Add missing LEX_STRING::strs for my_snprintf

[ParadoxV5]

• The Jira issue number for this PR is: MDEV-21978

Description

This commit backports a fix group from #3360 for 11.5.
These are the only mistakes introduced during 11.5.x.

What problem is the patch trying to solve?

When these members switched from plain char* to LEX_CSTRING, not all usages were converted. Specifically, in this commit are args of my_snprintf derivatives.
Because until MDEV-21978 (#3360), automated type checks were unavailable for those functions due to their incompatibility, so these tools didn’t catch them.

If some output changed that is not visible in a test case, what was it looking like before the change and how it's looking with this patch applied?

str is the initial member of LEX_CSTRING (typedef of struct st_mysql_const_lex_string), so this difference did not cause incompatibility on most platforms with seamless alignments;
howëver, my co-mentor notes that Visual Studio reads garbage.

Release Notes

• Fixed data size mismatches that were garbling error and debug outputs (or possibly even crashes) on problematic platforms
  • (I adapted this line from the mould I used for previous iterations of [MDEV-21978 part 2] Enable and Fix GCC -Wformat Checks on my_vsnprintf and Users #3360 (comment) backports.)

How can this PR be tested?

This difference only manifests when LEX_CSTRING::str does not align on the 0th index of the overall LEX_CSTRING.
Besides using Visual Studio as noted, another deception to disregard this leniency is to shift the str member away from the 0th alignment, such as adding a padding member before it or moving it after member length.

Basing the PR against the correct MariaDB version

• This is a new feature or a refactoring, and the PR is based against the main branch.
• This is a bug fix, and the PR is based against the earliest maintained branch in which the bug can be reproduced.

PR quality check

• I checked the CODING_STANDARDS.md file and my PR conforms to this where appropriate.
• For any trivial modifications to the PR, I am ok with the reviewer making the changes themselves.

[ParadoxV5]

Oh great, the sporadic failures came back to haunt us.

[grooverdan]

Looks all right. Thanks. 11.5 is now EOL - can you rebase to 11.6 please.

[ParadoxV5]

👀 Oh right, 11.5 is a rolling release and has been superceded by 11.6.
The Maintenance Policy links to the rolling release system announcement but doesn’t (explicitly) say much itself.