Re: [PATCH] ARM: amba: Fix leak of driver_override attribute value

[Greg KH <gregkh@linuxfoundation.org>]

On Thu, Sep 20, 2018 at 09:16:36AM +0200, Geert Uytterhoeven wrote:
> Hi Greg,
> 
> On Thu, Sep 20, 2018 at 9:09 AM Greg KH <gregkh@linuxfoundation.org> wrote:
> > On Thu, Sep 20, 2018 at 08:48:36AM +0200, Geert Uytterhoeven wrote:
> > > On Thu, Sep 20, 2018 at 1:48 AM Trent Piepho <tpiepho@impinj.com> wrote:
> > > > If driver_override was set when a device was released the string would
> > > > not be kfree'ed in amba_device_release and thus leaked when the amba
> > > > device was freed.
> > > >
> > > > Cc: Russell King <linux@armlinux.org.uk>
> > > > Cc: Todd Kjos <tkjos@google.com>
> > > > Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> > > > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > > Signed-off-by: Trent Piepho <tpiepho@impinj.com>
> > >
> > > Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
> > > Fixes: 3cf385713460eb2b ("ARM: 8256/1: driver coamba: add device
> > > binding path 'driver_override'")
> >
> > Then it should also have a cc: stable, right?
> 
> Perhaps. I usually leave that up to the maintainer, else git send-email sends
> it to stable immediately.

That's fine, no one ever complains about that.  In fact it is _good_ to
have that happen, as it gives us stable people a "heads up" that
something is coming to resolve a reported problems.

> The modern backporting AI will consider it anyway, due to the subject, and
> the Fixes tag, right?

Don't count on the "AI" to pick things up if you _know_ it resolves a
problem, like you have said here.

So please, just add it when you know it needs to be backported,
otherwise it might never get backported.

thanks,

greg k-h