What causes "resource temporarily unavailable" in v4l2 - linux

I have compiled adv7180 driver available here.
I am unloading the ov5642 cameradriver(which in my case is built-in) and loading the adv7180_tvin module and after I am loading mxcv4l2_capture module which creates video0 in /dev/.
(dmesg command says: "mxc camera on IPU2_CSI1 registered as video0")
But when I try to access video0 with v4l2-ctl I got a message "resource temporarily unavailable" or when I am using gstreamer I got message "Can not open /dev/video0" (but the device is really created).
Is that a problem in device tree settings or it can be caused by something else? Which tools should I use to find out what causes this issue?
My device tree settings look like below:
&i2c3{
adv7180: adv7180#20{
compatible = "adv,adv7180";
reg = <0x20>;
clocks = <&clks IMX6QDL_CLK_CKO2>;
clock-names = "csi_mclk";
pwn-gpios = <&gpio3 10 GPIO_ACTIVE_LOW>;
ipu_id = <1>;
csi_id = <1>;
mclk = <24000000>;
mclk_source = <0>;
pinctrl-names = "default";
pinctrl-0 = <&pinctrl_hummingboard2_parallel>;
cvbs = <1>;
};
};
I need to add that before adv7180 I was using above settings for ov5642 camera (excluding cvbs setting) and everything worked properly.
EDIT:
Ok I got one clue.
When I load modules in dmesg message "mxc_v4l2_master_attach: ipu(0:1)/csi(1:1)/mipi(0:0) doesn't match" shows.
But it only happens when ipu_id=<1> in v4l2_cap device tree settings and in adv7180 settings. When i change ipu_id to ipu_id=<0> in v4l2 settings and adv7180 dmesg now shows "parallel attach to IPU1 CSI1 and I can access the /dev/video0 succesfully with v4l2-ctl tool.
But In my case there is only one possibility to use IPU2_CSI1.
Why can't I set IPU2 to adv7180 when I was using it successfully to ov5642 ?

As per my knowledge i.MX6 having two IPUs. I think by default IPU1 parallel interface is not enabled in the board file. So you need to check the IOMUXC_GPR1 register setting (bit 19 and 20) for IPU/CSI1 and pass the csi_id in your camera driver.
As you are using the parallel interface so check your pin muxing setting as well in your device tree. (which is not required for serial interface.)
Edit:
There are two ways which you can follow to update the register setting from the kernel space (boardfile or camera driver) itself:
1. From the board file:
struct regmap *gpr
gpr = syscon_regmap_lookup_by_compatible("fsl,imx6q-iomuxc-gpr");
regmap_update_bits(gpr, IOMUXC_GPR1, 1 << 20, 1 << 20);
2. From the boardfile or camera driver
void __iomem *va_ipu2_address;
u32 reg_val;
va_ipu2_address = ioremap(0x20e0004,0xe0004);
reg_val = readl(va_ipu2_address);
/* Enable parallel interface to IPU2 CSI1. */
writel(reg_val | 1 << 20, va_ipu2_address);

Thanks for Your answer.
My pinmuxing looks like this:
&iomuxc{
hummingboard2{
pinctrl_hummingboard2_parallel: hummingboard2_parallel{
fsl,pins = <
MX6QDL_PAD_EIM_A24__IPU2_CSI1_DATA19 0x0b0b1
MX6QDL_PAD_EIM_A23__IPU2_CSI1_DATA18 0x0b0b1
MX6QDL_PAD_EIM_A22__IPU2_CSI1_DATA17 0x0b0b1
MX6QDL_PAD_EIM_A21__IPU2_CSI1_DATA16 0x0b0b1
MX6QDL_PAD_EIM_A20__IPU2_CSI1_DATA15 0x0b0b1
MX6QDL_PAD_EIM_A19__IPU2_CSI1_DATA14 0x0b0b1
MX6QDL_PAD_EIM_A18__IPU2_CSI1_DATA13 0x0b0b1
MX6QDL_PAD_EIM_A17__IPU2_CSI1_DATA12 0x0b0b1
MX6QDL_PAD_EIM_DA11__IPU2_CSI1_HSYNC 0x0b0b1
MX6QDL_PAD_EIM_DA12__IPU2_CSI1_VSYNC 0x0b0b1
MX6QDL_PAD_EIM_A16__IPU2_CSI1_PIXCLK 0x0b0b1
MX6QDL_PAD_EIM_DA10__GPIO3_IO10 0x400130b1
>;
};
};
};
and it's been working succesively with ov5642 camera.
No I see that the adv7180 driver does not take an ipu_id as an argument from device tree so I think it is using the default ipu which is (I think) IPU1.
I've been playing arround how to change settings in IOMUXC_GPR1. Bit 20 needs to be set ("enable parallel interface to IPU2 CSI1). But have got no more ideas how to do it in device tree.

