Mailman 3 June 2014 - pypy-dev

PyPy3 2.3.1 released
by Philip Jenvey 20 Jun '14

20 Jun '14

===================== PyPy3 2.3.1 - Fulcrum ===================== We're pleased to announce the first stable release of PyPy3. PyPy3 targets Python 3 (3.2.5) compatibility. We would like to thank all of the people who donated_ to the `py3k proposal`_ for supporting the work that went into this. You can download the PyPy3 2.3.1 release here: http://pypy.org/download.html#pypy3-2-3-1 Highlights ========== * The first stable release of PyPy3: support for Python 3! * The stdlib has been updated to Python 3.2.5 * Additional support for the u'unicode' syntax (`PEP 414`_) from Python 3.3 * Updates from the default branch, such as incremental GC and various JIT improvements * Resolved some notable JIT performance regressions from PyPy2: - Re-enabled the previously disabled collection (list/dict/set) strategies - Resolved performance of iteration over range objects - Resolved handling of Python 3's exception __context__ unnecessarily forcing frame object overhead .. _`PEP 414`: http://legacy.python.org/dev/peps/pep-0414/ What is PyPy? ============== PyPy is a very compliant Python interpreter, almost a drop-in replacement for CPython 2.7.6 or 3.2.5. It's fast due to its integrated tracing JIT compiler. This release supports x86 machines running Linux 32/64, Mac OS X 64, Windows, and OpenBSD, as well as newer ARM hardware (ARMv6 or ARMv7, with VFPv3) running Linux. While we support 32 bit python on Windows, work on the native Windows 64 bit python is still stalling, we would welcome a volunteer to `handle that`_. .. _`handle that`: http://doc.pypy.org/en/latest/windows.html#what-is-missing-for-a-full-64-bi… How to use PyPy? ================= We suggest using PyPy from a `virtualenv`_. Once you have a virtualenv installed, you can follow instructions from `pypy documentation`_ on how to proceed. This document also covers other `installation schemes`_. .. _donated: http://morepypy.blogspot.com/2012/01/py3k-and-numpy-first-stage-thanks-to.h… .. _`py3k proposal`: http://pypy.org/py3donate.html .. _`pypy documentation`: http://doc.pypy.org/en/latest/getting-started.html#installing-using-virtual… .. _`virtualenv`: http://www.virtualenv.org/en/latest/ .. _`installation schemes`: http://doc.pypy.org/en/latest/getting-started.html#installing-pypy Cheers, the PyPy team -- Philip Jenvey

1 0

pypy-stm: not dead
by Armin Rigo 16 Jun '14

16 Jun '14

Hi all, Just a quick note to tell you sorry: although the 2.3 release of PyPy said the release of PyPy-STM would occur very soon, it's still delayed. We have found some real use cases in which it is several times slower than PyPy (both with the JIT) and so we're busy setting up a few missing things with the GC, notably the card marking scheme, which we think is causing it. Armin

1 1

wierd test failure when running tests in parallel
by Matti Picus 15 Jun '14

15 Jun '14

on windows own build, tests that call ctypes2lltype with an Unsigned failed last night's build, for instance http://buildbot.pypy.org/summary/longrepr?testname=test_connect_ex&builder=… Running this test alone passes. Two questions 1. Should there be a fastpath like elif T is lltype.Unsigned: llobj = cobj 2. how can ctypes2lltype(Unsigned, 784L) fail? Note: I have started running two test processes in parallel, which may be creating many more open file handles. Matti

2 1

--sandbox woes. H ow do I install modules?
by Barry, John 14 Jun '14

14 Jun '14

Hoping someone here can point me in the right direction. I'm attempting to create a sandbox, load it up with our requirements via pip, and start figuring out how to execute this from our codebase. It all falls down every single time when I attempt to install distribute.py, because import time fails within the sandbox. So the question is twofold: (1) How can I get import time working (as we call datetime.datetime.utcnow() with astonishing regularity) (2) How do I install additional packages into the sandbox if I can't install pip (potentially not an issue if import time reoslves it) -- Confidentiality Notice: This e-mail transmission may contain confidential or legally privileged information that is intended only for the individual or entity named in the e-mail address. If you have received this communication in error, please notify me by return e-mail, and destroy this communication and all copies thereof, including any attachments. Apervita ® is a registered trademark of Pervasive Health Inc.

2 1

PYPYLOG and colon as a special character
by Matti Picus 13 Jun '14

13 Jun '14

if PYPYLOG is a filename, we log to it (except debug_print) if PYPYLOG is :filename, we log debug_print as well if PYPYLOG is section:filename, we log only that section But : is a valid char for filename on windows, so something like PYPYLOG=C:\TEMP\log.log does the wrong thing, and even worse, does so silently. Matti

3 3

failing test for on windows
by Matti Picus 12 Jun '14

12 Jun '14

We are down to the last 50 own tests on win32. Here is one I need help with: http://buildbot.pypy.org/summary/longrepr?testname=TestLowLevelType.%28%29.… The failure is that, like ARM, ssize == llmemory.sizeof(Signed) * 2 where ssize is 8 Is the test wrong or is there a problem with overallocating memory? Matti

1 0

test_round_up_for_allocation is failing on windows
by Matti Picus 12 Jun '14

12 Jun '14

1 0

my notes about trying to build unipycation
by nejaka osoba 12 Jun '14

12 Jun '14

https://gist.github.com/koo5/393746b19624debb2472

1 0

Numpy on pypy
by Gayathri J 12 Jun '14

12 Jun '14

Hi I am trying to do a matrix dot using numpy.dot the numpy.dot on pypy seems lot slower than bumpy.do on python is there an alternative to using numpy.dot - so that i can do matrix.dot on pypy faster than cpython.... suggestions would be very helpful thanks Gayathri

2 1

Matrix dot product
by Gayathri J 11 Jun '14

11 Jun '14

Following is my python code for matrix multiplication *import time* *a=random.rand(1**28,128,128,15)* *b=np.random.rand(15,15)* *c=np.zeros((128,128,128,15), dtype=np.float64)* *t=time.time()* *c=a.dot(b)* *print time.time()-t* Timing for *pypy -> 3.957* *python 2.7 -> .80* So find that the basic matrix dot operation is faster on python 2.7 than on pypy 2.2.1 why is this so? am i missing something? thanks gayathri

3 2