- disable the test suffering from Bug#41098 Query Cache returns wrong result with
                                            concurrent insert
- additional improvements

Details:
- Move the test for the historic bug
  Bug 28249 Query Cache returns wrong result with concurrent insert / certain lock
  into its own testscript query_cache_28249.test.
- query_cache.test:
  - replace error numbers with error names
  - remove trailing spaces, replace tabs with spaces
  - reset of @@global.log_bin_trust_function_creators to its original value
    at the end of the test
- query_cache_28249.test:
  - replace sleep 5 with poll routine
  - avoid random differences caused by concurrent_inserts effects
  - improved comments and formatting

(no conflicts)

symlink.test failed when run in an environment that has mysql-test/var
symlinked to elsewhere, e.g. a memory file system. This is the case
when running mysql-test-run --mem.

In this case the server does not detect that the directory specified
with a DATA/INDEX DIRECTORY clause is within its data home directory.
This problem was reported as Bug#39277 (Creation of table with data
and/or index files in data home directory succeeds). It was decided
that it will not be fixed in 5.1. Hence, the current behavior is
accepted for 5.1. It will be fixed in 6.0 though.

Fixed the test case so that it works in both environments. 1. When no
symbolic link is involved, the server notices that the data/index
directory is in its data hone directory and rejects the CREATE/ALTER
TABLE statement. 2. When the data home directory is symlinked, it
does not notice the problem and executes the statement sucessfully.

Problem: in 5.0 'check table for upgrade' doesn't detect
incompatible collation changes made in 5.0.48.

Fix: backport #39585 fix to 5.0

Altering how MTR checks global variable status to exclude timestamp
Changed SQL statements to update style.

Problem was usage of read_range_first with an empty key.

Solution was to not to give a key if it was empty.

ONLY_FULL_GROUP_BY

The check for non-aggregated columns in queries with aggregate function, but without
GROUP BY was treating all the parts of the query as if they are in the SELECT list.
Fixed by ignoring the non-aggregated fields in the WHERE clause.

Several system variables did not behave like system variables should do.
When trying to SET them or use them in SELECT, they were reported as
"unknown system variable". But they appeared in SHOW VARIABLES.

This has been fixed by removing the "fixed_vars" array of variables
and integrating the variables into the normal system variables chain.
All of these variables do now behave as read-only global-only
variables. Trying to SET them tells they are read-only, trying to
SELECT the session value tells they are global only. Selecting the
global value works. It delivers the same value as SHOW VARIABLES.

- restore original state of event_scheduler at the end of the test
- minor fixes around comments, formatting

the problem is the same as reported in bug#20835,
so the fix is backport of bug#20835 patch.

returns truncated results

Problem: performig conversion from {INT, DECIMAL, REAL} to CHAR
we incorrectly set its max length in some cases that may lead
to truncated results returned.

Fix: properly set CONVERT({INT, DECIMAL, REAL}, CHAR) result's
max length.