Bug #118083 optimize for invisible indexes ddl

Description:
Optimize DDL execution for invisible indexes by reducing traversal from twice to once

Previously, the native logic would traverse the index_visibility_list twice when handling DDL operations related to invisible indexes. This commit refactors the relevant logic to complete the operation with just a single traversal.

How to repeat:
just read the code

Suggested fix:
optimize as the below patch

a simple patch to describe the optimization

(*) I confirm the code being submitted is offered under the terms of the OCA, and that I am authorized to contribute it.

Contribution: alter_invisible_index_optimize.patch (application/octet-stream, text), 1.97 KiB.

Hello Alex Xing,

Thank you for the report and contribution.

regards,
umesh

Hi Alex,

The code change is logically sound because it avoids a second scan of alter_info->alter_index_visibility_list without altering the existing behavior.

This code path runs only once during ALTER TABLE preparation, and ALTER INDEX ... VISIBLE/INVISIBLE lists are usually very small. As a result, the expected performance improvement is negligible. The overall cost of ALTER TABLE is still dominated by storage-engine operations, data dictionary updates, and any required table rebuilds.

Thanks
Prabakaran

For these reasons, we are not moving forward with this patch.