➜
This issue tracker has been migrated to GitHub,
and is currently read-only.
For more information,
see the GitHub FAQs in the Python's Developer Guide.
Created on 2021-04-02 17:36 by corona10, last changed 2022-04-11 14:59 by admin. This issue is now closed.
| Files | ||||
|---|---|---|---|---|
| File name | Uploaded | Description | Edit | |
| bench_enumerate.py | corona10, 2021-04-02 17:36 | |||
| Pull Requests | |||
|---|---|---|---|
| URL | Status | Linked | Edit |
| PR 25154 | merged | corona10, 2021-04-02 17:53 | |
| PR 30904 | merged | JelleZijlstra, 2022-01-26 03:39 | |
| Messages (5) | |||
|---|---|---|---|
| msg390083 - (view) | Author: Dong-hee Na (corona10) *  |
Date: 2021-04-02 17:36 | |
Finally, I success to implement PEP 590 for enumerate type which is the well-used type in Python. Amazingly enhanced! +-----------------+------------------------+----------------------------+ | Benchmark | enumerate_bench_master | enumerate_bench_vectorcall | +=================+========================+============================+ | bench enumerate | 527 ns | 380 ns: 1.39x faster | +-----------------+------------------------+----------------------------+ |
|||
| msg404705 - (view) | Author: Dong-hee Na (corona10) * |
Date: 2021-10-21 23:19 | |
See my last benchmark https://github.com/python/cpython/pull/25154#issuecomment-949061515 Same result in both Linux VM and Macbook pro. |
|||
| msg404706 - (view) | Author: Dong-hee Na (corona10) * |
Date: 2021-10-21 23:20 | |
New changeset 83f202a802b0810e2f0900c6d298fd104390f2ba by Dong-hee Na in branch 'main': bpo-43706: Use PEP 590 vectorcall to speed up enumerate() (GH-25154) https://github.com/python/cpython/commit/83f202a802b0810e2f0900c6d298fd104390f2ba |
|||
| msg404707 - (view) | Author: Dong-hee Na (corona10) * |
Date: 2021-10-21 23:23 | |
Thanks Victor! |
|||
| msg404717 - (view) | Author: Dong-hee Na (corona10) * |
Date: 2021-10-22 01:59 | |
@vstinner I noticed that the msg390083 was based on a non-optimized build. https://github.com/python/cpython/pull/25154#issuecomment-949061515 ran with the same build option condition. Here is the PGO + LTO based benchmark. Still fast. +-----------------+----------+----------------------+ | Benchmark | base_opt | vectorcall_opt | +=================+==========+======================+ | bench enumerate | 384 ns | 277 ns: 1.39x faster | +-----------------+----------+----------------------+ |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2022-04-11 14:59:43 | admin | set | github: 87872 |
| 2022-01-26 03:39:17 | JelleZijlstra | set | nosy:
+ JelleZijlstra pull_requests: + pull_request29083 |
| 2021-10-22 01:59:19 | corona10 | set | messages: + msg404717 |
| 2021-10-21 23:29:20 | corona10 | set | versions: + Python 3.11, - Python 3.10 |
| 2021-10-21 23:23:14 | corona10 | set | status: open -> closed messages: + msg404707 stage: patch review -> resolved |
| 2021-10-21 23:20:51 | corona10 | set | messages: + msg404706 |
| 2021-10-21 23:19:55 | corona10 | set | nosy:
+ kj |
| 2021-10-21 23:19:38 | corona10 | set | messages: + msg404705 |
| 2021-04-02 17:53:45 | corona10 | set | keywords:
+ patch stage: patch review pull_requests: + pull_request23901 |
| 2021-04-02 17:36:26 | corona10 | set | type: performance versions: + Python 3.10 |
| 2021-04-02 17:36:13 | corona10 | create | |