Repository navigation
Caching the tuple hash calculation speeds up some code significantly #131525
Description
Proposal:
Back in 2013, it was determined that caching the result of tuple_hash did not have any significant speedup.
However, a lot has changed since then, and in a recent experiment to add a tuple hash cache back in, the mdp benchmark increased by 86%.
Admittedly, there was no measurable improvement on any other benchmark, but it also seems to have no downside, including for memory usage when measured with max_rss.
Has this already been discussed elsewhere?
No response given
Links to previous discussion of this feature:
- [ ]
Activity
pythongh-131525: Cache the result of tuple_hash
pythongh-131525: Cache the result of tuple_hash
rhettinger commentedon Mar 20, 2025
This would nicely simplify and speed-up the pure python implementation of
lru_cache. The_HashedSeqclass would no longer be needed, and a plain tuple would suffice.gh-131525: Cache the result of tuple_hash (#131529)
pythongh-131525: Remove `_HashedSeq` wrapper from `lru_cache`
pythongh-131525: Remove `_HashedSeq` wrapper from `lru_cache`
gh-131525: Remove `_HashedSeq` wrapper from `lru_cache` (gh-131922)
pythongh-131525: Cache the result of tuple_hash (python#131529)
erlend-aasland commentedon May 14, 2025
Is there more work to be done here?