Pinefeat cef168 lens control board driver
Subsystems (2)
v6 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 v6:
- removed "focus_range" custom control, instead modified the range of the standard "focus_absolute"
There were no changes in dt-bindings patch.
Link to v5: 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 | 47 +++ .../devicetree/bindings/vendor-prefixes.yaml | 2 + MAINTAINERS | 7 + drivers/media/i2c/Kconfig | 9 + drivers/media/i2c/Makefile | 1 + drivers/media/i2c/cef168.c | 317 ++++++++++++++++++ include/uapi/linux/v4l2-controls.h | 6 + 7 files changed, 389 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml create mode 100644 drivers/media/i2c/cef168.c
base-commit: 6093a688a07da07808f0122f9aa2a3eed250d853
Signature · 1 line
2.34.1
PATCH v6 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 | 47 +++++++++++++++++++ .../devicetree/bindings/vendor-prefixes.yaml | 2 + MAINTAINERS | 6 +++ 3 files changed, 55 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml
Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yaml new file +47 -0
@@ -0,0 +1,47 @@ +# 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. + +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
@@ -1241,6 +1241,8 @@ patternProperties: description: Picochip Ltd "^pine64,.*": description: Pine64 + "^pinefeat,.*": + description: Pinefeat LLP "^pineriver,.*": description: Shenzhen PineRiver Designs Co., Ltd. "^pixcir,.*":
MAINTAINERS +6 -0
@@ -20332,6 +20332,12 @@ 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
Re: [PATCH v6 1/2] dt-bindings: Pinefeat cef168 lens control board
Quoted 10 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.
Signed-off-by Aliaksandr Smirnou <[email protected]>
I acked v5, any reason for not applying it?
https://lore.kernel.org/all/20251007-geologic-epidemic-2f667f232cf6@spud/#t
Re: [PATCH v6 1/2] dt-bindings: Pinefeat cef168 lens control board
Quoted 2 lines · Conor Dooley
I acked v5, any reason for not applying it?
https://lore.kernel.org/all/20251007-geologic-epidemic-2f667f232cf6@spud/#t
Yes, the dt-bindings patch can be applied.
The v6 series does not change anything in the dt-bindings patch - only another patch in the series was updated, so v6 was sent.
Apologies - I wasn't aware that an acked patch could be omitted from the new revision.
Conor Dooley
Acked-byRe: [PATCH v6 1/2] dt-bindings: Pinefeat cef168 lens control board
Quoted 11 lines · Aliaksandr Smirnou
Quoted 2 lines · Conor Dooley
I acked v5, any reason for not applying it?
https://lore.kernel.org/all/20251007-geologic-epidemic-2f667f232cf6@spud/#tYes, the dt-bindings patch can be applied.
The v6 series does not change anything in the dt-bindings patch - only another patch in the series was updated, so v6 was sent.Apologies - I wasn't aware that an acked patch could be omitted from the new revision.
No no, it's not that it should have been omitted - it's that you should have applied my ack on v5 when submitting your v6. I guess timing worked against you because I sent it like 20 mins before you sent v6.
Re: [PATCH v6 1/2] dt-bindings: Pinefeat cef168 lens control board
Quoted 10 lines · Conor Dooley
Quoted 9 lines · Aliaksandr Smirnou
Quoted 2 lines · Conor Dooley
I acked v5, any reason for not applying it?
https://lore.kernel.org/all/20251007-geologic-epidemic-2f667f232cf6@spud/#tYes, the dt-bindings patch can be applied.
The v6 series does not change anything in the dt-bindings patch - only another patch in the series was updated, so v6 was sent.
Quoted 7 lines
Quoted 2 lines
Apologies - I wasn't aware that an acked patch could be omitted from the new revision.
No no, it's not that it should have been omitted - it's that you should have applied my ack on v5 when submitting your v6. I guess timing worked against you because I sent it like 20 mins before you sent v6.
Acked-by Conor Dooley <[email protected]>
Thanks for clarifying. I'll make sure to include your Acked-by tag in the next version.
