Avoid double hash lookup in torch._library.simple_registry by swolchok · Pull Request #161328 · pytorch/pytorch

swolchok · 2025-08-23T00:01:18Z

Stack from ghstack (oldest at bottom):

Not a huge cost, but free win is free.

Not a huge cost, but free win is free. [ghstack-poisoned]

pytorch-bot · 2025-08-23T00:01:21Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/161328

📄 Preview Python docs built from this PR
📄 Preview C++ docs built from this PR
❓ Need help or want to give feedback on the CI? Visit the bot commands wiki

Note: Links to docs will display an error until the docs builds have been completed.

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

ROCm MI2xx CI/CD workflows failing due to : download from https://api.github.com/repos/pytorch/pytorch timed out.

✅ No Failures

As of commit bd894e2 with merge base 05eeb29 ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

Not a huge cost, but free win is free. ghstack-source-id: e25f8d5 Pull Request resolved: #161328

Not a huge cost, but free win is free. [ghstack-poisoned]

pytorchmergebot · 2025-08-27T18:23:18Z

Starting merge as part of PR stack under #161329

swolchok · 2025-08-27T20:17:36Z

this broke python test/profiler/test_profiler_tree.py -k test_profiler_experimental_tree_with_stack_and_torch_dispatch

swolchok · 2025-08-27T20:21:04Z

this broke python test/profiler/test_profiler_tree.py -k test_profiler_experimental_tree_with_stack_and_torch_dispatch

looks like we simply need to resnapshot the test. trace is picking up the new get() call we added and wasn't including not in on a dictionary before. next ghstack update will do this

Not a huge cost, but free win is free. [ghstack-poisoned]

pytorchmergebot · 2025-08-30T00:30:51Z

Starting merge as part of PR stack under #161432

pytorchmergebot · 2025-08-30T05:51:21Z

Starting merge as part of PR stack under #161432

`auto` forces a copy. Confirmed this did something noticable with perf. Pull Request resolved: #161329 Approved by: https://github.com/zpcore, https://github.com/fduwjj, https://github.com/Skylion007, https://github.com/bdhirsh ghstack dependencies: #161301, #161292, #161304, #161308, #161315, #161317, #161328

If we want them interned, we should intern at callsites. (The numpy reference has bit rotted; see numpy/numpy@b222eb6#diff-6bdb6105198083838f51c57b55b3a49472ed23043bb40018f1ea41138e687163) Profiling a simple torchdispatch benchmark with perf before/after seems to show that time spent copying std::strings and interning Python strings is gone, though there is some noise and the improvement is very small. Pull Request resolved: #161432 Approved by: https://github.com/ezyang ghstack dependencies: #161301, #161292, #161304, #161308, #161315, #161317, #161328, #161329

) symbools are not identical with Py_True or PyFalse, so we can do those cheap checks first and at least get plain old bools to go fast. Pull Request resolved: #161455 Approved by: https://github.com/Skylion007 ghstack dependencies: #161301, #161292, #161304, #161308, #161315, #161317, #161328, #161329, #161432

…61328) Not a huge cost, but free win is free. Pull Request resolved: pytorch#161328 Approved by: https://github.com/Skylion007 ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317

`auto` forces a copy. Confirmed this did something noticable with perf. Pull Request resolved: pytorch#161329 Approved by: https://github.com/zpcore, https://github.com/fduwjj, https://github.com/Skylion007, https://github.com/bdhirsh ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317, pytorch#161328

…h#161432) If we want them interned, we should intern at callsites. (The numpy reference has bit rotted; see numpy/numpy@b222eb6#diff-6bdb6105198083838f51c57b55b3a49472ed23043bb40018f1ea41138e687163) Profiling a simple torchdispatch benchmark with perf before/after seems to show that time spent copying std::strings and interning Python strings is gone, though there is some noise and the improvement is very small. Pull Request resolved: pytorch#161432 Approved by: https://github.com/ezyang ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317, pytorch#161328, pytorch#161329

…rch#161455) symbools are not identical with Py_True or PyFalse, so we can do those cheap checks first and at least get plain old bools to go fast. Pull Request resolved: pytorch#161455 Approved by: https://github.com/Skylion007 ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317, pytorch#161328, pytorch#161329, pytorch#161432

…61328) Not a huge cost, but free win is free. Pull Request resolved: pytorch#161328 Approved by: https://github.com/Skylion007 ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317

