Add device tree support for Exynos4 SoC camera subsystem
Subsystems (5)
Sylwester Nawrocki
maintainerv4 cover Add device tree support for Exynos4 SoC camera subsystem
This series adds devicetree support for the front and rear camera of the Exynos4412 SoC Trats2 board. It converts related drivers to use the v4l2-async API. The SoC output clocks are provided to external image image sensors through the common clock API.
I'd appreciate a DT binding maintainer reviewed patches 2/10, 3/10. With an Ack I could finally push these things upstream.
Sylwester Nawrocki (10):
Documentation: dt: Add DT binding documentation for S5K6A3 image
sensor
Documentation: dt: Add DT binding documentation for S5C73M3 camera
Documentation: devicetree: Update Samsung FIMC DT binding
V4L: Add driver for s5k6a3 image sensor
V4L: s5c73m3: Add device tree support
exynos4-is: Use external s5k6a3 sensor driver
exynos4-is: Add clock provider for the SCLK_CAM clock outputs
exynos4-is: Add support for asynchronous subdevices registration
ARM: dts: Add rear camera nodes for Exynos4412 TRATS2 board
ARM: dts: exynos4: Update clk provider part of the camera subsystem
.../devicetree/bindings/media/samsung-fimc.txt | 36 +- .../devicetree/bindings/media/samsung-s5c73m3.txt | 97 +++++ .../devicetree/bindings/media/samsung-s5k6a3.txt | 33 ++ arch/arm/boot/dts/exynos4.dtsi | 6 +- arch/arm/boot/dts/exynos4412-trats2.dts | 86 ++++- drivers/media/i2c/Kconfig | 8 + drivers/media/i2c/Makefile | 1 + drivers/media/i2c/s5c73m3/s5c73m3-core.c | 207 ++++++++--- drivers/media/i2c/s5c73m3/s5c73m3-spi.c | 6 + drivers/media/i2c/s5c73m3/s5c73m3.h | 4 + drivers/media/i2c/s5k6a3.c | 388 ++++++++++++++++++++ drivers/media/platform/exynos4-is/fimc-is-regs.c | 2 +- drivers/media/platform/exynos4-is/fimc-is-sensor.c | 285 +------------- drivers/media/platform/exynos4-is/fimc-is-sensor.h | 49 +-- drivers/media/platform/exynos4-is/fimc-is.c | 97 ++--- drivers/media/platform/exynos4-is/fimc-is.h | 4 +- drivers/media/platform/exynos4-is/media-dev.c | 329 ++++++++++++----- drivers/media/platform/exynos4-is/media-dev.h | 32 +- 18 files changed, 1127 insertions(+), 543 deletions(-) create mode 100644 Documentation/devicetree/bindings/media/samsung-s5c73m3.txt create mode 100644 Documentation/devicetree/bindings/media/samsung-s5k6a3.txt create mode 100644 drivers/media/i2c/s5k6a3.c
--
1.7.9.5
Sylwester Nawrocki
maintainerv4 cover Add device tree support for Exynos4 SoC camera subsystem
This series adds devicetree support for the front and rear camera of the Exynos4412 SoC Trats2 board. It converts related drivers to use the v4l2-async API. The SoC output clocks are provided to external image image sensors through the common clock API.
I'd appreciate a DT binding maintainer reviewed patches 2/11, 3/11. With an Ack I could finally push these things upstream.
Sylwester Nawrocki (10):
Documentation: dt: Add DT binding documentation for S5K6A3 image
sensor
Documentation: dt: Add DT binding documentation for S5C73M3 camera
Documentation: devicetree: Update Samsung FIMC DT binding
V4L: Add driver for s5k6a3 image sensor
V4L: s5c73m3: Add device tree support
exynos4-is: Use external s5k6a3 sensor driver
exynos4-is: Add clock provider for the SCLK_CAM clock outputs
exynos4-is: Add support for asynchronous subdevices registration
ARM: dts: Add rear camera nodes for Exynos4412 TRATS2 board
ARM: dts: exynos4: Update clk provider part of the camera subsystem
.../devicetree/bindings/media/samsung-fimc.txt | 36 +- .../devicetree/bindings/media/samsung-s5c73m3.txt | 97 +++++ .../devicetree/bindings/media/samsung-s5k6a3.txt | 33 ++ arch/arm/boot/dts/exynos4.dtsi | 6 +- arch/arm/boot/dts/exynos4412-trats2.dts | 86 ++++- drivers/media/i2c/Kconfig | 8 + drivers/media/i2c/Makefile | 1 + drivers/media/i2c/s5c73m3/s5c73m3-core.c | 207 ++++++++--- drivers/media/i2c/s5c73m3/s5c73m3-spi.c | 6 + drivers/media/i2c/s5c73m3/s5c73m3.h | 4 + drivers/media/i2c/s5k6a3.c | 388 ++++++++++++++++++++ drivers/media/platform/exynos4-is/fimc-is-regs.c | 2 +- drivers/media/platform/exynos4-is/fimc-is-sensor.c | 285 +------------- drivers/media/platform/exynos4-is/fimc-is-sensor.h | 49 +-- drivers/media/platform/exynos4-is/fimc-is.c | 97 ++--- drivers/media/platform/exynos4-is/fimc-is.h | 4 +- drivers/media/platform/exynos4-is/media-dev.c | 329 ++++++++++++----- drivers/media/platform/exynos4-is/media-dev.h | 32 +- 18 files changed, 1127 insertions(+), 543 deletions(-) create mode 100644 Documentation/devicetree/bindings/media/samsung-s5c73m3.txt create mode 100644 Documentation/devicetree/bindings/media/samsung-s5k6a3.txt create mode 100644 drivers/media/i2c/s5k6a3.c
--
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 1/10 Documentation: dt: Add DT binding documentation for S5K6A3 image sensor
In reply to Sylwester NawrockiThis patch adds binding documentation for the Samsung S5K6A3(YX) raw image sensor.
---
Changes since v3:
-none.
Changes since v2:
- rephrased 'clocks' and 'clock-names' properties' description;
--- .../devicetree/bindings/media/samsung-s5k6a3.txt | 33 ++++++++++++++++++++ 1 file changed, 33 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/samsung-s5k6a3.txt
Documentation/devicetree/bindings/media/samsung-s5k6a3.txt new file +33 -0
@@ -0,0 +1,33 @@ +Samsung S5K6A3(YX) raw image sensor +--------------------------------- + +S5K6A3(YX) is a raw image sensor with MIPI CSI-2 and CCP2 image data interfaces +and CCI (I2C compatible) control bus. + +Required properties: + +- compatible : "samsung,s5k6a3"; +- reg : I2C slave address of the sensor; +- svdda-supply : core voltage supply; +- svddio-supply : I/O voltage supply; +- afvdd-supply : AF (actuator) voltage supply; +- gpios : specifier of a GPIO connected to the RESET pin; +- clocks : should contain list of phandle and clock specifier pairs + according to common clock bindings for the clocks described + in the clock-names property; +- clock-names : should contain "extclk" entry for the sensor's EXTCLK clock; + +Optional properties: + +- clock-frequency : the frequency at which the "extclk" clock should be + configured to operate, in Hz; if this property is not + specified default 24 MHz value will be used. + +The common video interfaces bindings (see video-interfaces.txt) should be +used to specify link to the image data receiver. The S5K6A3(YX) device +node should contain one 'port' child node with an 'endpoint' subnode. + +Following properties are valid for the endpoint node: + +- data-lanes : (optional) specifies MIPI CSI-2 data lanes as covered in + video-interfaces.txt. The sensor supports only one data lane.
Signature · 1 line
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 2/10 Documentation: dt: Add DT binding documentation for S5C73M3 camera
In reply to Sylwester NawrockiThis adds DT binding documentation for Samsung S5C73M3 camera sensor with an embedded ISP.
---
Changes since v3:
- DT binding documentation separated into this patch;
Changes since v2:
- rephrased 'clocks' and 'clock-names' properties' description;
--- .../devicetree/bindings/media/samsung-s5c73m3.txt | 97 ++++++++++++++++++++ 1 file changed, 97 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/samsung-s5c73m3.txt
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +97 -0
@@ -0,0 +1,97 @@ +Samsung S5C73M3 8Mp camera ISP +------------------------------ + +The S5C73M3 camera ISP supports MIPI CSI-2 and parallel (ITU-R BT.656) video +data busses. The I2C bus is the main control bus and additionally the SPI bus +is used, mostly for transferring the firmware to and from the device. Two +slave device nodes corresponding to these control bus interfaces are required +and should be placed under respective bus controller nodes. + +I2C slave device node +--------------------- + +Required properties: + +- compatible : "samsung,s5c73m3"; +- reg : I2C slave address of the sensor; +- vdd-int-supply : digital power supply (1.2V); +- vdda-supply : analog power supply (1.2V); +- vdd-reg-supply : regulator input power supply (2.8V); +- vddio-host-supply : host I/O power supply (1.8V to 2.8V); +- vddio-cis-supply : CIS I/O power supply (1.2V to 1.8V); +- vdd-af-supply : lens power supply (2.8V); +- xshutdown-gpios : specifier of GPIO connected to the XSHUTDOWN pin; +- standby-gpios : specifier of GPIO connected to the STANDBY pin; +- clocks : should contain list of phandle and clock specifier pairs + according to common clock bindings for the clocks described + in the clock-names property; +- clock-names : should contain "cis_extclk" entry for the CIS_EXTCLK clock; + +Optional properties: + +- clock-frequency : the frequency at which the "cis_extclk" clock should be + configured to operate, in Hz; if this property is not + specified default 24 MHz value will be used. + +The common video interfaces bindings (see video-interfaces.txt) should be used +to specify link from the S5C73M3 to an external image data receiver. The S5C73M3 +device node should contain one 'port' child node with an 'endpoint' subnode for +this purpose. The data link from a raw image sensor to the S5C73M3 can be +similarly specified, but it is optional since the S5C73M3 ISP and a raw image +sensor are usually inseparable and form a hybrid module. + +Following properties are valid for the endpoint node(s): + +endpoint subnode +---------------- + +- data-lanes : (optional) specifies MIPI CSI-2 data lanes as covered in + video-interfaces.txt. This sensor doesn't support data lane remapping + and physical lane indexes in subsequent elements of the array should + be only consecutive ascending values. + +SPI device node +--------------- + +Required properties: + +- compatible : "samsung,s5c73m3"; + +For more details see description of the SPI busses bindings +(../spi/spi-bus.txt) and bindings of a specific bus controller. + +Example: + +i2c at 138A000000 { + ... + s5c73m3 at 3c { + compatible = "samsung,s5c73m3"; + reg = <0x3c>; + vdd-int-supply = <&buck9_reg>; + vdda-supply = <&ldo17_reg>; + vdd-reg-supply = <&cam_io_reg>; + vddio-host-supply = <&ldo18_reg>; + vddio-cis-supply = <&ldo9_reg>; + vdd-af-supply = <&cam_af_reg>; + clock-frequency = <24000000>; + clocks = <&clk 0>; + clock-names = "cis_extclk"; + reset-gpios = <&gpf1 3 1>; + standby-gpios = <&gpm0 1 1>; + port { + s5c73m3_ep: endpoint { + remote-endpoint = <&csis0_ep>; + data-lanes = <1 2 3 4>; + }; + }; + }; +}; + +spi at 1392000 { + ... + s5c73m3_spi: s5c73m3 { + compatible = "samsung,s5c73m3"; + reg = <0>; + ... + }; +};
Signature · 1 line
1.7.9.5
PATCH v4 2/10 Documentation: dt: Add DT binding documentation for S5C73M3 camera
Quoted 22 lines · Sylwester Nawrocki
This adds DT binding documentation for Samsung S5C73M3 camera sensor with an embedded ISP.
