i2c2 | cssh like ssh tool for iTerm2 - based on i2cssh | Development Tools library

At first you need to understand how to deal with your dts file.

You need to decide whether you want to work with the official dts file that is provided by your linux kernel and patch it. Or, create your own dts file and use it instead.

To understand how to patch the dts or how to add custom one, I have 2 stackoverflow answers for that, here and here.

After you understand that, now lets talk about adding your I2C node.

First, you need to search if your sensor has an official kernel driver.

If it has a driver, then, it has a compatible name that is defined in it.

You take that name with the sensor address in that I2C bus and:

For example, adding the sensor of address 0x01 to I2C 1 bus:

This issue was resolved. After running the KMD driver for the NVDLA_IP_0 through the instruction mention in the comments and here by Ian Abbott, the node appeared in /pro/iomem.

Your program runs out of RAM. The ESP32 has 520 KiB of RAM, out of that 328 KiB is used for data and 192 KiB for instructions. This has to fit the Python interpreter (which is a memory-hungry program itself) and all that your script uses.

Here you're creating 3 huge arrays of 32-bit integers which together allocate 30% of the entire data RAM (each takes 128x64x4=32KiB to store just 1KiB of frame data). That's not how it's done. Also I have no idea what's inside the SSD1306 drivers you're creating 5 of - maybe frame buffers of their own?

If you're certain this is the way to go, get the ESP-WROVER-B module which has 8 MiB of added PSRAM.

The CPAR field of the DMA1_Channel4 structure most likely expects a value of type uint32_t, but you provide a value of type "pointer to uint32_t". Most likely you mean:

After some comments, there was more useful info, biggest part being that oscilloscope showed Start 1010100 0 1 Stop. While 1010100 is the 0x54 address, the W bit needs to be counted as well, so the 8-bit data was actually 10101000, instead of the expected 01010100. Address should not have been shifted left, as it was already correct.

This is a somewhat common mistake with I2C. Attention needs to be paid to the exact address and R/!W bits, as in some documents and APIs they're shifted left, in some they aren't.

A very helpful bit of info here was the oscilloscope trace, that really helped show what exactly is going on.

(Gathering the answer based on comments I have given earlier)

For sake of the clarification I have to say, that from your DSDT we can get the following information. There are 3 groups of PMICs, i.e. DSCx, CLPx and PMIx. I believe they are based on the model, like Desktop, Laptop, 2-in-1. And in each case all PMICs in the same group have different _UID. From the provided output of the grep -H 15 ... we have only 2 out of 10 enumerated with the instances INT3472:08 and INT3472:09 (exactly two last defined in DSDT). And they are PMIx, you may check this by grep -H . /sys/bus/acpi/devices/INT3472:*/path.

Your interest is the PMI1 which consumes three GPIO lines from Intel GPIO driver, i.e. pins 121, 122 and 143 (you may decode them as Community #2, Group #5 or GPP_F, relative to the group pins 1, 2 and 23, this may help you to understand _INI method that touches these lines via other methods in DSDT), and provides 3 + 7 = 10 pins according to its driver.

Now to the code. The _DEP ACPI method is used solely for linking power resources, and Linux kernel has other means how to hijack resources from other device, because what you are trying to do is to use the resource which is not related to the device you are creating driver for.

The method is to find device by ACPI HID:

Do you have camera PMIC drivers loaded? Your camera sensors likely powered off. The camera PMIC is TPS68470 chip which is represented by three drivers:

1. PMIC as Multi-Functional Device (MFD) driver with ACPI ID INT3472
2. GPIO driver for it, and
3. OpRegion driver to support access from ACPI code

All three must be enabled and loaded in order to get power gating work.

Side note regarding to I²C addresses. In data sheet the addresses most probably are in 8-bit format, means the real ones (7-bit) are 0x10 or 0x36.