[PATCH 0/2] hid-mosart: small patches

[PATCH 0/2] hid-mosart: small patches

[Benjamin Tissoires]

Hi guys,

Just a few patch before introducing hid-multitouch. It would be great if
this could be in 2.6.38 (or even before...) as hid-multitouch is a bit late.

This small patch-set enhance the support of hid-mosart in the kernel:
the first patch allows X to handle it as an absolute device, and the second
one introduce suspend/resume support.

The patch work is against the branch next of Dmitry's tree.

Henrik, I think that you added a device in your ppa.
Do you want me to send it?

Cheers,
Benjamin

[PATCH 1/2] hid-mosart: ignore buttons report

[Benjamin Tissoires]

This commit allows the device to be recognized as a touchscreen, and not a
touchpad by xf86-input-evdev.

The device has 2 modes. The first one is an emulation of a touchscreen by
sending left and right button, and the second mode is the one used in
dual-touch (sending trackingID, touch and else).

That's why there is a hid report containing left and right buttons
(9000001 and 9000002). The point is that xorg relies on these fields to
determine if it's a touchpad or a touchscreen.
Clearing the report (return -1) makes xorg detecting it out of the box
as a quite pleasant (dual)touchscreen.

Signed-off-by: Benjamin Tissoires <benjamin.tissoires@enac.fr>
---
 drivers/hid/hid-mosart.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/drivers/hid/hid-mosart.c b/drivers/hid/hid-mosart.c
index ac5421d..acd8a49 100644
--- a/drivers/hid/hid-mosart.c
+++ b/drivers/hid/hid-mosart.c
@@ -90,6 +90,10 @@ static int mosart_input_mapping(struct hid_device *hdev, struct hid_input *hi,
 	case 0xff000000:
 		/* ignore HID features */
 		return -1;
+
+	case HID_UP_BUTTON:
+		/* ignore buttons */
+		return -1;
 	}
 
 	return 0;
-- 
1.7.3.2

Re: [PATCH 1/2] hid-mosart: ignore buttons report

[Chase Douglas]

On 12/13/2010 06:59 AM, Benjamin Tissoires wrote:
> This commit allows the device to be recognized as a touchscreen, and not a
> touchpad by xf86-input-evdev.
> 
> The device has 2 modes. The first one is an emulation of a touchscreen by
> sending left and right button, and the second mode is the one used in
> dual-touch (sending trackingID, touch and else).
> 
> That's why there is a hid report containing left and right buttons
> (9000001 and 9000002). The point is that xorg relies on these fields to
> determine if it's a touchpad or a touchscreen.
> Clearing the report (return -1) makes xorg detecting it out of the box
> as a quite pleasant (dual)touchscreen.
> 
> Signed-off-by: Benjamin Tissoires <benjamin.tissoires@enac.fr>

Acked-by: Chase Douglas <chase.douglas@canonical.com>

Please add "Cc: stable@kernel.org" to the SOB area so this is picked up
for previous kernels too.

Thanks!

-- Chase

Re: [PATCH 1/2] hid-mosart: ignore buttons report

[Henrik Rydberg]

On 12/13/2010 03:59 PM, Benjamin Tissoires wrote:

> This commit allows the device to be recognized as a touchscreen, and not a
> touchpad by xf86-input-evdev.
> 
> The device has 2 modes. The first one is an emulation of a touchscreen by
> sending left and right button, and the second mode is the one used in
> dual-touch (sending trackingID, touch and else).
> 
> That's why there is a hid report containing left and right buttons
> (9000001 and 9000002). The point is that xorg relies on these fields to
> determine if it's a touchpad or a touchscreen.
> Clearing the report (return -1) makes xorg detecting it out of the box
> as a quite pleasant (dual)touchscreen.
> 
> Signed-off-by: Benjamin Tissoires <benjamin.tissoires@enac.fr>
> ---


Signed-off-by: Henrik Rydberg <rydberg@euromail.se>

Thanks,
Henrik

[PATCH 2/2] hid-mosart: support suspend/resume

[Benjamin Tissoires]

The device has 2 modes. The first one is an emulation of a touchscreen
by sending left and right button, and the second mode is the one used in
dual-touch (sending trackingID, touch and else).

In case of a suspend/resume, the device switch back to the first mode
described above (with left and right buttons).
This adds a hook in .reset_resume for the device to be switched to
the correct mode (I just copied the code in mosart_probe).