Signed-off-by Sylwester Nawrocki <[email protected]>Acked-by Kyungmin Park <[email protected]>---
Changes since v3:
- DT binding documentation separated into this patch;Changes since v2:
- rephrased 'clocks' and 'clock-names' properties' description;--- .../devicetree/bindings/media/samsung-s5c73m3.txt | 97 ++++++++++++++++++++ 1 file changed, 97 insertions(+) create mode 100644 Documentation/devicetree/bindings/media/samsung-s5c73m3.txtDocumentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +0
-0
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 8
Quoted 8 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +8
-0@@ -0,0 +1,97 @@ +Samsung S5C73M3 8Mp camera ISP +------------------------------ + +The S5C73M3 camera ISP supports MIPI CSI-2 and parallel (ITU-R BT.656) video +data busses. The I2C bus is the main control bus and additionally the SPI bus +is used, mostly for transferring the firmware to and from the device. Two +slave device nodes corresponding to these control bus interfaces are required +and should be placed under respective bus controller nodes.
So this has both an I2C interface and an SPI interface that are used at the same time?
Quoted 40 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +40
-0+ +I2C slave device node +--------------------- + +Required properties: + +- compatible : "samsung,s5c73m3"; +- reg : I2C slave address of the sensor; +- vdd-int-supply : digital power supply (1.2V); +- vdda-supply : analog power supply (1.2V); +- vdd-reg-supply : regulator input power supply (2.8V); +- vddio-host-supply : host I/O power supply (1.8V to 2.8V); +- vddio-cis-supply : CIS I/O power supply (1.2V to 1.8V); +- vdd-af-supply : lens power supply (2.8V); +- xshutdown-gpios : specifier of GPIO connected to the XSHUTDOWN pin; +- standby-gpios : specifier of GPIO connected to the STANDBY pin; +- clocks : should contain list of phandle and clock specifier pairs + according to common clock bindings for the clocks described + in the clock-names property; +- clock-names : should contain "cis_extclk" entry for the CIS_EXTCLK clock; + +Optional properties: + +- clock-frequency : the frequency at which the "cis_extclk" clock should be + configured to operate, in Hz; if this property is not + specified default 24 MHz value will be used. + +The common video interfaces bindings (see video-interfaces.txt) should be used +to specify link from the S5C73M3 to an external image data receiver. The S5C73M3 +device node should contain one 'port' child node with an 'endpoint' subnode for +this purpose. The data link from a raw image sensor to the S5C73M3 can be +similarly specified, but it is optional since the S5C73M3 ISP and a raw image +sensor are usually inseparable and form a hybrid module. + +Following properties are valid for the endpoint node(s): + +endpoint subnode +---------------- + +- data-lanes : (optional) specifies MIPI CSI-2 data lanes as covered in
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 58
Quoted 10 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +10
-0+ video-interfaces.txt. This sensor doesn't support data lane remapping + and physical lane indexes in subsequent elements of the array should + be only consecutive ascending values. + +SPI device node +--------------- + +Required properties: + +- compatible : "samsung,s5c73m3";
It might make sense to explicitly link these two nodes somehow, in case multiple instances appear somewhere. However, that can come later in the case of a multi-instance device, and isn't necessary now.
Quoted 24 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +24
-0+ +For more details see description of the SPI busses bindings +(../spi/spi-bus.txt) and bindings of a specific bus controller. + +Example: + +i2c at 138A000000 { + ... + s5c73m3 at 3c { + compatible = "samsung,s5c73m3"; + reg = <0x3c>; + vdd-int-supply = <&buck9_reg>; + vdda-supply = <&ldo17_reg>; + vdd-reg-supply = <&cam_io_reg>; + vddio-host-supply = <&ldo18_reg>; + vddio-cis-supply = <&ldo9_reg>; + vdd-af-supply = <&cam_af_reg>; + clock-frequency = <24000000>; + clocks = <&clk 0>; + clock-names = "cis_extclk"; + reset-gpios = <&gpf1 3 1>; + standby-gpios = <&gpm0 1 1>; + port { + s5c73m3_ep: endpoint {
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 92
Quoted 10 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +10
-0+ remote-endpoint = <&csis0_ep>; + data-lanes = <1 2 3 4>; + }; + }; + }; +}; + +spi at 1392000 { + ... + s5c73m3_spi: s5c73m3 {
Nit: this should have a 0 unit-address to match the reg.
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 97
Quoted 5 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +5
-0+ compatible = "samsung,s5c73m3"; + reg = <0>; + ... + }; +};
Otherwise I don't see anything problematic about the binding.
Cheers,
Mark.
Sylwester Nawrocki
maintainerPATCH v4 2/10 Documentation: dt: Add DT binding documentation for S5C73M3 camera
On 21/02/14 16:42, Mark Rutland wrote:
[...]
Quoted 5 lines
Quoted 5 lines
+++
b/Documentation/devicetree/bindings/media/samsung-s5c73m3.txtDocumentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +3
-0@@ -0,0 +1,97 @@ +Samsung S5C73M3 8Mp camera ISP +------------------------------ +
Quoted 8 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 8
Quoted 5 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +5
-0+The S5C73M3 camera ISP supports MIPI CSI-2 and parallel (ITU-R BT.656) video +data busses. The I2C bus is the main control bus and additionally the SPI bus +is used, mostly for transferring the firmware to and from the device. Two +slave device nodes corresponding to these control bus interfaces are required +and should be placed under respective bus controller nodes.So this has both an I2C interface and an SPI interface that are used at the same time?
Yes, both are needed. AFAIU SPI is added so the firmware upload is faster.
Quoted 45 lines
Quoted 45 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +45
-0+I2C slave device node +--------------------- + +Required properties: + +- compatible : "samsung,s5c73m3"; +- reg : I2C slave address of the sensor; +- vdd-int-supply : digital power supply (1.2V); +- vdda-supply : analog power supply (1.2V); +- vdd-reg-supply : regulator input power supply (2.8V); +- vddio-host-supply : host I/O power supply (1.8V to 2.8V); +- vddio-cis-supply : CIS I/O power supply (1.2V to 1.8V); +- vdd-af-supply : lens power supply (2.8V); +- xshutdown-gpios : specifier of GPIO connected to the XSHUTDOWN pin; +- standby-gpios : specifier of GPIO connected to the STANDBY pin; +- clocks : should contain list of phandle and clock specifier pairs + according to common clock bindings for the clocks described + in the clock-names property; +- clock-names : should contain "cis_extclk" entry for the CIS_EXTCLK clock; + +Optional properties: + +- clock-frequency : the frequency at which the "cis_extclk" clock should be + configured to operate, in Hz; if this property is not + specified default 24 MHz value will be used. + +The common video interfaces bindings (see video-interfaces.txt) should be used +to specify link from the S5C73M3 to an external image data receiver. The S5C73M3 +device node should contain one 'port' child node with an 'endpoint' subnode for +this purpose. The data link from a raw image sensor to the S5C73M3 can be +similarly specified, but it is optional since the S5C73M3 ISP and a raw image +sensor are usually inseparable and form a hybrid module. + +Following properties are valid for the endpoint node(s): + +endpoint subnode +---------------- + +- data-lanes : (optional) specifies MIPI CSI-2 data lanes as covered in + video-interfaces.txt. This sensor doesn't support data lane remapping + and physical lane indexes in subsequent elements of the array should + be only consecutive ascending values. + +SPI device node +---------------
Quoted 8 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 15
Quoted 4 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +4
-0+ +Required properties: + +- compatible : "samsung,s5c73m3";It might make sense to explicitly link these two nodes somehow, in case multiple instances appear somewhere. However, that can come later in the case of a multi-instance device, and isn't necessary now.
I guess a phandle at the I2C slave device node, pointing to the SPI node and/or the other way around would do. I don't expect these devices to be used in multiple instances though and would prefer to address that when necessary.
We could try and create a root node for this device with an interesting structure, if we wanted to go much into details. But it could get a bit complicated given the scheme the I2C/SPI bus binding are structured now. Presumably that's something that could be handled later with a different compatible string if required.
Quoted 27 lines
Quoted 27 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +27
-0+For more details see description of the SPI busses bindings +(../spi/spi-bus.txt) and bindings of a specific bus controller. + +Example: + +i2c at 138A000000 { + ... + s5c73m3 at 3c { + compatible = "samsung,s5c73m3"; + reg = <0x3c>; + vdd-int-supply = <&buck9_reg>; + vdda-supply = <&ldo17_reg>; + vdd-reg-supply = <&cam_io_reg>; + vddio-host-supply = <&ldo18_reg>; + vddio-cis-supply = <&ldo9_reg>; + vdd-af-supply = <&cam_af_reg>; + clock-frequency = <24000000>; + clocks = <&clk 0>; + clock-names = "cis_extclk"; + reset-gpios = <&gpf1 3 1>; + standby-gpios = <&gpm0 1 1>; + port { + s5c73m3_ep: endpoint { + remote-endpoint = <&csis0_ep>; + data-lanes = <1 2 3 4>; + }; + };
Quoted 8 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 92
Quoted 6 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +6
-0+ }; +}; + +spi at 1392000 { + ... + s5c73m3_spi: s5c73m3 {Nit: this should have a 0 unit-address to match the reg.
OK, I'll correct that.
Quoted 9 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt line 97
Quoted 5 lines
Documentation/devicetree/bindings/media/samsung-s5c73m3.txt new file +5
-0+ compatible = "samsung,s5c73m3"; + reg = <0>; + ... + }; +};Otherwise I don't see anything problematic about the binding.
Acked-by Mark Rutland <[email protected]>
Thanks for the review.
--
Regards,
Sylwester
Sylwester Nawrocki
maintainerPATCH v4 3/10 Documentation: devicetree: Update Samsung FIMC DT binding
In reply to Sylwester NawrockiThis patch documents following updates of the Exynos4 SoC camera subsystem devicetree binding:
- addition of #clock-cells property to 'camera' node - the #clock-cells property is needed when the sensor sub-devices use clock provided by the camera host interface;
- addition of an optional clock-output-names property;
- change of the clock-frequency at image sensor node from mandatory to an optional property - there should be no need to require this property by the camera host device binding, a default frequency value can ofen be used;
- addition of a requirement of specific order of values in clocks/ clock-names properties, so the first two entry in the clock-names property can be used as parent clock names for the camera master clock provider. It happens all in-kernel dts files list the clock in such order, thus there should be no regression as far as in-kernel dts files are concerned.
--- .../devicetree/bindings/media/samsung-fimc.txt | 36 +++++++++++++++----- 1 file changed, 28 insertions(+), 8 deletions(-)
Documentation/devicetree/bindings/media/samsung-fimc.txt +28 -8
@@ -20,6 +20,7 @@ Required properties: the clock-names property; - clock-names : must contain "sclk_cam0", "sclk_cam1", "pxl_async0", "pxl_async1" entries, matching entries in the clocks property. + First two entries must be "sclk_cam0", "sclk_cam1". The pinctrl bindings defined in ../pinctrl/pinctrl-bindings.txt must be used to define a required pinctrl state named "default" and optional pinctrl states:
@@ -32,6 +33,22 @@ way around. The 'camera' node must include at least one 'fimc' child node. +Optional properties (*: + +- #clock-cells: from the common clock bindings (../clock/clock-bindings.txt), + must be 1. A clock provider is associated with the 'camera' node and it should + be referenced by external sensors that use clocks provided by the SoC on + CAM_*_CLKOUT pins. The clock specifier cell stores an index of a clock. + The indices are 0, 1 for CAM_A_CLKOUT, CAM_B_CLKOUT clocks respectively. + +- clock-output-names: from the common clock bindings, should contain names of + clocks registered by the camera subsystem corresponding to CAM_A_CLKOUT, + CAM_B_CLKOUT output clocks, in this order. Parent clock of these clocks are + specified be first two entries of the clock-names property. + +(* #clock-cells and clock-output-names are mandatory properties if external +image sensor devices reference 'camera' device node as a clock provider. + 'fimc' device nodes -------------------
@@ -97,8 +114,8 @@ Image sensor nodes The sensor device nodes should be added to their control bus controller (e.g. I2C0) nodes and linked to a port node in the csis or the parallel-ports node, using the common video interfaces bindings, defined in video-interfaces.txt. -The implementation of this bindings requires clock-frequency property to be -present in the sensor device nodes. +An optional clock-frequency property needs to be present in the sensor device +nodes. Default value when this property is not present is 24 MHz. Example:
@@ -114,7 +131,7 @@ Example: vddio-supply = <...>; clock-frequency = <24000000>; - clocks = <...>; + clocks = <&camera 1>; clock-names = "mclk"; port {
@@ -135,7 +152,7 @@ Example: vddio-supply = <...>; clock-frequency = <24000000>; - clocks = <...>; + clocks = <&camera 0>; clock-names = "mclk"; port {
@@ -149,12 +166,15 @@ Example: camera { compatible = "samsung,fimc", "simple-bus"; - #address-cells = <1>; - #size-cells = <1>; - status = "okay"; - + clocks = <&clock 132>, <&clock 133>; + clock-names = "sclk_cam0", "sclk_cam1"; + #clock-cells = <1>; + clock-output-names = "cam_mclk_a", "cam_mclk_b"; pinctrl-names = "default"; pinctrl-0 = <&cam_port_a_clk_active>; + status = "okay"; + #address-cells = <1>; + #size-cells = <1>; /* parallel camera ports */ parallel-ports {
Signature · 1 line
1.7.9.5
PATCH v4 3/10 Documentation: devicetree: Update Samsung FIMC DT binding
Quoted 8 lines · Sylwester Nawrocki
This patch documents following updates of the Exynos4 SoC camera subsystem devicetree binding:
- addition of #clock-cells property to 'camera' node - the #clock-cells property is needed when the sensor sub-devices use clock provided by the camera host interface;
- addition of an optional clock-output-names property;
- change of the clock-frequency at image sensor node from mandatory to an optional property - there should be no need to require this property
Quoted 8 lines
by the camera host device binding, a default frequency value can ofen be used;
- addition of a requirement of specific order of values in clocks/ clock-names properties, so the first two entry in the clock-names property can be used as parent clock names for the camera master clock provider. It happens all in-kernel dts files list the clock in such order, thus there should be no regression as far as in-kernel dts files are concerned.
