Hi Bryan,
On 5/18/22 16:30, Bryan O'Donoghue wrote:
The IMX577 is on CCI1/CSI2 providing four lanes of camera data.
An example media-ctl pipeline is:
media-ctl --reset
media-ctl -v -d /dev/media0 -V '"imx412 '20-001a'":0[fmt:SRGGB10/4056x3040 field:none]'
media-ctl -V '"msm_csiphy2":0[fmt:SRGGB10/4056x3040]'
media-ctl -V '"msm_csid0":0[fmt:SRGGB10/4056x3040]'
media-ctl -V '"msm_vfe0_rdi0":0[fmt:SRGGB10/4056x3040]'
media-ctl -l '"msm_csiphy2":1->"msm_csid0":0[1]'
media-ctl -l '"msm_csid0":1->"msm_vfe0_rdi0":0[1]'
yavta -B capture-mplane -c -I -n 5 -f SRGGB10P -s 4056x3040 -F /dev/video0
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@xxxxxxxxxx>
---
arch/arm64/boot/dts/qcom/qrb5165-rb5.dts | 98 ++++++++++++++++++++++++
1 file changed, 98 insertions(+)
diff --git a/arch/arm64/boot/dts/qcom/qrb5165-rb5.dts b/arch/arm64/boot/dts/qcom/qrb5165-rb5.dts
index 0e63f707b911..48b31790c434 100644
--- a/arch/arm64/boot/dts/qcom/qrb5165-rb5.dts
+++ b/arch/arm64/boot/dts/qcom/qrb5165-rb5.dts
@@ -1203,6 +1203,43 @@ sdc2_card_det_n: sd-card-det-n {
function = "gpio";
bias-pull-up;
};
+
+ cam2_default: cam2-default {
+ rst {
+ pins = "gpio78";
+ function = "gpio";
+
+ drive-strength = <2>;
+ bias-disable;
+ };
+
+ mclk {
+ pins = "gpio96";
+ function = "cam_mclk";
+
+ drive-strength = <16>;
+ bias-disable;
+ };
+ };
+
+ cam2_suspend: cam2-suspend {
+ rst {
+ pins = "gpio78";
+ function = "gpio";
+
+ drive-strength = <2>;
+ bias-pull-down;
+ output-low;
+ };
+
+ mclk {
+ pins = "gpio96";
+ function = "cam_mclk";
+
+ drive-strength = <2>;
+ bias-disable;
+ };
+ };
I still stick to my opinion that the description of rst/mclk pins should
be uniformly added to the SoC specific .dtsi file. The pins and functions
in these device tree nodes are not changeable, a board file should just
select proper pairs.
Do you have any objections to it?
};
&uart12 {
@@ -1294,3 +1331,64 @@ &qup_spi0_data_clk {
drive-strength = <6>;
bias-disable;
};
+
+&camcc {
+ status = "okay";
+};
+
+&camss {
+ status = "okay";
+ vdda-phy-supply = <&vreg_l5a_0p88>;
+ vdda-pll-supply = <&vreg_l9a_1p2>;
+
+ ports {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ /* The port index denotes CSIPHY id i.e. csiphy2 */
+ port@2 {
+ reg = <2>;
+ csiphy2_ep: endpoint {
+ clock-lanes = <7>;
+ data-lanes = <0 1 2 3>;
+ remote-endpoint = <&imx412_ep>;
+ };
+
+ };
+ };
+};
+
+&cci1 {
+ status = "okay";
+};
+
+&cci1_i2c0 {
+ camera@1a {
+ compatible = "sony,imx412";
+ reg = <0x1a>;
+
+ reset-gpios = <&tlmm 78 GPIO_ACTIVE_LOW>;
+ pinctrl-names = "default", "suspend";
+ pinctrl-0 = <&cam2_default>;
+ pinctrl-1 = <&cam2_suspend>;
+
+ clocks = <&camcc CAM_CC_MCLK2_CLK>;
+ assigned-clocks = <&camcc CAM_CC_MCLK2_CLK>;
+ assigned-clock-rates = <24000000>;
+
+ power-domains = <&camcc TITAN_TOP_GDSC>;
Above 'power-domains' property is not needed, it shall be implied by CCI.
+ dovdd-supply = <&vreg_l7f_1p8>;
+ avdd-supply = <&vdc_5v>;
+ dvdd-supply = <&vdc_5v>;
+
+ status = "okay";
Here 'status' property is not needed.
+ port {
+ imx412_ep: endpoint {
+ clock-lanes = <1>;
+ link-frequencies = /bits/ 64 <600000000>;
+ data-lanes = <1 2 3 4>;
+ remote-endpoint = <&csiphy2_ep>;
+ };
+ };
+ };
+};
I run on you branch on top of linux-next, but switch build options from modules to built-in
CONFIG_I2C_QCOM_CCI=y
CONFIG_VIDEO_QCOM_CAMSS=y
I didn't get the sensor initialized and hence there is no /dev/media0 node:
[ 0.620205] i2c-qcom-cci ac50000.cci: Found 19200000 cci clk rate while 37500000 was expected
[ 0.620551] i2c 20-001a: Fixing up cyclic dependency with ac6a000.camss
[ 0.620754] imx412 20-001a: Looking up dovdd-supply from device tree
[ 0.620797] imx412 20-001a: Looking up avdd-supply from device tree
[ 0.620860] imx412 20-001a: Looking up dvdd-supply from device tree
[ 0.620876] duplicated lane 1 in clock-lanes, using defaults
[ 0.622789] imx412 20-001a: failed to find sensor: -5
[ 0.622880] imx412: probe of 20-001a failed with error -5
I believe the problem could be related to CCI, please remind me, are there I2C bus pull-ups?
--
Best wishes,
Vladimir
