CQ merged change w/out dependency

[Jeffrey Kardatzke]

I submitted a few CLs yesterday & the day before as follows:

A - https://chrome-internal-review.googlesource.com/477614

B - https://chrome-internal-review.googlesource.com/477653

C - https://chrome-internal-review.googlesource.com/479096

In A I have a CQ-DEPEND on B.

In B I have a CQ-DEPEND on A & C.

So effectively A & B need to go in together and B is dependent on C.

So I would have expected either C to go in first, and then A & B to go in together... or all 3 to go in at once. However, currently A is merged and B & C are both still in the commit queue (which keeps failing and retrying them repeatedly for some reason).

Am I mistaken on how this should have worked, or did something go wrong?

Thanks,

--

Jeffrey Kardatzke

jkard...@google.com

Google, Inc.

[Jason Clinton]

Git on Borg is currently failing most requests. A possible explanation is that all of your CL's were going to go in but only 1 of them managed to successfully get submitted. https://bugs.chromium.org/p/chromium/issues/detail?id=774368

--
--
Chromium OS Developers mailing list: chromiu...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-os-dev?hl=en

[Jeffrey Kardatzke]

Thanks, that would definitely explain the weirdness.

[Mike Frysinger]

if the tree is in a broken state due to incomplete dep merges like this, this is a situation where it's acceptable to chump CLs to get back in a coherent state.  just let the sheriffs know.

-mike

[Jeffrey Kardatzke]

What's happened so far doesn't actually break anything...but if the 2 that are left don't go in the right order it will then break the build, so if that happens I'll let the sheriff know.