Re: [PATCH v6 1/2] dt-bindings: Pinefeat cef168 lens control board
In reply to Aliaksandr SmirnouQuoted 10 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.
Signed-off-by Aliaksandr Smirnou <[email protected]>
You already got review at v4! Why are you forcing us to do the job multiple times?
<form letter>
This is a friendly reminder during the review process.
It looks like you received a tag and forgot to add it.
If you do not know the process, here is a short explanation:
Please add Acked-by/Reviewed-by/Tested-by tags when posting new versions of patchset, under or above your Signed-off-by tag, unless patch changed significantly (e.g. new properties added to the DT bindings). Tag is "received", when provided in a message replied to you on the mailing list. Tools like b4 can help here. However, there's no need to repost patches *only* to add the tags. The upstream maintainer will do that for tags received on the version they apply.
Please read:
https://elixir.bootlin.com/linux/v6.12-rc3/source/Documentation/process/submitting-patches.rst#L577
If a tag was not added on purpose, please state why and what changed. </form letter>
Best regards,
Krzysztof
Re: [PATCH v6 1/2] dt-bindings: Pinefeat cef168 lens control board
Quoted 2 lines · Krzysztof Kozlowski
You already got review at v4! Why are you forcing us to do the job multiple times?
Apologies for the oversight. There were no changes in the dt-bindings patch; I should have kept the Reviewed-by tag from the earlier one. I'll make sure to include it next time.
PATCH v6 2/2 media: i2c: Pinefeat cef168 lens control board driver
In reply to Aliaksandr SmirnouAdd 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 | 1 + drivers/media/i2c/Kconfig | 9 + drivers/media/i2c/Makefile | 1 + drivers/media/i2c/cef168.c | 317 +++++++++++++++++++++++++++++ include/uapi/linux/v4l2-controls.h | 6 + 5 files changed, 334 insertions(+) create mode 100644 drivers/media/i2c/cef168.c
MAINTAINERS +1 -0
@@ -20337,6 +20337,7 @@ 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 PLANTOWER PMS7003 AIR POLLUTION SENSOR DRIVER M: Tomasz Duszynski <[email protected]>
drivers/media/i2c/Kconfig +9 -0
@@ -823,6 +823,15 @@ 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" + select CRC8 + 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
@@ -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 +317 -0
317 lines not on this page · raw message
include/uapi/linux/v4l2-controls.h +6 -0
@@ -228,6 +228,12 @@ enum v4l2_colorfx {*/ #define V4L2_CID_USER_RKISP1_BASE (V4L2_CID_USER_BASE + 0x1220) +/* +* The base for Pinefeat CEF168 driver controls. +* We reserve 16 controls for this driver. +*/ +#define V4L2_CID_USER_CEF168_BASE (V4L2_CID_USER_BASE + 0x1230) +/* MPEG-class control IDs */ /* The MPEG controls are applicable to all codec controls * and the 'MPEG' part of the define is historical */
Signature · 1 line
2.34.1
Hi Aliaksandr,
I've been asked by Kieran Bingham to review your patches, I've reproduced your work on setup with a Pi 5 with a 6.18 kernel and I got a few questions
Quoted 7 lines · Aliaksandr Smirnou
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
I built the calibration application but it would not work out the box,
I can see that CEF168_V4L2_CID_CUSTOM is different between that is in the patch and in the application? Which is the correct one, does the patch need updating?
The tool to modify the device tree also did not work with the IMX477, I needed to manually modify my device tree to get the dtbo to build, there needs to be some further discussions on how we could improve the user experience when it comes to enabling the vcm with camera sensors.
Quoted 21 lines
Changes in v6:
- removed "focus_range" custom control, instead modified the range of the standard "focus_absolute"There were no changes in dt-bindings patch.
