Re: Lifetime of FmgrInfo - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Lifetime of FmgrInfo
Date
Msg-id 12213.1239944938@sss.pgh.pa.us
Whole thread Raw
In response to Lifetime of FmgrInfo  (Joshua Tolley <eggyknap@gmail.com>)
List pgsql-hackers
Joshua Tolley <eggyknap@gmail.com> writes:
> It then does some work to determine whether the result in "function" is
> valid or not. So I got to wondering, what's the lifetime of the
> FunctionCallInfoinfo object passed to the call handler function?

Query lifespan, usually.  There are counterexamples on both sides,
but as a rule it's expected that it's worthwhile for a function to
cache anything it can in the fcinfo->flinfo->fn_extra structure.
If it's not worthwhile, that's the caller's fault not the function's.
        regards, tom lane


pgsql-hackers by date:

Previous
From: KaiGai Kohei
Date:
Subject: [PATCH] unalias of ACL_SELECT_FOR_UPDATE
Next
From: Heikki Linnakangas
Date:
Subject: Re: [PATCH] unalias of ACL_SELECT_FOR_UPDATE