On Thu, 8 Dec 2022 10:55:07 GMT, Markus Grönlund <mgronlun@openjdk.org> wrote:
Greetings,
This change pre-calculates the event size for checkpoint / constant pool events (event id 1) before serialization to disk and circumvents the need to reserve a limited checkpoint event size up front, to be later filled in, using a padded write of the actual size. It lets us represent small checkpoint events more effectively, but more importantly, it allows for larger checkpoint / constant pool events because the event size pre-reservation no longer restricts them. The parser side will continue to read checkpoint event sizes as varints, but will expose the checkpoint event size as a long instead of an int.
Testing: jdk_jfr, stress testing
Thanks Markus
Marked as reviewed by egahlin (Reviewer). ------------- PR: https://git.openjdk.org/jdk/pull/11584