Link to v5: 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 | 47 +++ .../devicetree/bindings/vendor-prefixes.yaml | 2 + MAINTAINERS | 7 + drivers/media/i2c/Kconfig | 9 + drivers/media/i2c/Makefile | 1 + drivers/media/i2c/cef168.c | 317 ++++++++++++++++++ include/uapi/linux/v4l2-controls.h | 6 + 7 files changed, 389 insertions(+)
Quoted 5 lines
create mode 100644
Documentation/devicetree/bindings/media/i2c/pinefeat,cef168.yamlcreate mode 100644drivers/media/i2c/cef168.cbase-commit:
6093a688a07da07808f0122f9aa2a3eed250d853
Thank you,
Alen Karnil
Quoted 3 lines · Alen Karnil
I've been asked by Kieran Bingham to review your patches, I've reproduced your work on setup with a Pi 5 with a 6.18 kernel and I got a few questions
Hi Alen,
Thank you for resuming the work on the patch.
Quoted 4 lines
I built the calibration application but it would not work out the box, I can see that
CEF168_V4L2_CID_CUSTOMis different between that is in the patch and in the application? Which is the correct one, does the patch need updating?
The CEF168_V4L2_CID_CUSTOM value used in the patch is the correct one.
Although the repository currently defines it with a different value, this does not cause any issues for users because the driver is built and installed locally together with the calibration tool.
I can update the value in the cef168 repository at any time, so as long as users pull the latest version of the repository, everything will work correctly.
If the driver is eventually merged into the Linux kernel source tree, the repository will need to be updated accordingly anyway, because there will be no need to build the driver locally.
Quoted 3 lines
The tool to modify the device tree also did not work with the IMX477, I needed to manually modify my device tree to get the dtbo to build,
I just double-checked the setup on a Raspberry Pi 5 running the latest 6.18 kernel from the stock Raspberry Pi OS, and everything works correctly. Are you using a customized Linux distribution?
Could you provide the build error logs and the overlay files generated by the tool before you modified them?
The modifications applied by the tool are shown in this diff:
https://gist.github.com/pinefeat/1b3a258a193754f073e171b1ed33a0cc
Were the generated files different in your setup?
Quoted 3 lines
there needs to be some further discussions on how we could improve the user experience when it comes to enabling the vcm with camera sensors.
I agree. I started the discussion on the Raspberry Pi forum, but the driver needs to be merged first before moving forward with that.
https://forums.raspberrypi.com/viewtopic.php?p=2318070#p2318070
Quoted 29 lines · Aliaksandr Smirnou
Quoted 3 lines · Alen Karnil
I've been asked by Kieran Bingham to review your patches, I've reproduced your work on setup with a Pi 5 with a 6.18 kernel and I got a few questions
Hi Alen,
Thank you for resuming the work on the patch.
Quoted 4 lines
I built the calibration application but it would not work out the box, I can see that
CEF168_V4L2_CID_CUSTOMis different between that is in the patch and in the application? Which is the correct one, does the patch need updating?The
CEF168_V4L2_CID_CUSTOMvalue used in the patch is the correct one.Although the repository currently defines it with a different value, this does not cause any issues for users because the driver is built and installed locally together with the calibration tool.
I can update the value in the cef168 repository at any time, so as long as users pull the latest version of the repository, everything will work correctly.
If the driver is eventually merged into the Linux kernel source tree, the repository will need to be updated accordingly anyway, because there will be no need to build the driver locally.
Quoted 8 lines
Quoted 3 lines
The tool to modify the device tree also did not work with the IMX477, I needed to manually modify my device tree to get the dtbo to build,
I just double-checked the setup on a Raspberry Pi 5 running the latest 6.18 kernel from the stock Raspberry Pi OS, and everything works correctly. Are you using a customized Linux distribution?
Strange, I've taken the latest RPi lite trxie OS, but I replaced the kernel from the rpi-6.18.y branch.
