Description

kumaraditya303

opened on Aug 9, 2023

Issue body actions

Currently asyncio tasks are stored in a Weakset, this is inefficient and in some cases causes bugs because of thread safety (#80788). In terms of memory usage it requires maintaining a full set and their corresponding weakref callback to cleanup objects when deallocated and finalized by the gc. In applications where tasks are created at fast pace this becomes a bottle neck, to mitigate this now asyncio tasks will now be stored in a global double linked of tasks for cases where Task is a subclass of _asyncio.Task in other cases we still rely on the weakset. This reduces the work done by the gc speedups the execution and reduces memory usage. In some of my own benchmarks I have seen 15- 20% improvement and pyperformance benchmarks reflect roughly the same.

https://github.com/faster-cpython/benchmarking-public/blob/main/results/bm-20230805-3.13.0a0-1d32835/bm-20230805-linux-x86_64-kumaraditya303-linked_list-3.13.0a0-1d32835-vs-base.md

Updated: https://github.com/faster-cpython/benchmarking-public/tree/main/results/bm-20240622-3.14.0a0-4717aaa#vs-base

Linked PRs

• GH-107803: double linked list implementation for asyncio tasks #107804
• gh-107803: add whatsnew for asyncio double linked list implementation #120995
• gh-107803: fix thread safety issue in double linked list implementation  #121007
• GH-107803: use circular double linked list #126577

Reacted by Gregory P. Smith, Ruslan, siddish reddy, fruitoiz, Bakytzhan, Abraão Zaidan and J. Nick Koston

Reacted by Itamar Oren, Charlie Zhao, Jacob Bower, Kyriakos Psarakis, Gregory P. Smith, Maxim Martynov, Okeyo, ., benedikt-bartscher, fruitoiz and 4 more

Activity

1. kumaraditya303

   added

   performancePerformance or resource usagePerformance or resource usage

   topic-asyncio

   3.13only security fixesonly security fixes

   on Aug 9, 2023
2. kumaraditya303

   self-assigned this

   on Aug 9, 2023
3. github-project-automation

   added this to asyncioon Aug 9, 2023
4. github-project-automation

   moved this to Todo in asyncioon Aug 9, 2023
5. willingc

   added a commit that references this issue on Jun 22, 2024

   GH-107803: double linked list implementation for asyncio tasks (GH-10…

   4717aaa
6. kumaraditya303

   added

   3.14bugs and security fixesbugs and security fixes

   and removed

   3.13only security fixesonly security fixes

   on Jun 23, 2024
7. colesbury commented on Jun 24, 2024

   

   colesbury

   on Jun 24, 2024

   Contributor

   More actions

   I think there are some thread-safety issues with the linked list implementation:

   • It's not going to be thread-safe in the free-threaded build. We can probably add a lock or a critical section.
   • Even with the GIL, the linked list of tasks may be modified while it's being looped over because the GIL can be released in add_one_task.

   Reacted by Carol Willing
8. willingc commented on Jun 24, 2024

   

   willingc

   on Jun 24, 2024

   Contributor

   More actions

   Thanks @colesbury for flagging this.
9. kumaraditya303 commented on Jun 25, 2024

   

   kumaraditya303

   on Jun 25, 2024

   ContributorAuthor

   More actions

   I am closing this, will track free threaded issues in #120974

   Reacted by Carol Willing
10. kumaraditya303

    closed this as completedon Jun 25, 2024
11. github-project-automation

    moved this from Todo to Done in asyncioon Jun 25, 2024
12. colesbury commented on Jun 25, 2024

    

    colesbury

    on Jun 25, 2024

    Contributor

    More actions

    @kumaraditya303, can you fix the thread-safety/reentrancy issue that affects the default build (with the GIL)? head can be deallocated while it's being used. It needs to be protected by a Py_INCREF(). This is not specific to the free-threading build.

    cpython/Modules/_asynciomodule.c

    Lines 3711 to 3721 in 2d3187b

    	TaskObj *tail = &state->asyncio_tasks.tail;
    	while (head != tail)
    	{
    	if (add_one_task(state, tasks, (PyObject *)head, loop) < 0) {
    	Py_DECREF(tasks);
    	Py_DECREF(loop);
    	return NULL;
    	}
    	head = head->next;
    	assert(head != NULL);
    	}

    Reacted by Carol Willing
13. kumaraditya303

    added 2 commits that reference this issue on Jun 26, 2024

    gh-107803: fix thread safety issue in double linked list implementati…

    8223544

    gh-107803: add whatsnew for asyncio double linked list implementation (…

    58a3580
14. mrahtz

    added 3 commits that reference this issue on Jun 30, 2024

    pythonGH-107803: double linked list implementation for asyncio tasks (p…

    797d82f

    pythongh-107803: fix thread safety issue in double linked list implem…

    57e50d9

    pythongh-107803: add whatsnew for asyncio double linked list implemen…

    29e69e4
15. facebook-github-bot

    added a commit that references this issue on Jul 3, 2024

    Backport gh-107803: Double linked list implementation for asyncio tasks

    6f4de81
16. estyxx

    added 3 commits that reference this issue on Jul 17, 2024

    pythonGH-107803: double linked list implementation for asyncio tasks (p…

    c5a0305

    pythongh-107803: fix thread safety issue in double linked list implem…

    6970af9

    pythongh-107803: add whatsnew for asyncio double linked list implemen…

    b4ff32a
17. kumaraditya303

    added a commit that references this issue on Nov 8, 2024

    GH-107803: use circular double linked list for tasks in `_asyncio` (#…

    a3e8e7b
18. picnixz

    added a commit that references this issue on Dec 8, 2024

    pythonGH-107803: use circular double linked list for tasks in `_async…

    9eaa951
19. ebonnal

    added a commit that references this issue on Jan 12, 2025

    pythonGH-107803: use circular double linked list for tasks in `_async…

    1cd71e3

Issue actions