Skip to contentSkip to conversation

Pinefeat cef168 lens control board driver

PATCH Superseded Aliaksandr Smirnou 4 participants 10 messages Aug 17, 2025 – Aug 22, 2025

Subsystems (2)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

v3 cover Pinefeat cef168 lens control board driver

This patch series adds support for the Pinefeat adapter, which interfaces Canon EF and EF-S lenses to non-Canon camera bodies. The cef168 circuit control board provides an I2C interface for electronic focus and aperture control. The driver integrates with the V4L2 sub-device API.

For more information about the product, see:
https://github.com/pinefeat/cef168

Changes in v3:
- removed vcc-supply property and example
- fixed incorrect type in assignment
- fixed cast to restricted
- removed unreachable code
- changed comparison to NULL
- fixed indent in commit message

Link to v2: https://lore.kernel.org/all/[email protected]/

Patches:
dt-bindings: Pinefeat cef168 lens control board
media: i2c: Pinefeat cef168 lens control board driver

 .../bindings/media/i2c/pinefeat,cef168.yaml   |  48 +++
 .../devicetree/bindings/vendor-prefixes.yaml  |   2 +
 MAINTAINERS                                   |   8 +
 drivers/media/i2c/Kconfig                     |   8 +
 drivers/media/i2c/Makefile                    |   1 +
 drivers/media/i2c/cef168.c                    | 335 ++++++++++++++++++
 drivers/media/i2c/cef168.h                    |  51 +++
 7 files changed, 453 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
 create mode 100644 drivers/media/i2c/cef168.c
 create mode 100644 drivers/media/i2c/cef168.h

base-commit: 2b38afce25c4e1b8f943ff4f0a2b51d6c40f2ed2

Signature · 1 line
2.34.1
#1 lore raw

PATCH v3 1/2 dt-bindings: Pinefeat cef168 lens control board

Add the Device Tree schema and examples for the Pinefeat cef168 lens control board. This board interfaces Canon EF & EF-S lenses with non-Canon camera bodies, enabling electronic control of focus and aperture via V4L2.

Power supply is derived from fixed supplies via connector or GPIO header. Therefore, the driver does not manage any regulator, so representing any supply in the binding is redundant.

---
 .../bindings/media/i2c/pinefeat,cef168.yaml   | 48 +++++++++++++++++++
 .../devicetree/bindings/vendor-prefixes.yaml  |  2 +
 MAINTAINERS                                   |  6 +++
 3 files changed, 56 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml new file +48 -0
new file mode 100644
index 000000000000..80cdff40d175
@@ -0,0 +1,48 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+# Copyright (c) 2025 Pinefeat LLP
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/media/i2c/pinefeat,cef168.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Pinefeat cef168 lens driver
+
+maintainers:
+  - Aliaksandr Smirnou <[email protected]>
+
+description: |
+  Pinefeat produces an adapter designed to interface between
+  Canon EF & EF-S lenses and non-Canon camera bodies, incorporating
+  features for electronic focus and aperture adjustment. The cef168
+  circuit board, included with the adapter, provides a software
+  programming interface that allows control of lens focus and
+  aperture positions. This driver enables controlling the lens
+  focus and aperture via the V4L2 (Video4Linux2) API.
+
+properties:
+  compatible:
+    enum:
+      - pinefeat,cef168
+
+  reg:
+    maxItems: 1
+
+required:
+  - compatible
+  - reg
+
+additionalProperties: false
+
+examples:
+  - |
+    i2c {
+        #address-cells = <1>;
+        #size-cells = <0>;
+
+        camera-lens@d {
+            compatible = "pinefeat,cef168";
+            reg = <0x0d>;
+        };
+    };
+
+...
Documentation/devicetree/bindings/vendor-prefixes.yaml +2 -0
index 77160cd47f54..dab27f769b0a 100644
@@ -1195,6 +1195,8 @@ patternProperties:
     description: Picochip Ltd
   "^pine64,.*":
     description: Pine64
+  "^pinefeat,.*":
+    description: Pinefeat LLP
   "^pineriver,.*":
     description: Shenzhen PineRiver Designs Co., Ltd.
   "^pixcir,.*":
MAINTAINERS +6 -0
index fe168477caa4..811c6a150029 100644
@@ -19985,6 +19985,12 @@ S:	Supported
 F:	Documentation/devicetree/bindings/input/pine64,pinephone-keyboard.yaml
 F:	drivers/input/keyboard/pinephone-keyboard.c
 
+PINEFEAT CEF168 LENS DRIVER
+M:	Aliaksandr Smirnou <[email protected]>
+L:	[email protected]
+S:	Supported
+F:	Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
+
 PLANTOWER PMS7003 AIR POLLUTION SENSOR DRIVER
 M:	Tomasz Duszynski <[email protected]>
 S:	Maintained
Signature · 1 line
2.34.1
#2 lore raw

Re: [PATCH v3 1/2] dt-bindings: Pinefeat cef168 lens control board

Quoted 9 lines · Aliaksandr Smirnou

Add the Device Tree schema and examples for the Pinefeat cef168 lens control board. This board interfaces Canon EF & EF-S lenses with non-Canon camera bodies, enabling electronic control of focus and aperture via V4L2.

Power supply is derived from fixed supplies via connector or GPIO header. Therefore, the driver does not manage any regulator, so representing any supply in the binding is redundant.

