The chip is similar, but has status bits at different positions,
so use the correct bits.
Signed-off-by: Andreas Kemnade <andreas@kemnade.info>
---
drivers/input/touchscreen/ektf2127.c | 41 ++++++++++++++++++++++++----
1 file changed, 35 insertions(+), 6 deletions(-)
diff --git a/drivers/input/touchscreen/ektf2127.c b/drivers/input/touchscreen/ektf2127.c
index cc3103b9cbfba..028b605d1bb74 100644
--- a/drivers/input/touchscreen/ektf2127.c
+++ b/drivers/input/touchscreen/ektf2127.c
@@ -13,6 +13,7 @@
* Hans de Goede <hdegoede@redhat.com>
*/
+#include <linux/bits.h>
#include <linux/gpio/consumer.h>
#include <linux/interrupt.h>
#include <linux/i2c.h>
@@ -46,6 +47,11 @@ struct ektf2127_ts {
struct input_dev *input;
struct gpio_desc *power_gpios;
struct touchscreen_properties prop;
+ int status_shift;
+};
+
+struct ektf2127_i2c_chip_data {
+ int status_shift;
};
static void ektf2127_parse_coordinates(const u8 *buf, unsigned int touch_count,
@@ -112,8 +118,8 @@ static void ektf2127_report2_contact(struct ektf2127_ts *ts, int slot,
static void ektf2127_report2_event(struct ektf2127_ts *ts, const u8 *buf)
{
- ektf2127_report2_contact(ts, 0, &buf[1], !!(buf[7] & 2));
- ektf2127_report2_contact(ts, 1, &buf[4], !!(buf[7] & 4));
+ ektf2127_report2_contact(ts, 0, &buf[1], !!(buf[7] & BIT(ts->status_shift)));
+ ektf2127_report2_contact(ts, 1, &buf[4], !!(buf[7] & BIT(ts->status_shift + 1)));
input_mt_sync_frame(ts->input);
input_sync(ts->input);
@@ -246,7 +252,9 @@ static int ektf2127_query_dimension(struct i2c_client *client, bool width)
static int ektf2127_probe(struct i2c_client *client)
{
+ const struct i2c_device_id *id = i2c_client_get_device_id(client);
struct device *dev = &client->dev;
+ const struct ektf2127_i2c_chip_data *chip_data;
struct ektf2127_ts *ts;
struct input_dev *input;
u8 buf[4];
@@ -303,6 +311,17 @@ static int ektf2127_probe(struct i2c_client *client)
return error;
ts->input = input;
+
+ chip_data = device_get_match_data(&client->dev);
+ if (!chip_data)
+ chip_data = (const struct ektf2127_i2c_chip_data *)id->driver_data;
+ if (!chip_data) {
+ dev_err(&client->dev, "missing chip data\n");
+ return -EINVAL;
+ }
+
+ ts->status_shift = chip_data->status_shift;
+
input_set_drvdata(input, ts);
error = devm_request_threaded_irq(dev, client->irq,
@@ -325,18 +344,28 @@ static int ektf2127_probe(struct i2c_client *client)
return 0;
}
+static const struct ektf2127_i2c_chip_data ektf2127_data = {
+ .status_shift = 1,
+};
+
+static const struct ektf2127_i2c_chip_data ektf2232_data = {
+ .status_shift = 0,
+};
+
#ifdef CONFIG_OF
static const struct of_device_id ektf2127_of_match[] = {
- { .compatible = "elan,ektf2127" },
- { .compatible = "elan,ektf2132" },
+ { .compatible = "elan,ektf2127", .data = &ektf2127_data},
+ { .compatible = "elan,ektf2132", .data = &ektf2127_data},
+ { .compatible = "elan,ektf2232", .data = &ektf2232_data},
{}
};
MODULE_DEVICE_TABLE(of, ektf2127_of_match);
#endif
static const struct i2c_device_id ektf2127_i2c_id[] = {
- { "ektf2127", 0 },
- { "ektf2132", 0 },
+ { .name = "ektf2127", .driver_data = (long)&ektf2127_data },
+ { .name = "ektf2132", .driver_data = (long)&ektf2127_data },
+ { .name = "ektf2232", .driver_data = (long)&ektf2232_data },
{}
};
MODULE_DEVICE_TABLE(i2c, ektf2127_i2c_id);
--
2.39.2On Mon, May 6, 2024 at 12:48 AM Andreas Kemnade <andreas@kemnade.info> wrote:
>
> The chip is similar, but has status bits at different positions,
> so use the correct bits.
...
> @@ -46,6 +47,11 @@ struct ektf2127_ts {
> struct input_dev *input;
> struct gpio_desc *power_gpios;
> struct touchscreen_properties prop;
> + int status_shift;
> +};
> +
> +struct ektf2127_i2c_chip_data {
> + int status_shift;
> };
>
> static void ektf2127_parse_coordinates(const u8 *buf, unsigned int touch_count,
I'm wondering if you are using --histogram diff algo when preparing the patches.
...
> + chip_data = device_get_match_data(&client->dev);
> + if (!chip_data)
> + chip_data = (const struct ektf2127_i2c_chip_data *)id->driver_data;
> + if (!chip_data) {
> + dev_err(&client->dev, "missing chip data\n");
> + return -EINVAL;
First of all, there is a special combined API for this, please use it
instead. Second, use dev_err_probe().
> + }
...
> #ifdef CONFIG_OF
Side note: Ideally we should kill this ugliness. But it can be done later on.
--
With Best Regards,
Andy Shevchenko
On Mon, 6 May 2024 15:05:52 +0300
Andy Shevchenko <andy.shevchenko@gmail.com> wrote:
> From: Andy Shevchenko <andy.shevchenko@gmail.com>
> To: Andreas Kemnade <andreas@kemnade.info>
> Cc: dmitry.torokhov@gmail.com, robh@kernel.org, krzk+dt@kernel.org, conor+dt@kernel.org, u.kleine-koenig@pengutronix.de, hdegoede@redhat.com, siebren.vroegindeweij@hotmail.com, linux-input@vger.kernel.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
> Subject: Re: [PATCH v2 3/3] Input: ektf2127 - add ektf2232 support
> Date: Mon, 6 May 2024 15:05:52 +0300
>
> On Mon, May 6, 2024 at 12:48 AM Andreas Kemnade <andreas@kemnade.info> wrote:
> >
> > The chip is similar, but has status bits at different positions,
> > so use the correct bits.
>
> ...
>
> > @@ -46,6 +47,11 @@ struct ektf2127_ts {
> > struct input_dev *input;
> > struct gpio_desc *power_gpios;
> > struct touchscreen_properties prop;
> > + int status_shift;
> > +};
> > +
> > +struct ektf2127_i2c_chip_data {
> > + int status_shift;
> > };
> >
> > static void ektf2127_parse_coordinates(const u8 *buf, unsigned int touch_count,
>
> I'm wondering if you are using --histogram diff algo when preparing the patches.
No, I am not using that, it seems to not make that chunk nicer.
Yes, we want
+ int status_shift;
};
+
+struct ektf2127_i2c_chip_data {
+ int status_shift;
+};
But that is not shorter or simpler, just more readable.
Regards,
Andreas
On Mon, May 6, 2024 at 7:21 PM Andreas Kemnade <andreas@kemnade.info> wrote:
> On Mon, 6 May 2024 15:05:52 +0300
> Andy Shevchenko <andy.shevchenko@gmail.com> wrote:
> > From: Andy Shevchenko <andy.shevchenko@gmail.com>
> > To: Andreas Kemnade <andreas@kemnade.info>
> > Date: Mon, 6 May 2024 15:05:52 +0300
> > On Mon, May 6, 2024 at 12:48 AM Andreas Kemnade <andreas@kemnade.info> wrote:
...
> > I'm wondering if you are using --histogram diff algo when preparing the patches.
>
> No, I am not using that, it seems to not make that chunk nicer.
> Yes, we want
>
> + int status_shift;
> };
> +
> +struct ektf2127_i2c_chip_data {
> + int status_shift;
> +};
>
> But that is not shorter or simpler, just more readable.
And that's exactly what histogram is about. I suggest making it
default for the Linux kernel project (or globally in your
~/.gitconfig).
--
With Best Regards,
Andy Shevchenko
On Mon, 6 May 2024 21:43:14 +0300
Andy Shevchenko <andy.shevchenko@gmail.com> wrote:
> On Mon, May 6, 2024 at 7:21 PM Andreas Kemnade <andreas@kemnade.info> wrote:
> > On Mon, 6 May 2024 15:05:52 +0300
> > Andy Shevchenko <andy.shevchenko@gmail.com> wrote:
> > > From: Andy Shevchenko <andy.shevchenko@gmail.com>
> > > To: Andreas Kemnade <andreas@kemnade.info>
> > > Date: Mon, 6 May 2024 15:05:52 +0300
> > > On Mon, May 6, 2024 at 12:48 AM Andreas Kemnade <andreas@kemnade.info> wrote:
>
> ...
>
> > > I'm wondering if you are using --histogram diff algo when preparing the patches.
> >
> > No, I am not using that, it seems to not make that chunk nicer.
> > Yes, we want
> >
> > + int status_shift;
> > };
> > +
> > +struct ektf2127_i2c_chip_data {
> > + int status_shift;
> > +};
> >
> > But that is not shorter or simpler, just more readable.
>
> And that's exactly what histogram is about. I suggest making it
> default for the Linux kernel project (or globally in your
> ~/.gitconfig).
>
>
again, it does not do anything helpful in this case, I tried to run
git format-patch --histogram with
no improvements. But it might help in other places.
Regards,
Andreas
© 2016 - 2026 Red Hat, Inc.