It looks the importer already did revert changes from https://chromium.googlesource.com/chromium/src/+/8b4edde07c in two different CLs:
https://chromium-review.googlesource.com/602030 (manual by rakuco)
https://chromium-review.googlesource.com/602471 (automatic)

One additional thing to worry about:

https://chromium.googlesource.com/chromium/src/+/ed96173eaf is now exported as part of https://github.com/w3c/web-platform-tests/pull/6822, and I maintained the two commits separately. The exporter should therefore find these and not try to export them again. However, I could have squashed them into two, deliberate or my accident. Then, the exporter would keep trying to export ed96173eaf, and if at some point in the future it stops conflicting, it would be merged. That would be very surprising.

After the manual fixup in https://github.com/w3c/web-platform-tests/pull/6845, https://chromium.googlesource.com/chromium/src/+/8b4edde07c no longer conflicted when applied on wpt, and https://github.com/w3c/web-platform-tests/pull/6846 was automatically created, probably merged soon.

For the timeOrigin changes, the most recent export logs still complain that "Allow Date.now() rounding errors in timeOrigin test" can't be applied cleanly, meaning that it's trying to export it.

I think that's because the exporter expects each chromium change to go in one PR, but I had to put two commits in a single PR to get stable results and merge them. The exporter will keep trying to create a PR for that change until it goes out of the 5000 commit window. A fix for this would be to first look in the wpt commit history to determine what's been exported, and then look for PRs only for the things that remain.

First of all, this issue is a dup of  issue 743149 . Marking as so.

Some more information regarding the timeOrigin test:

> I maintained the two commits separately. The exporter should therefore find these and not try to export them again. However, I could have squashed them into two, deliberate or my accident. Then, the exporter would keep trying to export ed96173eaf, and if at some point in the future it stops conflicting, it would be merged. That would be very surprising.

Unfortunately this is not true. The exporter does not look at GitHub commits, but rather PRs. It would normally expect the two changes to go in two separate PRs, with the change IDs in description. Now that we have one PR, the remediation is to put the two change IDs into the PR description (and the exporter will link this PR with both changes), if I understand correctly. foolip@ has already modified the PR description in https://github.com/w3c/web-platform-tests/pull/6822.

> Now that we have one PR, the remediation is to put the two change IDs into the PR description (and the exporter will link this PR with both changes), if I understand correctly. foolip@ has already modified the PR description in https://github.com/w3c/web-platform-tests/pull/6822.

Unfortunately again, this remediation doesn't work -- the exporter only looks at the first "Change-Id". But it wouldn't cause more trouble than an error message in the logs, which looks like:

The following commit did not apply cleanly:
Subject: Allow Date.now() rounding errors in timeOrigin test (https://chromium.googlesource.com/chromium/src/+/ed96173eaf)

The error message will eventually go away after this commit goes out of the history window.