DHT11.ko (32-bit Bookworm Raspberry Pi OS with a 64-bit kernel) (raspberry pi 4b)

[Abdulrahman-Yasser]

Describe the bug

I think DHT11 module doesn't work with my kernel image

Steps to reproduce the behaviour

My aim is to read dht11 from the file system (kernel module)
Steps :

1. Changed my device tree bcm2711-rpi-4-b.dts and added the following lines

humidity-sensor {
             compatible = "dht11";
             gpios = <&gpio 17 GPIO_ACTIVE_HIGH>;
         }; 

2. connected my dht data to gpio17 - vcc to rpi.5v - gnd to rpi.gnd
3. i can find the file in ls -l /sys/bus/iio/devices/iio:device0
4. output of ls -l /sys/bus/iio/devices/iio:device0 is :

-rw-r--r-- 1 root root 4096 Jun  7 01:30 in_humidityrelative_input
-rw-r--r-- 1 root root 4096 Jun  7 01:30 in_temp_input
-r--r--r-- 1 root root 4096 Jun  7 01:29 name
lrwxrwxrwx 1 root root    0 Jun  7 01:30 of_node -> ../../../../firmware/devicetree/base/humidity-sensor
drwxr-xr-x 2 root root    0 Jun  7 01:30 power
lrwxrwxrwx 1 root root    0 Jun  7 01:30 subsystem -> ../../../../bus/iio
-rw-r--r-- 1 root root 4096 Jun  7 01:29 uevent
-r--r--r-- 1 root root 4096 Jun  7 01:30 waiting_for_supplier

5. I try to read the temperature using the following code :

sudo cat in_temp_input 

It shows

cat: in_temp_input: Input/output error

6. output of dmesg | tail is

[   59.967095] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 4
[   62.719064] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3
[   64.866734] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3
[  193.956733] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3

7. The reading in dmesg is correct. I tried to test it with a lighter and the temperature increased correctly.
8. I tried to connect dht11 on gpio4 and use a python script to read the data using adafruit-circuitpython-dht and it worked fine (following that article https://pimylifeup.com/raspberry-pi-dht11-sensor/)

Device (s)

Raspberry Pi 4 Mod. B

System

cat /etc/issue :

Raspbian GNU/Linux 12 \n \l

cat /etc/rpi-issue :

Raspberry Pi reference 2024-03-15
Generated using pi-gen, https://github.com/RPi-Distro/pi-gen, 11096428148f0f2be3985ef3126ee71f99c7f1c2, stage5

vcgencmd version :

May 24 2024 15:30:04 
Copyright (c) 2012 Broadcom
version 4942b7633c0ff1af1ee95a51a33b56a9dae47529 (clean) (release) (start)

uname -m :

 aarch64

Logs

[   59.967095] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 4
[   62.719064] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3
[   64.866734] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3
[  193.956733] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3

Additional context

It may be irrelevant. But I am also appended lcd and i2c-gepio-expander to my rpi device tree to work with the modules. Both of them work fine.

[fmthola]

I am having the same issue. After trying to fix the wiring, changing GPIO pins, it still wasn't resolving the issue.

I gave up, and just wrote a python script that would read the dmesg and I would parse it myself.
In your example: [ 193.956733] dht11 humidity-sensor: Don't know how to decode data: 42 0 31 3
42 is the humidity and 31 is the temperature (in C)

The code below only returns the humidity. WIth my sensor, I was getting values that would be 100+ the number. In your example, I would have values that would say 142 instead of just 42. There is logic in this code of if it's greater than 100, it will just ignore and try again.

Program flow:

• I have python open the /sys/bus/iio/devices/iio:device0/in_humidityrelative_input file, this creates an error, however that creates the dmesg error
• wait a moment after reading the file so dmesg has the log
• read the dmesg log files
• show only the most recent file

I know this is a hacky workaround. For my other application, I then simply imported and called this file to get the humidity

import subprocess
import time
import re

def read_humidity():
    try:
        with open('/sys/bus/iio/devices/iio:device0/in_humidityrelative_input', 'r') as file:
            humidity = file.read().strip()
            return humidity
    except Exception as e:
        # Do not print the error
        return None

def get_dmesg_messages():
    try:
        # Capture the output of dmesg
        result = subprocess.run(['dmesg'], stdout=subprocess.PIPE, text=True)
        # Return the output as a list of lines
        return result.stdout.splitlines()
    except Exception as e:
        print(f"Error reading dmesg: {e}")
        return []

def filter_dht11_messages(dmesg_lines):
    dht11_messages = []
    pattern = re.compile(r"dht11.*Don't know how to decode data:")
    for line in dmesg_lines:
        if pattern.search(line):
            dht11_messages.append(line)
    return dht11_messages

def extract_first_number(message):
    match = re.search(r"decode data: (\d+)", message)
    if match:
        return int(match.group(1))
    return None

def main():
    last_dht11_message = None
    last_extracted_number = None
    
    while True:
        # Read humidity
        humidity = read_humidity()
        if humidity is not None:
            print(f"Humidity: {humidity}%")
        
        # Wait for 0.1 seconds
        time.sleep(0.1)

        # Get the dmesg messages
        dmesg_lines = get_dmesg_messages()
        
        # Filter for DHT11-related messages
        dht11_messages = filter_dht11_messages(dmesg_lines)
        
        # Extract the most recent message if there are any new ones
        if dht11_messages:
            most_recent_message = dht11_messages[-1]
            if most_recent_message != last_dht11_message:
                last_dht11_message = most_recent_message
                extracted_number = extract_first_number(most_recent_message)
                if extracted_number is not None and extracted_number <= 100:
                    last_extracted_number = extracted_number
        
        # Print the most recent extracted number
        if last_extracted_number is not None:
            print(f"Most recent extracted number: {last_extracted_number}")
        else:
            print("No valid number extracted from DHT11 messages.")
        