Quoted 2 lines
Could you provide the build error logs and the overlay files generated by the tool before you modified them?
using the configure.sh tool with imx477imx477_378-overlay.dtsi
```
// SPDX-License-Identifier: GPL-2.0-only
// Definitions for IMX477 camera module on VC I2C bus
/{
compatible = "brcm,bcm2835";
fragment@0 {
target = <&i2c0if>;
__overlay__ {
status = "okay";
};
};
clk_frag: fragment@1 {
target = <&cam1_clk>;
cam_clk: __overlay__ {
clock-frequency = <24000000>;
status = "okay";
};
};
fragment@2 {
target = <&i2c0mux>;
__overlay__ {
status = "okay";
};
};
reg_frag: fragment@3 {
target = <&cam1_reg>;
cam_reg: __overlay__ {
startup-delay-us = <300000>;
};
};
fragment@4 {
target = <&cam_node>;
__overlay__ {
lens-focus = <&vcm_node>;
};
};
reg_alwayson_frag: fragment@99 {
target = <&cam1_reg>;
__dormant__ {
regulator-always-on;
};
};
i2c_frag: fragment@100 {
target = <&i2c_csi_dsi>;
__overlay__ {
#address-cells = <1>;
#size-cells = <0>;
status = "okay";
#include "imx477_378.dtsi"
};
};
csi_frag: fragment@101 {
target = <&csi1>;
csi: __overlay__ {
status = "okay";
port {
csi_ep: endpoint {
remote-endpoint = <&cam_endpoint>;
clock-lanes = <0>;
data-lanes = <1 2>;
clock-noncontinuous;
};
};
};
};
fragment@102 {
target = <&csi1>;
__dormant__ {
compatible = "brcm,bcm2835-unicam-legacy";
};
};
__overrides__ {
rotation = <&cam_node>,"rotation:0";
orientation = <&cam_node>,"orientation:0";
media-controller = <0>,"!102";
cam0 = <&i2c_frag>, "target:0=",<&i2c_csi_dsi0>,
<&csi_frag>, "target:0=",<&csi0>,
<&clk_frag>, "target:0=",<&cam0_clk>,
<®_frag>, "target:0=",<&cam0_reg>,
<®_alwayson_frag>, "target:0=",<&cam0_reg>,
<&cam_node>, "clocks:0=",<&cam0_clk>,
<&cam_node>, "VANA-supply:0=",<&cam0_reg>;
always-on = <0>, "+99";
link-frequency = <&cam_endpoint>,"link-frequencies#0";
vcm = <&vcm_node>, "status",
<0>, "=4";
};
};
&cam_node {
status = "okay";
};
&cam_endpoint {
remote-endpoint = <&csi_ep>;
};
&vcm_node {
status = "okay";
};
```And imx477_378.dtsi
```
cam_node: imx477@1a {
reg = <0x1a>;
status = "disabled";
clocks = <&cam1_clk>;
clock-names = "xclk";
VANA-supply = <&cam1_reg>; /* 2.8v */
VDIG-supply = <&cam_dummy_reg>; /* 1.05v */
VDDL-supply = <&cam_dummy_reg>; /* 1.8v */
rotation = <180>;
orientation = <2>;
port {
cam_endpoint: endpoint {
clock-lanes = <0>;
data-lanes = <1 2>;
clock-noncontinuous;
link-frequencies =
/bits/ 64 <450000000>;
};
};
};
vcm_node: cef168@d {
compatible = "pinefeat,cef168";
reg = <0x0d>;
status = "disabled";
vcc-supply = <&vdd_3v3_reg>;
};
```If put these into my kernel build I get:
$ make dtbs DTCO arch/arm64/boot/dts/overlays/camera-mux-2port.dtbo arch/arm64/boot/dts/overlays/imx477_378.dtsi:26.20-31.3: ERROR (duplicate_label): /fragment@200/__overlay__/pca@70/i2c@1/cef168@d: Duplicate label 'vcm_node' on /fragment@200/__overlay__/pca@70/i2c@1/cef168@d and /fragment@200/__overlay__/pca@70/i2c@0/cef168@d ERROR: Input tree has errors, aborting (use -f to force output) make[3]: *** [scripts/Makefile.dtbs:142: arch/arm64/boot/dts/overlays/camera-mux-2port.dtbo] Error 2 make[2]: *** [scripts/Makefile.build:544: arch/arm64/boot/dts/overlays] Error 2 make[1]: *** [/home/admin/linux/Makefile:1498: dtbs] Error 2 make: *** [Makefile:248: __sub-make] Error 2
Quoted 6 lines
The modifications applied by the tool are shown in this diff:
https://gist.github.com/pinefeat/1b3a258a193754f073e171b1ed33a0cc
Were the generated files different in your setup?
