Re: Fix to expose a GUC variable, Log_disconnections, to outside of postgres.c
| От | Satoshi Nagayasu | 
|---|---|
| Тема | Re: Fix to expose a GUC variable, Log_disconnections, to outside of postgres.c | 
| Дата | |
| Msg-id | 559E1A9D.1070802@uptime.jp обсуждение исходный текст  | 
		
| Ответ на | Re: Fix to expose a GUC variable, Log_disconnections, to outside of postgres.c (Heikki Linnakangas <hlinnaka@iki.fi>) | 
| Список | pgsql-hackers | 
On 2015/07/09 15:30, Heikki Linnakangas wrote: > On 07/09/2015 07:19 AM, Satoshi Nagayasu wrote: >> >> On 2015/07/09 13:06, Tom Lane wrote: >>> Satoshi Nagayasu <snaga@uptime.jp> writes: >>>> I just found that Log_disconnections value has not been >>>> exposed to outside of postgres.c. >>>> Despite that, Log_connections has already been exposed. >>> >>> Why would an external module need to touch either one? >> >> To check settings of GUC variables from a shared preload >> library. >> >> For example, I'd like to print "WARNING ...." in _PG_init() >> when some GUC variable is disabled on postmaster startup. > > I still have a hard time seeing why an extension would be interested in > Log_disconnections. But hey, extensions do strange things.. Definitely. :) > You could use GetConfigOption(). I'm not necessarily opposed to exposing > Log_disconnections, but with GetConfigOption() it will work with earlier > versions too. That's it! This is what I'm looking for. Thanks a lot. :) Regards, -- NAGAYASU Satoshi <snaga@uptime.jp>
В списке pgsql-hackers по дате отправления: