Hacker News

Favorites Setup
Comment by throwaway81523 | original | RISC-V is now officially supported by CPython
[−]throwaway81523 · 2026-08-30 Sun 08:22 UTC · link
CPython is written in C, did running it on Risc-V take more than a recompilation? Were there any surprises? Yes of course it needs testing and RISC-V in the CI stack, but I'd expect fairly smooth sailing.
[−]tyingq · 2026-08-30 Sun 08:50 UTC · link
There is a Configure type subsystem, so you'll see stuff like this: https://github.com/python/cpython/pull/156277/changes#diff-4...
[−]emmatyping · 2026-08-30 Sun 09:20 UTC · link
In short, we haven't found many bugs, but it also wasn't just recompile on a new platform.

We've seen test failures like https://github.com/python/cpython/issues/151040

And perf support needs to be tested and merged https://github.com/python/cpython/issues/121201

Overall though, fairly smooth sailing as you say.

The reason RISC-V wasn't already supported is a mix of lacking hardware access for build bots and committers willing to pledge time to support it.

[−]shash · 2026-08-30 Sun 09:52 UTC · link
We ran it 4 years ago (including numpy) and it wasn’t a massive uphill climb or anything. It’s gotten a little easier since then if anything.
[−]rzulasf · 2026-08-30 Sun 11:15 UTC · link
Yes, RISC-V usually compiles out of the box. They just need a new "milestone" so the corporations that hire the remaining mediocre core developers think they are doing something.

It's the same with WASI integration. Two Junta members already commit to that failed project in order to show presence and activity.

It is a cancel and buzzword driven project.

[−]tlk-22815 · 2026-08-30 Sun 14:06 UTC · link
We ran python-3.14 on RISC-V before switching to C++ only. It compiled without any errors and all tests passed.

I don't understand the significance of this announcement in the submission.