.c.

Good idea.

> That said if you see almost no differences in the pseudo C we ask GCC to
> compile you've probably got already the answer.
>
> Anyway IMO the big diff from -O0 -O1 here should be that values are not
> constantly loaded and stored into the stack, this might indeed make a
> difference on the GC.

Thanks. That matches what I can "see" in the raw assembler code in LLDB.

(Haven't looked yet at the jit code in the repo BTW to find out how I
can keep the the .o or .s files. Seems I can't compile GCC anyway
without jumping through hoops, on my system :-/).

And then... I saw by accident that it builds with -lmps instead of
-lmps-debug. That finished me off :-(.