[wpt-sync] Sync PR 27350 - [testharness.js] Revert assert-tracking code
Categories
(Testing :: web-platform-tests, task, P4)
Tracking
(Not tracked)
People
(Reporter: wpt-sync, Unassigned)
References
()
Details
(Whiteboard: [wptsync downstream])
Sync web-platform-tests PR 27350 into mozilla-central (this bug is closed when the sync is complete).
PR: https://github.com/web-platform-tests/wpt/pull/27350
Details from upstream follow.
b'Stephen McGruer <smcgruer@chromium.org>' wrote:
[testharness.js] Revert assert-tracking code
Despite a series of attempts to fix the performance regressions from
this code (largely by propagating the value of 'output' properly), we
have hit a wall with tests that run in workers. There is no way
currently for a testharness.js script loaded in a worker to know that it
shouldn't be tracking asserts, and this is causing a large performance
regression on Chromium's bots.This reverts commit 9c225de1bd1c353b8eb58f17880a3841fea07ac5.
This reverts commit 67d4437890c271410c891a59bb468d1f65501f77.
This reverts commit 9498e208823f66db0ab96c44207d362f0333188a.
This reverts commit ddb7079f2e7612db30e9cb3611e9662909a657f8.
This reverts commit 7217e443c66461350204832d6c4f05d02f0cfbe1.
This reverts commit 934077ad8594041ef09a19181434f32c10c34d20.Because I did not revert a commit in the middle of these (that added
unrelated debug logging), I had to do a few fixups for the reverts.
(I.e. they were not all pure reverts).Associated Chromium bug: https://crbug.com/1170588
| Assignee | ||
Comment 1•5 years ago
|
||
| Assignee | ||
Updated•5 years ago
|
Description
•