Shouldn't we be using openlog's LOG_NOWAIT option? - Mailing list pgsql-hackers

From Tom Lane
Subject Shouldn't we be using openlog's LOG_NOWAIT option?
Date
Msg-id 27872.1088113155@sss.pgh.pa.us
Whole thread Raw
List pgsql-hackers
In the Single Unix Spec man page for syslog, I read

LOG_NOWAIT        Do not wait for child processes that may have been created during the course of logging the message.
     This option should be used by processes that enable notification of child termination using SIGCHLD,        since
syslog()may otherwise block waiting for a child whose exit status has already been collected.
 

Well, the postmaster certainly enables SIGCHLD, so it seems pretty
dangerous that we don't use this option ...
        regards, tom lane


pgsql-hackers by date:

Previous
From: Thomas Hallgren
Date:
Subject: Re: warning missing
Next
From: "Greg Sabino Mullane"
Date:
Subject: Re: [Re] Re: PREPARE and transactions