On Thu, 1 Nov 2007, Marcin Ślusarz wrote:
> > > swapper/0 [HC1[1]:SC0[0]:HE0:SE1] takes:
> > > (&port->lock){+...}, at: [<ffffffff880e212c>]
> > > usb_serial_generic_read_bulk_callback+0x6c/0x1c0 [usbserial]
> > > {hardirq-on-W} state was registered at:
> > The fix for this one has already been submitted to Greg --
> > http://lkml.org/lkml/2007/10/28/38
> with this patch I still get:
> =================================
> [ INFO: inconsistent lock state ]
> 2.6.23 #9
> ---------------------------------
> inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> swapper/0 [HC1[1]:SC0[0]:HE0:SE1] takes:
> (&port->lock){+...}, at: [<ffffffff880e912d>]
> usb_serial_generic_read_bulk_callback+0x6d/0x1d0 [usbserial]
> {hardirq-on-W} state was registered at:
> [<ffffffff80251273>] __lock_acquire+0x703/0x1110
> [<ffffffff880e93b6>] usb_serial_generic_write+0x76/0x2e0 [usbserial]
What kernel version are you using? I should have already fixed this one in
current git -- commit acd2a847
--
Jiri Kosina