I'm not sure I follow the reasoning here. Why does this matter? Why can child nodes not get these by name if they have to?
Quoted 12 lines
Signed-off-by Sylwester Nawrocki <[email protected]>Acked-by Kyungmin Park <[email protected]>--- .../devicetree/bindings/media/samsung-fimc.txt | 36 +++++++++++++++----- 1 file changed, 28 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/media/samsung-fimc.txt +0
-0
Documentation/devicetree/bindings/media/samsung-fimc.txt Required properties: · line 23
Quoted 4 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +1
-0@@ -20,6 +20,7 @@ Required properties: the clock-names property; - clock-names : must contain "sclk_cam0", "sclk_cam1", "pxl_async0", "pxl_async1" entries, matching entries in the clocks property. + First two entries must be "sclk_cam0", "sclk_cam1".
I don't think this is a good idea.
Documentation/devicetree/bindings/media/samsung-fimc.txt way around. · line 36
Quoted 8 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +1
-0The pinctrl bindings defined in ../pinctrl/pinctrl-bindings.txt must be used to define a required pinctrl state named "default" and optional pinctrl states:@@ -32,6 +33,22 @@ way around. The 'camera' node must include at least one 'fimc' child node. +Optional properties (*:
Is that a smiley face?
Documentation/devicetree/bindings/media/samsung-fimc.txt way around. · line 47
Quoted 11 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +11
-0+ +- #clock-cells: from the common clock bindings (../clock/clock-bindings.txt), + must be 1. A clock provider is associated with the 'camera' node and it should + be referenced by external sensors that use clocks provided by the SoC on + CAM_*_CLKOUT pins. The clock specifier cell stores an index of a clock. + The indices are 0, 1 for CAM_A_CLKOUT, CAM_B_CLKOUT clocks respectively. + +- clock-output-names: from the common clock bindings, should contain names of + clocks registered by the camera subsystem corresponding to CAM_A_CLKOUT, + CAM_B_CLKOUT output clocks, in this order. Parent clock of these clocks are + specified be first two entries of the clock-names property.
Do you need this?
That's not how clock-names is supposed to work. The clock-names property is for the names of the _input_ clock lines on the device, not the output names on whichever parent clock they came from.
Any clock-names property description should define absolutely the set of names. As this does not, NAK.
Documentation/devicetree/bindings/media/samsung-fimc.txt way around. · line 50
Quoted 3 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +3
-0+ +(* #clock-cells and clock-output-names are mandatory properties if external +image sensor devices reference 'camera' device node as a clock provider.
s/(*/Note:/
Documentation/devicetree/bindings/media/samsung-fimc.txt Image sensor nodes · line 118
Quoted 12 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +3
-2+ 'fimc' device nodes -------------------@@ -97,8 +114,8 @@ Image sensor nodes The sensor device nodes should be added to their control bus controller (e.g. I2C0) nodes and linked to a port node in the csis or the parallel-ports node, using the common video interfaces bindings, defined in video-interfaces.txt. -The implementation of this bindings requires clock-frequency property to be -present in the sensor device nodes. +An optional clock-frequency property needs to be present in the sensor device +nodes. Default value when this property is not present is 24 MHz.
s/needs to/should/ ?
What is this the frequency of?
Thanks,
Mark.
PATCH v4 3/10 Documentation: devicetree: Update Samsung FIMC DT binding
Quoted 18 lines · Mark Rutland
Quoted 8 lines · Sylwester Nawrocki
This patch documents following updates of the Exynos4 SoC camera subsystem devicetree binding:
- addition of #clock-cells property to 'camera' node - the #clock-cells property is needed when the sensor sub-devices use clock provided by the camera host interface;
- addition of an optional clock-output-names property;
- change of the clock-frequency at image sensor node from mandatory to an optional property - there should be no need to require this propertyQuoted 8 lines
by the camera host device binding, a default frequency value can ofen be used;
- addition of a requirement of specific order of values in clocks/ clock-names properties, so the first two entry in the clock-names property can be used as parent clock names for the camera master clock provider. It happens all in-kernel dts files list the clock in such order, thus there should be no regression as far as in-kernel dts files are concerned.
Quoted 2 lines
I'm not sure I follow the reasoning here. Why does this matter? Why can child nodes not get these by name if they have to?
These input clocks won't be used directly by child nodes. More details below...
Quoted 23 lines
Quoted 11 lines
Signed-off-by Sylwester Nawrocki<[email protected]>Acked-by Kyungmin Park<[email protected]>--- .../devicetree/bindings/media/samsung-fimc.txt | 36 +++++++++++++++----- 1 file changed, 28 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/media/samsung-fimc.txt +0
-0Documentation/devicetree/bindings/media/samsung-fimc.txt Required properties: · line 23
Quoted 4 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +1
-0@@ -20,6 +20,7 @@ Required properties: the clock-names property; - clock-names : must contain "sclk_cam0", "sclk_cam1", "pxl_async0", "pxl_async1" entries, matching entries in the clocks property. + First two entries must be "sclk_cam0", "sclk_cam1".I don't think this is a good idea.
Quoted 5 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +0
-0The pinctrl bindings defined in ../pinctrl/pinctrl-bindings.txt must be used to define a required pinctrl state named "default" and optional pinctrl states:@@ -32,6 +33,22 @@ way around.
Quoted 5 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt way around. · line 36
Quoted 3 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +1
-0The 'camera' node must include at least one 'fimc' child node. +Optional properties (*:Is that a smiley face?
Sorry, it was supposed to be a reference mark, will remove this.
Quoted 11 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt way around. · line 47
Quoted 10 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +10
-0+- #clock-cells: from the common clock bindings (../clock/clock-bindings.txt), + must be 1. A clock provider is associated with the 'camera' node and it should + be referenced by external sensors that use clocks provided by the SoC on + CAM_*_CLKOUT pins. The clock specifier cell stores an index of a clock. + The indices are 0, 1 for CAM_A_CLKOUT, CAM_B_CLKOUT clocks respectively. + +- clock-output-names: from the common clock bindings, should contain names of + clocks registered by the camera subsystem corresponding to CAM_A_CLKOUT, + CAM_B_CLKOUT output clocks, in this order. Parent clock of these clocks are + specified be first two entries of the clock-names property.
Quoted 1 line
Do you need this?
All right, that might have been a bad idea, it mixes names of clocks registered
by the main clock controller with names of clock input lines at the device. It's a mistake I have been usually sensitive to and now made it myself. :/
My intention was to maintain the clock tree, since the camera block doesn't generate the clock itself, it merely passes through the clocks from the SoC main
clock controller (CMU). So clk parents need to be properly set and since there
is no clock-output-names property at the CMU DT node, of_clk_get_parent_name()
cannot be used.
So presumably the DT binding would be only specifying that the sclk_cam0, sclk_cam1 clock input entries are associated with output clocks named as in clock-output-names property.
And the driver could either:
1) hard code those (well defined) CMU clock (clk parent) names,
2) clk_get() its input clock, retrieve name with __clk_get_name() and
pass
it as parent name to clk_register() - it sounds a bit hacky though.
The output clock names could be also well defined by the binding per the IP's
compatible. Nevertheless using clock-output-names seems cleaner to me than defining character strings in the driver.
What do you think ?
Quoted 3 lines
That's not how clock-names is supposed to work. The clock-names property is for the names of the _input_ clock lines on the device, not the output names on whichever parent clock they came from.
Yes, I agree, that was a pretty bad abuse of the bindings.
Quoted 8 lines
Any clock-names property description should define absolutely the set of names. As this does not, NAK.
Documentation/devicetree/bindings/media/samsung-fimc.txt way around. · line 50
Quoted 3 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +3
-0+ +(* #clock-cells and clock-output-names are mandatory properties if external +image sensor devices reference 'camera' device node as a clock provider.s/(*/Note:/
Thanks, will fix that.
Quoted 13 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt Image sensor nodes · line 118
Quoted 11 lines
Documentation/devicetree/bindings/media/samsung-fimc.txt +2
-2'fimc' device nodes -------------------@@ -97,8 +114,8 @@ Image sensor nodes The sensor device nodes should be added to their control bus controller (e.g. I2C0) nodes and linked to a port node in the csis or the parallel-ports node, using the common video interfaces bindings, defined in video-interfaces.txt. -The implementation of this bindings requires clock-frequency property to be -present in the sensor device nodes. +An optional clock-frequency property needs to be present in the sensor device +nodes. Default value when this property is not present is 24 MHz.s/needs to/should/ ?
Sounds better, I'll change it.
Quoted 1 line
What is this the frequency of?
It's frequency of an external sensor's master clock, which is provided by the SoC.
Since the sensors have now possibility to control the clock themselves the host
interface (binding) doesn't have to care about that frequency. This is supposed
to be an assigned frequency, it has been discussed already on devicetree mailing
list a few months ago.
--
Regards,
Sylwester
PATCH v4 3/10 Documentation: devicetree: Update Samsung FIMC DT binding
[Ccing Mike]
Quoted 37 lines · Sylwester Nawrocki
Quoted 20 lines · Mark Rutland
Quoted 10 lines · Sylwester Nawrocki
+- #clock-cells: from the common clock bindings (
../clock/clock-bindings.txt),
+ must be 1. A clock provider is associated with the 'camera' node and it should
+ be referenced by external sensors that use clocks provided by the SoC on
+ CAM_*_CLKOUT pins. The clock specifier cell stores an index of a clock.
+ The indices are 0, 1 forCAM_A_CLKOUT,CAM_B_CLKOUTclocks respectively.Quoted 8 lines
+
+- clock-output-names: from the common clock bindings, should contain names of
+ clocks registered by the camera subsystem corresponding toCAM_A_CLKOUT,
+CAM_B_CLKOUToutput clocks, in this order. Parent clock of these clocks are
+ specified be first two entries of the clock-names property.Quoted 1 line
Do you need this?
All right, that might have been a bad idea, it mixes names of clocks registered
by the main clock controller with names of clock input lines at the device. It's a mistake I have been usually sensitive to and now made it myself. :/My intention was to maintain the clock tree, since the camera block doesn't generate the clock itself, it merely passes through the clocks from the SoC main
clock controller (CMU). So clk parents need to be properly set and since there
is no clock-output-names property at the CMU DT node,of_clk_get_parent_name()
cannot be used.
Quoted 6 lines
So presumably the DT binding would be only specifying that the
sclk_cam0,sclk_cam1clock input entries are associated with output clocks named as in clock-output-names property.And the driver could either:
1) hard code those (well defined) CMU clock (clk parent) names,
I don't think this would be a good idea, as those CMU clock names may vary between SoCs.
Quoted 3 lines
2)
clk_get()its input clock, retrieve name with__clk_get_name()and
pass
it as parent name toclk_register()- it sounds a bit hacky though.
This looks fine, at least until proper interface is added to CCF. Exynos audio subsystem clock driver does exactly the same.
However, the right thing would be to make it possible to use pointers to struct clk instead of strings to list parent(s). This could be done by adding .parents field (struct clk **) to clk_init_data struct and make clk_register() use it if available.
Quoted 7 lines
The output clock names could be also well defined by the binding per the IP's
compatible. Nevertheless using clock-output-names seems cleaner to me than defining character strings in the driver.What do you think ?
<RANT>
Personally, I don't like clock-output-names at all. The idea of having a global namespace for all clock providers looks flawed to me. This property only tries to work around the not-quite-right design by giving device tree the right to force CCF to use particular internal clock identifiers and avoid collisions if two providers happen to have the same internal clock names.
I believe there should be completely no need for clock-output-names, (other than a textual label for debugging purposes, analogically to regulator-name). If one clock provider needs a clock from another, then you should list it using clock bindings in node of the former, not rely on some static name assignments.
Then, after eliminating the need to use static names anymore, the namespace could be split into per-provider namespaces and we would have no collision issue anymore.
Of course it's probably to late already for such changes, as there are already systems relying on clock-output-names (i.e. without inter-IP clock dependencies listed using clock bindings) and we would have to keep backwards compatibility anyway...
</RANT>
Best regards,
Tomasz
PATCH v4 3/10 Documentation: devicetree: Update Samsung FIMC DT binding
Quoted 45 lines · Tomasz Figa
[Ccing Mike]
Quoted 42 lines · Sylwester Nawrocki
Quoted 20 lines · Mark Rutland
Quoted 10 lines · Sylwester Nawrocki
+- #clock-cells: from the common clock bindings (
../clock/clock-bindings.txt),
+ must be 1. A clock provider is associated with the 'camera' node and it should
+ be referenced by external sensors that use clocks provided by the SoC on
+ CAM_*_CLKOUT pins. The clock specifier cell stores an index of a clock.
+ The indices are 0, 1 forCAM_A_CLKOUT,CAM_B_CLKOUTclocks respectively.Quoted 8 lines
+
+- clock-output-names: from the common clock bindings, should contain names of
+ clocks registered by the camera subsystem corresponding toCAM_A_CLKOUT,
+CAM_B_CLKOUToutput clocks, in this order. Parent clock of these clocks are
+ specified be first two entries of the clock-names property.Quoted 1 line
Do you need this?
All right, that might have been a bad idea, it mixes names of clocks registered
by the main clock controller with names of clock input lines at the device.
It's a mistake I have been usually sensitive to and now made it myself. :/My intention was to maintain the clock tree, since the camera block doesn't
generate the clock itself, it merely passes through the clocks from the SoC main
clock controller (CMU). So clk parents need to be properly set and since there is no clock-output-names property at the CMU DT node,of_clk_get_parent_name()cannot be used.So presumably the DT binding would be only specifying that the
sclk_cam0,sclk_cam1clock input entries are associated with output clocks named as in clock-output-names property.
Quoted 5 lines
Quoted 2 lines
And the driver could either:
1) hard code those (well defined) CMU clock (clk parent) names,I don't think this would be a good idea, as those CMU clock names may vary between SoCs.
For the record, I'm not in favour of this approach, even though these clock names happen to be same for all relevant SoCs.
Quoted 6 lines
Quoted 3 lines
2)
clk_get()its input clock, retrieve name with__clk_get_name()and pass
it as parent name toclk_register()- it sounds a bit hacky though.This looks fine, at least until proper interface is added to CCF. Exynos audio subsystem clock driver does exactly the same.
Hmm, I looked at this driver and somehow missed that.
Quoted 4 lines
However, the right thing would be to make it possible to use pointers to struct clk instead of strings to list parent(s). This could be done by adding .parents field (struct clk **) to
clk_init_datastruct and makeclk_register()use it if available.
We would just need to ensure the locking is done properly.
--
Regards,
Sylwester
Sylwester Nawrocki
maintainerPATCH v4 4/10 V4L: Add driver for s5k6a3 image sensor
In reply to Sylwester NawrockiThis patch adds subdev driver for Samsung S5K6A3 raw image sensor. As it is intended at the moment to be used only with the Exynos FIMC-IS (camera ISP) subsystem it is pretty minimal subdev driver. It doesn't do any I2C communication since the sensor is controlled by the ISP and its own firmware.
This driver, if needed, can be updated in future into a regular subdev driver where the main CPU communicates with the sensor directly.
---
Changes since v3:
- clk_get() called moved from s_power() to driver probe() callback.
Changes since v1:
- added missing pm_runtime_disable(),
- removed subdev name overriding,
- s/S5K6A3_DEF_PIX/S5K6A3_DEFAULT_,
- merged patch adding v4l2-async API support,
- clock-frequency property is now optional and a default frequency value will be used when it is missing, rather than bailing out,
- reset GPIO made mandatory as it is required for proper power on/off sequences and all known boards use it,
- maximum image size is now 1412x1412 pixels, as specified in in the S5K6A3YX datasheet.
- regulator_enable()/disable() used instead of the regulator bulk API to ensure proper regulators enable/disable sequences.
--- drivers/media/i2c/Kconfig | 8 + drivers/media/i2c/Makefile | 1 + drivers/media/i2c/s5k6a3.c | 388 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 397 insertions(+) create mode 100644 drivers/media/i2c/s5k6a3.c
drivers/media/i2c/Kconfig +8 -0
@@ -579,6 +579,14 @@ config VIDEO_S5K6AA This is a V4L2 sensor-level driver for Samsung S5K6AA(FX) 1.3M camera sensor with an embedded SoC image signal processor. +config VIDEO_S5K6A3 + tristate "Samsung S5K6A3 sensor support" + depends on MEDIA_CAMERA_SUPPORT + depends on I2C && VIDEO_V4L2 && VIDEO_V4L2_SUBDEV_API && OF + ---help--- + This is a V4L2 sensor-level driver for Samsung S5K6A3 raw + camera sensor. + config VIDEO_S5K4ECGX tristate "Samsung S5K4ECGX sensor support" depends on I2C && VIDEO_V4L2 && VIDEO_V4L2_SUBDEV_API
drivers/media/i2c/Makefile +1 -0
@@ -66,6 +66,7 @@ obj-$(CONFIG_VIDEO_MT9V032) += mt9v032.o obj-$(CONFIG_VIDEO_SR030PC30) += sr030pc30.o obj-$(CONFIG_VIDEO_NOON010PC30) += noon010pc30.o obj-$(CONFIG_VIDEO_S5K6AA) += s5k6aa.o +obj-$(CONFIG_VIDEO_S5K6A3) += s5k6a3.o obj-$(CONFIG_VIDEO_S5K4ECGX) += s5k4ecgx.o obj-$(CONFIG_VIDEO_S5K5BAF) += s5k5baf.o obj-$(CONFIG_VIDEO_S5C73M3) += s5c73m3/
drivers/media/i2c/s5k6a3.c new file +388 -0
388 lines not on this page · raw message
Signature · 1 line
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 5/10 V4L: s5c73m3: Add device tree support
In reply to Sylwester NawrockiThis patch adds the V4L2 asynchronous subdev registration and device tree support. Common clock API is used to control the sensor master clock from within the subdev.
---
Changes since v3:
- move clk_get from subdev s_power() to driver probe() callback and make it devm_clk_get(),
- pass any errors from dev_clk_get() through, rather than overwriting with EPROBE_DEFER. Patch [1] is required for things to work correctly after this change.
[1] http://www.spinics.net/lists/arm-kernel/msg306072.html
--- drivers/media/i2c/s5c73m3/s5c73m3-core.c | 207 ++++++++++++++++++++++-------- drivers/media/i2c/s5c73m3/s5c73m3-spi.c | 6 + drivers/media/i2c/s5c73m3/s5c73m3.h | 4 + 3 files changed, 167 insertions(+), 50 deletions(-)
drivers/media/i2c/s5c73m3/s5c73m3-core.c +157 -50
@@ -15,7 +15,7 @@* GNU General Public License for more details. */ -#include <linux/sizes.h> +#include <linux/clk.h> #include <linux/delay.h> #include <linux/firmware.h> #include <linux/gpio.h>
@@ -23,7 +23,9 @@#include <linux/init.h> #include <linux/media.h> #include <linux/module.h> +#include <linux/of_gpio.h> #include <linux/regulator/consumer.h> +#include <linux/sizes.h> #include <linux/slab.h> #include <linux/spi/spi.h> #include <linux/videodev2.h>
@@ -33,6 +35,7 @@#include <media/v4l2-subdev.h> #include <media/v4l2-mediabus.h> #include <media/s5c73m3.h> +#include <media/v4l2-of.h> #include "s5c73m3.h"
@@ -46,6 +49,8 @@ static int update_fw;module_param (update_fw, int,0644 );#define S5C73M3_EMBEDDED_DATA_MAXLEN SZ_4K +#define S5C73M3_MIPI_DATA_LANES 4 +#define S5C73M3_CLK_NAME "cis_extclk"static const char *const s5c73m3_supply_names[S5C73M3_MAX_SUPPLIES ] = {"vdd-int" ,/* Digital Core supply (1.2V), CAM_ISP_CORE_1.2V */
@@ -1355,9 +1360,20 @@ static int __s5c73m3_power_on(struct s5c73m3 *state)for (i =0 ; i <S5C73M3_MAX_SUPPLIES ; i++) { ret =regulator_enable (state->supplies[i].consumer);if (ret) - goto err; + goto err_reg_dis; } + ret =clk_set_rate (state->clock, state->mclk_frequency); +if (ret <0 ) + goto err_reg_dis; + + ret =clk_prepare_enable (state->clock); +if (ret <0 ) + goto err_reg_dis; + +v4l2_dbg (1 , s5c73m3_dbg, &state->oif_sd,"clock frequency: %ld\n" , +clk_get_rate (state->clock)); +s5c73m3_gpio_deassert (state,STBY );usleep_range (100 ,200 );
@@ -1365,7 +1381,8 @@ static int __s5c73m3_power_on(struct s5c73m3 *state)usleep_range (50 ,100 );return 0 ; -err: + +err_reg_dis:for (--i; i >=0 ; i--)regulator_disable (state->supplies[i].consumer);return ret;
@@ -1380,6 +1397,9 @@ static int __s5c73m3_power_off(struct s5c73m3 *state)if (s5c73m3_gpio_assert (state,STBY ))usleep_range (100 ,200 ); + +clk_disable_unprepare (state->clock); + state->streaming =0 ; state->isp_ready =0 ;
@@ -1388,6 +1408,7 @@ static int __s5c73m3_power_off(struct s5c73m3 *state)if (ret) goto err; } +return 0 ; err:for (++i; i <S5C73M3_MAX_SUPPLIES ; i++) {
