On 11 September 2015 at 17:35, Mark Brown [off-list ref] wrote:

On Fri, Sep 11, 2015 at 04:06:07PM +0200, Tomeu Vizoso wrote:

quoted

Once a platform device (with the platform bus as its parent) is
retrieved from the deferred queue, both the parent and the device in
question are locked (because of the USB stuff mentioned below). If
that device depends on another device whose parent is the platform bus
and we try to probe it (useless, but I don't see a good way of
avoiding it) then we'll deadlock when device_attach locks that device.

Ugh, that's nasty.  Trying to fix this would most likely devolve into
trying to shove things onto the deferred list in the right order but
that's definitely a very different solution with problems.

Well, that's effectively what my "ordered probing" series did (and
indeed didn't have this problem and parents were made sure to have
probed before their children), but if we want to have dependency
information before the probe starts, then we cannot reuse the
implementation of the DT bindings that we already have in the resource
getters as this series do.

Regards,

Tomeu