To Andrii to triage.

Marja@ for some reason I can't find what https://chromium-review.googlesource.com/c/440984/ depended on (I tried checking all patchsets, but Gerrit now returns nothing :(). Can you help me?

I suspect this is this bug https://b.corp.google.com/issues/35190185 (internal). Can you check this https://b.corp.google.com/issues/35190185#comment3 and see if the events were as described there
(note: I am not asking you to agree with rational for WAI there).

I guess it used to depend on this one: https://chromium-review.googlesource.com/c/440246/

Maybe rebasing it cleared the dependency?

(I'll check out the bug you linked later.)

I don't think it's the same bug as https://b.corp.google.com/issues/35190185 . In that one, the user rebases A, and then B is no longer landable.

I didn't rebase A. I just landed A, and then tried to land B. Plus, in my case, rebasing B succeeded (in the other bug, it failed).

Yep, this is something else then.

Igor also hit the same issue today and I've filed a bug since on his two CLs dependnecies still show: https://b.corp.google.com/issues/35377658

If Gerrit team can't resolve it short-term, I will call "rebase" in CQ whenever this exception is hit.

Gerrit has done substantial work, including changes to upstream Git, but things are still not ready. 

tandrii@: what's the status of this? Both in terms of the external dependency and the CQ-side workaround? Can/Should I bump this to -Launch, or do we expect it to be resolved by the time dogfood emails go out?

The linked bug above is marked as fixed. I've also filed a potential CL (cl/151749669) which may help avoid some user confusion when a change needs to be rebased. Marking this as fixed based on state of internal tracking bug.