[lworld] RFR: 8370951: [lworld] Value record ClassCircularityError

[Frederic Parain]

On Fri, 31 Oct 2025 13:37:17 GMT, Paul Hübner <phubner at openjdk.org> wrote:

> Hi all,
> 
> When we encounter situations where we can't preload, for example if two value classes refer to each other, **may** not be able to do flattening as we cannot compute the field layout. We have log messages as a hint to application or library developers to let them know of that fact. They are very useful for debugging. However, retaining them at the `warning` level means they are shown by default. This may be misleading, as seeing e.g. `ClassCircularityError` in a log may indicate that something has gone gravely wrong, which is not the case with preload failures. This PR relegates all preload failure logs to level `info`, such that the information is opt-in and will not unnecessarily cause confusion.
> 
> Testing: tier 1.

LGTM.

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

Marked as reviewed by fparain (Committer).

PR Review: https://git.openjdk.org/valhalla/pull/1710#pullrequestreview-3416285535