I will include my working dts* for the imx477
imx477_378-overlay.dtsi
```
// SPDX-License-Identifier: GPL-2.0-only
// Definitions for IMX477 camera module on VC I2C bus
/{
compatible = "brcm,bcm2835";
fragment@0 {
target = <&i2c0if>;
__overlay__ {
status = "okay";
};
};
clk_frag: fragment@1 {
target = <&cam1_clk>;
cam_clk: __overlay__ {
clock-frequency = <24000000>;
status = "okay";
};
};
fragment@2 {
target = <&i2c0mux>;
__overlay__ {
status = "okay";
};
};
reg_frag: fragment@3 {
target = <&cam1_reg>;
cam_reg: __overlay__ {
startup-delay-us = <300000>;
};
};
reg_alwayson_frag: fragment@99 {
target = <&cam1_reg>;
__dormant__ {
regulator-always-on;
};
};
i2c_frag: fragment@100 {
target = <&i2c_csi_dsi>;
__overlay__ {
#address-cells = <1>;
#size-cells = <0>;
status = "okay";
#include "imx477_378.dtsi"
vcm: cef168@d {
compatible = "pinefeat,cef168";
reg = <0x0d>;
status = "disabled";
vcc-supply = <&vdd_3v3_reg>;
};
};
};
csi_frag: fragment@101 {
target = <&csi1>;
csi: __overlay__ {
status = "okay";
port {
csi_ep: endpoint {
remote-endpoint = <&cam_endpoint>;
clock-lanes = <0>;
data-lanes = <1 2>;
clock-noncontinuous;
};
};
};
};
fragment@102 {
target = <&csi1>;
__dormant__ {
compatible = "brcm,bcm2835-unicam-legacy";
};
};
__overrides__ {
rotation = <&cam_node>,"rotation:0";
orientation = <&cam_node>,"orientation:0";
media-controller = <0>,"!102";
cam0 = <&i2c_frag>, "target:0=",<&i2c_csi_dsi0>,
<&csi_frag>, "target:0=",<&csi0>,
<&clk_frag>, "target:0=",<&cam0_clk>,
<®_frag>, "target:0=",<&cam0_reg>,
<®_alwayson_frag>, "target:0=",<&cam0_reg>,
<&cam_node>, "clocks:0=",<&cam0_clk>,
<&cam_node>, "VANA-supply:0=",<&cam0_reg>;
always-on = <0>, "+99";
vcm = <&vcm>, "status=okay",
<&cam_node>,"lens-focus:0=", <&vcm>;
link-frequency = <&cam_endpoint>,"link-frequencies#0";
};
};
&cam_node {
status = "okay";
};
&cam_endpoint {
remote-endpoint = <&csi_ep>;
};
```Also I needed to add vcm to my dtoverlay for the pi's config.txt
Quoted 7 lines
Quoted 4 lines
there needs to be some further discussions on how we could improve the user experience when it comes to enabling the vcm with camera sensors.
I agree. I started the discussion on the Raspberry Pi forum, but the driver needs to be merged first before moving forward with that.
Sure that makes sense, lets focus on the driver for now.
Also you will most likely need to update the patch since the line offsets have moved to work 6.18
Quoted 5 lines · Alen Karnil
Quoted 2 lines
Could you provide the build error logs and the overlay files generated by the tool before you modified them?
using the
configure.shtool with imx477imx477_378-overlay.dtsi
...
The files generated after running the configure.sh tool look correct and match mine exactly.
