Re: [PATCH v3] driver core: Allow device link operations inside sync_state()

[Saravana Kannan]

On Thu, Nov 14, 2019 at 2:45 PM Rafael J. Wysocki <rafael@xxxxxxxxxx> wrote:
>
> On Thu, Nov 14, 2019 at 11:23 PM Saravana Kannan <saravanak@xxxxxxxxxx> wrote:
> >
> > Some sync_state() implementations might need to call APIs that in turn
> > make calls to device link APIs. So, do the sync_state() callbacks
> > without holding the device link lock.
> >
> > Signed-off-by: Saravana Kannan <saravanak@xxxxxxxxxx>
> > ---
> >  drivers/base/core.c | 80 ++++++++++++++++++++++++++++++++++++++++-----
> >  1 file changed, 72 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/base/core.c b/drivers/base/core.c
> > index e6d3e6d485da..9a88bcfd5d33 100644
> > --- a/drivers/base/core.c
> > +++ b/drivers/base/core.c
> > @@ -695,7 +695,26 @@ int device_links_check_suppliers(struct device *dev)
> >         return ret;
> >  }
> >
> > -static void __device_links_supplier_sync_state(struct device *dev)
> > +/**
> > + * __device_links_queue_sync_state - Queue a device for sync_state() callback
> > + * @dev: Device to call sync_state() on
> > + * @list: List head to queue the @dev on
> > + *
> > + * Queues a device for a sync_state() callback when the device links write lock
> > + * isn't held. This allows the sync_state() execution flow to use device links
> > + * APIs.  The caller must ensure this function is called with
> > + * device_links_write_lock() held.
> > + *
> > + * This function does a get_device() to make sure the device is not freed while
> > + * on this list.
> > + *
> > + * So the caller must also ensure that device_links_flush_sync_list() is called
> > + * as soon as the caller releases device_links_write_lock().  This is necessary
> > + * to make sure the sync_state() is called in a timely fashion and the
> > + * put_device() is called on this device.
> > + */
> > +static void __device_links_queue_sync_state(struct device *dev,
> > +                                           struct list_head *list)
> >  {
> >         struct device_link *link;
> >
> > @@ -709,12 +728,46 @@ static void __device_links_supplier_sync_state(struct device *dev)
> >                         return;
> >         }
> >
> > -       if (dev->bus->sync_state)
> > -               dev->bus->sync_state(dev);
> > -       else if (dev->driver && dev->driver->sync_state)
> > -               dev->driver->sync_state(dev);
> > -
> > +       /*
> > +        * Set the flag here to avoid adding the same device to a list more
> > +        * than once. This can happen if new consumers get added to the device
> > +        * and probed before the list is flushed.
> > +        */
> >         dev->state_synced = true;
> > +
> > +       if (list_empty(&dev->links.defer_sync)) {
> > +               get_device(dev);
> > +               list_add_tail(&dev->links.defer_sync, list);
> > +       } else {
> > +               WARN_ON(1);
> > +       }
>
> Now you can do:
>
>         if (WARN_ON(!list_empty(&dev->links.defer_sync)))
>                 return;
>
>         get_device(dev);
>         list_add_tail(&dev->links.defer_sync, list);
>
> The rest of the patch LGTM.

Sending the patch right away in hopes of getting a reviewed-by or ack :)

-Saravana