`auto` forces a copy. Confirmed this did something noticable with perf. Pull Request resolved: pytorch#161329 Approved by: https://github.com/zpcore, https://github.com/fduwjj, https://github.com/Skylion007, https://github.com/bdhirsh ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317, pytorch#161328

…h#161432) If we want them interned, we should intern at callsites. (The numpy reference has bit rotted; see numpy/numpy@b222eb6#diff-6bdb6105198083838f51c57b55b3a49472ed23043bb40018f1ea41138e687163) Profiling a simple torchdispatch benchmark with perf before/after seems to show that time spent copying std::strings and interning Python strings is gone, though there is some noise and the improvement is very small. Pull Request resolved: pytorch#161432 Approved by: https://github.com/ezyang ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317, pytorch#161328, pytorch#161329

…rch#161455) symbools are not identical with Py_True or PyFalse, so we can do those cheap checks first and at least get plain old bools to go fast. Pull Request resolved: pytorch#161455 Approved by: https://github.com/Skylion007 ghstack dependencies: pytorch#161301, pytorch#161292, pytorch#161304, pytorch#161308, pytorch#161315, pytorch#161317, pytorch#161328, pytorch#161329, pytorch#161432

Avoid double hash lookup in torch._library.simple_registry

375940d

Not a huge cost, but free win is free. [ghstack-poisoned]

swolchok mentioned this pull request Aug 23, 2025

Fix OpSchema equality check #161231

Closed

swolchok added a commit that referenced this pull request Aug 23, 2025

Avoid double hash lookup in torch._library.simple_registry

9a598b9

Not a huge cost, but free win is free. ghstack-source-id: e25f8d5 Pull Request resolved: #161328

swolchok mentioned this pull request Aug 23, 2025

Fix accidental copy in pushPyOutToStack #161329

Closed

swolchok requested review from XilunWu, ezyang and zpcore August 23, 2025 01:43

swolchok added the topic: not user facing topic category label Aug 23, 2025

Skylion007 approved these changes Aug 23, 2025

View reviewed changes

Update on "Avoid double hash lookup in torch._library.simple_registry"

672bc18

Not a huge cost, but free win is free. [ghstack-poisoned]

This was referenced Aug 27, 2025

Avoid redundant PyTuple_GetSize call in _maybe_handle_torch_function #161633

Closed

fastpath type Tensor in THPVariable_NewWithVar #161634

Closed

PythonArgParser::symintlist{,Optional} should use SymDimVector #161635

Closed

Update on "Avoid double hash lookup in torch._library.simple_registry"

78b2fc9

Not a huge cost, but free win is free. [ghstack-poisoned]

swolchok added 2 commits August 29, 2025 14:32

rebase on "Avoid double hash lookup in torch._library.simple_registry"

404598d

Not a huge cost, but free win is free. [ghstack-poisoned]

Update on "Avoid double hash lookup in torch._library.simple_registry"

bd894e2

Not a huge cost, but free win is free. [ghstack-poisoned]

pytorchmergebot added the Merged label Aug 30, 2025

pytorchmergebot closed this in eb9526a Aug 30, 2025

github-actions bot deleted the gh/swolchok/802/head branch September 30, 2025 02:08

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Avoid double hash lookup in torch._library.simple_registry#161328

Avoid double hash lookup in torch._library.simple_registry#161328
swolchok wants to merge 5 commits intogh/swolchok/802/basefrom
gh/swolchok/802/head

swolchok commented Aug 23, 2025 •

edited

Loading

Uh oh!

pytorch-bot bot commented Aug 23, 2025 •

edited

Loading

Uh oh!

pytorchmergebot commented Aug 27, 2025

Uh oh!

swolchok commented Aug 27, 2025

Uh oh!

swolchok commented Aug 27, 2025

Uh oh!

pytorchmergebot commented Aug 30, 2025

Uh oh!

pytorchmergebot commented Aug 30, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

swolchok commented Aug 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-bot bot commented Aug 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/161328

❗ 1 Active SEVs

✅ No Failures

Uh oh!

pytorchmergebot commented Aug 27, 2025

Uh oh!

swolchok commented Aug 27, 2025

Uh oh!

swolchok commented Aug 27, 2025

Uh oh!

pytorchmergebot commented Aug 30, 2025

Uh oh!

pytorchmergebot commented Aug 30, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

swolchok commented Aug 23, 2025 •

edited

Loading

pytorch-bot bot commented Aug 23, 2025 •

edited

Loading