Wut? This doesn't make sense, you have supplies so they should be documented. The fact that they're shared with a bunch of other things on the SBC you're aiming the product at doesn't matter. What if someone doesn't use this sensor with an RPi and there is a dedicated regulator?

Quoted 1 line

Is that actually your email address?

Quoted 33 lines
---
 .../bindings/media/i2c/pinefeat,cef168.yaml   | 48 +++++++++++++++++++
 .../devicetree/bindings/vendor-prefixes.yaml  |  2 +
 MAINTAINERS                                   |  6 +++
 3 files changed, 56 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml new file +19 -0
new file mode 100644
index 000000000000..80cdff40d175
@@ -0,0 +1,48 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+# Copyright (c) 2025 Pinefeat LLP
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/media/i2c/pinefeat,cef168.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Pinefeat cef168 lens driver
+
+maintainers:
+  - Aliaksandr Smirnou <[email protected]>
+
+description: |
+  Pinefeat produces an adapter designed to interface between
+  Canon EF & EF-S lenses and non-Canon camera bodies, incorporating
+  features for electronic focus and aperture adjustment. The cef168
+  circuit board, included with the adapter, provides a software
+  programming interface that allows control of lens focus and
+  aperture positions.
Quoted 2 lines

This driver enables controlling the lens
+ focus and aperture via the V4L2 (Video4Linux2) API.

Don't mention drivers in bindings, how linux handles things is not relevant to a hardware description.

#3 lore raw

Re: [PATCH v3 1/2] dt-bindings: Pinefeat cef168 lens control board

Quoted 4 lines · Conor Dooley

Wut? This doesn't make sense, you have supplies so they should be documented. The fact that they're shared with a bunch of other things on the SBC you're aiming the product at doesn't matter. What if someone doesn't use this sensor with an RPi and there is a dedicated regulator?

The supply was introduced in the second patch version but later removed following feedback from another maintainer. I’m fine with adding it back if needed - I just want to make sure we’re consistent.

https://lore.kernel.org/all/[email protected]/

Would you prefer it described as before?

Quoted 1 line

Is that actually your email address?

will change to a named one

Quoted 1 line

Don't mention drivers in bindings, how linux handles things is not

relevant to a hardware description.

will remove driver mention

#4 lore raw

Re: [PATCH v3 1/2] dt-bindings: Pinefeat cef168 lens control board

In reply to Conor Dooley
Quoted 7 lines · Conor Dooley
Quoted 5 lines · Aliaksandr Smirnou

Add the Device Tree schema and examples for the Pinefeat cef168 lens control board. This board interfaces Canon EF & EF-S lenses with non-Canon camera bodies, enabling electronic control of focus and aperture via V4L2.

Quoted 7 lines
Quoted 3 lines

Power supply is derived from fixed supplies via connector or GPIO header. Therefore, the driver does not manage any regulator, so representing any supply in the binding is redundant.

Wut? This doesn't make sense, you have supplies so they should be documented. The fact that they're shared with a bunch of other things on the SBC you're aiming the product at doesn't matter. What if someone

There is also some explanation at v2 discussion. I asked for that because there is no known design (neither RPi or other boards having compatible hat/connectors) which would have these supplies controllable. Adding them now would mean you should make them also required (because in fact they are), but since these are non-controllable there would be just regulator-fixed with voltage and that's it. It's just bloat.

Quoted 1 line

doesn't use this sensor with an RPi and there is a dedicated regulator?

Unlikely but if ever such hardware appears, we can always add regulators.

Best regards,
Krzysztof

#5 lore raw

PATCH v3 2/2 media: i2c: Pinefeat cef168 lens control board driver

In reply to Aliaksandr Smirnou

Add support for the Pinefeat cef168 lens control board that provides electronic focus and aperture control for Canon EF & EF-S lenses on non-Canon camera bodies.

---
 MAINTAINERS                |   2 +
 drivers/media/i2c/Kconfig  |   8 +
 drivers/media/i2c/Makefile |   1 +
 drivers/media/i2c/cef168.c | 335 +++++++++++++++++++++++++++++++++++++
 drivers/media/i2c/cef168.h |  51 ++++++
 5 files changed, 397 insertions(+)
 create mode 100644 drivers/media/i2c/cef168.c
 create mode 100644 drivers/media/i2c/cef168.h
MAINTAINERS +2 -0
index 811c6a150029..922efc000722 100644
@@ -19990,6 +19990,8 @@ M:	Aliaksandr Smirnou <[email protected]>
 L:	[email protected]
 S:	Supported
 F:	Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
+F:	drivers/media/i2c/cef168.c
+F:	drivers/media/i2c/cef168.h
 
 PLANTOWER PMS7003 AIR POLLUTION SENSOR DRIVER
 M:	Tomasz Duszynski <[email protected]>
drivers/media/i2c/Kconfig +8 -0
index 6237fe804a5c..c4c3b03a0b98 100644
@@ -791,6 +791,14 @@ config VIDEO_AK7375
 	  capability. This is designed for linear control of
 	  voice coil motors, controlled via I2C serial interface.
 
