[PATCH] Input: elants_i2c - Do not bind to i2c-hid compatible ACPI instantiated devices

Several users have been reporting that elants_i2c gives several errors
during probe and that their touchscreen does not work on their Lenovo AMD
based laptops with a touchscreen with a ELAN0001 ACPI hardware-id:

[    0.550596] elants_i2c i2c-ELAN0001:00: i2c-ELAN0001:00 supply vcc33 not found, using dummy regulator
[    0.551836] elants_i2c i2c-ELAN0001:00: i2c-ELAN0001:00 supply vccio not found, using dummy regulator
[    0.560932] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.562427] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.595925] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.597974] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.621893] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.622504] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.632650] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (4d 61 69 6e): -121
[    0.634256] elants_i2c i2c-ELAN0001:00: boot failed: -121
[    0.699212] elants_i2c i2c-ELAN0001:00: invalid 'hello' packet: 00 00 ff ff
[    1.630506] elants_i2c i2c-ELAN0001:00: Failed to read fw id: -121
[    1.645508] elants_i2c i2c-ELAN0001:00: unknown packet 00 00 ff ff

Despite these errors, the elants_i2c driver stays bound to the device
(it returns 0 from its probe method despite the errors), blocking the
i2c-hid driver from binding.

Manually unbinding the elants_i2c driver and binding the i2c-hid driver
makes the touchscreen work.

Check if the ACPI-fwnode for the touchscreen contains one of the i2c-hid
compatiblity-id strings and if it has the I2C-HID spec's DSM to get the
HID descriptor address, If it has both then make elants_i2c not bind,
so that the i2c-hid driver can bind.

This assumes that non of the (older) elan touchscreens which actually
need the elants_i2c driver falsely advertise an i2c-hid compatiblity-id
+ DSM in their ACPI-fwnodes. If some of them actually do have this
false advertising, then this change may lead to regressions.

Cc: Benjamin Tissoires <redacted>
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=207759
Signed-off-by: Hans de Goede <redacted>
---
 drivers/input/touchscreen/elants_i2c.c | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)

diff --git a/drivers/input/touchscreen/elants_i2c.c b/drivers/input/touchscreen/elants_i2c.c
index 4c2b579f6c8b..510638e5ba5a 100644
--- a/drivers/input/touchscreen/elants_i2c.c
+++ b/drivers/input/touchscreen/elants_i2c.c

@@ -1334,6 +1334,12 @@ static void elants_i2c_power_off(void *_data)
 	}
 }
 
+static const struct acpi_device_id i2c_hid_ids[] = {
+	{"ACPI0C50", 0 },
+	{"PNP0C50", 0 },
+	{ },
+};
+
 static int elants_i2c_probe(struct i2c_client *client,
 			    const struct i2c_device_id *id)
 {

@@ -1342,6 +1348,25 @@ static int elants_i2c_probe(struct i2c_client *client,
 	unsigned long irqflags;
 	int error;
 
+#ifdef CONFIG_ACPI
+	/* Don't bind to i2c-hid compatible devices, these are handled by the i2c-hid drv. */
+	if (acpi_match_device_ids(ACPI_COMPANION(&client->dev), i2c_hid_ids) == 0) {
+		static const guid_t i2c_hid_guid =
+			GUID_INIT(0x3CDFF6F7, 0x4267, 0x4555,
+				  0xAD, 0x05, 0xB3, 0x0A, 0x3D, 0x89, 0x38, 0xDE);
+		acpi_handle handle = ACPI_HANDLE(&client->dev);
+		union acpi_object *obj;
+
+		obj = acpi_evaluate_dsm_typed(handle, &i2c_hid_guid, 1, 1, NULL,
+					      ACPI_TYPE_INTEGER);
+		if (obj) {
+			dev_warn(&client->dev, "elants_i2c: This device appears to be an I2C-HID device, not binding\n");
+			ACPI_FREE(obj);
+			return -ENODEV;
+		}
+	}
+#endif
+
 	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
 		dev_err(&client->dev,
 			"%s: i2c check functionality error\n", DEVICE_NAME);

-- 
2.30.1

Hi Hans,

On Tue, Mar 2, 2021 at 3:51 PM Hans de Goede [off-list ref] wrote:

