Re: Kerberos test suite
От | Michael Paquier |
---|---|
Тема | Re: Kerberos test suite |
Дата | |
Msg-id | 20180306020855.GI1878@paquier.xyz обсуждение исходный текст |
Ответ на | Re: Kerberos test suite (Peter Eisentraut <peter.eisentraut@2ndquadrant.com>) |
Ответы |
Re: Kerberos test suite
|
Список | pgsql-hackers |
On Mon, Mar 05, 2018 at 02:45:07PM -0500, Peter Eisentraut wrote: > On 2/27/18 00:21, Michael Paquier wrote: >> +my ($stdout, $krb5_version); >> +IPC::Run::run [ 'krb5-config', '--version' ], '>', \$stdout or die >> "could not execute krb5-config"; >> +$stdout =~ m/Kerberos 5 release ([0-9]+\.[0-9]+)/ or die "could not get >> Kerberos version"; >> +$krb5_version = $1; >> Time for a new routine command_log which executes the command, then >> returns stdout and stderr to the caller? > > I didn't do anything about this. Do we have any more places where that > would be needed right now? I don't mind if this happens later on. I can send a patch as well if necessary. There are a couple of places where things can be consolidated using a single entry point: - 010_pg_archivecleanup.pl - PostgresNode::psql, where we could merge things. - PostgresNode::poll_query_until - PostgresNode::pg_recvlogical_upto - TestLib::check_pg_config - TestLib::program_help_ok - TestLib::program_version_ok - TestLib::program_options_handling_ok - TestLib::command_like - TestLib::command_like_safe - TestLib::command_fails_like - TestLib::command_checks_all +if ($ENV{with_gssapi} eq 'yes') +{ + plan tests => 4; +} +else +{ + plan skip_all => 'GSSAPI/Kerberos not supported by this build'; +} Thanks for adding this sanity check. +my $kdc_port = int(rand() * 16384) + 49152; That may not be worth worrying as that's an ephemeral port range but it can make the test a bit unstable... Perhaps the tests should be skipped on Windows or just produce an error? Like LDAP tests, libraries are supported on Windows but the hardcoded paths make things harder to handle there. -- Michael
Вложения
В списке pgsql-hackers по дате отправления: