1901363 - [wpt-sync] Sync PR 46662 - streams: underlyingSource.cancel() should called when abort, event with pending pull

Status: RESOLVED FIXED

(Core :: DOM: Streams, task, P4)

Description

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

Sync web-platform-tests PR 46662 into mozilla-central (this bug is closed when the sync is complete).

PR: https://github.com/web-platform-tests/wpt/pull/46662
Details from upstream follow.

Milly <[e-mail]> wrote:

streams: underlyingSource.cancel() should called when abort, event with pending pull

The spec ReadableStreamPipeTo and ReadableStreamCancel define that the signal aborted algorithm calls underlyingSource.cancel(). However, Webkit does not call underlyingSource.cancel() if signal aborted occurs while underlyingSource.pull() is pending.

The Webkit issue only occurs when canceling with a signal, but I added a similar test for reader.cancel() just to be sure.

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=28c2e3984c6d96088bc5ae979304a71289b3b96d

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 10 tests and 11 subtests

Status Summary

Firefox

OK : 8
PASS : 172
FAIL : 4
ERROR: 2

Chrome

OK : 8
PASS : 176
ERROR: 2

Safari

OK : 8
PASS : 160
FAIL : 16
ERROR: 2

Links

Gecko CI (Treeherder)
GitHub PR Head
GitHub PR Base

Details

Firefox-only Failures

• /streams/piping/abort.any.serviceworker.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL
• /streams/piping/abort.any.sharedworker.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL
• /streams/piping/abort.any.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL
• /streams/piping/abort.any.worker.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL

New Tests That Don't Pass

• /streams/piping/abort.any.serviceworker.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL (Chrome: PASS, Safari: PASS)
• /streams/piping/abort.any.sharedworker.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL (Chrome: PASS, Safari: PASS)
• /streams/piping/abort.any.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL (Chrome: PASS, Safari: PASS)
• /streams/piping/abort.any.shadowrealm.html [wpt.fyi]: ERROR (Chrome: ERROR, Safari: ERROR)
• /streams/piping/abort.any.worker.html [wpt.fyi]
  • a rejection from underlyingSink.abort() should be preferred to one from underlyingSource.cancel(): FAIL (Chrome: PASS, Safari: PASS)
• /streams/readable-streams/cancel.any.shadowrealm.html [wpt.fyi]: ERROR (Chrome: ERROR, Safari: ERROR)

Comment 3

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/c9df9486cd3c
[wpt PR 46662] - streams: underlyingSource.cancel() should called when abort, even with pending pull, a=testonly

Comment 4

[Sandor Molnar[:smolnar]]

https://hg.mozilla.org/mozilla-central/rev/c9df9486cd3c