Several users have been reporting that elants_i2c gives several errors
during probe and that their touchscreen does not work on their Lenovo AMD
based laptops with a touchscreen with a ELAN0001 ACPI hardware-id:

[    0.550596] elants_i2c i2c-ELAN0001:00: i2c-ELAN0001:00 supply vcc33 not found, using dummy regulator
[    0.551836] elants_i2c i2c-ELAN0001:00: i2c-ELAN0001:00 supply vccio not found, using dummy regulator
[    0.560932] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.562427] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.595925] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.597974] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.621893] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.622504] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.632650] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (4d 61 69 6e): -121
[    0.634256] elants_i2c i2c-ELAN0001:00: boot failed: -121
[    0.699212] elants_i2c i2c-ELAN0001:00: invalid 'hello' packet: 00 00 ff ff
[    1.630506] elants_i2c i2c-ELAN0001:00: Failed to read fw id: -121
[    1.645508] elants_i2c i2c-ELAN0001:00: unknown packet 00 00 ff ff

Despite these errors, the elants_i2c driver stays bound to the device
(it returns 0 from its probe method despite the errors), blocking the
i2c-hid driver from binding.

Manually unbinding the elants_i2c driver and binding the i2c-hid driver
makes the touchscreen work.

Check if the ACPI-fwnode for the touchscreen contains one of the i2c-hid
compatiblity-id strings and if it has the I2C-HID spec's DSM to get the
HID descriptor address, If it has both then make elants_i2c not bind,
so that the i2c-hid driver can bind.

This assumes that non of the (older) elan touchscreens which actually
need the elants_i2c driver falsely advertise an i2c-hid compatiblity-id
+ DSM in their ACPI-fwnodes. If some of them actually do have this
false advertising, then this change may lead to regressions.

I like the v2 much better, and I believe the regressions should be
fairly small now.

One nitpick below:

Andy in https://patchwork.kernel.org/project/linux-input/patch/20210226193225.47129-3-andriy.shevchenko@linux.intel.com/
proposed a few cleanups for i2c-hid.
I believe a similar pacth should be squashed here (the include is the
one I am more worried about).

Besides that:
Acked-by: Benjamin Tissoires <redacted>

Cheers,
Benjamin

+               acpi_handle handle = ACPI_HANDLE(&client->dev);
+               union acpi_object *obj;
+
+               obj = acpi_evaluate_dsm_typed(handle, &i2c_hid_guid, 1, 1, NULL,
+                                             ACPI_TYPE_INTEGER);
+               if (obj) {
+                       dev_warn(&client->dev, "elants_i2c: This device appears to be an I2C-HID device, not binding\n");
+                       ACPI_FREE(obj);
+                       return -ENODEV;
+               }
+       }
+#endif
+
        if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
                dev_err(&client->dev,
                        "%s: i2c check functionality error\n", DEVICE_NAME);
--
2.30.1

On Tue, Mar 02, 2021 at 03:50:57PM +0100, Hans de Goede wrote:

This ideally needs to be protected by CONFIG_ACPI.

No need for "elants_i2c" prefix as dev_warn already gives driver info I
believe.

+			ACPI_FREE(obj);
+			return -ENODEV;
+		}
+	}
+#endif

Could we tuck this away into "elants_acpi_is_hid_device" and have #ifdef
protecting that and have a complementing stub?

+
 	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
 		dev_err(&client->dev,
 			"%s: i2c check functionality error\n", DEVICE_NAME);

As a cleanup should probably drop device prefix from this message as
well.

Thanks.

-- 
Dmitry

Hi,

On 3/2/21 8:42 PM, Dmitry Torokhov wrote:

On Tue, Mar 02, 2021 at 03:50:57PM +0100, Hans de Goede wrote:

quoted

Several users have been reporting that elants_i2c gives several errors
during probe and that their touchscreen does not work on their Lenovo AMD
based laptops with a touchscreen with a ELAN0001 ACPI hardware-id:

[    0.550596] elants_i2c i2c-ELAN0001:00: i2c-ELAN0001:00 supply vcc33 not found, using dummy regulator
[    0.551836] elants_i2c i2c-ELAN0001:00: i2c-ELAN0001:00 supply vccio not found, using dummy regulator
[    0.560932] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.562427] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.595925] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.597974] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.621893] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (77 77 77 77): -121
[    0.622504] elants_i2c i2c-ELAN0001:00: software reset failed: -121
[    0.632650] elants_i2c i2c-ELAN0001:00: elants_i2c_send failed (4d 61 69 6e): -121
[    0.634256] elants_i2c i2c-ELAN0001:00: boot failed: -121
[    0.699212] elants_i2c i2c-ELAN0001:00: invalid 'hello' packet: 00 00 ff ff
[    1.630506] elants_i2c i2c-ELAN0001:00: Failed to read fw id: -121
[    1.645508] elants_i2c i2c-ELAN0001:00: unknown packet 00 00 ff ff

Despite these errors, the elants_i2c driver stays bound to the device
(it returns 0 from its probe method despite the errors), blocking the
i2c-hid driver from binding.

Manually unbinding the elants_i2c driver and binding the i2c-hid driver
makes the touchscreen work.

Check if the ACPI-fwnode for the touchscreen contains one of the i2c-hid
compatiblity-id strings and if it has the I2C-HID spec's DSM to get the
HID descriptor address, If it has both then make elants_i2c not bind,
so that the i2c-hid driver can bind.

This assumes that non of the (older) elan touchscreens which actually
need the elants_i2c driver falsely advertise an i2c-hid compatiblity-id
+ DSM in their ACPI-fwnodes. If some of them actually do have this
false advertising, then this change may lead to regressions.

Cc: Benjamin Tissoires <redacted>
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=207759
Signed-off-by: Hans de Goede <redacted>
---
 drivers/input/touchscreen/elants_i2c.c | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)

diff --git a/drivers/input/touchscreen/elants_i2c.c b/drivers/input/touchscreen/elants_i2c.c
index 4c2b579f6c8b..510638e5ba5a 100644
--- a/drivers/input/touchscreen/elants_i2c.c
+++ b/drivers/input/touchscreen/elants_i2c.c

@@ -1334,6 +1334,12 @@ static void elants_i2c_power_off(void *_data)
 	}
 }
 
+static const struct acpi_device_id i2c_hid_ids[] = {
+	{"ACPI0C50", 0 },
+	{"PNP0C50", 0 },
+	{ },
+};

This ideally needs to be protected by CONFIG_ACPI.

quoted

+
 static int elants_i2c_probe(struct i2c_client *client,
 			    const struct i2c_device_id *id)
 {

@@ -1342,6 +1348,25 @@ static int elants_i2c_probe(struct i2c_client *client,
 	unsigned long irqflags;
 	int error;
 
+#ifdef CONFIG_ACPI
+	/* Don't bind to i2c-hid compatible devices, these are handled by the i2c-hid drv. */
+	if (acpi_match_device_ids(ACPI_COMPANION(&client->dev), i2c_hid_ids) == 0) {
+		static const guid_t i2c_hid_guid =
+			GUID_INIT(0x3CDFF6F7, 0x4267, 0x4555,
+				  0xAD, 0x05, 0xB3, 0x0A, 0x3D, 0x89, 0x38, 0xDE);
+		acpi_handle handle = ACPI_HANDLE(&client->dev);
+		union acpi_object *obj;
+
+		obj = acpi_evaluate_dsm_typed(handle, &i2c_hid_guid, 1, 1, NULL,
+					      ACPI_TYPE_INTEGER);
+		if (obj) {
+			dev_warn(&client->dev, "elants_i2c: This device appears to be an I2C-HID device, not binding\n");

No need for "elants_i2c" prefix as dev_warn already gives driver info I
believe.

Right, I will fix this for v2.

quoted

+			ACPI_FREE(obj);
+			return -ENODEV;
+		}
+	}
+#endif

Could we tuck this away into "elants_acpi_is_hid_device" and have #ifdef
protecting that and have a complementing stub?

That is a good idea, I'll do that for v2.

quoted

+
 	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
 		dev_err(&client->dev,
 			"%s: i2c check functionality error\n", DEVICE_NAME);

As a cleanup should probably drop device prefix from this message as
well.

Ack.

Regards,

Hans