Re: [PATCH for 3.19.y to 3.10.y] net: sysctl_net_core: remove unused variable one

[Greg KH <gregkh@linuxfoundation.org>]

On Sun, Apr 26, 2015 at 10:09:23AM +0200, Dirk Behme wrote:
> On 26.04.2015 09:19, Greg KH wrote:
> >On Sun, Apr 26, 2015 at 09:11:03AM +0200, Dirk Behme wrote:
> >>Back porting the upstream commit b1cb59cf2efe7 ("net: sysctl_net_core: check
> >>SNDBUF and RCVBUF for min length") to -stable introduces a compiler warning
> >>
> >>net/core/sysctl_net_core.c:26:12: warning: ‘one’ defined but not used [-Wunused-variable]
> >>  static int one = 1;
> >>
> >>While upstream has an additional user of the variable 'one' after that commit
> >>("tstamp_allow_data"), -stable doesn't have any user of 'one' any more, resulting
> >>in above compiler warning. Or in other words: -stable doesn't have the
> >>commit b245be1f4db1a0 ("net-timestamp: no-payload only sysctl") introducing the
> >>additional user of 'one'.
> >>
> >>Fix the compiler warning by removing the unused variable.
> >
> >No, this was done on purpose, sorry, I can't take this patch.  The
> >network maintainer wants it this way to be easier to backport things.
> >
> >We can live with the compiler warning.
> 
> Would it be an option to pick
> 
> b245be1f4db1a0 ("net-timestamp: no-payload only sysctl")
> 
> then, too?

That's a new feature, which isn't ok for stable kernels, right?

Also, the networking mailing list and maintainer needs to be asked about
networking stable patches, not me, it's up to him what he wants to see
applied to the stable kernels.

thanks,

greg k-h