On Wed, Jan 07, 2015 at 10:42:14AM -0500, Alan Stern wrote:
> On Thu, 18 Dec 2014, Greg Kroah-Hartman wrote:
> > That seems reasonable to me, unbinding when a reset is happening is
> > going to be a rare condition, but if we get rid of it, and we try to
> > queue a reset for a device that is gone,
On Thu, 18 Dec 2014, Greg Kroah-Hartman wrote:
> On Thu, Dec 18, 2014 at 02:40:48PM -0500, Alan Stern wrote:
> > On Thu, 18 Dec 2014, Russell King - ARM Linux wrote:
> >
> > > While unplugging a Logitek Keyboard/mouse micro-receiver, I got the
> > > lockdep splat below.
> > >
> > > However, I do
On Thu, Dec 18, 2014 at 02:40:48PM -0500, Alan Stern wrote:
> On Thu, 18 Dec 2014, Russell King - ARM Linux wrote:
>
> > While unplugging a Logitek Keyboard/mouse micro-receiver, I got the
> > lockdep splat below.
> >
> > However, I don't fully understand this splat - I see nothing in
> > flush_w
On Thu, 18 Dec 2014, Russell King - ARM Linux wrote:
> While unplugging a Logitek Keyboard/mouse micro-receiver, I got the
> lockdep splat below.
>
> However, I don't fully understand this splat - I see nothing in
> flush_work() nor process_one_work() making use of "intf->reset_ws" -
> which seem
While unplugging a Logitek Keyboard/mouse micro-receiver, I got the
lockdep splat below.
However, I don't fully understand this splat - I see nothing in
flush_work() nor process_one_work() making use of "intf->reset_ws" -
which seems to be a USB thing. I guess lockdep is being re-used to
validate