[HACKERS] Re: Authentication tests, and plain 'password' authentication with aSCRAM verifier

Поиск
Список
Период
Сортировка
От Heikki Linnakangas
Тема [HACKERS] Re: Authentication tests, and plain 'password' authentication with aSCRAM verifier
Дата
Msg-id bb5cf245-a05e-d50e-82c6-9f6fa2aa36a8@iki.fi
обсуждение исходный текст
Ответ на [HACKERS] Authentication tests, and plain 'password' authentication with aSCRAM verifier  (Heikki Linnakangas <hlinnaka@iki.fi>)
Ответы [HACKERS] Re: Authentication tests, and plain 'password' authentication with aSCRAM verifier  (Michael Paquier <michael.paquier@gmail.com>)
Список pgsql-hackers
On 03/14/2017 03:43 PM, Michael Paquier wrote:
> +       /*
> +        * The password looked like a SCRAM verifier, but could not be
> +        * parsed.
> +        */
> +       elog(LOG, "invalid SCRAM verifier for user \"%s\"", username);
> This would be sent back to the client, no? I think that you should use
> *logdetail as well in scram_verify_plain_password.

No, LOG messages are never sent to the client. Well, unless you have 
client_min_messages='log', but then all the LOG messages with details 
would be sent to the clients anyway. (We don't process the GUCs from the 
startup packet until after authentication, so an unauthenticated user 
cannot set client_min_messages='log').

Committed, thanks.

- Heikki



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Heikki Linnakangas
Дата:
Сообщение: [HACKERS] Re: BUG #13755: pgwin32_is_service not checking ifSECURITY_SERVICE_SID is disabled
Следующее
От: vinayak
Дата:
Сообщение: Re: [HACKERS] ANALYZE command progress checker