Ok. I found it !
I couldn't set bit 20 in IOMUXC_GPR1 register using mach-imx6q.c file so I did it this way:
in console:
sudo devmem2 0x20e0004
and read the existing value (which was in my case 0x48643005). Then I set bit 20 to one ("1") so I got 0x48743005 and I put this value into the register:
sudo devmem2 0x20e0004 w 0x48743005
next I loaded adv7180_tvin and mxc_v4l2_capture modules and captured frames using gsreamer:
gst-launch-1.0 imxv4l2videosrc device=/dev/video0 ! imxipuvideotransform ! autovideosink deinterlace=true
Everything works great ! Thanks for help !

Related

trouble configuring timings for LVDS display (linux on rk3288)

I'm trying to hook up an LVDS display to my dev board based on rockchip rk3288 and have trouble getting graphics to show on the display.
The display is the GWTS80MNFG1E0 by SGD. The datasheet is here http://www.datadisplay-group.com/fileadmin/pdf/produkte/Displays/SGD/GWTS80MNFG1E0_Datasheet.pdf
Here is the .dtsi file I created for the display:
/ {
disp_timings: display-timings {
native-mode = <&timing0>;
timing0: timing0 {
screen-type = <SCREEN_LVDS>;
lvds-format = <LVDS_8BIT_1>;
out-face = <OUT_P888>;
clock-frequency = <48690000 52590000 60830000>;
hactive = <1656 1660 1760>;
vactive = <490 528 576>;
hback-porch = <5 16 141>;
hfront-porch = <19 44 155>;
vback-porch = <5 5 91>;
vfront-porch = <5 43 91>;
hsync-len = <1 2 140>;
vsync-len = <1 2 90>;
hsync-active = <0>;
vsync-active = <0>;
de-active = <1>;
pixelclk-active = <0>;
swap-rb = <0>;
swap-rg = <0>;
swap-gb = <0>;
};
};
};
The problem is that the display turns on but I can't see any graphics. In the .config file for the kernel I have CONFIG_RK32_LVDS=y. The .dts file for the board includes my .dtsi file.
Hardware details:
On my breadboard, interfacing my dev board to the display, I've pulled STBYB, RESET, SELB, RL, and TB high via a 10k pullup to 3.3v.
For backlight I'm using the adafruit tft friend - https://www.adafruit.com/product/1932
I'm taking 5v from my dev board to feed the backlight driver and I've hooked up LED- and LED+ of the display to the backlight driver. The driver is configured to output 75mA at the moment. The adafruit board has specs of outputting up to 125mA at 24V while the SGD datasheet mentions 25.6V. I'm not sure if this is a problem or not.
I have the VSDN/VSDP hooked up to a TPS65132WEVM-669 (texas instruments) which I've programmed to provide +/- 5.5v. This EVM is powered from the same 5v as the backlight driver. I've verified it's outputting the correct voltages.
I've hooked up the display inputs RXIN[0123]+/- to board lvds port outputs D[0123]P/N. Display clock inputs RXCLKIN+/- are hooked up to board lvds port clock lines CLK0P/N.
My questions:
1) is the backlight driver the problem here?
2) is the .dtsi file I created for the display correct?
3) is there anything else I can check w/r/t my kernel / dts config or the hardware itself?
Thanks
B
There are five items which need to be verified:
LVDS settings in dts:
Check the IPU channel which is correctly mapped on the lvds device
data-width (18bit/24bit) of the lvds channel.
LVDS output format for serializer (VESA or Non-VESA format )
Frame buffer settings in dts:
Check your frame buffer settings
- bit pet pixels
- Pixel format
- ipu clock
Kernel command:
Check whether any of the kernel command which is overwrite the dtsi settings during kernel startup.
Backlight:
In order to make the backlight, measure the display current.
Lock:
Check the physical connection between serializer and deserializer. Make sure the LOCK signal.
I had as similar issue after a kernel update, and this kernel "fix" was the problem. Newer kernels (after 2013) fall back on not having any LVDS screen, which mean that you don't see any graphics.
http://marc.info/?l=git-commits-head&m=138449380916013&w=2

Pin Muxing in linux kernel

I want to do following pin muxing.
i.e USART Tx -> to GPIO -> Back to USART Tx pin inside Linux kernel for some purpose.
I tried to make the PIN as GPIO using gpio_request and gpio_direction_output, so i am able to make that pin as GPIO . But as i want to switch back from GPIO to USART Tx pin, it is not working, I tried following at91_set_A_periph to that pin, still no luck.
Working on kernel : 3.18 and at91 atmel board.
You can have a look at the i2c-imx driver. It does exactly that.
You need to use pinctrl_lookup_state to retrieve the different pinctrl states (one of those being USART Tx and the other one GPIO). Then you can switch between both with pinctrl_select_state.
To sum it up, you'd have something like that in your uart node:
usart3: serial#fc00c000 {
pinctrl-names = "default","gpio";
pinctrl-0 = <&pinctrl_usart3>;
pinctrl-1 = <&pinctrl_usart3_gpio>;
tx-gpio = <&pioE 4 GPIO_ACTIVE_LOW>;
status = "okay";
};
In the driver code:
pinctrl_pins_default = pinctrl_lookup_state(pinctrl, PINCTRL_STATE_DEFAULT);
pinctrl_pins_gpio = pinctrl_lookup_state(pinctrl, "gpio");
tx_gpio = of_get_named_gpio(pdev->dev.of_node, "tx-gpio", 0);
Then, you can use pinctrl_select_state to switch back and forth between pinctrl_pins_default and pinctrl_pins_gpio. tx_gpio is your gpio.

Specifying GPIO Numbers for IO Expander in Linux Device Tree

I'm trying to add a PCA9557 I/O expander to an arm-based system on an I2C bus. The system already has another I/O expander on a different I2C bus. I am trying to figure out how to specify which GPIO numbers the pins on the new expander get, and how to get both working.
Here's the device tree section for the existing expander, under I2C bus 2:
i2c2: i2c#e8007000 {
status = "ok";
pca9539: pca9539#74 {
compatible = "nxp,pca9539";
reg = <0x74>;
interrupt-parent = <&gpio>;
interrupts = <9 0x0>;
gpio-controller;
#gpio-cells = <2>;
interrupt-controller;
#interrupt-cells = <2>;
};
};
Using the above, the existing I/O expander (with 16 GPIOs) appears in linux as /sys/class/gpio/gpiochip128, exposing GPIO numbers 128 - 143. GPIOs 0-127 are built in to the host processor.
I added the following for the new expander on I2C bus 0:
i2c0: i2c#e8003000 {
status = "ok";
pca9557: pca9557#18 {
compatible = "nxp,pca9557";
reg = <0x18>;
gpio-controller;
#gpio-cells = <2>;
};
I also modified the kernel config to build the GPIO_PCA953X driver, which should support the PCA9557.
When I compile and boot with the above added to the device tree, I now see the NEW expander (PCA9557) mapped as /sys/class/gpio/gpiochip136, and it works (I can set its IO 0 pin using GPIO128).
However, there are no longer any GPIO pins for the other expander. It still appears as a device on the appropriate bus under /sys/devices/... but a directory listing doesn't show the "driver..." and "i2c-2" items which were there before.
So how do I get BOTH expanders to appear in /sys/class/gpio/ with different ranges of GPIO numbers, so I can use both?
I guess the "128" as the base GPIO for the original expander was just the next available GPIO? But why does the new expander end up starting at GPIO 136?
I've seen several references to this page: GPIO bindings documentation, but it was fairly generic and didn't help much.

