Date: Mon, 1 Jul 2013 16:33:13 +0200 From: Svatopluk Kraus <onwahe@gmail.com> To: Sergey Kandaurov <pluknet@gmail.com> Cc: freebsd-current@freebsd.org Subject: Re: too many warnigns for variable 'cons_next_next' Message-ID: <CAFHCsPVvXRL-bPaWMbXcVTe9gfS96NmMb4Wooj6K85JDQy0pHg@mail.gmail.com> In-Reply-To: <CAE-mSOLRK_aSNoY%2BUFLZVADxz0m_ynKJB=JX42CLbNf5E5L2Pg@mail.gmail.com> References: <CAFHCsPUGfK4RiwYWb75qLCkDhzBjN45w182WS1HrYUiYHAKMbg@mail.gmail.com> <CAE-mSOLRK_aSNoY%2BUFLZVADxz0m_ynKJB=JX42CLbNf5E5L2Pg@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Jul 1, 2013 at 3:40 PM, Sergey Kandaurov <pluknet@gmail.com> wrote:
> On 1 July 2013 17:11, Svatopluk Kraus <onwahe@gmail.com> wrote:
> > Hi,
> >
> > could somebody help me, please, with too many following warnings:
> > sys/buf_ring.h:189:33: warning: variable 'cons_next_next' set but not
> used
> >
> > As the warning is produced by header which is included a lot, the
> important
> > warnings are visible poorly among them. I think that
> buf_ring_dequeue_sc()
> > in sys/buf_ring.h could be change to suppress the warning this way:
> > static __inline void *
> > buf_ring_dequeue_sc(struct buf_ring *br)
> > {
> > - uint32_t cons_head, cons_next, cons_next_next;
> > + uint32_t cons_head, cons_next, cons_next_next __unused;
> > uint32_t prod_tail;
> > void *buf;
> >
> > cons_head = br->br_cons_head;
> > prod_tail = br->br_prod_tail;
> >
> > + if (cons_head == prod_tail)
> > + return (NULL);
> >
> > cons_next = (cons_head + 1) & br->br_cons_mask;
> > - cons_next_next = (cons_head + 2) & br->br_cons_mask;
> >
> > - if (cons_head == prod_tail)
> > - return (NULL);
> > #ifdef PREFETCH_DEFINED
> > + cons_next_next = (cons_head + 2) & br->br_cons_mask;
> > if (cons_next != prod_tail) {
> > prefetch(br->br_ring[cons_next]);
> > if (cons_next_next != prod_tail)
> > prefetch(br->br_ring[cons_next_next]);
> > }
> > #endif
> >
>
> Why not put its declaration under #ifdef as well? YMMV.
>
> --
> wbr,
> pluknet
>
It's possible too. But, why not to do some optimalization at the same time?
Svata
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAFHCsPVvXRL-bPaWMbXcVTe9gfS96NmMb4Wooj6K85JDQy0pHg>