On Fri, 11 Aug 2017, Michal Suchánek wrote:
> On 2017-08-10 18:30, Jason Gunthorpe wrote:
> > On Thu, Aug 10, 2017 at 12:18:11PM +0200, Michal Suchánek wrote:
> > > > Existing bus implementations do properly chain to driver shutdown (eg
> > > > look at mmc_bus_shutdown) and it appears to have been written like
> > >
> > > Neither isa nor ibmebus does. These are two random buses I tried to
> > > look at.
> >
> > I'm not following, I see this:
> >
> > static void ibmebus_bus_device_shutdown(struct device *dev)
> > {
> > struct platform_device *of_dev = to_platform_device(dev);
> > struct platform_driver *drv = to_platform_driver(dev->driver);
> >
> > if (dev->driver && drv->shutdown)
> > drv->shutdown(of_dev);
> > }
> >
> > It looks to me like in this case the struct device_driver shutdown is
> > not used, and instead the struct platform_driver shutdown is called.
>
> And it is not used even if a device driver sets it and expects it to run.
Which is the kind of landmine it is best avoided in drivers/, so it
would be nice to get WARN_ON() during device register when
dev->shutdown() methods *that are going to be ignored* because of
class/bus handlers are non-NULL...
Either that, or the device->shutdown() methods should always be called,
and drivers that should not/need not have them should be fixed...
--
Henrique Holschuh