+config VIDEO_CEF168
+	tristate "CEF168 lens control support"
+	help
+	  This is a driver for the CEF168 lens control board.
+	  The board provides an I2C interface for electronic focus
+	  and aperture control of EF and EF-S lenses. The driver
+	  integrates with the V4L2 sub-device API.
+
 config VIDEO_DW9714
 	tristate "DW9714 lens voice coil support"
 	depends on GPIOLIB
drivers/media/i2c/Makefile +1 -0
index 5873d29433ee..75a95f850f18 100644
@@ -25,6 +25,7 @@ obj-$(CONFIG_VIDEO_BT856) += bt856.o
 obj-$(CONFIG_VIDEO_BT866) += bt866.o
 obj-$(CONFIG_VIDEO_CCS) += ccs/
 obj-$(CONFIG_VIDEO_CCS_PLL) += ccs-pll.o
+obj-$(CONFIG_VIDEO_CEF168) += cef168.o
 obj-$(CONFIG_VIDEO_CS3308) += cs3308.o
 obj-$(CONFIG_VIDEO_CS5345) += cs5345.o
 obj-$(CONFIG_VIDEO_CS53L32A) += cs53l32a.o
drivers/media/i2c/cef168.c new file +335 -0

335 lines not on this page · raw message

drivers/media/i2c/cef168.h new file +51 -0

51 lines not on this page · raw message

Signature · 1 line
2.34.1
#6 lore raw

Re: [PATCH v3 2/2] media: i2c: Pinefeat cef168 lens control board driver

Hi Aliaksandr
thanks for the patch

just a quick review, I'm not a v4l2-control expert so you might want to wait for more reviews on that part..

Quoted 65 lines · Aliaksandr Smirnou

Add support for the Pinefeat cef168 lens control board that provides electronic focus and aperture control for Canon EF & EF-S lenses on non-Canon camera bodies.

---
 MAINTAINERS                |   2 +
 drivers/media/i2c/Kconfig  |   8 +
 drivers/media/i2c/Makefile |   1 +
 drivers/media/i2c/cef168.c | 335 +++++++++++++++++++++++++++++++++++++
 drivers/media/i2c/cef168.h |  51 ++++++
 5 files changed, 397 insertions(+)
 create mode 100644 drivers/media/i2c/cef168.c
 create mode 100644 drivers/media/i2c/cef168.h
MAINTAINERS +2 -0
index 811c6a150029..922efc000722 100644
@@ -19990,6 +19990,8 @@ M:	Aliaksandr Smirnou <[email protected]>
 L:	[email protected]
 S:	Supported
 F:	Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
+F:	drivers/media/i2c/cef168.c
+F:	drivers/media/i2c/cef168.h
 
 PLANTOWER PMS7003 AIR POLLUTION SENSOR DRIVER
 M:	Tomasz Duszynski <[email protected]>
drivers/media/i2c/Kconfig +8 -0
index 6237fe804a5c..c4c3b03a0b98 100644
@@ -791,6 +791,14 @@ config VIDEO_AK7375
 	  capability. This is designed for linear control of
 	  voice coil motors, controlled via I2C serial interface.
 
+config VIDEO_CEF168
+	tristate "CEF168 lens control support"
+	help
+	  This is a driver for the CEF168 lens control board.
+	  The board provides an I2C interface for electronic focus
+	  and aperture control of EF and EF-S lenses. The driver
+	  integrates with the V4L2 sub-device API.
+
 config VIDEO_DW9714
 	tristate "DW9714 lens voice coil support"
 	depends on GPIOLIB
drivers/media/i2c/Makefile +1 -0
index 5873d29433ee..75a95f850f18 100644
@@ -25,6 +25,7 @@ obj-$(CONFIG_VIDEO_BT856) += bt856.o
 obj-$(CONFIG_VIDEO_BT866) += bt866.o
 obj-$(CONFIG_VIDEO_CCS) += ccs/
 obj-$(CONFIG_VIDEO_CCS_PLL) += ccs-pll.o
+obj-$(CONFIG_VIDEO_CEF168) += cef168.o
 obj-$(CONFIG_VIDEO_CS3308) += cs3308.o
 obj-$(CONFIG_VIDEO_CS5345) += cs5345.o
 obj-$(CONFIG_VIDEO_CS53L32A) += cs53l32a.o
drivers/media/i2c/cef168.c new file +0 -0
new file mode 100644
index 000000000000..563251a54835

drivers/media/i2c/cef168.c line 4

Quoted 4 lines
drivers/media/i2c/cef168.c new file +4 -0
@@ -0,0 +1,335 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (c) 2025 Pinefeat LLP
+
+#include <linux/crc8.h>

Do you need to

select CRC8

in Kconfig then ?

drivers/media/i2c/cef168.c line 12

Quoted 8 lines
drivers/media/i2c/cef168.c new file +8 -0
+#include <linux/delay.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/pm_runtime.h>
+#include <media/v4l2-ctrls.h>
+#include <media/v4l2-device.h>
+#include <media/v4l2-event.h>
+#include "cef168.h"

Why an header file ? This isn't going to be included by any other .c file, right ?

Quoted 20 lines
drivers/media/i2c/cef168.c new file +20 -0
+
+/*
+ * cef168 device structure
+ */
+struct cef168_device {
+	struct v4l2_ctrl_handler ctrls;
+	struct v4l2_subdev sd;
+};
+
+static inline struct cef168_device *to_cef168(struct v4l2_ctrl *ctrl)
+{
+	return container_of(ctrl->handler, struct cef168_device, ctrls);
+}
+
+static inline struct cef168_device *sd_to_cef168(struct v4l2_subdev *subdev)
+{
+	return container_of(subdev, struct cef168_device, sd);
+}
+
+static int cef168_i2c_write(struct cef168_device *cef168_dev, u8 cmd, u16 val)

drivers/media/i2c/cef168.c line 42

Quoted 10 lines
drivers/media/i2c/cef168.c new file +10 -0
+{
+	struct i2c_client *client = v4l2_get_subdevdata(&cef168_dev->sd);
+	int retry, ret;
+
+	__le16 le_data = cpu_to_le16(val);
+	char tx_data[4] = { cmd, ((u8 *)&le_data)[0], ((u8 *)&le_data)[1] };
+
+	tx_data[3] = crc8(cef168_crc8_table, tx_data, 3, CRC8_INIT_VALUE);
+
+	for (retry = 0; retry < 3; retry++) {

This seems a bit random, why do you need to retry three times ?

Quoted 66 lines
drivers/media/i2c/cef168.c new file +66 -0
+		ret = i2c_master_send(client, tx_data, sizeof(tx_data));
+		if (ret == sizeof(tx_data))
+			return 0;
+		else if (ret != -EIO && ret != -EREMOTEIO)
+			break;
+	}
+
+	dev_err(&client->dev, "I2C write fail after %d retries, ret=%d\n",
+		retry, ret);
+	return -EIO;
+}
+
+static int cef168_i2c_read(struct cef168_device *cef168_dev,
+			   struct cef168_data *rx_data)
+{
+	struct i2c_client *client = v4l2_get_subdevdata(&cef168_dev->sd);
+
+	int ret = i2c_master_recv(client, (char *)rx_data,
+				  sizeof(struct cef168_data));
+	if (ret != sizeof(struct cef168_data)) {
+		dev_err(&client->dev, "I2C read fail, ret=%d\n", ret);
+		return -EIO;
+	}
+
+	u8 computed_crc = crc8(cef168_crc8_table, (const u8 *)rx_data,
+			       sizeof(struct cef168_data) - 1, CRC8_INIT_VALUE);
+	if (computed_crc != rx_data->crc8) {
+		dev_err(&client->dev,
+			"CRC mismatch calculated=0x%02X read=0x%02X\n",
+			computed_crc, rx_data->crc8);
+		return -EIO;
+	}
+
+	rx_data->moving_time = le16_to_cpup((__le16 *)&rx_data->moving_time);
+	rx_data->focus_position_min = le16_to_cpup((__le16 *)&rx_data->focus_position_min);
+	rx_data->focus_position_max = le16_to_cpup((__le16 *)&rx_data->focus_position_max);
+	rx_data->focus_position_cur = le16_to_cpup((__le16 *)&rx_data->focus_position_cur);
+	rx_data->focus_distance_min = le16_to_cpup((__le16 *)&rx_data->focus_distance_min);
+	rx_data->focus_distance_max = le16_to_cpup((__le16 *)&rx_data->focus_distance_max);
+
+	return 0;
+}
+
+static int cef168_set_ctrl(struct v4l2_ctrl *ctrl)
+{
+	struct cef168_device *dev = to_cef168(ctrl);
+	u8 cmd;
+
+	switch (ctrl->id) {
+	case V4L2_CID_FOCUS_ABSOLUTE:
+		return cef168_i2c_write(dev, INP_SET_FOCUS, ctrl->val);
+	case V4L2_CID_FOCUS_RELATIVE:
+		cmd = ctrl->val < 0 ? INP_SET_FOCUS_N : INP_SET_FOCUS_P;
+		return cef168_i2c_write(dev, cmd, abs(ctrl->val));
+	case V4L2_CID_IRIS_ABSOLUTE:
+		return cef168_i2c_write(dev, INP_SET_APERTURE, ctrl->val);
+	case V4L2_CID_IRIS_RELATIVE:
+		cmd = ctrl->val < 0 ? INP_SET_APERTURE_N : INP_SET_APERTURE_P;
+		return cef168_i2c_write(dev, cmd, abs(ctrl->val));
+	case CEF168_V4L2_CID_CUSTOM(calibrate):
+		return cef168_i2c_write(dev, INP_CALIBRATE, 0);
+	}
+
+	return -EINVAL;
+}
+

drivers/media/i2c/cef168.c line 118

Quoted 10 lines
drivers/media/i2c/cef168.c new file +10 -0
+static int cef168_get_ctrl(struct v4l2_ctrl *ctrl)
+{
+	struct cef168_device *dev = to_cef168(ctrl);
+	int rval;
+
+	if (ctrl->id != V4L2_CID_FOCUS_ABSOLUTE &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(data) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(focus_range) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(lens_id))
+		return -EINVAL;

If you mark them WRITE_ONLY wouldn't the core take care of this ?

drivers/media/i2c/cef168.c line 120

Quoted 2 lines
drivers/media/i2c/cef168.c new file +2 -0
+
+	struct cef168_data data;

I thought the compiler would complain for variables declared not at the beginning of the function

Quoted 163 lines
drivers/media/i2c/cef168.c new file +163 -0
+
+	rval = cef168_i2c_read(dev, &data);
+	if (rval < 0)
+		return rval;
+
+	switch (ctrl->id) {
+	case V4L2_CID_FOCUS_ABSOLUTE:
+		ctrl->val = data.focus_position_cur;
+		return 0;
+	case CEF168_V4L2_CID_CUSTOM(focus_range):
+		ctrl->p_new.p_u32[0] = ((u32)data.focus_position_min << 16) |
+				       (u32)data.focus_position_max;
+		return 0;
+	case CEF168_V4L2_CID_CUSTOM(lens_id):
+		ctrl->p_new.p_u8[0] = data.lens_id;
+		return 0;
+	case CEF168_V4L2_CID_CUSTOM(data):
+		memcpy(ctrl->p_new.p_u8, &data, sizeof(data));
+		return 0;
+	}
+
+	return -EINVAL;
+}
+
+static const struct v4l2_ctrl_ops cef168_ctrl_ops = {
+	.g_volatile_ctrl = cef168_get_ctrl,
+	.s_ctrl = cef168_set_ctrl,
+};
+
+static const struct v4l2_ctrl_config cef168_lens_id_ctrl = {
+	.ops = &cef168_ctrl_ops,
+	.id = CEF168_V4L2_CID_CUSTOM(lens_id),
+	.type = V4L2_CTRL_TYPE_U8,
+	.name = "Lens ID",
+	.min = 0,
+	.max = U8_MAX,
+	.step = 1,
+	.def = 0,
+	.flags = V4L2_CTRL_FLAG_VOLATILE | V4L2_CTRL_FLAG_READ_ONLY,
+};
+
+static const struct v4l2_ctrl_config cef168_focus_range_ctrl = {
+	.ops = &cef168_ctrl_ops,
+	.id = CEF168_V4L2_CID_CUSTOM(focus_range),
+	.type = V4L2_CTRL_TYPE_U32,
+	.name = "Focus Range",
+	.min = 0,
+	.max = U32_MAX,
+	.step = 1,
+	.def = 0,
+	.flags = V4L2_CTRL_FLAG_VOLATILE | V4L2_CTRL_FLAG_READ_ONLY,
+};
+
+static const struct v4l2_ctrl_config cef168_data_ctrl = {
+	.ops = &cef168_ctrl_ops,
+	.id = CEF168_V4L2_CID_CUSTOM(data),
+	.type = V4L2_CTRL_TYPE_U8,
+	.name = "Data",
+	.min = 0,
+	.max = U8_MAX,
+	.step = 1,
+	.def = 0,
+	.dims = { sizeof(struct cef168_data) / sizeof(u8) },
+	.elem_size = sizeof(u8),
+	.flags = V4L2_CTRL_FLAG_VOLATILE | V4L2_CTRL_FLAG_READ_ONLY,
+};
+
+static const struct v4l2_ctrl_config cef168_calibrate_ctrl = {
+	.ops = &cef168_ctrl_ops,
+	.id = CEF168_V4L2_CID_CUSTOM(calibrate),
+	.type = V4L2_CTRL_TYPE_BUTTON,
+	.name = "Calibrate",
+};
+
+static int cef168_open(struct v4l2_subdev *sd, struct v4l2_subdev_fh *fh)
+{
+	return pm_runtime_resume_and_get(sd->dev);
+}
+
+static int cef168_close(struct v4l2_subdev *sd, struct v4l2_subdev_fh *fh)
+{
+	pm_runtime_put(sd->dev);
+	return 0;
+}
+
+static const struct v4l2_subdev_internal_ops cef168_int_ops = {
+	.open = cef168_open,
+	.close = cef168_close,
+};
+
+static const struct v4l2_subdev_core_ops cef168_core_ops = {
+	.log_status = v4l2_ctrl_subdev_log_status,
+	.subscribe_event = v4l2_ctrl_subdev_subscribe_event,
+	.unsubscribe_event = v4l2_event_subdev_unsubscribe,
+};
+
+static const struct v4l2_subdev_ops cef168_ops = {
+	.core = &cef168_core_ops,
+};
+
+static void cef168_subdev_cleanup(struct cef168_device *cef168_dev)
+{
+	v4l2_async_unregister_subdev(&cef168_dev->sd);
+	v4l2_ctrl_handler_free(&cef168_dev->ctrls);
+	media_entity_cleanup(&cef168_dev->sd.entity);
+}
+
+static int cef168_init_controls(struct cef168_device *dev)
+{
+	struct v4l2_ctrl *ctrl;
+	struct v4l2_ctrl_handler *hdl = &dev->ctrls;
+	const struct v4l2_ctrl_ops *ops = &cef168_ctrl_ops;
+
+	v4l2_ctrl_handler_init(hdl, 8);
+
+	ctrl = v4l2_ctrl_new_std(hdl, ops, V4L2_CID_FOCUS_ABSOLUTE, 0, S16_MAX,
+				 1, 0);
+	if (ctrl)
+		ctrl->flags |= V4L2_CTRL_FLAG_VOLATILE |
+			       V4L2_CTRL_FLAG_EXECUTE_ON_WRITE;
+	v4l2_ctrl_new_std(hdl, ops, V4L2_CID_FOCUS_RELATIVE, S16_MIN, S16_MAX,
+			  1, 0);
+	ctrl = v4l2_ctrl_new_std(hdl, ops, V4L2_CID_IRIS_ABSOLUTE, 0, S16_MAX,
+				 1, 0);
+	if (ctrl)
+		ctrl->flags |= V4L2_CTRL_FLAG_WRITE_ONLY |
+			       V4L2_CTRL_FLAG_EXECUTE_ON_WRITE;
+	v4l2_ctrl_new_std(hdl, ops, V4L2_CID_IRIS_RELATIVE, S16_MIN, S16_MAX, 1,
+			  0);
+	v4l2_ctrl_new_custom(hdl, &cef168_calibrate_ctrl, NULL);
+	v4l2_ctrl_new_custom(hdl, &cef168_focus_range_ctrl, NULL);
+	v4l2_ctrl_new_custom(hdl, &cef168_data_ctrl, NULL);
+	v4l2_ctrl_new_custom(hdl, &cef168_lens_id_ctrl, NULL);
+
+	if (hdl->error)
+		dev_err(dev->sd.dev, "%s fail error: 0x%x\n", __func__,
+			hdl->error);
+	dev->sd.ctrl_handler = hdl;
+	return hdl->error;
+}
+
+static int cef168_probe(struct i2c_client *client)
+{
+	struct cef168_device *cef168_dev;
+	int rval;
+
+	cef168_dev = devm_kzalloc(&client->dev, sizeof(*cef168_dev),
+				  GFP_KERNEL);
+	if (!cef168_dev)
+		return -ENOMEM;
+
+	v4l2_i2c_subdev_init(&cef168_dev->sd, client, &cef168_ops);
+	cef168_dev->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE |
+				V4L2_SUBDEV_FL_HAS_EVENTS;
+	cef168_dev->sd.internal_ops = &cef168_int_ops;
+
+	rval = cef168_init_controls(cef168_dev);
+	if (rval)
+		goto err_cleanup;
+
+	rval = media_entity_pads_init(&cef168_dev->sd.entity, 0, NULL);
+	if (rval < 0)
+		goto err_cleanup;

