Bugs item #1216562, was opened at 2005-06-07 19:31 Message generated for change (Comment added) made by loewis You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=105470&aid=1216562&group_id=5470
Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Windows Group: Feature Request >Status: Closed >Resolution: Wont Fix Priority: 5 Submitted By: Niall Douglas (ned14) Assigned to: Nobody/Anonymous (nobody) Summary: Replace MSVC memory allocator with ptmalloc2 Initial Comment: You can get a port of ptmalloc2 to win32/win64 at http:// www.nedprod.com/programs/Win32/ptmalloc2/. ptmalloc2 is very considerably faster than the MSVC CRT allocator in small object allocations such as those python would frequently make. If one were to replace python's allocator with ptmalloc2, one should expect to see a sizeable speed increase. As to my knowledge python extension modules don't pass malloced blocks between themselves and the core interpreter, there shouldn't be a problem with calling the wrong allocator with the wrong block. However if there were problems, ptmalloc2 allocated blocks are very easily recognisable from MSVC allocated ones and so the core interpreter could easily free any kind of allocated block. The reverse, unfortunately, is not the case unless one forced all python extension modules to use a python provided allocator. ---------------------------------------------------------------------- >Comment By: Martin v. Löwis (loewis) Date: 2005-06-15 22:54 Message: Logged In: YES user_id=21627 I'm rejecting the idea. There is no proof that there would indeed be a speed improvement, and I very much doubt such a claim. Python uses obmalloc for most objects, so malloc performance is mostly irrelevant. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=105470&aid=1216562&group_id=5470 _______________________________________________ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
