1820674 - [wpt-sync] Sync PR 38836 - Ensure we never fire a navigateerror event without a navigate event

Status: RESOLVED FIXED

(Testing :: web-platform-tests, task, P4)

Description

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Sync web-platform-tests PR 38836 into mozilla-central (this bug is closed when the sync is complete).

PR: https://github.com/web-platform-tests/wpt/pull/38836
Details from upstream follow.

Nate Chapin <[e-mail]> wrote:

Ensure we never fire a navigateerror event without a navigate event

1. Currently, we try to error any upcoming traversals on frame detach,
   rather than leaving the promises unfulfilled. That's inconsistent
   with what we do for an ongoing navigation. Leave them unfulfilled,
   and move the cleanup logic to ContextDestroyed(), where the
   cleanup for the ongoing navigation case is.
2. Now the only place we error an upcoming traversals is
   TraverseCancelled(). Directly reject the promise and remove the
   traversal from upcoming_traversals_, rather than going through
   RejectPromisesAndFireNavigateErrorEvent(). This fixes the issue
   where we would fire a navigateerror event for a traversal that
   had not yet fired a navigate event.
3. Now RejectPromisesAndFireNavigateErrorEvent() and
   FinalizeWithAbortedNavigationError() are now only used for the
   ongoing navigation, so there's no need for them to take a
   NavigationApiNaviagtion* parameter.
4. Every codepath that ends a navigation now knows whether it is an
   ongoing navigation or an upcoming traversal. Clear the
   appropriate state directly rather than needing
   CleanupApiNavigation().

Bug: 1421647
Change-Id: I959b9f7b47e962c41aa57a49dcbf0e4bfce9a15e
Reviewed-on: https://chromium-review.googlesource.com/4313039
WPT-Export-Revision: bc3a6864038842dbafbaf3b4170140893fe0688b

Comment 1

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Pushed to try (stability) https://treeherder.mozilla.org/#/jobs?repo=try&revision=cc2812d334d2229e07bdaac73c481e2b9566444f

Comment 2

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

CI Results

Ran 9 Firefox configurations based on mozilla-central, and Firefox, Chrome, and Safari on GitHub CI

Total 5 tests and 1 subtests

Status Summary

Firefox

OK : 5
FAIL: 5

Chrome

OK : 5
PASS: 2
FAIL: 3

Safari

OK : 5
FAIL: 5

Links

Gecko CI (Treeherder)
GitHub PR Head
GitHub PR Base

Details

New Tests That Don't Pass

• /navigation-api/navigate-event/navigation-traverseTo-in-iframe-same-document-preventDefault.html [wpt.fyi]
  • navigation.traverseTo() in an iframe with same-document preventDefault in its parent: FAIL (Chrome: FAIL, Safari: FAIL)
• /navigation-api/navigation-methods/return-value/traverseTo-detach-cross-document-before-navigate-event.html [wpt.fyi]
  • traverseTo() promise rejections when detaching an iframe before onnavigate (cross-document): FAIL (Chrome: FAIL, Safari: FAIL)
• /navigation-api/navigation-methods/return-value/traverseTo-detach-same-document-before-navigate-event.html [wpt.fyi]
  • traverseTo() promise rejections when detaching an iframe before onnavigate (same-document): FAIL (Chrome: FAIL, Safari: FAIL)
• /navigation-api/navigation-methods/sandboxing-back-parent.html [wpt.fyi]
  • A sandboxed iframe cannot navigate its parent via its own navigation object by using back(): FAIL (Chrome: PASS, Safari: FAIL)
• /navigation-api/navigation-methods/sandboxing-back-sibling.html [wpt.fyi]
  • A sandboxed iframe cannot navigate its sibling via its own navigation object by using back(): FAIL (Chrome: PASS, Safari: FAIL)

Comment 3

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/b921f9c29e8a
[wpt PR 38836] - Ensure we never fire a navigateerror event without a navigate event, a=testonly
https://hg.mozilla.org/integration/autoland/rev/359a63b55a31
[wpt PR 38836] - Update wpt metadata, a=testonly

Comment 4

[Narcis Beleuzu [:NarcisB]]

https://hg.mozilla.org/mozilla-central/rev/b921f9c29e8a
https://hg.mozilla.org/mozilla-central/rev/359a63b55a31