What is a "pixel valve" in bcm2835 chip

In open source driver code of vc4 authored by Eric, there's something called "pixel valve". Apparently it has something to do with display output. But I would really want have more information about what exactly it is? How does it work?
https://github.com/peyo-hd/kernel_rpi/blob/vc4/arch/arm/boot/dts/bcm2708_common.dtsi
pixelvalve#7e807000 {
compatible = "brcm,bcm2835-pixelvalve2";
reg = <0x7e807000 0x100>;
interrupts = <2 10>; /* pixelvalve */
};

TinyX shows display using builtin fbtft touchscreen driver but touch doesn't work

I'm using an "adafruitts" touchscreen with a raspi to control a usb peripheral.
The full raspbian kernel takes forever to boot (50 seconds), and part of that is due to the touchscreen driver loading (by modprobe/udev) and initializing.
During the first 20-30 seconds of boot, the display is not loaded, so it is blank. I need this to be a user-friendly item that cannot be blank for 30 seconds each time it is turned on, so I've used buildroot to build a small kernel with the touchscreen driver built-in. (I am on a steep learning curve with buildroot and kernel building in general).
The display driver is fbtft_device.c patched to include the adafruitts display. This patch defines the "touch" half:
/* Touch device spi-half of adafruit touchscreen */
.name = "adafruitts",
.spi = &(struct spi_board_info) {
.modalias = "stmpe610",
.max_speed_hz = 500000,
.mode = SPI_MODE_0,
.chip_select = 1,
.platform_data = &(struct stmpe_platform_data) {
.blocks = STMPE_BLOCK_TOUCHSCREEN | STMPE_BLOCK_GPIO,
.irq_over_gpio = 1,
.irq_gpio = 24,
.irq_trigger = IRQF_TRIGGER_FALLING,
.irq_base = GPIO_IRQ_START + GPIO_IRQS,
.ts = &(struct stmpe_ts_platform_data) {
.sample_time = 4,
.mod_12b = 1,
.ref_sel = 0,
.adc_freq = 2,
.ave_ctrl = 3,
.touch_det_delay = 4,
.settling = 2,
.fraction_z = 7,
.i_drive = 0,
},
}
},
.is_support = 1,
.gpio_settings = (struct gpio_setting []) {
{
.gpio = 24,
.pull = pull_up,
}
},
.gpio_num_settings = 1,
},
and the LCD half:
}, {
/* LCD component of adafruit touchscreen */
.name = "adafruitts",
.spi = &(struct spi_board_info) {
.modalias = "fb_ili9340",
.max_speed_hz = 16000000,
.mode = SPI_MODE_0,
.chip_select = 0,
.platform_data = &(struct fbtft_platform_data) {
.display = {
.buswidth = 8,
.backlight = 1,
},
.bgr = true,
.gpios = (const struct fbtft_gpio []) {
{ "dc", 25 },
{},
},
}
}
}, {
by including:
fbtft_device.name=adafruitts
in the cmdline.txt for the boot loader, I've gotten the display half of the system to work (it boots in ~ 5 seconds :) ) with tinyX/matchbox desktop showing the desktop, but I cannot get the touchscreen part to work (the cursor does not move when I touch the screeen).
Somehow I have to bind the touch part of the touchscreen to tinyX, but I have not been able to figure out how to do this.
I have tried to specify the keyboard (and mouse) when launching tinyX:
X -keybd smpte610 (for example)
but X reports it cannot find the driver.
How can I verify the touch screen input device was successfully loaded?
The boot log has these messages about fbtft_device:
fbtft_device: SPI devices registered:
fbtft_device: spidev spi0.0 500kHz 8 bits mode=0x00
fbtft_device: spidev spi0.1 500kHz 8 bits mode=0x00
