Re: Improve error reporting in 027_stream_regress test
От | Nazir Bilal Yavuz |
---|---|
Тема | Re: Improve error reporting in 027_stream_regress test |
Дата | |
Msg-id | CAN55FZ3B=cWkj3NhLwY8NiDhXvs34RCUk5i-nZA8TqO5+KWnWA@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: Improve error reporting in 027_stream_regress test (Michael Paquier <michael@paquier.xyz>) |
Список | pgsql-hackers |
Hi, On Tue, 29 Jul 2025 at 04:46, Michael Paquier <michael@paquier.xyz> wrote: > > > - 0002 is the actual patch that improves error reporting in the > > 027_stream_regress test by using the read_file_ends() function. It > > adds a regression_log_helper() function, which reads the > > PG_TEST_FILE_READ_LINES environment variable and then calls > > read_file_ends() with force_line_count set to true. This approach > > avoids any potential race condition where the environment variable > > might be modified after being read in the regression_log_helper() and > > before used in the read_file_ends(). > > Why do you think the environment variable could be changed mid-flight > in the context of a single test run? I mean it is a run-time variable, so it is up to the user. The timing between these two operations are extremely tight, so it is very unlikely but not impossible. Do you think this edge case is not worth accounting for? -- Regards, Nazir Bilal Yavuz Microsoft
В списке pgsql-hackers по дате отправления: