Re: New flex warnings - Mailing list pgsql-patches

From Tom Lane
Subject Re: New flex warnings
Date
Msg-id 14275.1210345952@sss.pgh.pa.us
Whole thread Raw
In response to New flex warnings  (Peter Eisentraut <peter_e@gmx.net>)
List pgsql-patches
Peter Eisentraut <peter_e@gmx.net> writes:
> With GCC 4.3, I get warnings from every flex scanner that 'input' is defined
> but not used.  This can be solved by adding %option noinput.  I tested this
> option with a current flex and with the old 2.5.4a; both accept it.  See
> attached patch.  Does anyone see problems with this?

Hm, I wonder why we didn't see those before ... [ looks at code... ]
Oh: yyinput() recurses internally, so even though it's not called
from anywhere else, older gcc's don't realize it's really unreferenced.

I confirm 2.5.4 has the noinput option.  Patch seems ok from here,

            regards, tom lane

pgsql-patches by date:

Previous
From: Simon Riggs
Date:
Subject: Re: Verified fix for Bug 4137
Next
From: Tom Lane
Date:
Subject: Re: Verified fix for Bug 4137