fbtft_device: 'fb' Platform devices registered:
fbtft_device: bcm2708_fb id=-1 pdata? no
fbtft_device: Deleting spi0.1 (spi0.1)
fbtft_device: Looking at item 0
fbtft_device: Setting pin 24 to 2
stmpe-spi: probe of spi0.1 failed with error -22
fbtft_device: Deleting spi0.0 (spi0.0)
Console: switching to colour frame buffer device 40x30
graphics fb0: fb_ili9340 frame buffer, 320x240, 150 KiB video memory, 16 KiB buffer memory, fps=20, spi0.0 at 16 MHz
fbtft_device: GPIOS used by 'adafruitts':
fbtft_device: 'dc' = GPIO25
fbtft_device: SPI devices registered:
fbtft_device: stmpe610 spi0.1 48000kHz 8 bits mode=0x00
fbtft_device: fb_ili9340 spi0.0 16000kHz 8 bits mode=0x00
kgdb: Registered I/O driver kgdboc.
Is the kgdb message associated with fbtft_device or something else?
If I look in /dev/input I see: event0, event1, and mice. event0 and event1 are associated with an attached keyboard (according to the boot log) and I have no mouse attached. Should there be some other items in input?
If the touch screen input device IS loaded, how to I specify the correct driver for tinyX?
Thanks
What I learned:
By comparing the boot messages in my modprobe/udev/module loading kernel with the fast built-in kernel, it shows:
stmpe-spi: probe of spi0.1 failed with error -22
is a "bad" thing.
A successful driver load will say (something like):
bcm2708_spi.0: registered child spi0.0
and then later:
input: stmpe-ts as /devices/virtual/input/input0
I fixed the "probe" failure by making these changes to my kernel configuration file. (Sorry, I don't want to include the whole thing, so these are the changes from when I had the issue to when the driver successfully loaded according to the syslog):
< Touch Did Not respond > Touch Did respond
> CONFIG_INPUT_FF_MEMLESS=y
< CONFIG_INPUT_POLLDEV=m > CONFIG_INPUT_POLLDEV=y
< CONFIG_INPUT_EVDEV=m > CONFIG_INPUT_EVDEV=y
< CONFIG_TOUCHSCREEN_STMPE=m > CONFIG_TOUCHSCREEN_STMPE=y
> CONFIG_KEYBOARD_STMPE=y
< CONFIG_SERIO=m > CONFIG_SERIO_SERPORT=m
> CONFIG_SPI_DEBUG=y
< CONFIG_SPI_SPIDEV=y
> CONFIG_SPI_GPIO=y
My main objective with these changes was to try to make sure that the dependent drivers were also built-in, and I enabled the debug. (Some of these were magically set by menuconfig, and this is diff from the "non-default" values from buildroot, so the diff is - different)
With this config, I now have event0, event1, event2, mice, and mouse0. The syslog says event1 and event2 are associated with the usb keyboard I have attached. I have no extra mouse attached.
I could use "evtest" to see events from /dev/input/event0 whenever I touched the display. evtest'ing /dev/input/mouse0 threw "Inappropriate ioctl for device"
I restarted X (tinyX) using:
X -mouse mouse,,/dev/input/mouse0
and touches worked, but the touch axis is rotated from the display axis.
I could not figure out a way to fix this in tinyX, so I'm going with a full blown Xorg implementation.
Make sure you have enabled the evdev input support in tinyx (BR2_PACKAGE_XSERVER_XORG_SERVER_KDRIVE_EVDEV) and use the syntax specified in hw/kdrive/src/kinput.c:
/*
* You can call your kdriver server with something like:
* $ ./hw/kdrive/yourserver/X :1 -mouse evdev,,device=/dev/input/event4 -keybd
* evdev,,device=/dev/input/event1,xkbmodel=abnt2,xkblayout=br
*/

Resources