Pyjion – A JIT for Python based upon CoreCLR (opens in new tab)

(github.com)

72 pointsLofkin10y ago23 comments

23 comments

One of the most promissing projects :

  * it keeps CPython as is.
  * hence retains compat with extensions.
  * but still has the potential to improve performance like pypy.
  * and yet can be shipped just as an additional compiled lib inside the Python dist.

If it ever works, it will be wonderful.

mozumder10y ago

It would also be nice if it worked outside of Windows.

jsherer10y ago

I had the same though and looked into it. It seems that CoreCLR isn't the limitation there. It supports a variety of platforms: https://github.com/dotnet/coreclr#build-status

Looks like the authors are focused on Windows for their initial prototype development.

1 more reply

smortaz10y ago

if the speedups pan out, that is the plan! the work on performance has not been started yet however.

2 more replies

andor10y ago

Here's their list of passing and failing CPython tests:

https://github.com/Microsoft/Pyjion/blob/master/Tests/python...

brettcannon10y ago

One thing to note about the list of failing tests is the vast majority of them are due to a single bug (or two): https://github.com/Microsoft/Pyjion/issues/103 . And we currently only have 11 open bugs so the test failures make it look worse than it is.

brettcannon10y ago

Thanks to all of the great questions we got, we have now written an FAQ for the project: https://github.com/Microsoft/Pyjion#faq

the_gigi10y ago

How does Pygion compare with IronPython?

ericfrederich10y ago

From reading the description, and from what little I know about IronPython...

IronPython is a re-implementation of Python in C# which allows it to play nice with .NET. This is similar to Jython which is a re-implementation of Python in Java that plays nice with the JVM.

This project on the other had is not a re-implementation at all, but aims to augment the existing CPython implementation with JIT APIs and is using CoreCLR for the JIT.

brettcannon10y ago

What ericfrederich said is all accurate. We are also considering playing with ChakraCore as a JIT back-end as well once we are sure that our abstract interpreter code that drives the JIT emission code is compatible with Python semantics.

1 more reply

ericfrederich10y ago

Is there a PEP for this, or is it too early, more of a proof of concept?

brettcannon10y ago

It's too early. We want to prove to ourselves that we can get the performance we want before proposing the changes we think we want to CPython's C API.

j / k navigate · click thread line to collapse

23 comments

sametmax10y ago

One of the most promissing projects :

  * it keeps CPython as is.
  * hence retains compat with extensions.
  * but still has the potential to improve performance like pypy.
  * and yet can be shipped just as an additional compiled lib inside the Python dist.

If it ever works, it will be wonderful.

mozumder10y ago

It would also be nice if it worked outside of Windows.

jsherer10y ago

I had the same though and looked into it. It seems that CoreCLR isn't the limitation there. It supports a variety of platforms: https://github.com/dotnet/coreclr#build-status

Looks like the authors are focused on Windows for their initial prototype development.

1 more reply

smortaz10y ago

if the speedups pan out, that is the plan! the work on performance has not been started yet however.

2 more replies

andor10y ago

Here's their list of passing and failing CPython tests:

https://github.com/Microsoft/Pyjion/blob/master/Tests/python...

brettcannon10y ago

Thanks to all of the great questions we got, we have now written an FAQ for the project: https://github.com/Microsoft/Pyjion#faq

the_gigi10y ago

How does Pygion compare with IronPython?

ericfrederich10y ago

From reading the description, and from what little I know about IronPython...

IronPython is a re-implementation of Python in C# which allows it to play nice with .NET. This is similar to Jython which is a re-implementation of Python in Java that plays nice with the JVM.

This project on the other had is not a re-implementation at all, but aims to augment the existing CPython implementation with JIT APIs and is using CoreCLR for the JIT.

brettcannon10y ago

1 more reply

ericfrederich10y ago

Is there a PEP for this, or is it too early, more of a proof of concept?

brettcannon10y ago

It's too early. We want to prove to ourselves that we can get the performance we want before proposing the changes we think we want to CPython's C API.

j / k navigate · click thread line to collapse