Re: [PATCH 2/3] driver core: Use rwsem for kill_device() serialization

[Lukas Wunner <lukas@xxxxxxxxx>]

On Thu, Jul 30, 2020 at 08:53:26AM +0200, Greg Kroah-Hartman wrote:
> On Wed, Jul 08, 2020 at 03:27:02PM +0200, Lukas Wunner wrote:
> > kill_device() is currently serialized with driver probing by way of the
> > device_lock().  We're about to serialize it with device_add() as well
> > to prevent addition of children below a device which is going away.
> 
> Why?  Who does this?  Shouldn't the bus that is trying to do this know
> this is happening?

AFAICS, at least spi and i2c are affected.

I first thought that pci is affected as well but it seems the global
pci_lock_rescan_remove() performs the required serialization.

I've yet to take a closer look at acpi and usb.  Any bus which
creates a device hierarchy with dynamic addition & removal needs
to make sure no new children are added after removal of the parent
has begun.


> So, why are you pushing this down into the driver core, can't this be
> done in whatever crazy bus wants to do this, like is done here?

I guess it can.  Let me try to perform the locking at the bus level then.

Thanks,

Lukas