Use freelist only on CPython#349
Merged
Merged
Conversation
This optimization seems working on the GraalPy, but looks useless:
collatz0(871): Mean +- std dev: [nocache-graalpy] 2.19 ms +- 1.14 ms -> [graalpy] 5.12 ms +- 3.54 ms: 2.34x slower
collatz0((1<<128)+31): Mean +- std dev: [nocache-graalpy] 25.6 ms +- 13.5 ms -> [graalpy] 17.6 ms +- 14.4 ms: 1.45x faster
collatz1(97): Mean +- std dev: [nocache-graalpy] 4.40 ms +- 3.17 ms -> [graalpy] 2.15 ms +- 1.98 ms: 2.05x faster
collatz1(871): Mean +- std dev: [nocache-graalpy] 2.49 ms +- 2.22 ms -> [graalpy] 4.84 ms +- 3.92 ms: 1.94x slower
collatz2(871): Mean +- std dev: [nocache-graalpy] 2.24 ms +- 1.11 ms -> [graalpy] 5.06 ms +- 3.67 ms: 2.26x slower
collatz2((1<<128)+31): Mean +- std dev: [nocache-graalpy] 11.6 ms +- 8.2 ms -> [graalpy] 25.9 ms +- 15.1 ms: 2.23x slower
Benchmark hidden because not significant (3): collatz0(97), collatz1((1<<128)+31), collatz2(97)
Geometric mean: 1.26x slower
While on CPython:
collatz0(97): Mean +- std dev: [nocache-cpython] 122 us +- 2 us -> [cpython] 87.4 us +- 1.6 us: 1.40x faster
collatz0(871): Mean +- std dev: [nocache-cpython] 184 us +- 3 us -> [cpython] 131 us +- 4 us: 1.40x faster
collatz0((1<<128)+31): Mean +- std dev: [nocache-cpython] 968 us +- 17 us -> [cpython] 699 us +- 32 us: 1.38x faster
collatz1(97): Mean +- std dev: [nocache-cpython] 152 us +- 4 us -> [cpython] 105 us +- 3 us: 1.45x faster
collatz1(871): Mean +- std dev: [nocache-cpython] 231 us +- 13 us -> [cpython] 158 us +- 1 us: 1.47x faster
collatz1((1<<128)+31): Mean +- std dev: [nocache-cpython] 1.22 ms +- 0.05 ms -> [cpython] 847 us +- 14 us: 1.43x faster
collatz2(97): Mean +- std dev: [nocache-cpython] 124 us +- 2 us -> [cpython] 87.4 us +- 0.4 us: 1.41x faster
collatz2(871): Mean +- std dev: [nocache-cpython] 187 us +- 7 us -> [cpython] 131 us +- 1 us: 1.43x faster
collatz2((1<<128)+31): Mean +- std dev: [nocache-cpython] 988 us +- 24 us -> [cpython] 703 us +- 2 us: 1.41x faster
Geometric mean: 1.42x faster
This approach doesn't work on PyPy (triggers "Invalid usage of a dying
CPython object" error), but hardly it does make sense here. For
instance, following pure-Python benchmark shows same numbers on PyPy,
regardless on freelist size:
```
import os
import pyperf
from operator import add
_freelist = []
_max_freelist_size = int(os.getenv("FREELIST_SIZE"))
class xyz:
def __new__(cls, *args, **kwargs):
if _freelist:
# Reuse an existing instance from the pool
instance = _freelist.pop()
return instance
# Fall back to standard memory allocation if pool is empty
return super().__new__(cls)
def __init__(self, value):
self.value = value
def __del__(self):
if len(_freelist) < _max_freelist_size:
# Resurrect 'self' by adding it to the free-list
_freelist.append(self)
def __add__(self, other):
return xyz(self.value + other.value)
def __repr__(self):
return f"xyz('{self.value}')"
x, y = map(xyz, [123, 321])
runner = pyperf.Runner()
s = repr(x) + " + " + repr(y)
runner.bench_func(s, add, x, y)
```
Closes diofant#301
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
closes #301