drivers/media/i2c/cef168.c line 293

Quoted 10 lines
drivers/media/i2c/cef168.c new file +10 -0
+
+	cef168_dev->sd.entity.function = MEDIA_ENT_F_LENS;
+
+	rval = v4l2_async_register_subdev(&cef168_dev->sd);
+	if (rval < 0)
+		goto err_cleanup;
+
+	crc8_populate_msb(cef168_crc8_table, CEF_CRC8_POLYNOMIAL);
+
+	pm_runtime_set_active(&client->dev);

Is the device powered up at this point ?

drivers/media/i2c/cef168.c line 295

Quoted 2 lines
drivers/media/i2c/cef168.c new file +2 -0
+	pm_runtime_enable(&client->dev);
+	pm_runtime_idle(&client->dev);

If you depend on the pm_runtime_resume_and_get() call in open() to power the device up, then you need to depend on PM in KConfig ?

Quoted 50 lines
drivers/media/i2c/cef168.c new file +40 -0
+
+	return 0;
+
+err_cleanup:
+	v4l2_ctrl_handler_free(&cef168_dev->ctrls);
+	media_entity_cleanup(&cef168_dev->sd.entity);
+
+	return rval;
+}
+
+static void cef168_remove(struct i2c_client *client)
+{
+	struct v4l2_subdev *sd = i2c_get_clientdata(client);
+	struct cef168_device *cef168_dev = sd_to_cef168(sd);
+
+	pm_runtime_disable(&client->dev);
+	pm_runtime_set_suspended(&client->dev);
+	cef168_subdev_cleanup(cef168_dev);
+}
+
+static const struct of_device_id cef168_of_table[] = {
+	{ .compatible = "pinefeat,cef168" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, cef168_of_table);
+
+static struct i2c_driver cef168_i2c_driver = {
+	.driver = {
+		.name = CEF168_NAME,
+		.of_match_table = cef168_of_table,
+	},
+	.probe = cef168_probe,
+	.remove = cef168_remove,
+};
+
+module_i2c_driver(cef168_i2c_driver);
+
+MODULE_AUTHOR("[email protected]>");
+MODULE_DESCRIPTION("CEF168 lens driver");
+MODULE_LICENSE("GPL");
drivers/media/i2c/cef168.h new file +4 -0
new file mode 100644
index 000000000000..cdce1a19bda0
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Pinefeat cef168 lens driver
+ *

drivers/media/i2c/cef168.h line 14

Quoted 10 lines
drivers/media/i2c/cef168.h new file +10 -0
+ * Copyright (c) 2025 Pinefeat LLP
+ */
+
+#ifndef CEF168_CEF168_H
+#define CEF168_CEF168_H
+
+#define CEF168_NAME "cef168"
+
+#define CEF168_V4L2_CID_CUSTOM(ctrl) \
+	((V4L2_CID_USER_BASE | 168) + custom_##ctrl)

I think you need to reserve space for your controls in include/uapi/linux/v4l2-controls.h

otherwise this will never be visible to applications ?

drivers/media/i2c/cef168.h line 18

Quoted 4 lines
drivers/media/i2c/cef168.h new file +4 -0
+
+enum { custom_lens_id, custom_data, custom_focus_range, custom_calibrate };
+
+/**

No need to kerneldoc unless you properly document all fields and include the file in some of the Documentation/

