Merge V8 version 8.8.278.17

[dahlstrom-g]

This is a small update to the latest 8.8 version:
https://chromium.googlesource.com/v8/v8/+log/8.8-lkgr
There were no merge conflicts.

[kaidokert]

Mostly looks like a clean update to me. A few test jobs failed - just for peace of mind, would be good to retry only the failing ones.

Lets work look at how to get that to 24lts as well once merged - or we can leave it trunk only as well.

Would be good for another reviewer to take a quick look.

[johnxwork]

The code changes look good, I'm excited about the next steps!

We recently switched to github workflow so please wait for review from andrewsavage@ to ensure we are doing rebase in the right way(to keep commit history e.t.c)

The lint pre-submit check is complaining about missing bug ID. The other test failures might be flakes(usually we don't see so many flakes, maybe it's due to recently enabled tests). You can re-run the failed test steps to check.

[kaidokert]

The lint pre-submit check is complaining about missing bug ID.

This one is currently expected, because subtree squashes produce merge commits that won't have bug ID in them. In the longer run, we'll redesign the Bug ID check to run this per-PR, not per individual commit.

Also the PR that updates subtrees needs to be submitted with merge, not the default squash and merge strategy.

[dahlstrom-g]

The code changes look good, I'm excited about the next steps!

Thanks! I'm sure it will get harder. ; )

We recently switched to github workflow so please wait for review from andrewsavage@ to ensure we are doing rebase in the right way(to keep commit history e.t.c)

That part I think I understand; it should be good to go.
The git-subtree.sh contrib script uses these trailers:

git-subtree-dir: third_party/v8
git-subtree-split: f7f4327516912dfca0d7401044e890ca0ce885ad

The lint pre-submit check is complaining about missing bug ID.

Yeah—it seems weird to add a Buganizer link to the upstream commit metadata
but I suppose I could. As long as the check can be ignored, though, it's OK with me.

The other test failures might be flakes(usually we don't see so many flakes, maybe it's due to recently enabled tests). You can re-run the failed test steps to check.

I did that; the android-arm check passed but the evergreen ones are still in progress.

Finally (maybe it's obvious) the commits in this PR shouldn't be squashed together.

[kaidokert]

it seems weird to add a Buganizer link to the upstream commit metadata but I suppose I could.

No let's not do that - the problem is with Bug ID check as is't currently set up. We'll bypass the check when merging. We'll just have to let people know internally as well that this is OK for now.

[dahlstrom-g]

[mistake while experimenting with GitHub UI]

[dahlstrom-g]

The evergreen checks succeeded on the third try (after running 82 min + 94 min +54 min ≈ 230 min, not including retry lag).

[kaidokert]

You are getting failed checks on Raspi-2 here because base commit was already broken ( metrics_test was deleted in trunk). It's now fixed in trunk - we can go ahead and merge this as is, if it's otherwise ready. Let me know - merging with merge button requires admin powers for now.

[dahlstrom-g]

You are getting failed checks on Raspi-2 here because base commit was already broken ( metrics_test was deleted in trunk). It's now fixed in trunk - we can go ahead and merge this as is, if it's otherwise ready. Let me know - merging with merge button requires admin powers for now.

LGTM