During startup, InnoDB crash recovery must write a FILE_CHECKPOINT record. However, in MDEV-12353 the redo log record format was changed. When we are upgrading from a logically empty 10.2, 10.3, or 10.4 redo log, we must not write such a record, because it is encoded differently from the MLOG_CHECKPOINT record. We must simply create a new logically empty log file and replace the old ib_logfile0 with that.

Writing to the old log file would corrupt the log file. If the server is killed during the upgrade, we may end up with having a corrupted log file, and both the old and new version of the server may refuse to start up.