[issue1541] py3k: Re-enable rpython int sized app-level int
New submission from Philip Jenvey <pjenvey@underboss.org>: This is a significant performance degradation on the py3k branch, basically intobject.py was disabled due to various obscure breakages. I'm intending to kill int/long/bool's multimethods (unless someone else does first ;] ) on default before mucking around with these types again ---------- assignedto: pjenvey messages: 5938 nosy: pypy-issue priority: performance bug status: unread title: py3k: Re-enable rpython int sized app-level int ________________________________________ PyPy bug tracker <tracker@bugs.pypy.org> <https://bugs.pypy.org/issue1541> ________________________________________
Philip Jenvey <pjenvey@underboss.org> added the comment: This is pretty much done ---------- status: unread -> resolved ________________________________________ PyPy bug tracker <tracker@bugs.pypy.org> <https://bugs.pypy.org/issue1541> ________________________________________
Philip Jenvey <pjenvey@underboss.org> added the comment: An IRC user yay_ reported the follow still seeming slow: for i in range(10**8): j = 2*i ---------- status: resolved -> chatting ________________________________________ PyPy bug tracker <tracker@bugs.pypy.org> <https://bugs.pypy.org/issue1541> ________________________________________
Philip Jenvey <pjenvey@underboss.org> added the comment: Seems there's probably a couple things hurting us on this basic example: o Broken strategies (#1471). In particular dict strats, like the lack of CellDict affects this. One or two other specialized dicts are also broken and can affect other situations. I've already began working on this o Py3k's range may be hurting us too because it somewhat differs from py2's xrange in that it handles the equiv. of py2 longs. We'll have to double check this ________________________________________ PyPy bug tracker <tracker@bugs.pypy.org> <https://bugs.pypy.org/issue1541> ________________________________________
Philip Jenvey <pjenvey@underboss.org> added the comment: The follow up issue, the regression of the small range example, should be solved now that issue1471 is closed (092f39d9ab1c). I tried this example out about a week ago on an older version of the py3k-fix- strategies branch, and the performance was greatly improved. IIRC it was very close to default's. That means py3k's range + the JIT (with working strategies) isn't too much of a problem here ---------- status: chatting -> resolved ________________________________________ PyPy bug tracker <tracker@bugs.pypy.org> <https://bugs.pypy.org/issue1541> ________________________________________
participants (1)
-
Philip Jenvey