On Mon, Mar 13, 2006 at 04:43:32PM -0500, Alan Stern wrote: > On Mon, 13 Mar 2006, Paul Fulghum wrote: > > > This patch correct the flaw identified by Pete Zaitcev. > > > > Plus I now fully understand the sysfs oops. > > It is not a problem with sysfs, so GregKH rightly ignored me. > > > > A pointer to control_interface->dev is saved in the tty class device > > object created by tty_register_device(). This reference is used to > > create/destroy sysfs symlinks between the tty class device object and the > > control_interface->dev. On disconnect control_interface->dev is released. > > tty_unregister_device() is called later when the last tty file desc > > is closed. tty_unregister_device() then passes the stale pointer > > to sysfs_remove_link() causing the oops if the slab debug is turned on. > > > > So the below patch seems the cleanest way of fixing this. > > Wouldn't it be better to do get_device() on control_interface->dev just > before calling tty_register_device(), so that the data structure isn't > released too soon? (And of course, call put_device() after the tty > device is unregistered.) That way all the sysfs links remain intact, as > intended.
The sysfs core does the 'get' when it creates the symlink. I'm guessing that the acm driver isn't properly handing the reference counting of it, but I haven't looked into it much just yet... Again, I don't think this is the correct change, as we want that symlink there. thanks, greg k-h ------------------------------------------------------- This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
