Re: [PATCH v3] input : wacom_w8001 - report resolution to userland

On Mon, Jan 31, 2011 at 9:26 AM, Dmitry Torokhov
[off-list ref] wrote:

On Mon, Jan 31, 2011 at 12:50:14PM +0100, Henrik Rydberg wrote:

quoted

Hi Ping,

On Fri, Jan 28, 2011 at 10:37:42PM -0800, Ping Cheng wrote:

quoted

Serial devices send both pen and touch data through the same logical
port. Since we scaled touch to pen maximum, we use pen resolution
for touch as well here.

This is under the assumption that pen and touch share the same physical
surface. In the case when a small physical dimensional difference occurs
between pen and touch, we assume the tolerance for touch point precision
is higher than pen and the difference is within touch point tolerance.

A per-MT tool based resolution mechanism should be introduced if the
above assumption does not hold true for the pen and touch devices any
more. A good candidate to resolve this issue without introducing design
change could be to report touch resolution through ABS_MT_POSITION_X/Y.

This still sounds a bit like mixing apples and oranges. If this case
would ever happen, and I doubt it since the whole idea is to use
different tools on the same surface, then splitting into separate
devices would be appropriate.

quoted

This version corrected an oversimplified if statement in v2.

Signed-off-by: Ping Cheng <redacted>
---

quoted

 drivers/input/touchscreen/wacom_w8001.c |   20 +++++++++++++++-----
 1 files changed, 15 insertions(+), 5 deletions(-)

diff --git a/drivers/input/touchscreen/wacom_w8001.c b/drivers/input/touchscreen/wacom_w8001.c
index 5cb8449..68dcca0 100644
--- a/drivers/input/touchscreen/wacom_w8001.c
+++ b/drivers/input/touchscreen/wacom_w8001.c

@@ -51,6 +51,10 @@ MODULE_LICENSE("GPL");

 #define W8001_PKTLEN_TPCCTL        11      /* control packet */
 #define W8001_PKTLEN_TOUCH2FG      13

+/* resolution in points/mm */
+#define W8001_PEN_RESOLUTION    100
+#define W8001_TOUCH_RESOLUTION  10
+
 struct w8001_coord {
    u8 rdy;
    u8 tsw;

@@ -198,7 +202,7 @@ static void parse_touchquery(u8 *data, struct w8001_touch_query *query)

            query->y = 1024;
            if (query->panel_res)
                    query->x = query->y = (1 << query->panel_res);
-           query->panel_res = 10;
+           query->panel_res = W8001_TOUCH_RESOLUTION;
    }
 }

@@ -394,6 +398,8 @@ static int w8001_setup(struct w8001 *w8001)

            input_set_abs_params(dev, ABS_X, 0, coord.x, 0, 0);
            input_set_abs_params(dev, ABS_Y, 0, coord.y, 0, 0);
+           input_abs_set_res(dev, ABS_X, W8001_PEN_RESOLUTION);
+           input_abs_set_res(dev, ABS_Y, W8001_PEN_RESOLUTION);
            input_set_abs_params(dev, ABS_PRESSURE, 0, coord.pen_pressure, 0, 0);
            if (coord.tilt_x && coord.tilt_y) {
                    input_set_abs_params(dev, ABS_TILT_X, 0, coord.tilt_x, 0, 0);

@@ -418,15 +424,19 @@ static int w8001_setup(struct w8001 *w8001)

            w8001->max_touch_x = touch.x;
            w8001->max_touch_y = touch.y;

So max_touch is set here

quoted

-           /* scale to pen maximum */
+           /* pen supported */
            if (w8001->max_pen_x && w8001->max_pen_y) {
+                   /* scale to pen maximum */
                    touch.x = w8001->max_pen_x;
                    touch.y = w8001->max_pen_y;

And possibly redefined here

quoted

+           } else { /* touch only */
+                   /* set axes with touch data */
+                   input_set_abs_params(dev, ABS_X, 0, touch.x, 0, 0);
+                   input_set_abs_params(dev, ABS_Y, 0, touch.y, 0, 0);
+                   input_abs_set_res(dev, ABS_X, touch.panel_res);
+                   input_abs_set_res(dev, ABS_Y, touch.panel_res);
            }

-           input_set_abs_params(dev, ABS_X, 0, touch.x, 0, 0);
-           input_set_abs_params(dev, ABS_Y, 0, touch.y, 0, 0);
-
            switch (touch.sensor_id) {
            case 0:
            case 2:

And then just used to set MT later on, which will lead to
ABS_MT_POSITION_X values different from ABS_X. This is just getting
messier. In fact, the whole setup function is messy, with unclear
handling of error paths. Any chance one could simplify the logic
involved here? It should be possible to use a single code path for
input_set_abs_params() and input_abs_set_res(), for instance.

I have the following for the resolution reporting:

It works for me. Although the end result is the same as the v3 since
params and res have already been set in pen response blok if pen is
supported, it is more straight forward in logical to anyone who tries
to understand the code.

Thank you, Dmitry.

Ping

--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html