Repository navigation
Stackrefs (#118450) introduced a large performance regression on Windows #121263
Description
Activity
@mdboom Thanks for bringing this up! In the past this was usually due to MSVC not being able to inline in the eval loop because its too big of a function. I can change all the static inlines to macros and try benchmarking again?
I can change all the static inlines to macros and try benchmarking again?
Sounds like a good first thing to try. Thanks!
Bug description:
#118450 "Convert the evaluation stack to stack refs" (by @Fidget-Spinner) unfortunately seems to have introduced a 7% - 11% performance regression overall (though as much as 25% for some benchmarks) on Windows only. Linux (both x86_64 and aarch64) and macOS (arm) seem to have a change beneath the noise threshold. To be clear, @Fidget-Spinner's work is great and the lesson here is that for these sorts of sweeping low-level changes, we should have benchmarked on all of the Tier 1 platforms to be certain.
I don't know what the cause is -- it is probably MSVC not being able to "optimize through" something and introducing unnecessary overhead, but that's just a theory.
Cc: @brandtbucher, @markshannon
CPython versions tested on:
CPython main branch
Operating systems tested on:
Windows
Linked PRs