Fix SSL errors on the container Fedora cases.

[junaruga]

This PR fixes #1289 . Could you review this PR? Note I wrote the context in the commit message. Thanks.

---

This commit fixes the SSL related failures in the CI container Fedora cases.

The MariaDB 10.5.18 mysqld failed to start in the CI, container "fedora:rawhide" (Fedora 38) and "fedora:latest" (Fedora 37) cases with the SSL error below. So, we set the SSL configurations manually as well as MacOSX case.

 + /usr/libexec/mysqld --user=root --log-error=/build/mysql.log --ssl
2022-12-16 17:49:58 0 [Note] /usr/libexec/mysqld (mysqld 10.5.18-MariaDB) starting as process 724 ...
+ cat /build/mysql.log
2022-12-16 17:49:58 0 [Note] InnoDB: Uses event mutexes
2022-12-16 17:49:58 0 [Note] InnoDB: Compressed tables use zlib 1.2.12
2022-12-16 17:49:58 0 [Note] InnoDB: Number of pools: 1
2022-12-16 17:49:58 0 [Note] InnoDB: Using crc32 + pclmulqdq instructions
2022-12-16 17:49:58 0 [Note] mysqld: O_TMPFILE is not supported on /var/tmp (disabling future attempts)
2022-12-16 17:49:58 0 [Note] InnoDB: Using Linux native AIO
2022-12-16 17:49:58 0 [Note] InnoDB: Initializing buffer pool, total size = 134217728, chunk size = 134217728
2022-12-16 17:49:58 0 [Note] InnoDB: Completed initialization of buffer pool
2022-12-16 17:49:58 0 [Note] InnoDB: 128 rollback segments are active.
2022-12-16 17:49:58 0 [Note] InnoDB: Creating shared tablespace for temporary tables
2022-12-16 17:49:58 0 [Note] InnoDB: Setting file './ibtmp1' size to 12 MB. Physically writing the file full; Please wait ...
2022-12-16 17:49:58 0 [Note] InnoDB: File './ibtmp1' size is now 12 MB.
2022-12-16 17:49:58 0 [Note] InnoDB: 10.5.18 started; log sequence number 45079; transaction id 20
2022-12-16 17:49:58 0 [Note] Plugin 'FEEDBACK' is disabled.
2022-12-16 17:49:58 0 [Note] InnoDB: Loading buffer pool(s) from /var/lib/mysql/ib_buffer_pool
2022-12-16 17:49:58 0 [Note] InnoDB: Buffer pool(s) load completed at 221216 17:49:58
SSL error: Private key does not match the certificate public key
Error: -16 17:49:58 0 [ERROR] Failed to setup SSL
Error: -16 17:49:58 0 [ERROR] SSL error: Private key does not match the certificate public key
Error: -16 17:49:58 0 [ERROR] Aborting

After fixing the issue, we also saw the following failing tests with the error below. So, we recreated the db user.

expected no Exception, got #<Mysql2::Error: Access denied for user 'root'@'localhost'> with backtrace:

Note this issue didn't happened on the container CentOS case. But as we applied the fix to all the container cases to simplify the logic.

[junaruga]

The MacOS case failed. But I don't think the failure is related to this PR.
https://github.com/brianmario/mysql2/actions/runs/3716062019/jobs/6301955057

[sodabrew]

The set of CI scripts might be due for minor refactoring, but let's land this to unblock now. Thank you for the PR!!

[junaruga]

The set of CI scripts might be due for minor refactoring, but let's land this to unblock now. Thank you for the PR!!

Thanks for reviewing! Right. We might be able to refactor the set of the CI scripts. What can we refactor in the CI scripts in your opinion?

[voxik]

After fixing the issue, we also saw the following failing tests with the error below. So, we recreated the db user.

expected no Exception, got #<Mysql2::Error: Access denied for user 'root'@'localhost'> with backtrace:

Just guessing, but isn't the key in this commit?

https://src.fedoraproject.org/fork/mschorm/rpms/rubygem-mysql2/c/c7e03e581c0ea3c8a8e38d59fc6df13352434c6b

[junaruga]

Perhaps, it improves the testing logic. In this PR, I used the logic used in non-container test cases to align.

mysql2/ci/setup.sh

Lines 133 to 138 in 7f6f33a

	sudo mysql ${MYSQL_OPTS} -u "${DB_SYS_USER}" <<SQL
	DROP USER 'root'@'localhost';
	CREATE USER 'root'@'localhost' IDENTIFIED BY '';
	GRANT ALL PRIVILEGES ON *.* TO 'root'@'localhost' WITH GRANT OPTION;
	FLUSH PRIVILEGES;
	SQL