@@ -1396,6 +1417,8 @@ err:v4l2_err (&state->oif_sd,"Failed to reenable %s: %d\n" , state->supplies[i].supply, r); } + +clk_prepare_enable (state->clock);return ret; }
@@ -1451,17 +1474,6 @@ static int s5c73m3_oif_registered(struct v4l2_subdev *sd)S5C73M3_JPEG_PAD , &state->oif_sd.entity,OIF_JPEG_PAD , MEDIA_LNK_FL_IMMUTABLE | MEDIA_LNK_FL_ENABLED); -mutex_lock (&state->lock); - ret =__s5c73m3_power_on (state); -if (ret ==0 ) -s5c73m3_get_fw_version (state); - -__s5c73m3_power_off (state); -mutex_unlock (&state->lock); - -v4l2_dbg (1 , s5c73m3_dbg, sd,"%s: Booting %s (%d)\n" , - __func__, ret ?"failed" :"succeeded" , ret); -return ret; }
@@ -1519,41 +1531,112 @@ static const struct v4l2_subdev_ops oif_subdev_ops = { .video = &s5c73m3_oif_video_ops, }; -static int s5c73m3_configure_gpios (struct s5c73m3 *state, -const struct s5c73m3_platform_data *pdata) +static int s5c73m3_configure_gpios (struct s5c73m3 *state) +{ +static const char *const gpio_names[] = { +"S5C73M3_STBY" ,"S5C73M3_RST" + }; +struct i2c_client *c = state->i2c_client; +struct s5c73m3_gpio *g = state->gpio; +int ret, i; + +for (i =0 ; i <GPIO_NUM ; ++i) { +unsigned int flags =GPIOF_DIR_OUT ; +if (g[i].level) + flags |=GPIOF_INIT_HIGH ; + ret =devm_gpio_request_one (&c->dev, g[i].gpio, flags, + gpio_names[i]); +if (ret) { +v4l2_err (c,"failed to request gpio %s\n" , + gpio_names[i]); +return ret; + } + } +return 0 ; +} + +static int s5c73m3_parse_gpios (struct s5c73m3 *state) +{ +static const char *const prop_names[] = { +"standby-gpios" ,"xshutdown-gpios" , + }; +struct device *dev = &state->i2c_client->dev; +struct device_node *node = dev->of_node; +int ret, i; + +for (i =0 ; i <GPIO_NUM ; ++i) { +enum of_gpio_flags of_flags; + + ret =of_get_named_gpio_flags (node, prop_names[i], +0 , &of_flags); +if (ret <0 ) { +dev_err (dev,"failed to parse %s DT property\n" , + prop_names[i]); +return -EINVAL ; + } + state->gpio[i].gpio = ret; + state->gpio[i].level = !(of_flags &OF_GPIO_ACTIVE_LOW ); + } +return 0 ; +} + +static int s5c73m3_get_platform_data (struct s5c73m3 *state) {struct device *dev = &state->i2c_client->dev; -const struct s5c73m3_gpio *gpio; -unsigned long flags; +const struct s5c73m3_platform_data *pdata = dev->platform_data; +struct device_node *node = dev->of_node; +struct device_node *node_ep; +struct v4l2_of_endpoint ep;int ret; - state->gpio[STBY ].gpio = -EINVAL ; - state->gpio[RST ].gpio = -EINVAL ; +if (!node) { +if (!pdata) { +dev_err (dev,"Platform data not specified\n" ); +return -EINVAL ; + } - gpio = &pdata->gpio_stby; -if (gpio_is_valid (gpio->gpio)) { - flags = (gpio->level ?GPIOF_OUT_INIT_HIGH :GPIOF_OUT_INIT_LOW ) - |GPIOF_EXPORT ; - ret =devm_gpio_request_one (dev, gpio->gpio, flags, -"S5C73M3_STBY" ); -if (ret <0 ) -return ret; + state->mclk_frequency = pdata->mclk_frequency; + state->gpio[STBY ] = pdata->gpio_stby; + state->gpio[RST ] = pdata->gpio_reset; +return 0 ; + } + + state->clock =devm_clk_get (dev,S5C73M3_CLK_NAME ); +if (IS_ERR (state->clock)) +return PTR_ERR (state->clock); - state->gpio[STBY ] = *gpio; +if (of_property_read_u32 (node,"clock-frequency" , + &state->mclk_frequency)) { + state->mclk_frequency =S5C73M3_DEFAULT_MCLK_FREQ ; +dev_info (dev,"using default %u Hz clock frequency\n" , + state->mclk_frequency); } - gpio = &pdata->gpio_reset; -if (gpio_is_valid (gpio->gpio)) { - flags = (gpio->level ?GPIOF_OUT_INIT_HIGH :GPIOF_OUT_INIT_LOW ) - |GPIOF_EXPORT ; - ret =devm_gpio_request_one (dev, gpio->gpio, flags, -"S5C73M3_RST" ); -if (ret <0 ) -return ret; + ret =s5c73m3_parse_gpios (state); +if (ret <0 ) +return -EINVAL ; - state->gpio[RST ] = *gpio; + node_ep =v4l2_of_get_next_endpoint (node,NULL ); +if (!node_ep) { +dev_warn (dev,"no endpoint defined for node: %s\n" , + node->full_name); +return 0 ; } +v4l2_of_parse_endpoint (node_ep, &ep); +of_node_put (node_ep); + +if (ep.bus_type !=V4L2_MBUS_CSI2 ) { +dev_err (dev,"unsupported bus type\n" ); +return -EINVAL ; + } +/* +* Number of MIPI CSI-2 data lanes is currently not configurable, +* always a default value of 4 lanes is used. +*/ +if (ep.bus.mipi_csi2.num_data_lanes !=S5C73M3_MIPI_DATA_LANES ) +dev_info (dev,"falling back to 4 MIPI CSI-2 data lanes\n" ); +return 0 ; }
@@ -1561,21 +1644,20 @@ static int s5c73m3_probe(struct i2c_client *client, const struct i2c_device_id *id) {struct device *dev = &client->dev; -const struct s5c73m3_platform_data *pdata = client->dev.platform_data;struct v4l2_subdev *sd;struct v4l2_subdev *oif_sd;struct s5c73m3 *state;int ret, i; -if (pdata ==NULL ) { -dev_err (&client->dev,"Platform data not specified\n" ); -return -EINVAL ; - } - state =devm_kzalloc (dev,sizeof (*state),GFP_KERNEL );if (!state)return -ENOMEM ; + state->i2c_client = client; + ret =s5c73m3_get_platform_data (state); +if (ret <0 ) +return ret; +mutex_init (&state->lock); sd = &state->sensor_sd; oif_sd = &state->oif_sd;
@@ -1613,11 +1695,7 @@ static int s5c73m3_probe(struct i2c_client *client,if (ret <0 )return ret; - state->mclk_frequency = pdata->mclk_frequency; - state->bus_type = pdata->bus_type; - state->i2c_client = client; - - ret =s5c73m3_configure_gpios (state, pdata); + ret =s5c73m3_configure_gpios (state);if (ret) goto out_err;
@@ -1651,9 +1729,29 @@ static int s5c73m3_probe(struct i2c_client *client,if (ret <0 ) goto out_err; + oif_sd->dev = dev; + + ret =__s5c73m3_power_on (state); +if (ret <0 ) + goto out_err1; + + ret =s5c73m3_get_fw_version (state); +__s5c73m3_power_off (state); + +if (ret <0 ) { +dev_err (dev,"Device detection failed: %d\n" , ret); + goto out_err1; + } + + ret =v4l2_async_register_subdev (oif_sd); +if (ret <0 ) + goto out_err1; +v4l2_info (sd,"%s: completed successfully\n" , __func__);return 0 ; +out_err1: +s5c73m3_unregister_spi_driver (state); out_err:media_entity_cleanup (&sd->entity);return ret;
@@ -1665,7 +1763,7 @@ static int s5c73m3_remove(struct i2c_client *client)struct s5c73m3 *state =oif_sd_to_s5c73m3 (oif_sd);struct v4l2_subdev *sensor_sd = &state->sensor_sd; -v4l2_device_unregister_subdev (oif_sd); +v4l2_async_unregister_subdev (oif_sd);v4l2_ctrl_handler_free (oif_sd->ctrl_handler);media_entity_cleanup (&oif_sd->entity);
@@ -1684,8 +1782,17 @@ static const struct i2c_device_id s5c73m3_id[] = { };MODULE_DEVICE_TABLE (i2c, s5c73m3_id); +#ifdef CONFIG_OF +static const struct of_device_id s5c73m3_of_match[] = { + { .compatible ="samsung,s5c73m3" }, + { } +}; +MODULE_DEVICE_TABLE (of, s5c73m3_of_match); +#endif +static struct i2c_driver s5c73m3_i2c_driver = { .driver = { + .of_match_table =of_match_ptr (s5c73m3_of_match), .name =DRIVER_NAME , }, .probe = s5c73m3_probe,
drivers/media/i2c/s5c73m3/s5c73m3-spi.c +6 -0
@@ -27,6 +27,11 @@#define S5C73M3_SPI_DRV_NAME "S5C73M3-SPI" +static const struct of_device_id s5c73m3_spi_ids[] = { + { .compatible ="samsung,s5c73m3" }, + { } +}; +enum spi_direction {SPI_DIR_RX ,SPI_DIR_TX
@@ -146,6 +151,7 @@ int s5c73m3_register_spi_driver(struct s5c73m3 *state) spidrv->driver.name =S5C73M3_SPI_DRV_NAME ; spidrv->driver.bus = &spi_bus_type; spidrv->driver.owner =THIS_MODULE ; + spidrv->driver.of_match_table = s5c73m3_spi_ids;return spi_register_driver (spidrv); }
drivers/media/i2c/s5c73m3/s5c73m3.h +4 -0
@@ -17,6 +17,7 @@#ifndef S5C73M3_H_ #define S5C73M3_H_ +#include <linux/clk.h> #include <linux/kernel.h> #include <linux/regulator/consumer.h> #include <media/v4l2-common.h>
@@ -321,6 +322,7 @@ enum s5c73m3_oif_pads {#define S5C73M3_MAX_SUPPLIES 6 +#define S5C73M3_DEFAULT_MCLK_FREQ 24000000Ustruct s5c73m3_ctrls {struct v4l2_ctrl_handler handler;
@@ -391,6 +393,8 @@ struct s5c73m3 {struct regulator_bulk_data supplies[S5C73M3_MAX_SUPPLIES ];struct s5c73m3_gpio gpio[GPIO_NUM ]; +struct clk *clock; +/* External master clock frequency */ u32 mclk_frequency;/* Video bus type - MIPI-CSI2/parallel */
Signature · 1 line
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 6/10 exynos4-is: Use external s5k6a3 sensor driver
In reply to Sylwester NawrockiThis patch removes the common fimc-is-sensor driver for image sensors that are normally controlled by the FIMC-IS firmware. The FIMC-IS driver now contains only a table of properties specific to each sensor. The sensor properties required for the ISP's firmware are parsed from device tree and retrieved from the internal table, which is selected based on the compatible property of an image sensor.
To use the Exynos4x12 internal ISP the S5K6A3 sensor driver (drivers/ media/i2c/s5k6a3.c) is now required.
--- drivers/media/platform/exynos4-is/fimc-is-regs.c | 2 +- drivers/media/platform/exynos4-is/fimc-is-sensor.c | 285 +------------------- drivers/media/platform/exynos4-is/fimc-is-sensor.h | 49 +--- drivers/media/platform/exynos4-is/fimc-is.c | 97 +++---- drivers/media/platform/exynos4-is/fimc-is.h | 4 +- 5 files changed, 57 insertions(+), 380 deletions(-)
drivers/media/platform/exynos4-is/fimc-is-regs.c +1 -1
@@ -112,7 +112,7 @@ void fimc_is_hw_set_sensor_num(struct fimc_is *is)mcuctl_write (IH_REPLY_DONE , is,MCUCTL_REG_ISSR (0 ));mcuctl_write (is->sensor_index, is,MCUCTL_REG_ISSR (1 ));mcuctl_write (IHC_GET_SENSOR_NUM , is,MCUCTL_REG_ISSR (2 )); -mcuctl_write (FIMC_IS_SENSOR_NUM , is,MCUCTL_REG_ISSR (3 )); +mcuctl_write (FIMC_IS_SENSORS_NUM , is,MCUCTL_REG_ISSR (3 )); }void fimc_is_hw_close_sensor (struct fimc_is *is,unsigned int index)
drivers/media/platform/exynos4-is/fimc-is-sensor.c +7 -278
310 lines not on this page · raw message
drivers/media/platform/exynos4-is/fimc-is-sensor.h +8 -41
77 lines not on this page · raw message
drivers/media/platform/exynos4-is/fimc-is.c +39 -58
159 lines not on this page · raw message
drivers/media/platform/exynos4-is/fimc-is.h +2 -2
@@ -39,7 +39,7 @@#define FIMC_IS_FW_LOAD_TIMEOUT 1000/* ms */ #define FIMC_IS_POWER_ON_TIMEOUT 1000/* us */ -#define FIMC_IS_SENSOR_NUM 2 +#define FIMC_IS_SENSORS_NUM 2/* Memory definitions */ #define FIMC_IS_CPU_MEM_SIZE (0xa00000)
@@ -253,7 +253,7 @@ struct fimc_is {struct firmware *f_w;struct fimc_isp isp; -struct fimc_is_sensor *sensor; +struct fimc_is_sensor sensor[FIMC_IS_SENSORS_NUM ];struct fimc_is_setfile setfile;struct vb2_alloc_ctx *alloc_ctx;
Signature · 1 line
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 7/10 exynos4-is: Add clock provider for the SCLK_CAM clock outputs
In reply to Sylwester NawrockiThis patch adds clock provider so the the SCLK_CAM0/1 output clocks can be accessed by image sensor devices through the clk API.
---
Changes since v3:
- use clock-output-names DT property instead of hard coding names of registered clocks in the driver; first two entries of the clock-names property value are used to specify parent clocks of cam_{a,b}_clkout clocks, rather than hard coding it to sclk_cam{0,1} in the driver.
- addressed issues pointed out in review by Mauro.
Changes since v2:
- use 'camera' DT node drirectly as clock provider node, rather than creating additional clock-controller child node.
- DT binding documentation moved to a separate patch.
--- drivers/media/platform/exynos4-is/media-dev.c | 110 +++++++++++++++++++++++++ drivers/media/platform/exynos4-is/media-dev.h | 19 ++++- 2 files changed, 128 insertions(+), 1 deletion(-)
drivers/media/platform/exynos4-is/media-dev.c +110 -0
@@ -11,6 +11,8 @@*/ #include <linux/bug.h> +#include <linux/clk.h> +#include <linux/clk-provider.h> #include <linux/device.h> #include <linux/errno.h> #include <linux/i2c.h>
@@ -1437,6 +1439,103 @@ static int fimc_md_get_pinctrl(struct fimc_md *fmd)return 0 ; } +#ifdef CONFIG_OF +static int cam_clk_prepare (struct clk_hw *hw) +{ +struct cam_clk *camclk =to_cam_clk (hw); +int ret; + +if (camclk->fmd->pmf ==NULL ) +return -ENODEV ; + + ret =pm_runtime_get_sync (camclk->fmd->pmf); +return ret <0 ? ret :0 ; +} + +static void cam_clk_unprepare (struct clk_hw *hw) +{ +struct cam_clk *camclk =to_cam_clk (hw); + +if (camclk->fmd->pmf ==NULL ) +return ; + +pm_runtime_put_sync (camclk->fmd->pmf); +} + +static const struct clk_ops cam_clk_ops = { + .prepare = cam_clk_prepare, + .unprepare = cam_clk_unprepare, +}; + +static void fimc_md_unregister_clk_provider (struct fimc_md *fmd) +{ +struct cam_clk_provider *cp = &fmd->clk_provider; +unsigned int i; + +if (cp->of_node) +of_clk_del_provider (cp->of_node); + +for (i =0 ; i < cp->num_clocks; i++) +clk_unregister (cp->clks[i]); +} + +static int fimc_md_register_clk_provider (struct fimc_md *fmd) +{ +struct cam_clk_provider *cp = &fmd->clk_provider; +struct device *dev = &fmd->pdev->dev; +int i, ret; + +for (i =0 ; i <ARRAY_SIZE (cp->clks); i++) { +struct cam_clk *camclk = &cp->camclk[i]; +struct clk_init_data init; + + ret =of_property_read_string_index (dev->of_node, +"clock-output-names" , i, &init.name); +if (ret <0 ) +break ; + + ret =of_property_read_string_index (dev->of_node, +"clock-names" , i, init.parent_names); +if (ret <0 ) +break ; + + init.num_parents =1 ; + init.ops = &cam_clk_ops; + init.flags =CLK_SET_RATE_PARENT ; + camclk->hw.init = &init; + camclk->fmd = fmd; + + cp->clks[i] =clk_register (NULL , &camclk->hw); +if (IS_ERR (cp->clks[i])) { +dev_err (dev,"failed to register clock: %s (%ld)\n" , + init.name,PTR_ERR (cp->clks[i])); + ret =PTR_ERR (cp->clks[i]); + goto err; + } + cp->num_clocks++; + } + +if (cp->num_clocks ==0 ) { +dev_warn (dev,"clk provider not registered\n" ); +return 0 ; + } + + cp->clk_data.clks = cp->clks; + cp->clk_data.clk_num = cp->num_clocks; + cp->of_node = dev->of_node; + ret =of_clk_add_provider (dev->of_node, of_clk_src_onecell_get, + &cp->clk_data); +if (ret ==0 ) +return 0 ; +err: +fimc_md_unregister_clk_provider (fmd); +return ret; +} +#else +#define fimc_md_register_clk_provider (fmd) (0) +#define fimc_md_unregister_clk_provider (fmd) (0) +#endif +static int fimc_md_probe (struct platform_device *pdev) {struct device *dev = &pdev->dev;
@@ -1464,16 +1563,24 @@ static int fimc_md_probe(struct platform_device *pdev) fmd->use_isp =fimc_md_is_isp_available (dev->of_node); + ret =fimc_md_register_clk_provider (fmd); +if (ret <0 ) { +v4l2_err (v4l2_dev,"clock provider registration failed\n" ); +return ret; + } + ret =v4l2_device_register (dev, &fmd->v4l2_dev);if (ret <0 ) {v4l2_err (v4l2_dev,"Failed to register v4l2_device: %d\n" , ret);return ret; } + ret =media_device_register (&fmd->media_dev);if (ret <0 ) {v4l2_err (v4l2_dev,"Failed to register media device: %d\n" , ret); goto err_md; } + ret =fimc_md_get_clocks (fmd);if (ret) goto err_clk;
@@ -1507,6 +1614,7 @@ static int fimc_md_probe(struct platform_device *pdev) ret =fimc_md_create_links (fmd);if (ret) goto err_unlock; + ret =v4l2_device_register_subdev_nodes (&fmd->v4l2_dev);if (ret) goto err_unlock;
@@ -1527,6 +1635,7 @@ err_clk:media_device_unregister (&fmd->media_dev); err_md:v4l2_device_unregister (&fmd->v4l2_dev); +fimc_md_unregister_clk_provider (fmd);return ret; }
@@ -1537,6 +1646,7 @@ static int fimc_md_remove(struct platform_device *pdev)if (!fmd)return 0 ; +fimc_md_unregister_clk_provider (fmd);v4l2_device_unregister (&fmd->v4l2_dev);device_remove_file (&pdev->dev, &dev_attr_subdev_conf_mode);fimc_md_unregister_entities (fmd);
drivers/media/platform/exynos4-is/media-dev.h +18 -1
@@ -10,6 +10,7 @@#define FIMC_MDEVICE_H_ #include <linux/clk.h> +#include <linux/clk-provider.h> #include <linux/platform_device.h> #include <linux/mutex.h> #include <linux/of.h>
@@ -89,6 +90,12 @@ struct fimc_sensor_info {struct fimc_dev *host; }; +struct cam_clk { +struct clk_hw hw; +struct fimc_md *fmd; +}; +#define to_cam_clk (_hw) container_of(_hw, struct cam_clk, hw) +/** * struct fimc_md - fimc media device information * @csis: MIPI CSIS subdevs data
@@ -105,6 +112,7 @@ struct fimc_sensor_info {* @pinctrl: camera port pinctrl handle * @state_default: pinctrl default state handle * @state_idle: pinctrl idle state handle +* @cam_clk_provider: CAMCLK clock provider structure * @user_subdev_api: true if subdevs are not configured by the host driver * @slock: spinlock protecting @sensor array */
@@ -122,13 +130,22 @@ struct fimc_md {struct media_device media_dev;struct v4l2_device v4l2_dev;struct platform_device *pdev; +struct fimc_pinctrl {struct pinctrl *pinctrl;struct pinctrl_state *state_default;struct pinctrl_state *state_idle; } pinctl; -bool user_subdev_api; +struct cam_clk_provider { +struct clk *clks[FIMC_MAX_CAMCLKS ]; +struct clk_onecell_data clk_data; +struct device_node *of_node; +struct cam_clk camclk[FIMC_MAX_CAMCLKS ]; +int num_clocks; + } clk_provider; + +bool user_subdev_api;spinlock_t slock;struct list_head pipelines; };
Signature · 1 line
1.7.9.5
PATCH v4 7/10 exynos4-is: Add clock provider for the SCLK_CAM clock outputs
Quoted 22 lines · Sylwester Nawrocki
This patch adds clock provider so the the SCLK_CAM0/1 output clocks can be accessed by image sensor devices through the clk API.
Signed-off-by Sylwester Nawrocki <[email protected]>Acked-by Kyungmin Park <[email protected]>---
Changes since v3:
- use clock-output-names DT property instead of hard coding names of registered clocks in the driver; first two entries of the clock-names property value are used to specify parent clocks of cam_{a,b}_clkout clocks, rather than hard coding it to sclk_cam{0,1} in the driver.
- addressed issues pointed out in review by Mauro.Changes since v2:
- use 'camera' DT node drirectly as clock provider node, rather than creating additional clock-controller child node.
- DT binding documentation moved to a separate patch.--- drivers/media/platform/exynos4-is/media-dev.c | 110 +++++++++++++++++++++++++ drivers/media/platform/exynos4-is/media-dev.h | 19 ++++- 2 files changed, 128 insertions(+), 1 deletion(-)
[...]
drivers/media/platform/exynos4-is/media-dev.c static int fimc_md_get_pinctrl(struct fimc_md *fmd) · line 1493
Quoted 12 lines
drivers/media/platform/exynos4-is/media-dev.c +12
-0+static int fimc_md_register_clk_provider (struct fimc_md *fmd) +{ +struct cam_clk_provider *cp = &fmd->clk_provider; +struct device *dev = &fmd->pdev->dev; +int i, ret; + +for (i =0 ; i <ARRAY_SIZE (cp->clks); i++) { +struct cam_clk *camclk = &cp->camclk[i]; +struct clk_init_data init; + + ret =of_property_read_string_index (dev->of_node, +"clock-output-names" , i, &init.name);
Are there not well-defined names for the clock outputs of the block?
drivers/media/platform/exynos4-is/media-dev.c static int fimc_md_get_pinctrl(struct fimc_md *fmd) · line 1498
Quoted 5 lines
drivers/media/platform/exynos4-is/media-dev.c +5
-0+if (ret <0 ) +break ; + + ret =of_property_read_string_index (dev->of_node, +"clock-names" , i, init.parent_names);
This shouldn't be a parent name. It should be the input line name.
I don't think this makes sense.
Why do you need the name of the parent clock?
Quoted 19 lines
drivers/media/platform/exynos4-is/media-dev.c +19
-0+if (ret <0 ) +break ; + + init.num_parents =1 ; + init.ops = &cam_clk_ops; + init.flags =CLK_SET_RATE_PARENT ; + camclk->hw.init = &init; + camclk->fmd = fmd; + + cp->clks[i] =clk_register (NULL , &camclk->hw); +if (IS_ERR (cp->clks[i])) { +dev_err (dev,"failed to register clock: %s (%ld)\n" , + init.name,PTR_ERR (cp->clks[i])); + ret =PTR_ERR (cp->clks[i]); + goto err; + } + cp->num_clocks++; + } +
drivers/media/platform/exynos4-is/media-dev.c static int fimc_md_get_pinctrl(struct fimc_md *fmd) · line 1527
Quoted 10 lines
drivers/media/platform/exynos4-is/media-dev.c +10
-0+if (cp->num_clocks ==0 ) { +dev_warn (dev,"clk provider not registered\n" ); +return 0 ; + } + + cp->clk_data.clks = cp->clks; + cp->clk_data.clk_num = cp->num_clocks; + cp->of_node = dev->of_node; + ret =of_clk_add_provider (dev->of_node, of_clk_src_onecell_get, + &cp->clk_data);
Are _all_ of the input clock lines available to children in hardware?
The binding and commit message(s) implied only two clocks were, so what's the point in exporting clocks which aren't available?
Thanks,
Mark.
