Re: Fwd: [PATCH 2/3] usb: type-c: USB Type-C Connector System Software Interface

[Heikki Krogerus <heikki.krogerus@linux.intel.com>]

On Thu, Feb 11, 2016 at 10:13:11AM +0200, Andy Shevchenko wrote:
> ---------- Forwarded message ----------
> From: Andy Shevchenko <andy.shevchenko@gmail.com>
> Date: Thu, Feb 11, 2016 at 10:10 AM
> Subject: Re: [PATCH 2/3] usb: type-c: USB Type-C Connector System
> Software Interface
> To: Oliver Neukum <oneukum@suse.de>
> 
> 
> On Wed, Feb 10, 2016 at 5:08 PM, Oliver Neukum <oneukum@suse.de> wrote:
> > On Wed, 2016-02-10 at 16:24 +0200, Andy Shevchenko wrote:
> >> On Wed, Feb 10, 2016 at 4:15 PM, Oliver Neukum <oneukum@suse.com> wrote:
> >> > On Wed, 2016-02-10 at 13:56 +0200, Andy Shevchenko wrote:
> >> >> > +err:
> >> >> > +       if (i > 0)
> >> >> > +               for (; i >= 0; i--, con--)
> >> >> > +                       typec_unregister_port(con->port);
> >> >>
> >> >> Perhaps
> >> >>
> >> >> while (--i >= 0) {
> >> >>  ...
> >> >> }
> >> >
> >> > While we are at it. No we should not change the semantics
> >> > of conditionals for the sake of appearance.
> >>
> >> I'm sorry I didn't get you.
> >> How this more or less standard pattern to clean up stuff on error path
> >> does with conditional semantics?
> >
> > You change a postdecrement to a predecrement. The highest
> > number the loop is executed for is changed.
> 
> I still didn't get.
> Variable i is just counter here,
> 
> And it seems there is a bug, since when i == 1, we will have
> 
> i = 1, con == connector[0]:
> typec_unregister_port(con->port);
> 
> i = 0, con == connector[1]:
> typec_unregister_port(con->port); <<< It wasn't registered yet!
> 
> The correct code should be something like
> if (i > 0)
>  for (--i; i >= 0; i--) {}
> 
> Which
> a) makes conditional redundant;
> b) classical pattern of while (--i >= 0) {}
> 
> So where am I wrong?

I think Andy has a point here.

Thanks,

-- 
heikki