Quoted 37 lines
drivers/media/i2c/cef168.h new file +33 -1
+ * cef168 data structure
+ */
+struct cef168_data {
+	__u8 lens_id;
+	__u8 moving : 1;
+	__u8 calibrating : 2;
+	__u16 moving_time;
+	__u16 focus_position_min;
+	__u16 focus_position_max;
+	__u16 focus_position_cur;
+	__u16 focus_distance_min;
+	__u16 focus_distance_max;
+	__u8 crc8;
+} __packed;
+
+/*
+ * cef168 I2C protocol commands
+ */
+#define INP_CALIBRATE 0x22
+#define INP_SET_FOCUS 0x80
+#define INP_SET_FOCUS_P 0x81
+#define INP_SET_FOCUS_N 0x82
+#define INP_SET_APERTURE 0x7A
+#define INP_SET_APERTURE_P 0x7B
+#define INP_SET_APERTURE_N 0x7C
+
+#define CEF_CRC8_POLYNOMIAL 168
+
+#ifdef DECLARE_CRC8_TABLE
+DECLARE_CRC8_TABLE(cef168_crc8_table);
+#endif
+
+#endif //CEF168_CEF168_H
--

2.34.1

#7 lore raw

Re: [PATCH v3 2/2] media: i2c: Pinefeat cef168 lens control board driver

Hi Jacopo,

Thank you for the review. Your remarks are very helpful. While I'll apply most of them, could you clarify the one regarding pm_runtime_ for me?

Quoted 3 lines · Jacopo Mondi
Quoted 1 line

+#include <linux/crc8.h>

Do you need to "select CRC8" in Kconfig then ?

Yes, I'll include it.

Quoted 3 lines
Quoted 1 line

+#include "cef168.h"

Why an header file ?

Ok, I'll remove the header moving everying in the .c file.

Quoted 3 lines

drivers/media/i2c/cef168.c line 42

Quoted 1 line
drivers/media/i2c/cef168.c new file +1 -0
+	for (retry = 0; retry < 3; retry++) {

This seems a bit random, why do you need to retry three times ?

The driver retries writes to work around an issue in the Raspberry Pi's I2C hardware, where the BCM2835 mishandles clock stretching. When the slave stretches the clock, the Pi can misread the SCL line or sample data too early, making it think the write failed. To improve reliability, the kernel driver automatically retries the write, effectively compensating for the hardware's timing bug.

Quoted 6 lines

drivers/media/i2c/cef168.c line 118

Quoted 4 lines
drivers/media/i2c/cef168.c new file +4 -0
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(data) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(focus_range) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(lens_id))
+		return -EINVAL;

If you mark them WRITE_ONLY wouldn't the core take care of this ?

These controls are read-only. The data they return depens on the lens.

Quoted 4 lines

drivers/media/i2c/cef168.c line 120

Quoted 1 line
drivers/media/i2c/cef168.c new file +1 -0
+	struct cef168_data data;

I thought the compiler would complain for variables declared not at the beginning of the function

Ok, I'll move the variable at the beginning.

Quoted 5 lines

drivers/media/i2c/cef168.c line 293

Quoted 1 line
drivers/media/i2c/cef168.c new file +1 -0
+	pm_runtime_set_active(&client->dev);

Is the device powered up at this point ?
If you depend on the pm_runtime_resume_and_get() call in open() to power the device up, then you need to depend on PM in KConfig ?

Yes, the device powers from 3v3 rail of a SBC, which makes it powered up as soon as the SBC is up. Given that, should I remove all code around Power Management Runtime (pm_runtime_*) as redundant?

Quoted 7 lines

drivers/media/i2c/cef168.h line 14

Quoted 2 lines
drivers/media/i2c/cef168.h new file +2 -0
+#define CEF168_V4L2_CID_CUSTOM(ctrl) \
+	((V4L2_CID_USER_BASE | 168) + custom_##ctrl)

I think you need to reserve space for your controls in include/uapi/linux/v4l2-controls.h

otherwise this will never be visible to applications ?

I found there is no need for that. Custom control become available automatically by name via the v4l2-ctl utility. For example, the focus range can be read directly in the terminal as follows:

v4l2-ctl -d $DEV_LENS -C focus_range

Quoted 5 lines
Quoted 2 lines

+/**
+ * cef168 data structure

No need to kerneldoc unless you properly document all fields and include the file in some of the Documentation/

OK, I'll remove the comment above the structure.

Kind regards,
Aliaksandr

#8 lore raw

Re: [PATCH v3 2/2] media: i2c: Pinefeat cef168 lens control board driver

Hello
+cc Hans

Quoted 22 lines · Aliaksandr Smirnou

Hi Jacopo,

Thank you for the review. Your remarks are very helpful. While I'll apply most of them, could you clarify the one regarding pm_runtime_ for me?

Quoted 3 lines · Jacopo Mondi
Quoted 1 line

+#include <linux/crc8.h>

Do you need to "select CRC8" in Kconfig then ?

Yes, I'll include it.

Quoted 3 lines
Quoted 1 line

+#include "cef168.h"

Why an header file ?

Ok, I'll remove the header moving everying in the .c file.

Quoted 3 lines

drivers/media/i2c/cef168.c line 42

Quoted 1 line
drivers/media/i2c/cef168.c new file +1 -0
+	for (retry = 0; retry < 3; retry++) {

This seems a bit random, why do you need to retry three times ?

