Mailman 3 December 2020 - pypy-dev

PyPy Winter Sprint + cpyext
by Stefan Beyer Dec. 21, 2022

Dec. 21, 2022

Hey everyone! Those of you who attended or followed one of the last two winter sprints in Leysin might know me, the others probably won't. I'm (still) a Master's student writing my thesis on PyPy's current issue with cross-heap cycles when using cpyext. The main point is, that they are not collected and stay around as floating garbage, even if they are not reachable any more. Correct me if I'm wrong, but I didn't notice anyone else working on that topic (and/or committing something) since I've picked it up two years ago (yeah, I know, that long ago...). I saw that you would be working on "cpyext performance and completeness" during the current winter sprint this week and thought that this might also concern my thesis. So I thought I'll give you an update. I recently pushed my current (non-optimized, breaking-some-tests, but more or less working) implementation of the "CPython-style" GC-extension for cpyext. It is still in an early stage and not all cases (legacy and non-legacy-finalizers, weakrefs) are handled. However, I picked up some pace during the last couple of weeks and I'm determined to finish this implementation during the following weeks (some quirks with the tests have been haunting me, but I think I figured most of them out by now). After that, I will implement a second alternative implementation (to compare to, mostly for the sake of my thesis), which will take some more time. I also added a couple of fancy test cases (the so called "dot-tests"), so that we can test complex object graphs a little bit easier (and also because it was kind of cool to have another language inside PyPy/RPython, even if it was only for the tests...), with more test cases to come (the current ones are a bit messy). None of the new changes concerning low-latency applications are currently integrated, but that should not be too hard. I guess it won't make much sense for me to join you at this year's winter sprint, especially as I won't be able to get there before Thursday, but I might be able to join you over the IRC channel (or some other form of communication if you like). If there is anything that is worth discussing please let me know! Also feel free to comment on my code, but beware that I might change some things once I try to do some optimizations (probably not many, but at least fix the worst issues) and make it a little bit more readable. You can find my code on the cpyext-gc-cycle branch.<https://bitbucket.org/pypy/pypy/commits/branch/cpyext-gc-cycle> Looking forward to hearing from you! Greetings, Stefan

2 2

Contributing Polyhedral Optimisations in PyPy
by muke101 Jan. 19, 2021

Jan. 19, 2021

I'm doing a computer science masters and am looking for an appropriate project to take on for a dissertation related to Polyhedral optimisations. Talking to my professor, we both think trying to implement the model and it's loop transformations in PyPy's JIT optimiser could be a good project to pursue, but before committing to anything I wanted to run this idea by the devs here who might be able to point out any hurdles I'd be likely to quickly come across that could prove difficult to solve at just a masters level, or whether or not these optimisations are actually already implemented in the first place (I have tried to google if this is the case and hadn't found anything, but can't be sure). I think this could have some good real world impact too as a lot of scientific code is written in Python and run on PyPy, and the Polyhedral model can offer substantial performance improvements in the form of auto-parallelization for these types of codes, which is why I'm interested in working on this for PyPy rather than CPython, although if anyone has good reason that I might want to look at CPython for this over PyPy please let me know. Appreciate any and all advice, thanks.

4 6

I’m happy to donate an M1 Mac to you
by Peter Vessenes Dec. 31, 2020

Dec. 31, 2020

Hey, Happy New Year — pypy is great. Please let me know where I can send an M1 Mac Mini, and I’ll ship it to you ASAP. A quick thanks to me as a followup edit would be nice, but not required. :) Peter Vessenes Sent from my iPad

1 1

Differences performance Julia / PyPy on very similar codes
by PIERRE AUGIER Dec. 26, 2020

Dec. 26, 2020

Hi, I post on this list a message written in PyPy issue tracker (https://foss.heptapod.net/pypy/pypy/-/issues/3349#note_150255). It is about some experiments I did on writing efficient implementations of the NBody problem https://github.com/paugier/nbabel to potentially answer to this article https://arxiv.org/pdf/2009.11295.pdf. I get from a PR an [interesting optimized implementation in Julia](https://github.com/paugier/nbabel/blob/master/julia/nbabel4_serial.j…. It is very fast (even slightly faster than in Pythran). One idea is to store the 3 floats of a 3d physical vector, (x, y, z), in a struct `Point4D` containing 4 floats to better use simd instructions. I added a pure Python implementation inspired by this new Julia implementation (but with a simple `Point3D` with 3 floats because with PyPy, the `Point4D` does not make the code faster) and good news it is with PyPy a bit faster than our previous PyPy implementations (only 3 times slower than the old C++ implementation). However, it is much slower than with Julia (while the code is very similar). I coded a simplified version in Julia with nearly nothing else that what can be written in pure Python (in particular, no `@inbounds` and `@simd` macros). It seems to me that the comparison of these 2 versions could be interesting. So I again simplified these 2 versions to keep only what is important for performance, which gives - https://github.com/paugier/nbabel/blob/master/py/microbench_pypy4.py - https://github.com/paugier/nbabel/blob/master/py/microbench_ju4.jl The results are summarized in https://github.com/paugier/nbabel/blob/master/py/microbench.md An important point is that with `Point3D` (a mutable class in Python and an immutable struct in Julia), Julia is 3.6 times faster than PyPy. Same code and nothing really fancy in Julia so I guess that PyPy might be missing some optimization opportunities. At least it would be interesting to understand what is slower in PyPy (and why). I have to admit that I don't know how to get interesting information on timing and what is happening with PyPy JIT in a particular case. I only used cProfile and it's of course clearly not enough. I can run vmprof but I'm not able to visualize the data because the website http://vmprof.com/ is down. I don't know if I can trust values given by IPython `%timeit` for particular instructions since I don't know if PyPy JIT does the same thing in `%timeit` and in the function `compute_accelerations`. I also feel that I really miss in pure Python an efficient fixed size homogeneous mutable sequence (a "Vector" in Julia words) that can contain basic numerical types (as Python `array.array`) but also instances of user-defined classes and instances of Vectors. The Python code uses a [pure Python implementation using a list](https://github.com/paugier/nbabel/blob/master/py/vector.py). I think it would be reasonable to have a good implementation highly compatible with PyPy (and potentially other Python implementations) in a package on PyPI. It would really help to write PyPy compatible numerical codes. What would be the good tool to implement such package? HPy? I wonder whether we can get some speedup compared to the pure Python version with lists. For very simple classes like `Point3d` and `Point4d`, I wonder if the data could be saved continuously in memory and if some operations could be done without boxing/unboxing. However, I really don't know what is slower in PyPy / faster in Julia. I would be very interested to get the points of view of people knowing well PyPy. Pierre

6 13

Pickling generators
by Ioannis Foufoulas Dec. 11, 2020

Dec. 11, 2020

Hi, While in PyPy was possible to pickle a generator function and resume after unpicking this does not happen with PyPy3: File "/snap/pypy3/72/lib-python/3/pickle.py", line 942, in save_global (obj, module_name, name)) pickle.PicklingError: Can't pickle <class 'generator'>: it's not found as builtins.generator Is this a bug or this feature is removed in PyPy3? Thanks, Yannis

4 3