[jdk8u-dev] RFR: 8335552: [8u] JDK-8303466 backport to 8u requires 3 ::Identity signature fixes [v2]

[Francisco Ferrari Bihurriet]

On Wed, 3 Jul 2024 16:17:39 GMT, Francisco Ferrari Bihurriet <fferrari at openjdk.org> wrote:

>> Hi, here is a followup fix for #529, where @wkia has noted a problem with the `MaxLNode::Identity`, `MinLNode::Identity` and `ConvI2LNode::Identity` signatures: https://github.com/openjdk/jdk8u-dev/pull/529#discussion_r1661486052.
>> 
>> A complimentary [JDK-8075511](https://bugs.openjdk.org/browse/JDK-8075511 "Enable -Woverloaded-virtual C++ warning for HotSpot build") backport can be found in #534.
>
> Francisco Ferrari Bihurriet has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Back out JDK-8075511 backport
>   
>   This reverts commit cbd483afa413f006c991f1b4d92c83463cc11baa.

The original backport hash [is obtained](https://github.com/openjdk/skara/blob/273c362ab262134584830fc99adfd23f6f7ee0f7/bots/pr/src/main/java/org/openjdk/skara/bots/pr/CheckablePullRequest.java#L322-L333) from a hidden comment in [this message](https://github.com/openjdk/jdk8u-dev/pull/532#issuecomment-2206284623) with a [specific format](https://github.com/openjdk/skara/blob/273c362ab262134584830fc99adfd23f6f7ee0f7/bots/pr/src/main/java/org/openjdk/skara/bots/pr/CheckablePullRequest.java#L43) (we can see the hidden part in the [raw message body](https://api.github.com/repos/openjdk/jdk8u-dev/issues/comments/2206284623)).

The `/integrate` command then [passes this backport hash as `original`](https://github.com/openjdk/skara/blob/273c362ab262134584830fc99adfd23f6f7ee0f7/bots/pr/src/main/java/org/openjdk/skara/bots/pr/IntegrateCommand.java#L218-L226) when creating the commit and, [unlike the `/clean` command](https://github.com/openjdk/skara/blob/273c362ab262134584830fc99adfd23f6f7ee0f7/bots/pr/src/main/java/org/openjdk/skara/bots/pr/CleanCommand.java#L66), the `backport` label isn't checked. The commit message's `original` [finally ends up in the `Backport-of:` footer](https://github.com/openjdk/skara/blob/273c362ab262134584830fc99adfd23f6f7ee0f7/vcs/src/main/java/org/openjdk/skara/vcs/openjdk/CommitMessageFormatters.java#L94).

So even if we managed to remove the `backport` label, the `Backport-of:` footer would still have been generated, as long as [this message](https://github.com/openjdk/jdk8u-dev/pull/532#issuecomment-2206284623) is readable by the bot.

Apparently, a PR that starts or becomes a backport, is totally unexpected to return to a non-backport state. Knowing about this irreversible process is enough for me, to be more careful next time. In the unlikely case we need to return to a non-backport PR in the future, we can just start a new PR.

-------------

PR Comment: https://git.openjdk.org/jdk8u-dev/pull/532#issuecomment-2209001817