mysqltest disconnect/connect-combo could be so
quick that connect would hit the server before
it had processed the disconnect. Since that
resulted in one more concurrent connection than
we meant to have, global or per-user
max-user-connections could be exceeded.
This could lead to "random" failures in tests
that set those limits.
The problem is that relying on the output of the 'ls' command is not
portable as its behavior is not the same between systems and it might
even not be available at all in (Windows).
So I added list_files that relies on the portable mysys library instead.
(and also list_files_write_file and list_files_append_file,
since the test was using '--exec ls' in that way.)
Another problem is that the backtrace facility wasn't being
enabled for non-Linux targets even if the target OS has the
backtrace functions. Also, the stacktrace functions inside
mysqltest were being used without proper checks for their
presence in the build.
The problem was that when a embedded linked version of mysqltest
crashed there was no way to obtain a stack trace if no core file
is available. Another problem is that the embedded version of
libmysql was not behaving (crash) the same as the non-embedded with
respect to sending commands to a explicitly closed connection.
The solution is to generate a mysqltest's stack trace on crash
and to enable "reconnect" if the connection handle was explicitly
closed so the behavior matches the non-embedded one.
The problem is a hack in mysqltest.c::append_field that modifies
the exponential notation of floating point numbers by removing a
zero after the the symbol 'e' (eg: 00001.2e+018 is converted to
00001.2e+18) but does not take into account the zerofill affect
in the start of the string.
The solution is to check if the field was zero filled and insert
a zero at the start of the string if a zero after the exponential
notation symbol is removed.
Bug#36028 - Test funcs_1.<engine>_trig_03e failing on Windows
Bug#36029 - Test funcs_1.<engine>_trig_0407 failing on Windows
Bug#36030 - Test funcs_1.<engine>_trig_08 failing on Windows
Adding $MASTER_MYSOCK to init_win_path()
When path names are short, master.sock ends up in MYSQL_TMP_DIR, but with longer path names,
master.sock ends up in /tmp/<random_string>/ and these tests will fail due to path delimiter difference.
New changeset to start with 5.0 -- Not all of these tests are present in 5.0, but want to keep mysqltest the same
100% effectively on Windows
The mysqltest docs state that the 'replace_result' command
doesn't perform any escape processing.
However the current implementation was processing backslash
escapes in the from/to strings.
This prevents replacing e.g. patch on windows (where backslash
is used as a path separator).
Fixed by removing the backslash escape processing from
'replace_result'.
The problem was that 'connect' command didn't set mysql_errno
variable, thus the script was unable to determine whether connection
was opened or not.
The fix is to set this variable.
Test cases will be added in the scope of Bug33507
into connect.test file.
--local-infile=0 checks can be bypassed by sending a FETCH LOCAL FILE response
Add a check for CLIENT_LOCAL_FILES before sending a local file.
Beware, that all binary distributions enable sending of local files and it's up
to the programs which use libmysql to disable it, if they don't use this functionality.
Otherwise they are not safe.
value" error even though the value was correct): a C function in my_getopt.c
was taking bool* in parameter and was called from C++ sql_plugin.cc,
but on some Mac OS X sizeof(bool) is 1 in C and 4 in C++, giving funny
mismatches. Fixed, all other occurences of bool in C are removed, future
ones are blocked by a "C-bool-catcher" in my_global.h (use my_bool).
