Hi Sergei, On Wed, Nov 25, 2015 at 02:35:59PM +0100, Sergei Golubchik wrote:
Hi, Sergey!
On Nov 25, Sergey Vojtovich wrote:
revision-id: 6019fee7d84e8ec7d64337ad080a04f9c106bb68 (mariadb-10.1.8-74-g6019fee) parent(s): 2c29e85eafbf5bafe9c431accea9c25db7699eeb committer: Sergey Vojtovich timestamp: 2015-11-25 13:58:15 +0400 message:
MDEV-7780 - Support for faking server version
Added --version=str (optional argument sets server version string).
diff --git a/mysql-test/r/mysqld--help.result b/mysql-test/r/mysqld--help.result index 7a6541b..c109ad2 100644 --- a/mysql-test/r/mysqld--help.result +++ b/mysql-test/r/mysqld--help.result @@ -1124,7 +1124,8 @@ The following options may be given as the first argument: CLIENT_STATISTICS, INDEX_STATISTICS and TABLE_STATISTICS tables in the INFORMATION_SCHEMA -v, --verbose Used with --help option for detailed help. - -V, --version Output version information and exit. + -V, --version[=name] + Output version information and exit.
This is pretty weird. That --version and --version=str have completely different behavior.
On the other hand, the server variable is @@version, so --version=str is kind of expected. So --set-version=str or --fake-version=str aren't very intuitive either.
I don't know a good solution to this. Use --version? Use --set-version or --fake-version? I'm not sure either, I'm just following: "We can simply make @@version writable (from the command-line, not run-time)".
Originally it looked weird to me too, but then I got used to it.
--wait-timeout=# The number of seconds the server waits for activity on a connection before closing it
diff --git a/mysql-test/suite/sys_vars/t/version.test b/mysql-test/suite/sys_vars/t/version.test new file mode 100644 index 0000000..3bc3f2b --- /dev/null +++ b/mysql-test/suite/sys_vars/t/version.test @@ -0,0 +1 @@ +SELECT @@version; diff --git a/sql/mysqld.cc b/sql/mysqld.cc
would be nice to have a test of the version as sent in the protocol. You can use, for example, mysql client to show that version.
Ok. Thanks, Sergey