Re: "unix_socket_directories" should be GUC_LIST_INPUT?

From: Ian Lawrence Barwick <barwick(at)gmail(dot)com>
To: Michael Paquier <michael(at)paquier(dot)xyz>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: "unix_socket_directories" should be GUC_LIST_INPUT?
Date: 2020-10-23 03:23:28
Message-ID: CAB8KJ=ixUyxMJLb6-mrQv0_TK-JHh=Eiqtazs2h8pvWCWUyYNg@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

2020年10月23日(金) 12:12 Michael Paquier <michael(at)paquier(dot)xyz>:
>
> On Fri, Oct 23, 2020 at 11:34:06AM +0900, Ian Lawrence Barwick wrote:
> > Not that I've ever had to do this (or would want to do it on a production
> > system), but this error message seems incorrect:
> >
> > postgres=# ALTER SYSTEM SET unix_socket_directories =
> > '/tmp/sock1','/tmp/sock2';
> > ERROR: SET unix_socket_directories takes only one argument
> >
> > Trivial patch attached.
>
> I have never seen that case, but I think that you are right. Still,
> that's not the end of it, see by yourself what the following command
> generates with only your patch, which is fancy:
> ALTER SYSTEM SET unix_socket_directories = '/tmp/sock1','/tmp/, sock2';
>
> We need an extra GUC_LIST_QUOTE on top of what you are proposing.

Ah yes, good point.

Updated version attached.

Regards

Ian Barwick

--
EnterpriseDB: https://www.enterprisedb.com

Attachment Content-Type Size
unix_socket_directories-guc_list_input.v2.patch text/x-patch 557 bytes

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2020-10-23 03:29:19 Re: Resetting spilled txn statistics in pg_stat_replication
Previous Message bucoo@sohu.com 2020-10-23 03:20:31 Re: Re: parallel distinct union and aggregate support patch