Quoted 16 lines
If put these into my kernel build I get:
$ make dtbs DTCO arch/arm64/boot/dts/overlays/camera-mux-2port.dtbo arch/arm64/boot/dts/overlays/imx477_378.dtsi:26.20-31.3: ERROR (duplicate_label): /fragment@200/__overlay__/pca@70/i2c@1/cef168@d: Duplicate label 'vcm_node' on /fragment@200/__overlay__/pca@70/i2c@1/cef168@d and /fragment@200/__overlay__/pca@70/i2c@0/cef168@d ERROR: Input tree has errors, aborting (use -f to force output) make[3]: *** [scripts/Makefile.dtbs:142: arch/arm64/boot/dts/overlays/camera-mux-2port.dtbo] Error 2 make[2]: *** [scripts/Makefile.build:544: arch/arm64/boot/dts/overlays] Error 2 make[1]: *** [/home/admin/linux/Makefile:1498: dtbs] Error 2 make: *** [Makefile:248: __sub-make] Error 2
It looks like a different Makefile is being used. In the cef168 repository's Makefile, there is no dtbs target. Also, our Makefile is only 27 lines long, while the log output indicates the error occurs at line 1498.
Could you please ensure you are using the Makefile provided in the cef168 repository?
Quoted 4 lines
Quoted 4 lines
Quoted 3 lines
there needs to be some further discussions on how we could improve the user experience when it comes to enabling the vcm with camera sensors.
Quoted 8 lines
Quoted 2 lines
I agree. I started the discussion on the Raspberry Pi forum, but the driver needs to be merged first before moving forward with that.
Sure that makes sense, lets focus on the driver for now.
Also you will most likely need to update the patch since the line offsets have moved to work 6.18
Will do. Should I update the patch before or after your review?
Quoted 28 lines · Aliaksandr Smirnou
Quoted 5 lines · Alen Karnil
Quoted 2 lines
Could you provide the build error logs and the overlay files generated by the tool before you modified them?
using the
configure.shtool with imx477imx477_378-overlay.dtsi
...The files generated after running the
configure.shtool look correct and match mine exactly.Quoted 16 lines
If put these into my kernel build I get:
$ make dtbs DTCO arch/arm64/boot/dts/overlays/camera-mux-2port.dtbo arch/arm64/boot/dts/overlays/imx477_378.dtsi:26.20-31.3: ERROR (duplicate_label): /fragment@200/__overlay__/pca@70/i2c@1/cef168@d: Duplicate label 'vcm_node' on /fragment@200/__overlay__/pca@70/i2c@1/cef168@d and /fragment@200/__overlay__/pca@70/i2c@0/cef168@d ERROR: Input tree has errors, aborting (use -f to force output) make[3]: *** [scripts/Makefile.dtbs:142: arch/arm64/boot/dts/overlays/camera-mux-2port.dtbo] Error 2 make[2]: *** [scripts/Makefile.build:544: arch/arm64/boot/dts/overlays] Error 2 make[1]: *** [/home/admin/linux/Makefile:1498: dtbs] Error 2 make: *** [Makefile:248: __sub-make] Error 2
Quoted 5 lines
It looks like a different
Makefileis being used. In the cef168 repository'sMakefile, there is no dtbs target. Also, ourMakefileis only 27 lines long, while the log output indicates the error occurs at line 1498.
I was using the kernel's makefile, basically following
https://www.raspberrypi.com/documentation/computers/linux_kernel.html
The kernel I am running has a LOCALVERSION set so it is 6.18.32-v8-afk+ (for reference)
Quoted 2 lines
Could you please ensure you are using the
Makefileprovided in the cef168 repository?
``` $ make make -C "/lib/modules/`uname -r`/build" M="/home/admin/work/cef168" DTC_INCLUDE="/usr/src/linux-headers-`uname -r | sed 's/rpt-.*/rpt-common-rpi/'`/include" make[1]: Entering directory '/home/admin/linux' make[2]: Entering directory '/home/admin/work/cef168' CC [M] cef168.o make[4]: *** No rule to make target 'imx477.dtbo', needed by 'modules.order'. Stop. make[3]: *** [/home/admin/linux/Makefile:2024: .] Error 2 make[2]: *** [/home/admin/linux/Makefile:248: __sub-make] Error 2 make[2]: Leaving directory '/home/admin/work/cef168' make[1]: *** [Makefile:248: __sub-make] Error 2 make[1]: Leaving directory '/home/admin/linux' make: *** [Makefile:16: default] Error 2 ```
This is what I got hung up on, which is why I moved onto copying the .dts* and building it from the kernel directory, which lead into the previous issue and me needing to roll my own .dts*
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 4 lines
there needs to be some further discussions on how we could improve the user experience when it comes to enabling the vcm with camera sensors.
