my_snprintf fixes for 10.11+

MDEV-21978 found many erroneous my_snprintf formattings when it reenabled GCC printf checks.
Patches to these errors should apply to the oldest applicable version in maintenance and up.

In MariaDB 10.11 and up, the following strings may display their fields corrupted depending on the compiler and platform:

Missed when DYNAMIC_ARRAY switched to size_t (MDEV-26221)

• (DBUG) To big array idx: <index>, array size is <size> (from DYNAMIC_ARRAY’s code itself 😅)
• (DBUG) nothing to do, <count> owners, <count> waiters
• (DBUG) chunk: #<count> first_page: <index> (<index>) full_pages: <count> (<size>) Left <size>
• (DBUG) chunk0_pages: <count> groups <count> groups per full page: <count> Group on last page: <count>

others

• Hashicorp server error: <HTTP code>, response: <response>
• ALTER query started at <Sequence Number>-0-0 could not be completed because of unexpected master server or its binlog change (should be a GTID: <Domain ID>-<Server ID>-<Sequence Number>) (from MDEV-11675)
• (DBUG) unlocking <count> plugins
• InnoDB: Ignoring data file '<file name>' with space ID <ID>. Another data file called <name> exists with the same space ID.
• (DBUG) space_id: <ID>, page_no: <offset>
• InnoDB: compression algorithm <name> (<index>) is not available. Please, load the corresponding provider plugin. (big-endian CPUs only)
• InnoDB: innodb_open_files=<A> is not greater than the number of system tablespace files, temporary tablespace files, innodb_undo_tablespaces=<B>; adjusting to innodb_open_files=<C> (<A> not affected)
• innodb_log_file_size must be at least innodb_log_buffer_size=<size> (big-endian CPUs only)
• InnoDB: Cannot replay rename of tablespace <id> to '<name>': <error message> (<id> was not filled in, offsetting all fields to read undefined) (missed in 28b27b96d0)
• InnoDB: Expected tablespace flags 0x<hex> but found 0x<hex> in the file <name>