Re: [PATCH] driver core: fix shutdown races with probe/remove

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Tue, Jun 05, 2012 at 01:09:45PM -0400, Alan Stern wrote:
> On Tue, 5 Jun 2012, Ming Lei wrote:
> 
> > On Tue, Jun 5, 2012 at 10:47 PM, Alan Stern <stern@rowland.harvard.edu> wrote:
> > >
> > > Avoid running probe, that's fine. ?But avoiding remove can lead to
> > > problems, because the subsystem and the driver will no longer agree on
> > > who should manage the device.
> > 
> > After device_shutdown() has been called, the whole system will enter power down
> > or reset later, so it doesn't matter if who should manage the device.
> 
> Maybe.  But there might be quite some time between the shutdown call 
> and the eventual power-off or reboot.
> 
> > Also once shutdown callback is called for the device, looks its other callbacks
> > should not touch the device any more.
> 
> You shouldn't depend on that.  Shutdown methods generally put the 
> device into a state suitable for power-off or reboot; they don't often 
> guarantee that the driver won't change the state later on.
> 
> On the whole, it might be easier just to hold the device lock during 
> the shutdown call.

That sounds much simpler to me.

greg k-h