Apparently(?) this also needs to be attached to the function declarator and does not work as a function specifier: `static void *__preserve_none slowpath();` and not `__preserve_none static void *slowpath();` (unlike GCC attribute syntax, which tends to be fairly gung-ho about this sort of thing, sometimes with confusing results).
Yay to getting undocumented MSVC features disclosed if Microsoft thinks you’re important enough :/
Important enough, or benefits them directly? I have no good guesses how improving Python's performance would benefit them, but I would guess that's the real reason.
I guess there are some Python workloads on Azure, Microsoft provides a lot of data analysis and LLM tools as a service (not paid by CPU minutes). Saving CPU cycles there directly translates to financial savings.
Im a bit out of the loop with this, but hope its not like that time with python 3.14, when it was claimed a geometric mean speedup of about 9-15% over the standard interpreter when built with Clang 19. It turned out the results were inflated due to a bug in LLVM 19 that prevented proper "tail duplication" optimization in the baseline interpreter's dispatch loop. Actual gains was aprox 4%.
Edit: Read through it and have come to the conclusion that the post is 100% OK and properly framed: He explicitly says his approach is to "sharing early and making a fool of myself," prioritizing transparency and rapid iteration over ironclad verification upfront.
One could make an argument that he should have cross-compiler checks, independent audits, or delayed announcements until results are bulletproof across all platforms. But given that he is 100% transparent with his thinking and how he works, it's all good in the hood.
Python’s goal is never really to be fast. If that were its goal, it would’ve had a JIT long ago instead of toying with optimizing the interpreter. Guido prioritized code simplicity over speed. A lot of speed improvements including the JIT (PEP 744 – JIT Compilation) came about after he stepped down.
TLDR: The tail-calling interpreter is slightly faster than computed goto.
> I used to believe the the tailcalling interpreters get their speedup from better register use. While I still believe that now, I suspect that is not the main reason for speedups in CPython.
> My main guess now is that tail calling resets compiler heuristics to sane levels, so that compilers can do their jobs.
> Let me show an example, at the time of writing, CPython 3.15’s interpreter loop is around 12k lines of C code. That’s 12k lines in a single function for the switch-case and computed goto interpreter.
> […] In short, this overly large function breaks a lot of compiler heuristics.
> One of the most beneficial optimisations is inlining. In the past, we’ve found that compilers sometimes straight up refuse to inline even the simplest of functions in that 12k loc eval loop.
I think in the protobuf example the musttail did in fact benefit from better register use. All the functions are called with the same arguments, so there is no need to shuffle the registers. The same six register-passed arguments are reused from one function to the next.
The Python interpreter core loop sounds like the perfect problem for AlphaEvolve. Or it's open source equivalent OpenEvolve if DeepMind doesn't want to speed up Python for the competition.
The money shot (wish this were included in the blog post):
https://github.com/python/cpython/pull/143068/files#diff-45b...Apparently(?) this also needs to be attached to the function declarator and does not work as a function specifier: `static void *__preserve_none slowpath();` and not `__preserve_none static void *slowpath();` (unlike GCC attribute syntax, which tends to be fairly gung-ho about this sort of thing, sometimes with confusing results).
Yay to getting undocumented MSVC features disclosed if Microsoft thinks you’re important enough :/
Important enough, or benefits them directly? I have no good guesses how improving Python's performance would benefit them, but I would guess that's the real reason.
I guess there are some Python workloads on Azure, Microsoft provides a lot of data analysis and LLM tools as a service (not paid by CPU minutes). Saving CPU cycles there directly translates to financial savings.
I wonder if this is related to Python in Excel. You'll have lots of people running numerical stuff written in Python, running on Microsoft servers.
A lot of commercial engineering and scientific software runs on windows.
I don't understand what the scene in a porno when they cum has to do with python
That's not where the term comes from https://en.wikipedia.org/wiki/Money_shot
I don't understand why your mind went immediately to pornography. Kind of creepy tbh.
You're overreacting, see the wikipedia article posted above.
> By 1977[2][3] the phrase had entered American usage as slang for the cum shot in a pornographic film
https://en.wikipedia.org/wiki/Money_shot
Im a bit out of the loop with this, but hope its not like that time with python 3.14, when it was claimed a geometric mean speedup of about 9-15% over the standard interpreter when built with Clang 19. It turned out the results were inflated due to a bug in LLVM 19 that prevented proper "tail duplication" optimization in the baseline interpreter's dispatch loop. Actual gains was aprox 4%.
Edit: Read through it and have come to the conclusion that the post is 100% OK and properly framed: He explicitly says his approach is to "sharing early and making a fool of myself," prioritizing transparency and rapid iteration over ironclad verification upfront.
One could make an argument that he should have cross-compiler checks, independent audits, or delayed announcements until results are bulletproof across all platforms. But given that he is 100% transparent with his thinking and how he works, it's all good in the hood.
This seems like very low hanging fruit. How is the core loop not already hyper optimized?
I'd have expected it to be hand rolled assembly for the major ISAs, with a C backup for less common ones.
How much energy has been wasted worldwide because of a relatively unoptimized interpreter?
Python’s goal is never really to be fast. If that were its goal, it would’ve had a JIT long ago instead of toying with optimizing the interpreter. Guido prioritized code simplicity over speed. A lot of speed improvements including the JIT (PEP 744 – JIT Compilation) came about after he stepped down.
Software has gotten so slow we've forgotten how fast computers are
If you want fast just use pypy and forget about cpython.
> This has caused many issues for compilers in the past, too many to list in fact. I have a EuroPython 2025 talk about this.
Looks like it refers to this:
https://youtu.be/pUj32SF94Zw
(wish it's a link in the article)
TLDR: The tail-calling interpreter is slightly faster than computed goto.
> I used to believe the the tailcalling interpreters get their speedup from better register use. While I still believe that now, I suspect that is not the main reason for speedups in CPython.
> My main guess now is that tail calling resets compiler heuristics to sane levels, so that compilers can do their jobs.
> Let me show an example, at the time of writing, CPython 3.15’s interpreter loop is around 12k lines of C code. That’s 12k lines in a single function for the switch-case and computed goto interpreter.
> […] In short, this overly large function breaks a lot of compiler heuristics.
> One of the most beneficial optimisations is inlining. In the past, we’ve found that compilers sometimes straight up refuse to inline even the simplest of functions in that 12k loc eval loop.
I think in the protobuf example the musttail did in fact benefit from better register use. All the functions are called with the same arguments, so there is no need to shuffle the registers. The same six register-passed arguments are reused from one function to the next.
Does MSVC support computed goto?
MSVC mostly generates slower code than gcc/clang, so maybe this trick reduces the gap.
Is this backed by real evidence?
My experience is 10%-15% slower than GCC. That was 10 years ago though.
if the author of this blog reads this: can we can an RSS, please?
Got it. I'll try to set one up this weekend.
The Python interpreter core loop sounds like the perfect problem for AlphaEvolve. Or it's open source equivalent OpenEvolve if DeepMind doesn't want to speed up Python for the competition.