PATCH v4 7/10 exynos4-is: Add clock provider for the SCLK_CAM clock outputs
Quoted 4 lines · Mark Rutland
Quoted 4 lines
--- drivers/media/platform/exynos4-is/media-dev.c | 110 +++++++++++++++++++++++++ drivers/media/platform/exynos4-is/media-dev.h | 19 ++++- 2 files changed, 128 insertions(+), 1 deletion(-)
...
Quoted 15 lines
drivers/media/platform/exynos4-is/media-dev.c static int fimc_md_get_pinctrl(struct fimc_md *fmd)
Quoted 13 lines
drivers/media/platform/exynos4-is/media-dev.c +12
-0+static int fimc_md_register_clk_provider (struct fimc_md *fmd) +{ +struct cam_clk_provider *cp =&fmd->clk_provider; +struct device *dev =&fmd->pdev->dev; +int i, ret; + +for (i =0 ; i<ARRAY_SIZE (cp->clks); i++) { +struct cam_clk *camclk =&cp->camclk[i]; +struct clk_init_data init; + + ret =of_property_read_string_index (dev->of_node, +"clock-output-names" , i,&init.name);Are there not well-defined names for the clock outputs of the block?
There are, but this driver handles multiple SoCs so I thought it's better to define these names in devicetree. They are supposed to be unique. Isn't that what the clock-output-names property is for ?
Quoted 11 lines
drivers/media/platform/exynos4-is/media-dev.c static int fimc_md_get_pinctrl(struct fimc_md *fmd) · line 1498
Quoted 5 lines
drivers/media/platform/exynos4-is/media-dev.c +5
-0+if (ret<0 ) +break ; + + ret =of_property_read_string_index (dev->of_node, +"clock-names" , i, init.parent_names);This shouldn't be a parent name. It should be the input line name.
I don't think this makes sense.
Why do you need the name of the parent clock?
As explained in previous e-mail, it is needed to maintain the clock tree, the parent clock names must be set properly. As the above is not correct I could hard code these names or retrieve with __clk_get_name() from proper input clocks. of_clk_get_parent_name() cannot be used.
Quoted 23 lines
Quoted 23 lines
drivers/media/platform/exynos4-is/media-dev.c +23
-0+if (ret<0 ) +break ; + + init.num_parents =1 ; + init.ops =&cam_clk_ops; + init.flags =CLK_SET_RATE_PARENT ; + camclk->hw.init =&init; + camclk->fmd = fmd; + + cp->clks[i] =clk_register (NULL ,&camclk->hw); +if (IS_ERR (cp->clks[i])) { +dev_err (dev,"failed to register clock: %s (%ld)\n" , + init.name,PTR_ERR (cp->clks[i])); + ret =PTR_ERR (cp->clks[i]); + goto err; + } + cp->num_clocks++; + } + +if (cp->num_clocks ==0 ) { +dev_warn (dev,"clk provider not registered\n" ); +return 0 ; + }
Quoted 8 lines
drivers/media/platform/exynos4-is/media-dev.c static int fimc_md_get_pinctrl(struct fimc_md *fmd) · line 1527
Quoted 6 lines
drivers/media/platform/exynos4-is/media-dev.c +6
-0+ + cp->clk_data.clks = cp->clks; + cp->clk_data.clk_num = cp->num_clocks; + cp->of_node = dev->of_node; + ret =of_clk_add_provider (dev->of_node, of_clk_src_onecell_get, + &cp->clk_data);Are _all_ of the input clock lines available to children in hardware?
This code is only for two clocks, ARRAY_SIZE(cp->clks) == 2 and cp->num_clocks can't be greater than 2. Nevertheless, for some SoC variants this driver handles (e.g. S5PV210) there can be only up to two clocks listed in the camera node.
The camera interface IP block doesn't generate the clock itself, it just passes it through when is active, i.e. its other clocks are enabled and the power domain activated.
Quoted 2 lines
The binding and commit message(s) implied only two clocks were, so what's the point in exporting clocks which aren't available?
I'm not sure what makes you think there is more than two ?
--
Regards,
Sylwester
Sylwester Nawrocki
maintainerPATCH v4 8/10 exynos4-is: Add support for asynchronous subdevices registration
In reply to Sylwester NawrockiAdd support for registering external sensor subdevs using v4l2-async API. The async API is used only for sensor subdevs and only for DT platforms.
--- drivers/media/platform/exynos4-is/media-dev.c | 233 ++++++++++++++----------- drivers/media/platform/exynos4-is/media-dev.h | 13 +- 2 files changed, 144 insertions(+), 102 deletions(-)
drivers/media/platform/exynos4-is/media-dev.c +132 -101
@@ -26,6 +26,7 @@#include <linux/pm_runtime.h> #include <linux/types.h> #include <linux/slab.h> +#include <media/v4l2-async.h> #include <media/v4l2-ctrls.h> #include <media/v4l2-of.h> #include <media/media-device.h>
@@ -220,6 +221,7 @@ static int __fimc_pipeline_open(struct exynos_media_pipeline *ep,if (ret <0 )return ret; } + ret =fimc_md_set_camclk (sd, true);if (ret <0 ) goto err_wbclk;
@@ -380,77 +382,18 @@ static void fimc_md_unregister_sensor(struct v4l2_subdev *sd)struct i2c_client *client =v4l2_get_subdevdata (sd);struct i2c_adapter *adapter; -if (!client) +if (!client || client->dev.of_node)return ;v4l2_device_unregister_subdev (sd); -if (!client->dev.of_node) { - adapter = client->adapter; -i2c_unregister_device (client); -if (adapter) -i2c_put_adapter (adapter); - } + adapter = client->adapter; +i2c_unregister_device (client); +if (adapter) +i2c_put_adapter (adapter); }#ifdef CONFIG_OF -/* Register I2C client subdev associated with @node. */ -static int fimc_md_of_add_sensor (struct fimc_md *fmd, -struct device_node *node,int index) -{ -struct fimc_sensor_info *si; -struct i2c_client *client; -struct v4l2_subdev *sd; -int ret; - -if (WARN_ON (index >=ARRAY_SIZE (fmd->sensor))) -return -EINVAL ; - si = &fmd->sensor[index]; - - client =of_find_i2c_device_by_node (node); -if (!client) -return -EPROBE_DEFER ; - -device_lock (&client->dev); - -if (!client->dev.driver || - !try_module_get (client->dev.driver->owner)) { - ret = -EPROBE_DEFER ; -v4l2_info (&fmd->v4l2_dev,"No driver found for %s\n" , - node->full_name); - goto dev_put; - } - -/* Enable sensor's master clock */ - ret =__fimc_md_set_camclk (fmd, &si->pdata, true); -if (ret <0 ) - goto mod_put; - sd =i2c_get_clientdata (client); - - ret =v4l2_device_register_subdev (&fmd->v4l2_dev, sd); -__fimc_md_set_camclk (fmd, &si->pdata, false); -if (ret <0 ) - goto mod_put; - -v4l2_set_subdev_hostdata (sd, &si->pdata); -if (si->pdata.fimc_bus_type ==FIMC_BUS_TYPE_ISP_WRITEBACK ) - sd->grp_id =GRP_ID_FIMC_IS_SENSOR ; -else - sd->grp_id =GRP_ID_SENSOR ; - - si->subdev = sd; -v4l2_info (&fmd->v4l2_dev,"Registered sensor subdevice: %s (%d)\n" , - sd->name, fmd->num_sensors); - fmd->num_sensors++; - -mod_put: -module_put (client->dev.driver->owner); -dev_put: -device_unlock (&client->dev); -put_device (&client->dev); -return ret; -} -/* Parse port node and register as a sub-device any sensor specified there. */ static int fimc_md_parse_port_node(struct fimc_md *fmd, struct device_node *port,
@@ -459,7 +402,6 @@ static int fimc_md_parse_port_node(struct fimc_md *fmd,struct device_node *rem, *ep, *np;struct fimc_source_info *pd;struct v4l2_of_endpoint endpoint; -int ret;u32 val; pd = &fmd->sensor[index].pdata;
@@ -487,6 +429,8 @@ static int fimc_md_parse_port_node(struct fimc_md *fmd,if (!of_property_read_u32 (rem,"clock-frequency" , &val)) pd->clk_frequency = val; +else + pd->clk_frequency =DEFAULT_SENSOR_CLK_FREQ ;if (pd->clk_frequency ==0 ) { v4l2_err(&fmd->v4l2_dev, "Wrong clock frequency at node %s\n",
@@ -526,10 +470,17 @@ static int fimc_md_parse_port_node(struct fimc_md *fmd,else pd->fimc_bus_type = pd->sensor_bus_type; - ret =fimc_md_of_add_sensor (fmd, rem, index); -of_node_put (rem); +if (WARN_ON (index >=ARRAY_SIZE (fmd->sensor))) +return -EINVAL ; -return ret; + fmd->sensor[index].asd.match_type =V4L2_ASYNC_MATCH_OF ; + fmd->sensor[index].asd.match.of.node = rem; + fmd->async_subdevs[index] = &fmd->sensor[index].asd; + + fmd->num_sensors++; + +of_node_put (rem); +return 0 ; }/* Register all SoC external sub-devices */
@@ -885,11 +836,13 @@ static void fimc_md_unregister_entities(struct fimc_md *fmd)v4l2_device_unregister_subdev (fmd->csis[i].sd); fmd->csis[i].sd =NULL ; } -for (i =0 ; i < fmd->num_sensors; i++) { -if (fmd->sensor[i].subdev ==NULL ) -continue ; -fimc_md_unregister_sensor (fmd->sensor[i].subdev); - fmd->sensor[i].subdev =NULL ; +if (fmd->pdev->dev.of_node ==NULL ) { +for (i =0 ; i < fmd->num_sensors; i++) { +if (fmd->sensor[i].subdev ==NULL ) +continue ; +fimc_md_unregister_sensor (fmd->sensor[i].subdev); + fmd->sensor[i].subdev =NULL ; + } }if (fmd->fimc_is)
@@ -1224,6 +1177,14 @@ static int __fimc_md_set_camclk(struct fimc_md *fmd,struct fimc_camclk_info *camclk;int ret =0 ; +/* +* When device tree is used the sensor drivers are supposed to +* control the clock themselves. This whole function will be +* removed once S5PV210 platform is converted to the device tree. +*/ +if (fmd->pdev->dev.of_node) +return 0 ; +if (WARN_ON (si->clk_id >=FIMC_MAX_CAMCLKS ) || !fmd || !fmd->pmf)return -EINVAL ;
@@ -1536,6 +1497,56 @@ err:#define fimc_md_unregister_clk_provider (fmd) (0)#endif +static int subdev_notifier_bound (struct v4l2_async_notifier *notifier, +struct v4l2_subdev *subdev, +struct v4l2_async_subdev *asd) +{ +struct fimc_md *fmd =notifier_to_fimc_md (notifier); +struct fimc_sensor_info *si =NULL ; +int i; + +/* Find platform data for this sensor subdev */ +for (i =0 ; i <ARRAY_SIZE (fmd->sensor); i++) +if (fmd->sensor[i].asd.match.of.node == subdev->dev->of_node) + si = &fmd->sensor[i]; + +if (si ==NULL ) +return -EINVAL ; + +v4l2_set_subdev_hostdata (subdev, &si->pdata); + +if (si->pdata.fimc_bus_type ==FIMC_BUS_TYPE_ISP_WRITEBACK ) + subdev->grp_id =GRP_ID_FIMC_IS_SENSOR ; +else + subdev->grp_id =GRP_ID_SENSOR ; + + si->subdev = subdev; + +v4l2_info (&fmd->v4l2_dev,"Registered sensor subdevice: %s (%d)\n" , + subdev->name, fmd->num_sensors); + + fmd->num_sensors++; + +return 0 ; +} + +static int subdev_notifier_complete (struct v4l2_async_notifier *notifier) +{ +struct fimc_md *fmd =notifier_to_fimc_md (notifier); +int ret; + +mutex_lock (&fmd->media_dev.graph_mutex); + + ret =fimc_md_create_links (fmd); +if (ret <0 ) + goto unlock; + + ret =v4l2_device_register_subdev_nodes (&fmd->v4l2_dev); +unlock: +mutex_unlock (&fmd->media_dev.graph_mutex); +return ret; +} +static int fimc_md_probe (struct platform_device *pdev) {struct device *dev = &pdev->dev;
@@ -1563,12 +1574,6 @@ static int fimc_md_probe(struct platform_device *pdev) fmd->use_isp =fimc_md_is_isp_available (dev->of_node); - ret =fimc_md_register_clk_provider (fmd); -if (ret <0 ) { -v4l2_err (v4l2_dev,"clock provider registration failed\n" ); -return ret; - } - ret =v4l2_device_register (dev, &fmd->v4l2_dev);if (ret <0 ) {v4l2_err (v4l2_dev,"Failed to register v4l2_device: %d\n" , ret);
@@ -1578,64 +1583,87 @@ static int fimc_md_probe(struct platform_device *pdev) ret =media_device_register (&fmd->media_dev);if (ret <0 ) {v4l2_err (v4l2_dev,"Failed to register media device: %d\n" , ret); - goto err_md; + goto err_v4l2_dev; } ret =fimc_md_get_clocks (fmd);if (ret) - goto err_clk; + goto err_md; fmd->user_subdev_api = (dev->of_node !=NULL ); -/* Protect the media graph while we're registering entities */ -mutex_lock (&fmd->media_dev.graph_mutex); - ret =fimc_md_get_pinctrl (fmd);if (ret <0 ) {if (ret !=EPROBE_DEFER )dev_err (dev,"Failed to get pinctrl: %d\n" , ret); - goto err_unlock; + goto err_clk; } +platform_set_drvdata (pdev, fmd); + +/* Protect the media graph while we're registering entities */ +mutex_lock (&fmd->media_dev.graph_mutex); +if (dev->of_node) ret =fimc_md_register_of_platform_entities (fmd, dev->of_node);else ret =bus_for_each_dev (&platform_bus_type,NULL , fmd, fimc_md_pdev_match); -if (ret) - goto err_unlock; +if (ret) { +mutex_unlock (&fmd->media_dev.graph_mutex); + goto err_clk; + }if (dev->platform_data || dev->of_node) { ret =fimc_md_register_sensor_entities (fmd); -if (ret) - goto err_unlock; +if (ret) { +mutex_unlock (&fmd->media_dev.graph_mutex); + goto err_m_ent; + } } - ret =fimc_md_create_links (fmd); -if (ret) - goto err_unlock; +mutex_unlock (&fmd->media_dev.graph_mutex); - ret =v4l2_device_register_subdev_nodes (&fmd->v4l2_dev); + ret =device_create_file (&pdev->dev, &dev_attr_subdev_conf_mode);if (ret) - goto err_unlock; + goto err_m_ent; +/* +* FIMC platform devices need to be registered before the sclk_cam +* clocks provider, as one of these devices needs to be activated +* to enable the clock. +*/ + ret =fimc_md_register_clk_provider (fmd); +if (ret <0 ) { +v4l2_err (v4l2_dev,"clock provider registration failed\n" ); + goto err_attr; + } - ret =device_create_file (&pdev->dev, &dev_attr_subdev_conf_mode); + fmd->subdev_notifier.subdevs = fmd->async_subdevs; + fmd->subdev_notifier.num_subdevs = fmd->num_sensors; + fmd->subdev_notifier.bound = subdev_notifier_bound; + fmd->subdev_notifier.complete = subdev_notifier_complete; + fmd->num_sensors =0 ; + + ret =v4l2_async_notifier_register (&fmd->v4l2_dev, + &fmd->subdev_notifier);if (ret) - goto err_unlock; + goto err_clk_p; -platform_set_drvdata (pdev, fmd); -mutex_unlock (&fmd->media_dev.graph_mutex);return 0 ; -err_unlock: -mutex_unlock (&fmd->media_dev.graph_mutex); +err_clk_p: +fimc_md_unregister_clk_provider (fmd); +err_attr: +device_remove_file (&pdev->dev, &dev_attr_subdev_conf_mode); err_clk:fimc_md_put_clocks (fmd); +err_m_ent:fimc_md_unregister_entities (fmd);media_device_unregister (&fmd->media_dev); err_md: +media_device_unregister (&fmd->media_dev); +err_v4l2_dev:v4l2_device_unregister (&fmd->v4l2_dev); -fimc_md_unregister_clk_provider (fmd);return ret; }
@@ -1647,12 +1675,15 @@ static int fimc_md_remove(struct platform_device *pdev)return 0 ;fimc_md_unregister_clk_provider (fmd); +v4l2_async_notifier_unregister (&fmd->subdev_notifier); +v4l2_device_unregister (&fmd->v4l2_dev);device_remove_file (&pdev->dev, &dev_attr_subdev_conf_mode);fimc_md_unregister_entities (fmd);fimc_md_pipelines_free (fmd);media_device_unregister (&fmd->media_dev);fimc_md_put_clocks (fmd); +return 0 ; }
drivers/media/platform/exynos4-is/media-dev.h +12 -1
@@ -32,8 +32,9 @@#define PINCTRL_STATE_IDLE "idle" -#define FIMC_MAX_SENSORS 8 +#define FIMC_MAX_SENSORS 4#define FIMC_MAX_CAMCLKS 2 +#define DEFAULT_SENSOR_CLK_FREQ 24000000U/* LCD/ISP Writeback clocks (PIXELASYNCMx) */ enum {
@@ -79,6 +80,7 @@ struct fimc_camclk_info {/** * struct fimc_sensor_info - image data source subdev information * @pdata: sensor's atrributes passed as media device's platform data +* @asd: asynchronous subdev registration data structure * @subdev: image sensor v4l2 subdev * @host: fimc device the sensor is currently linked to *
@@ -86,6 +88,7 @@ struct fimc_camclk_info {*/ struct fimc_sensor_info {struct fimc_source_info pdata; +struct v4l2_async_subdev asd;struct v4l2_subdev *subdev;struct fimc_dev *host; };
@@ -145,6 +148,9 @@ struct fimc_md {int num_clocks; } clk_provider; +struct v4l2_async_notifier subdev_notifier; +struct v4l2_async_subdev *async_subdevs[FIMC_MAX_SENSORS ]; +bool user_subdev_api;spinlock_t slock;struct list_head pipelines;
@@ -162,6 +168,11 @@ static inline struct fimc_md *entity_to_fimc_mdev(struct media_entity *me) container_of(me->parent, struct fimc_md, media_dev); } +static inline struct fimc_md *notifier_to_fimc_md (struct v4l2_async_notifier *n) +{ + return container_of(n, struct fimc_md, subdev_notifier); +} +static inline void fimc_md_graph_lock (struct exynos_video_entity *ve) {mutex_lock (&ve->vdev.entity.parent->graph_mutex);
Signature · 1 line
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 9/10 ARM: dts: Add rear camera nodes for Exynos4412 TRATS2 board
In reply to Sylwester NawrockiThis patch enables the rear facing camera (s5c73m3) on TRATS2 board by adding the I2C0 bus controller, s5c73m3 sensor, MIPI CSI-2 receiver and the sensor's voltage regulator supply nodes.
--- arch/arm/boot/dts/exynos4412-trats2.dts | 86 +++++++++++++++++++++++++++++-- 1 file changed, 81 insertions(+), 5 deletions(-)
arch/arm/boot/dts/exynos4412-trats2.dts +81 -5
@@ -71,7 +71,33 @@ enable-active-high; }; - /* More to come */ + cam_af_reg: voltage-regulator-2 { + compatible = "regulator-fixed"; + regulator-name = "CAM_AF"; + regulator-min-microvolt = <2800000>; + regulator-max-microvolt = <2800000>; + gpio = <&gpm0 4 0>; + enable-active-high; + }; + + cam_isp_core_reg: voltage-regulator-3 { + compatible = "regulator-fixed"; + regulator-name = "CAM_ISP_CORE_1.2V_EN"; + regulator-min-microvolt = <1200000>; + regulator-max-microvolt = <1200000>; + gpio = <&gpm0 3 0>; + enable-active-high; + regulator-always-on; + }; + + lcd_vdd3_reg: voltage-regulator-4 { + compatible = "regulator-fixed"; + regulator-name = "LCD_VDD_2.2V"; + regulator-min-microvolt = <2200000>; + regulator-max-microvolt = <2200000>; + gpio = <&gpc0 1 0>; + enable-active-high; + }; }; gpio-keys {
@@ -106,6 +132,38 @@ }; }; + i2c_0: i2c at 13860000 { + samsung,i2c-sda-delay = <100>; + samsung,i2c-slave-addr = <0x10>; + samsung,i2c-max-bus-freq = <400000>; + pinctrl-0 = <&i2c0_bus>; + pinctrl-names = "default"; + status = "okay"; + + s5c73m3 at 3c { + compatible = "samsung,s5c73m3"; + reg = <0x3c>; + standby-gpios = <&gpm0 1 1>; /* ISP_STANDBY */ + xshutdown-gpios = <&gpf1 3 1>; /* ISP_RESET */ + vdd-int-supply = <&buck9_reg>; + vddio-cis-supply = <&ldo9_reg>; + vdda-supply = <&ldo17_reg>; + vddio-host-supply = <&ldo18_reg>; + vdd-af-supply = <&cam_af_reg>; + vdd-reg-supply = <&cam_io_reg>; + clock-frequency = <24000000>; + /* CAM_A_CLKOUT */ + clocks = <&camera 0>; + clock-names = "cis_extclk"; + port { + s5c73m3_ep: endpoint { + remote-endpoint = <&csis0_ep>; + data-lanes = <1 2 3 4>; + }; + }; + }; + }; + i2c at 13890000 { samsung,i2c-sda-delay = <100>; samsung,i2c-slave-addr = <0x10>;
@@ -511,8 +569,8 @@ }; }; - camera { - pinctrl-0 = <&cam_port_b_clk_active>; + camera: camera { + pinctrl-0 = <&cam_port_a_clk_active &cam_port_b_clk_active>; pinctrl-names = "default"; status = "okay";
@@ -532,6 +590,23 @@ status = "okay"; }; + csis_0: csis at 11880000 { + status = "okay"; + vddcore-supply = <&ldo8_reg>; + vddio-supply = <&ldo10_reg>; + clock-frequency = <176000000>; + + /* Camera C (3) MIPI CSI-2 (CSIS0) */ + port at 3 { + reg = <3>; + csis0_ep: endpoint { + remote-endpoint = <&s5c73m3_ep>; + data-lanes = <1 2 3 4>; + samsung,csis-hs-settle = <12>; + }; + }; + }; + csis_1: csis at 11890000 { vddcore-supply = <&ldo8_reg>; vddio-supply = <&ldo10_reg>;
@@ -572,10 +647,11 @@ reg = <0x10>; svdda-supply = <&cam_io_reg>; svddio-supply = <&ldo19_reg>; + afvdd-supply = <&ldo19_reg>; clock-frequency = <24000000>; /* CAM_B_CLKOUT */ - clocks = <&clock_cam 1>; - clock-names = "mclk"; + clocks = <&camera 1>; + clock-names = "extclk"; samsung,camclk-out = <1>; gpios = <&gpm1 6 0>;
Signature · 1 line
1.7.9.5
Sylwester Nawrocki
maintainerPATCH v4 10/10 ARM: dts: exynos4: Update clk provider part of the camera subsystem
In reply to Sylwester NawrockiRemove unused /camera/clock-controller node and required clock properties to the camera node. This is required for a clock provider that will be referenced by image sensor devices.
--- arch/arm/boot/dts/exynos4.dtsi | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
arch/arm/boot/dts/exynos4.dtsi +2 -4
@@ -109,12 +109,10 @@ status = "disabled"; #address-cells = <1>; #size-cells = <1>; + #clock-cells = <1>; + clock-output-names = "cam_mclk_a", "cam_mclk_b"; ranges; - clock_cam: clock-controller { - #clock-cells = <1>; - }; - fimc_0: fimc at 11800000 { compatible = "samsung,exynos4210-fimc"; reg = <0x11800000 0x1000>;
Signature · 1 line
1.7.9.5