I agree. I started the discussion on the Raspberry Pi forum, but the driver needs to be merged first before moving forward with that.
Quoted 8 lines
Quoted 6 lines
Sure that makes sense, lets focus on the driver for now.
Also you will most likely need to update the patch since the line offsets have moved to work 6.18
Will do. Should I update the patch before or after your review?
I would prefer if you were to update it before just so we are in sync, thank you.
Quoted 20 lines · Alen Karnil
Quoted 2 lines
Could you please ensure you are using the
Makefileprovided in the cef168 repository?``` $ make make -C "/lib/modules/`uname -r`/build" M="/home/admin/work/cef168" DTC_INCLUDE="/usr/src/linux-headers-`uname -r | sed 's/rpt-.*/rpt-common-rpi/'`/include" make[1]: Entering directory '/home/admin/linux' make[2]: Entering directory '/home/admin/work/cef168' CC [M] cef168.o make[4]: *** No rule to make target 'imx477.dtbo', needed by 'modules.order'. Stop. make[3]: *** [/home/admin/linux/Makefile:2024: .] Error 2 make[2]: *** [/home/admin/linux/Makefile:248: __sub-make] Error 2 make[2]: Leaving directory '/home/admin/work/cef168' make[1]: *** [Makefile:248: __sub-make] Error 2 make[1]: Leaving directory '/home/admin/linux' make: *** [Makefile:16: default] Error 2 ```
Quoted 3 lines
This is what I got hung up on, which is why I moved onto copying the .dts* and building it from the kernel directory, which lead into the previous issue and me needing to roll my own .dts*
Could you pull the latest version of the main branch from the cef168 repository? This issue was fixed recently, so it looks like the code you pulled was from before that fix was merged.
The fix involved using the dtb-y target instead of obj-m in the Makefile.
Quoted 4 lines
Quoted 1 line
Will do. Should I update the patch before or after your review?
I would prefer if you were to update it before just so we are in sync, thank you.
Will do.
Quoted 26 lines · Aliaksandr Smirnou
Quoted 20 lines · Alen Karnil
Quoted 2 lines
Could you please ensure you are using the
Makefileprovided in the cef168 repository?``` $ make make -C "/lib/modules/`uname -r`/build" M="/home/admin/work/cef168" DTC_INCLUDE="/usr/src/linux-headers-`uname -r | sed 's/rpt-.*/rpt-common-rpi/'`/include" make[1]: Entering directory '/home/admin/linux' make[2]: Entering directory '/home/admin/work/cef168' CC [M] cef168.o make[4]: *** No rule to make target 'imx477.dtbo', needed by 'modules.order'. Stop. make[3]: *** [/home/admin/linux/Makefile:2024: .] Error 2 make[2]: *** [/home/admin/linux/Makefile:248: __sub-make] Error 2 make[2]: Leaving directory '/home/admin/work/cef168' make[1]: *** [Makefile:248: __sub-make] Error 2 make[1]: Leaving directory '/home/admin/linux' make: *** [Makefile:16: default] Error 2 ```Quoted 3 lines
This is what I got hung up on, which is why I moved onto copying the .dts* and building it from the kernel directory, which lead into the previous issue and me needing to roll my own .dts*
Quoted 5 lines
Could you pull the latest version of the main branch from the cef168 repository? This issue was fixed recently, so it looks like the code you pulled was from before that fix was merged.
The fix involved using the dtb-y target instead of obj-m in the
Makefile.
Just tested it, now it works, thanks.
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
Will do. Should I update the patch before or after your review?
I would prefer if you were to update it before just so we are in sync, thank you.
Will do.
I will continue the review once it's received.