Message-Id: <201002150744.51730.edt@aei.ca>
Date:	Mon, 15 Feb 2010 07:44:51 -0500
From:	Ed Tomlinson <edt@....ca>
To:	Dmitry Torokhov <dmitry.torokhov@...il.com>
Cc:	Jiri Kosina <jkosina@...e.cz>, Michael Poole <mdpoole@...ilus.org>,
	linux-input@...r.kernel.org, Marcel Holtmann <marcel@...tmann.org>,
	linux-bluetooth@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 0/2] Provide a driver for the Apple Magic Mouse - opps

On Monday 15 February 2010 02:11:46 Dmitry Torokhov wrote:
> On Sun, Feb 14, 2010 at 09:22:41AM -0500, Ed Tomlinson wrote:
> > On Sunday 14 February 2010 03:03:44 Dmitry Torokhov wrote:
> > > On Sat, Feb 13, 2010 at 02:29:29PM -0500, Ed Tomlinson wrote:
> > > > On Wednesday 10 February 2010 08:57:37 Jiri Kosina wrote:
> > > > > On Tue, 9 Feb 2010, Michael Poole wrote:
> > > > > 
> > > > > > I think this patch is ready for real review.  The Magic Mouse requires
> > > > > > that a driver send an unlock Report(Feature) command, similar to the
> > > > > > Wacom wireless tablet and Sixaxis controller quirks.  This turns on an
> > > > > > Input Report that isn't published in the input Report descriptor that
> > > > > > contains touch data (and usually overrides the normal motion and click
> > > > > > Report).
> > > > > > 
> > > > > > Because the mouse has only one switch and no scroll wheel, the driver
> > > > > > (under control of parameters) emulates a middle button and scroll wheel.
> > > > > > User space could also ignore and/or re-synthesize those events based on
> > > > > > the reported events.
> > > > > > 
> > > > > > The first patch exports hid_register_report() so the driver can turn on
> > > > > > the multitouch report.  The second patch adds the device ID and the
> > > > > > driver.  Some user-space tools to talk to the mouse directly (that is,
> > > > > > when it is not associated with the host's HIDP stack) are at
> > > > > > http://github.com/entrope/linux-magicmouse .
> > > > > 
> > > > > I have applied the driver into apple_magic_mouse branch and merged this 
> > > > > branch into for-next, so it should appear in the upcoming linux-next.
> > > > 
> > > 
> > > > This driver (or the hid changes) can triggers an opps.  What I did was
> > > > start X.  Turn on the magic mouse.  It connected on input7&8.  Then I
> > > > powered it off and on.  This time it conneced on input9&10.  Then I
> > > > exited X and got the opps.  Note my X does not hotplug the magic
> > > > mouse.  I've also included a trace of the udev events that generated
> > > > the log below (if there was a remove after X stopped it would not be
> > > > included).  To my eyes it looks like we leak an input device (there is
> > > > not a remove event for input8).
> > > >
> > > 
> > > Indeed, we seem to be missing call to input_unregister_device() in
> > > magicmouse_remove().
> > 
> > How does this look?  With this udevadm shows input8 being removed and
> > there is no more opps.
> 
> Almost... you need to do hid_hw_stop() first and only then unregister
> input device, Otherwise if you unload the module while moving the mouse
> it is likely to still oops.

How about this?  It applies on top of yesterdays patch.

Ed

---
diff --git a/drivers/hid/hid-magicmouse.c b/drivers/hid/hid-magicmouse.c
index 7d252d2..46fdeee 100644
--- a/drivers/hid/hid-magicmouse.c
+++ b/drivers/hid/hid-magicmouse.c
@@ -430,8 +430,8 @@ static void magicmouse_remove(struct hid_device *hdev)
 {
 	struct magicmouse_sc *msc;
 	msc = hid_get_drvdata(hdev);
-	input_unregister_device(msc->input);
 	hid_hw_stop(hdev);
+	input_unregister_device(msc->input);
 	kfree(msc);
 }
 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists