Re: useless assignment pointer argument - Mailing list pgsql-hackers

From Gaetano Mendola
Subject Re: useless assignment pointer argument
Date
Msg-id CAJycT5o4iL+1JwzkwyUfv9YVYiPOCzO1b2O+auxp-Zq93DzKjw@mail.gmail.com
Whole thread
In response to Re: useless assignment pointer argument  (Andres Freund <andres@anarazel.de>)
List pgsql-hackers
If the compiler is good the assignment is elided indeed, that's not what I meant to point out.

On Thu, 28 May 2015 at 22:17 Andres Freund <andres@anarazel.de> wrote:
On 2015-05-28 20:14:33 +0000, Gaetano Mendola wrote:
> src/backend/commands/explain.c:1692
> src/backend/commands/explain.c:1874
> src/backend/commands/explain.c:1986
>
> there is the following assignment:
>
>    ancestors = list_delete_first(ancestors);
>
> but it has no effect at all being that a function parameter and not used
> anymore after the assignment itself.

So? It costs little to nothing, and it'll make it much less likely that
a stale version of 'ancestors' is used when the code is expanded.

Greetings,

Andres Freund

pgsql-hackers by date:

Previous
From: Gaetano Mendola
Date:
Subject: Re: Possible pointer dereference
Next
From: Peter Eisentraut
Date:
Subject: Re: psql tabcomplete - minor bugfix - tabcomplete for SET ROLE TO xxx