Renaming a table from `mysql.*` causes panic.

What did you do?

upstream:

create table mysql.a (pk bigint primary key);
rename table mysql.a to test.a;

no need to create any changefeed.

What did you expect to see?

nothing.

if any changefeed is listening on test.*, failing it because of ErrSyncRenameTableFailed (table's old name is not in filter rule)

What did you see instead?

every node in the entire cluster panicked:

[2025/11/06 14:45:56.718 +08:00] [PANIC] [persist_storage_ddl_handlers.go:505] ["table not found"] [tableID=116] 
[stack="github.com/pingcap/ticdc/logservice/schemastore.getSchemaID
	github.com/pingcap/ticdc/logservice/schemastore/persist_storage_ddl_handlers.go:505
github.com/pingcap/ticdc/logservice/schemastore.buildPersistedDDLEventForRenameTable
	github.com/pingcap/ticdc/logservice/schemastore/persist_storage_ddl_handlers.go:613
github.com/pingcap/ticdc/logservice/schemastore.(*persistentStorage).handleDDLJob
	github.com/pingcap/ticdc/logservice/schemastore/persist_storage.go:726
github.com/pingcap/ticdc/logservice/schemastore.(*keyspaceSchemaStore).tryUpdateResolvedTs
	github.com/pingcap/ticdc/logservice/schemastore/schema_store.go:150
github.com/pingcap/ticdc/logservice/schemastore.(*schemaStore).RegisterKeyspace.func1
	github.com/pingcap/ticdc/logservice/schemastore/schema_store.go:520"]

Versions of the cluster

v8.5.4-release.2