[PATCH] HID: magicmouse: Don't report REL_{X,Y} for Magic Trackpad

[PATCH] HID: magicmouse: Don't report REL_{X,Y} for Magic Trackpad

[Chase Douglas]

With the recent switch to having the hid layer handle standard axis
initialization, the Magic Trackpad now reports relative axes. This would
be fine in the normal mode, but the driver puts the device in multitouch
mode where no relative events are generated. Also, userspace software
depends on accurate axis information for device type detection. Thus,
ignoring the relative axes from the Magic Trackpad is best.

Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
---
 drivers/hid/hid-magicmouse.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/drivers/hid/hid-magicmouse.c b/drivers/hid/hid-magicmouse.c
index e6dc151..ed732b7 100644
--- a/drivers/hid/hid-magicmouse.c
+++ b/drivers/hid/hid-magicmouse.c
@@ -433,6 +433,11 @@ static int magicmouse_input_mapping(struct hid_device *hdev,
 	if (!msc->input)
 		msc->input = hi->input;
 
+	/* Magic Trackpad does not give relative data after switching to MT */
+	if (hi->input->id.product == USB_DEVICE_ID_APPLE_MAGICTRACKPAD &&
+	    field->flags & HID_MAIN_ITEM_RELATIVE)
+		return -1;
+
 	return 0;
 }
 
-- 
1.7.1

Re: [PATCH] HID: magicmouse: Don't report REL_{X,Y} for Magic Trackpad

[Jiri Kosina]

On Wed, 8 Dec 2010, Chase Douglas wrote:

> With the recent switch to having the hid layer handle standard axis
> initialization, the Magic Trackpad now reports relative axes. This would
> be fine in the normal mode, but the driver puts the device in multitouch
> mode where no relative events are generated. Also, userspace software
> depends on accurate axis information for device type detection. Thus,
> ignoring the relative axes from the Magic Trackpad is best.

Applied, thank you. I take it this should better go into .37 still, right?

Thanks,

-- 
Jiri Kosina
SUSE Labs, Novell Inc.

Re: [PATCH] HID: magicmouse: Don't report REL_{X,Y} for Magic Trackpad

[Chase Douglas]

On 12/09/2010 05:41 AM, Jiri Kosina wrote:
> On Wed, 8 Dec 2010, Chase Douglas wrote:
> 
>> With the recent switch to having the hid layer handle standard axis
>> initialization, the Magic Trackpad now reports relative axes. This would
>> be fine in the normal mode, but the driver puts the device in multitouch
>> mode where no relative events are generated. Also, userspace software
>> depends on accurate axis information for device type detection. Thus,
>> ignoring the relative axes from the Magic Trackpad is best.
> 
> Applied, thank you. I take it this should better go into .37 still, right?

Yes, please.

Thanks,

-- Chase

Re: [PATCH] HID: magicmouse: Don't report REL_{X,Y} for Magic Trackpad

[Chase Douglas]

On 12/09/2010 10:13 AM, Chase Douglas wrote:
> On 12/09/2010 05:41 AM, Jiri Kosina wrote:
>> On Wed, 8 Dec 2010, Chase Douglas wrote:
>>
>>> With the recent switch to having the hid layer handle standard axis
>>> initialization, the Magic Trackpad now reports relative axes. This would
>>> be fine in the normal mode, but the driver puts the device in multitouch
>>> mode where no relative events are generated. Also, userspace software
>>> depends on accurate axis information for device type detection. Thus,
>>> ignoring the relative axes from the Magic Trackpad is best.
>>
>> Applied, thank you. I take it this should better go into .37 still, right?
> 
> Yes, please.

Jiri,

It looks like this never got pushed to .37. Can you add "Cc:
stable@kernel.org" to the SOB area and push it?

Thanks,

-- Chase

Re: [PATCH] HID: magicmouse: Don't report REL_{X,Y} for Magic Trackpad

[Jiri Kosina]

On Mon, 3 Jan 2011, Chase Douglas wrote:

> >>> With the recent switch to having the hid layer handle standard axis
> >>> initialization, the Magic Trackpad now reports relative axes. This would
> >>> be fine in the normal mode, but the driver puts the device in multitouch
> >>> mode where no relative events are generated. Also, userspace software
> >>> depends on accurate axis information for device type detection. Thus,
> >>> ignoring the relative axes from the Magic Trackpad is best.
> >>
> >> Applied, thank you. I take it this should better go into .37 still, right?
> > 
> > Yes, please.
> 
> Jiri,
> 
> It looks like this never got pushed to .37. Can you add "Cc:
> stable@kernel.org" to the SOB area and push it?

Hi Chase,

I am not rebasing my tree, so it's not easy to add Cc: stable@kernel.org 
to the commit once it has been mirrored out.

But I have this commit in my list to send it to -stable once it goes to 
Linus, manually.

Thanks,

-- 
Jiri Kosina
SUSE Labs, Novell Inc.