Mailman 3 September 2017 - pypy-dev

About gurobipy
by Luis José Novoa 31 May '18

31 May '18

Hi All, My name is Luis Novoa. Im working on a large scale optimization problem using python/gurobipy for my implementation. I was wondering if any of you in this list has used gurobipy with pypy, or if it is even possible at this time. Thanks in advance, Luis.

9 9

PhD Positions in programming language runtimes (Python)
by David Bremner 02 May '18

02 May '18

1 Position I: Python on OMR =========================== The candidate will explore the benefits and challenges of using OMR [1] components (e.g. JIT, GC) in Python [2,3] runtimes. This may involve extending OMR in novel ways to better support Python, or vice-versa. If questions like how to balance fast startup time for scripting applications against good performance for compute-intensive workloads sound interesting, this might an opportunity for you. 2 Position II: Python and PGAS ============================== The candidate will implement support for parallel programming with Python, using the PGAS [4] (parallel global address space) model. The PGAS model can significantly simplify the implementation of parallel algorithms on distributed and shared memory computers. Challenges include enabling parallelism while dealing with Python's Global Interpreter Lock (GIL), and implementing distributed garbage collection efficiently. 3 Qualifications ================ Normally applicants to the Computer Science PhD program at UNB require a Master's degree in computer science or a closely related area. This project will require substantial hands-on and experimental work, and being comfortable with Python and C++. Familiarity with Linux and Open Source Software development would be a plus. For position I, exposure to programming language concepts like GC, JIT, and bytecode would be helpful. For position II, experience with parallel programming frameworks and/or languages such as MPI, OpenMP, X10 is an asset. 4 Environment ============= This project is supported by the IBM Centre for Advanced Studies (Atlantic) at the University of New Brunswick [5]. We have a lively group of students, postdocs, and research assistants working on related problems. Students work closely with both UNB researchers and technical experts from IBM, gaining valuable perspective and experience. 5 Support ========= These positions are "fully funded"; the exact funding package is subject to finalization but will be consistent with other fully funded students in Computer Science at UNB. You can find some information about expenses on the UNB web site [6]. 6 Contact ========= Send your resumé and unofficial transcripts by email to Stephen MacKay <smackay3(a)unb.ca> Please specify which position(s) you are interested in. [1]: [http://www.eclipse.org/omr/] [2]: [http://www.python.org] [3]: [http://www.pypy.org] [4]: [http://www.pgas.org/] [5]: [http://www.unb.ca/research/casatlantic/] [6]: [http://www.unb.ca/fredericton/studentservices/international/new_internation…]

1 1

Delimited Continuations in PyPy?
by Timothy Baldridge 15 Dec '17

15 Dec '17

I recently came across this paper: https://www.microsoft.com/en-us/research/publication/implementing-algebraic… The gist is that they use setjmp/longjmp to create delimited continuations in compliant C99. It seems like this method is fairly close to what PyPy already does with stacklets. Any reason why this couldn't be ported to PyPy? Of course there'd have to be some sort of stack integration as with stacklets, but it seems that that mostly boils down to making sure the GC has a way to trace the bits of the stack contained in the continuations. I'm considering trying my hand at using their library via rffi, and copying what I need from _stacklet_shadowstack.py. But perhaps there's something I'm missing that makes this a no-go from the start. Thanks, Timothy Baldridge

3 3

Automated binding generation (and maintenance)
by Shaheed Haque 11 Oct '17

11 Oct '17

Hi, I have been looking at the problem of automated binding generation (and maintenance) for large C++ code bases for a little while now [1], but am new to cppyy. One issue I am struggling to find a good solution for is to generate an accurate list of the objects (classes, functions, variables etc) in a given header file in order to populate the selection .XML. Ideally, I'd like to be able to say "all objects in this translation unit". I tried the wildcard "*" but I believe that selects the transitive fanout (and runs into errors). Short of running Clang directly to generate the names, what options do I have? (Currently, I'm working around this by manually specifying a narrower wildcard such as "KJS*"). Also, as I looked around for approaches to this issue, I noted that the cppyy backend v610 5 source code has a "getAllClasses" whereas PyPI has 6.10.0.2. I'm not sure of the mapping of versions, but what is the cadence for updates to PyPI? [1] https://marc.info/?l=kde-core-devel&m=150464598710128&w=2

2 8

Tkinter ?
by Stuart Axon 30 Sep '17

30 Sep '17

Hi all, Is TkInter pypy still a thing ? I was trying some demos from aubio from pypy but they didn't work [1].My first thought was to try and install tkinter-pypy but that doens't seem to exist any more. I'm testing with pypy-nightly trunk on Ubuntu 17.04 64 bit. [1]$ python demo_keyboard.py Traceback (most recent call last): File "demo_keyboard.py", line 60, in <module> import matplotlib.pyplot as plt File "/home/stu/.virtualenvs/pypy-nightly-trunk/site-packages/matplotlib/pyplot.py", line 115, in <module> _backend_mod, new_figure_manager, draw_if_interactive, _show = pylab_setup() File "/home/stu/.virtualenvs/pypy-nightly-trunk/site-packages/matplotlib/backends/__init__.py", line 32, in pylab_setup globals(),locals(),[backend_name],0) File "/home/stu/.virtualenvs/pypy-nightly-trunk/site-packages/matplotlib/backends/backend_tkagg.py", line 13, in <module> import matplotlib.backends.tkagg as tkagg File "/home/stu/.virtualenvs/pypy-nightly-trunk/site-packages/matplotlib/backends/tkagg.py", line 9, in <module> from matplotlib.backends import _tkagg RuntimeError: Cannot dlopen tkinter module file S++

4 5

Numpy or Numpypy
by Stuart Axon 29 Sep '17

29 Sep '17

Hi, I'm trying out aubio and pysoundcard with pypy nightly. They use numpy, should I still be using numpypy or is upstream numpy the way to go ? S++

2 1

Explicit typing in RPython
by joannah nanjekye 28 Sep '17

28 Sep '17

Hello, Did someone happen to work on this project? If not then is it one we can break down in small tasks so I can work on it with guidance of course. Also I will be at pyconZA in south Africa so if anyone on the team is coming we can chat about this a bit. Kind regards, -- Joannah Nanjekye +256776468213 F : Nanjekye Captain Joannah S : joannah.nanjekye T : @Captain_Joannah SO : joannah *"You think you know when you learn, are more sure when you can write, even more when you can teach, but certain when you can program." Alan J. Perlis*

2 1

Pip Install Fails Due to cppyy-backend setup.py Bug
by Parker Boyce 27 Sep '17

27 Sep '17

I recently had reason to try to install cppyy using Python 3 and pip. With the use of the wheels for cppyy-cling the process went smoothly enough until pip tried to build cppyy-backend <https://pypi.python.org/pypi/cppyy-backend/0.2.0>. Building this package by itself results in the following error message; $ python3 setup.py build running build running build_ext x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -g -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -Ib'/usr/local/lib/python3.5/dist-packages/cppyy_backend/include' -I/usr/include/python3.5m -c src/clingwrapper.cxx -o build/temp.linux-x86_64-3.5/src/clingwrapper.o -std=c++11 -O2 cc1plus: warning: command line option ‘-Wstrict-prototypes’ is valid for C/ObjC but not for C++ In file included from src/clingwrapper.cxx:4:0: src/callcontext.h:11:20: fatal error: Rtypes.h: No such file or directory compilation terminated. error: command 'x86_64-linux-gnu-gcc' failed with exit status 1 A careful look at the command line argument for gcc reveals a mangled include argument, -Ib'/usr/local/lib/python3.5/dist-packages/cppyy_backend/include'. This is a binary literal string introduced by the output of a system call, "cling-config --cppflags". The functions in the subprocess module return 'bytes' objects in Python 3. In this case the return value from 'check_output' was appended wholesale into the 'include_dirs' argument of your extension, and the representation thereof was used in turn to build the function call to gcc. (The representation being a binary literal string) To fix this the bytes object returned by subprocess.check_output needs to be decoded to either 'utf-8' or 'ascii', preferably the former as Python 3 uses unicode literals. This can be accomplished by appending '.decode("utf-8")' to line 28; --- /home/parker/Documents/cppyy/cppyy-backend-0.2.0/setup.py Fri Sep 22 12:25:29 2017 +++ /home/parker/Documents/cppyy/cppyy-backend-0.2.0/setup.py.patched Tue Sep 26 19:37:50 2017 @@ -25,7 +25,7 @@ if root_install: return os.path.join(root_install, 'include') cli_arg = subprocess.check_output(['cling-config', '--cppflags']) - return cli_arg[2:-1] + return cli_arg[2:-1].decode("utf-8") class my_build_cpplib(_build_ext): def build_extension(self, ext): With this patch pip will be able to build the cppyy-backend package from source on Python 2 or 3, and pip installing cppyy will not require user intervention. P.S. The "Comments and Bugs" section of the cppyy docs directs the reader to a repository with issue tracking turned off, which is why you're getting this email instead.

2 1

macosx buildbot is very slow
by Matti Picus 25 Sep '17

25 Sep '17

http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64 Is the macosx buildbot OK? It seems to be running very slow, translations take over 6 hours. I just noticed this while trying some extra translations for the 5.9 release cycle. Also it seems the buildbot is running PyPy 2.5.1, you might want to upgrade that if possible. Matti

3 2

PyPy 5.9 release cycle has begun
by Matti Picus 25 Sep '17

25 Sep '17

Looking at the unfinished release notes http://pypy.readthedocs.io/en/latest/release-v5.9.0.html shows a general maturity in PyPy2.7 and strides forward in PyPy3.5 beta, available only for linux64. Any suggestions for "killer features" to appear at the top of the release notes, please chime in. We are down to a handful of corner cases failing in pandas and numpy on PyPy2.7, are there more achievements? Please try out the beta release versions 92449-c2437cf9b7f1 available here http://buildbot.pypy.org/nightly/release-pypy2.7-5.x Matti

3 3