Signed-off-by: Benjamin Tissoires <benjamin.tissoires@enac.fr>
---
 drivers/hid/hid-mosart.c |   16 ++++++++++++++++
 1 files changed, 16 insertions(+), 0 deletions(-)

diff --git a/drivers/hid/hid-mosart.c b/drivers/hid/hid-mosart.c
index acd8a49..15d0381 100644
--- a/drivers/hid/hid-mosart.c
+++ b/drivers/hid/hid-mosart.c
@@ -234,6 +234,19 @@ static int mosart_probe(struct hid_device *hdev, const struct hid_device_id *id)
 	return ret;
 }
 
+#ifdef CONFIG_PM
+static int mosart_reset_resume(struct hid_device *hdev)
+{
+	struct hid_report_enum *re = hdev->report_enum
+						+ HID_FEATURE_REPORT;
+	struct hid_report *r = re->report_id_hash[7];
+
+	r->field[0]->value[0] = 0x02;
+	usbhid_submit_report(hdev, r, USB_DIR_OUT);
+	return 0;
+}
+#endif
+
 static void mosart_remove(struct hid_device *hdev)
 {
 	hid_hw_stop(hdev);
@@ -262,6 +275,9 @@ static struct hid_driver mosart_driver = {
 	.input_mapped = mosart_input_mapped,
 	.usage_table = mosart_grabbed_usages,
 	.event = mosart_event,
+#ifdef CONFIG_PM
+	.reset_resume = mosart_reset_resume,
+#endif
 };
 
 static int __init mosart_init(void)
-- 
1.7.3.2

Re: [PATCH 2/2] hid-mosart: support suspend/resume

[Chase Douglas]

On 12/13/2010 06:59 AM, Benjamin Tissoires wrote:
> The device has 2 modes. The first one is an emulation of a touchscreen
> by sending left and right button, and the second mode is the one used in
> dual-touch (sending trackingID, touch and else).
> 
> In case of a suspend/resume, the device switch back to the first mode
> described above (with left and right buttons).
> This adds a hook in .reset_resume for the device to be switched to
> the correct mode (I just copied the code in mosart_probe).
> 
> Signed-off-by: Benjamin Tissoires <benjamin.tissoires@enac.fr>

Acked-by: Chase Douglas <chase.douglas@canonical.com>

Please add "Cc: stable@kernel.org" here too.

Thanks!

-- Chase

Re: [PATCH 2/2] hid-mosart: support suspend/resume

[Henrik Rydberg]

On 12/13/2010 03:59 PM, Benjamin Tissoires wrote:

> The device has 2 modes. The first one is an emulation of a touchscreen
> by sending left and right button, and the second mode is the one used in
> dual-touch (sending trackingID, touch and else).
> 
> In case of a suspend/resume, the device switch back to the first mode
> described above (with left and right buttons).
> This adds a hook in .reset_resume for the device to be switched to
> the correct mode (I just copied the code in mosart_probe).


Defining a static function for the mode switch would be nice.

Thanks,
Henrik

Re: [PATCH 0/2] hid-mosart: small patches

[Henrik Rydberg]

Benjamin,

Thanks for the patches, I will comment them separately.

> Just a few patch before introducing hid-multitouch. It would be great if

> this could be in 2.6.38 (or even before...) as hid-multitouch is a bit late.


The first patch looks like 2.6.37 material. I know both patches have been tested
for quite some time.

> Henrik, I think that you added a device in your ppa.

> Do you want me to send it?


Yes, please do. There is a hid-core change to go with it as well.

Jiri, I assume these will eventually go through your tree.

Thanks,
Henrik

Re: [PATCH 0/2] hid-mosart: small patches

[Jiri Kosina]

On Mon, 13 Dec 2010, Henrik Rydberg wrote:

> > Henrik, I think that you added a device in your ppa.
> 
> > Do you want me to send it?
> 
> Yes, please do. There is a hid-core change to go with it as well.
> 
> Jiri, I assume these will eventually go through your tree.

Thanks.

I have applied both Benjamin's patches now.

-- 
Jiri Kosina
SUSE Labs, Novell Inc.

Re: [PATCH 0/2] hid-mosart: small patches

[Chase Douglas]

On 12/14/2010 02:41 AM, Jiri Kosina wrote:
> I have applied both Benjamin's patches now.

Jiri,

If you haven't sent a pull request to Linus yet, can you add 'Cc:
stable@kernel.org'?

Thanks,

-- Chase