Quoted 7 lines

The driver retries writes to work around an issue in the Raspberry Pi's I2C hardware, where the BCM2835 mishandles clock stretching. When the slave stretches the clock, the Pi can misread the SCL line or sample data too early, making it think the write failed. To improve reliability, the kernel driver automatically retries the write, effectively compensating for the hardware's timing bug.

mmm, usually a mainline driver is not the right place for introducing in workarounds for a specific host/soc, as the assumption is that the driver should work on all platforms equally.

Your case is certainly different as this product is pi-specific...

With a comment I wouldn't mind too much. Maybe others do :)

Quoted 9 lines
Quoted 6 lines

drivers/media/i2c/cef168.c line 118

Quoted 4 lines
drivers/media/i2c/cef168.c new file +4 -0
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(data) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(focus_range) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(lens_id))
+		return -EINVAL;

If you mark them WRITE_ONLY wouldn't the core take care of this ?

These controls are read-only. The data they return depens on the lens.

Sorry, I wasn't clear.

If you mark as WO the controls you don't accept here, will the core handle this for you ?

Quoted 13 lines
Quoted 4 lines

drivers/media/i2c/cef168.c line 120

Quoted 1 line
drivers/media/i2c/cef168.c new file +1 -0
+	struct cef168_data data;

I thought the compiler would complain for variables declared not at the beginning of the function

Ok, I'll move the variable at the beginning.

Quoted 5 lines

drivers/media/i2c/cef168.c line 293

Quoted 1 line
drivers/media/i2c/cef168.c new file +1 -0
+	pm_runtime_set_active(&client->dev);

Is the device powered up at this point ?
If you depend on the pm_runtime_resume_and_get() call in open() to power the device up, then you need to depend on PM in KConfig ?

Quoted 4 lines

Yes, the device powers from 3v3 rail of a SBC, which makes it powered up as soon as the SBC is up. Given that, should I remove all code around Power Management Runtime (pm_runtime_*) as redundant?

Yes, I overlooked the fact you don't regiser any resume/suspend routine, so your calls to pm_runtime_set_active/pm_runtime_put are nops (they do at least provide usage counting, but I would anyway remove them).

Quoted 6 lines
Quoted 6 lines

drivers/media/i2c/cef168.h line 14

Quoted 2 lines
drivers/media/i2c/cef168.h new file +2 -0
+#define CEF168_V4L2_CID_CUSTOM(ctrl) \
+	((V4L2_CID_USER_BASE | 168) + custom_##ctrl)

I think you need to reserve space for your controls in include/uapi/linux/v4l2-controls.h

Quoted 8 lines
Quoted 1 line

otherwise this will never be visible to applications ?

I found there is no need for that. Custom control become available automatically by name via the v4l2-ctl utility. For example, the focus range can be read directly in the terminal as follows:

v4l2-ctl -d $DEV_LENS -C focus_range

Yes the driver enuemrates them, but you need to add them to the main header, otherwise USER_BASE | 168 will be take by someone else.

See in example 7c8c957ef12c41968adb66d785ce1dd5fb2f96e7 which is the latest commit that adds a custom control space.

Hans, do we need a uapi header file with the definition of the custom controls registered by this driver ?

Thanks
j

Quoted 10 lines
Quoted 5 lines
Quoted 2 lines

+/**
+ * cef168 data structure

No need to kerneldoc unless you properly document all fields and include the file in some of the Documentation/

OK, I'll remove the comment above the structure.

Kind regards,
Aliaksandr

#9 lore raw

Re: [PATCH v3 2/2] media: i2c: Pinefeat cef168 lens control board driver

Quoted 7 lines · Jacopo Mondi
Quoted 7 lines
Quoted 6 lines

drivers/media/i2c/cef168.c line 118

Quoted 4 lines
drivers/media/i2c/cef168.c new file +4 -0
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(data) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(focus_range) &&
+	    ctrl->id != CEF168_V4L2_CID_CUSTOM(lens_id))
+		return -EINVAL;

If you mark them WRITE_ONLY wouldn't the core take care of this ?

Quoted 7 lines
Quoted 2 lines

These controls are read-only. The data they return depens on the lens.

Sorry, I wasn't clear.

If you mark as WO the controls you don't accept here, will the core handle this for you ?

I see what you mean now. Indeed, the other controls are alredy WO, so the core will not let them pass here. I'll remove this check as redundant.

Quoted 12 lines
Quoted 12 lines
Quoted 7 lines

drivers/media/i2c/cef168.h line 14

Quoted 2 lines
drivers/media/i2c/cef168.h new file +2 -0
+#define CEF168_V4L2_CID_CUSTOM(ctrl) \
+	((V4L2_CID_USER_BASE | 168) + custom_##ctrl)

I think you need to reserve space for your controls in include/uapi/linux/v4l2-controls.h

otherwise this will never be visible to applications ?

I found there is no need for that. Custom control become available automatically by name via the v4l2-ctl utility. For example, the focus range can be read directly in the terminal as follows:

Quoted 6 lines
Quoted 2 lines

v4l2-ctl -d $DEV_LENS -C focus_range

Yes the driver enuemrates them, but you need to add them to the main header, otherwise USER_BASE | 168 will be take by someone else.

I see, ok, I'll reserve 16 controls for this driver in v4l2-controls.h.

Thank you for your help.

Kind regards,
Aliaksandr