-
Notifications
You must be signed in to change notification settings - Fork 70
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add an RFC for add_top_level_completion_handler
#168
base: master
Are you sure you want to change the base?
Add an RFC for add_top_level_completion_handler
#168
Conversation
… handler https://bugs.webkit.org/show_bug.cgi?id=259409 rdar://problem/112683033 Reviewed by Jonathan Bedard. Previously, we were considering the test as having to run to completion when any testharness.js completion handler first ran (which was sometimes that of the frame within the opened window; the nondeterminism here just adds flakiness to the already bad behaviour). Instead, only output anything for the top-level completion handler, as all results should be passed up to it. Note wptrunner with the WebDriver or Marionette executors only ever pays attention to the top-level completion handler (as they only pay attention to the top-level frame & window), thus they don't have any flakiness like this. Ideally testharness.js would have an API we could use for this; I've filed an RFC at web-platform-tests/rfcs#168 for this, but there's no reason not to do the simple fix ourselves, as at least for now this is a strict progression. (It would stop being a strict progression if at some point testharness.js started allowing fetch_tests_from_window() with a window with noopener.) It seems likely we have other tests that are marked as flaky because of it, but alas there's no easy way to find what tests will have been fixed by this. * LayoutTests/TestExpectations: * LayoutTests/http/tests/security/contentSecurityPolicy/upgrade-insecure-requests/basic-upgrade-cors.https-expected.txt: * LayoutTests/http/tests/security/contentSecurityPolicy/upgrade-insecure-requests/basic-upgrade-cors.https.html: * LayoutTests/imported/w3c/web-platform-tests/cookies/partitioned-cookies/partitioned-cookies.tentative.https-expected.txt: * LayoutTests/imported/w3c/web-platform-tests/html/browsers/browsing-the-web/navigating-across-documents/source/navigate-child-function-parent-then-fragment-expected.txt: * LayoutTests/platform/wk2/TestExpectations: * LayoutTests/resources/testharnessreport.js: Canonical link: https://commits.webkit.org/269483@main
…and we've had no substantive disagreement in two weeks, so I guess that means this is accepted despite nobody having reviewed it? 🤨 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No obvious objections from my side
Co-authored-by: Ms2ger <[email protected]>
This was discussed during the November 2023 meeting: This RFC may need some updates |
Per the discussion in https://github.com/web-platform-tests/wpt-notes/blob/47c3ea1a941c0b0f78869ff2d00a2e2f93594636/minutes/2023-11-07.md?plain=1#L24 we might want to try a different implementation approach. We likely need to try with existing tests to see which approach is most robust. |
Rendered