        # Wait before the next iteration
        time.sleep(2)

if __name__ == "__main__":
    main()

[ashley-b]

Am also seeing this error

Software details

Distro: Debain 12.1
kernel version: Linux version 6.6.28-v8+ (dom@buildbot) (aarch64-linux-gnu-gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #1757 SMP PREEMPT Thu Apr 18 11:15:41 BST 2024

lsb_release -a

Distributor ID:	Raspbian
Description:	Raspbian GNU/Linux 12 (bookworm)
Release:	12
Codename:	bookworm

Hardware details

• Raspberry Pi 5 Model B Rev 1.0
• /boot/firmware/config.txt dtoverlay=dht11,gpiopin=4

Gives this error % 50 of the time otherwise it gives a reasonable value as below

$ cat /sys/bus/iio/devices/iio:device0/in_temp_input
cat: '/sys/bus/iio/devices/iio:device0/in_temp_input': Input/output error

Works some times

$ cat /sys/bus/iio/devices/iio:device0/in_temp_input
31000

dmesg dump

[505002.065451] dht11 dht11@4: Don't know how to decode data: 140 0 15 1
[505012.064921] dht11 dht11@4: Don't know how to decode data: 24 0 30 3
[505012.064929] dht11 dht11@4: Don't know how to decode data: 140 0 15 1
[505022.063812] dht11 dht11@4: Don't know how to decode data: 24 0 30 3
[505022.063821] dht11 dht11@4: Don't know how to decode data: 140 0 15 1
[505032.064904] dht11 dht11@4: Don't know how to decode data: 24 0 30 3
[505032.064914] dht11 dht11@4: Don't know how to decode data: 140 0 15 1

[fmthola]

Just an update, I gave the code above that decodes the dmesg, but after a few days that "workaround" wasn't working as dmesg was then reporting it wasn't getting anything at all.

I ordered replacement sensors and when I swapped the existing sensor with a new one, this issue went away and I didn't need my workaround anymore

[VDV2000]

I have the same problem. RPi 3B, Bookworm 64bit.
after sudo dtoverlay dht11 gpiopin=4 it works unreliably.

dmesg:
[2024-11-03 23:35:32+01] dht11 dht11@4: Only 78 signal edges detected
[2024-11-03 23:35:33+01] dht11 dht11@4: Don't know how to decode data: 39 0 22 1
[2024-11-03 23:37:26+01] dht11 dht11@4: Don't know how to decode data: 39 0 22 1
[2024-11-03 23:37:29+01] dht11 dht11@4: Don't know how to decode data: 40 0 21 8
[2024-11-03 23:37:29+01] dht11 dht11@4: Don't know how to decode data: 148 0 10 132

The /sys files:
in_humidityrelative_input
in_temp_input
sometimes give only Input/output error,
sometimes read good, but without decimals like:
41000
21000

It seems that dtoverlay can't read nothing more than first two numbers from the bus.

[pelwell]

sometimes read good, but without decimals

That's expected behaviour - the value is in milli-percent, 41000 means 41%

It seems that dtoverlay can't read nothing more than first two numbers from the bus.

It's not the overlay doing the reading, it's the dht11 driver - the overlay just activates the driver. The driver itself is pure upstream code - we've not changed it at all.

[pelwell]

The datasheet for the DHT11 could be clearer IMO, but I think the bytes between the main values should be treated as tenths of a percent or tenths of a degree. However, the driver expects them to be zeroes. Perhaps greater accuracy was added later, and perhaps only to the temperature value.

There's a PR (#6454) that tries to interpret them as decimals, accepting any value between 0 and 9. If the values returned do happen to be zeroes then this patch will have no effect.

If you want to try a test build, wait about 40 minutes for the auto-builds to complete, then run sudo rpi-update pulls/6454, making sure to back up any important data first.

[pelwell]

The trial builds are good to go now.

[VDV2000]

If you want to try a test build, wait about 40 minutes for the auto-builds to complete, then run sudo rpi-update pulls/6454, making sure to back up any important data first.

I've performed the update and it's working without errors now. Device tree files are readable in short interval:

$ while true; do cat "/sys/bus/iio/devices/iio:device1/in_temp_input"; sleep 1; done
24200
24100
24100
24300
$ cat /sys/bus/iio/devices/iio:device1/in_humidityrelative_input
35000

The DHT11 is made by ASAIR and it seems that it is enhanced variants of original DHT11 because in the documentation and on the body of the device is written temp range -20 - 60°C (not from 0 degrees), thus another question is how and if it will work properly on subzero temperatures. I'm going to freeze something in fridge and get the results.
Update:

[2024-11-07 14:24:10+01] dht11 dht11@4: Don't know how to decode data: 59 0 3 3                                                                                                                                      
[2024-11-07 14:24:43+01] dht11 dht11@4: Don't know how to decode data: 64 0 2 4                                                                                                                                      
[2024-11-07 14:24:59+01] dht11 dht11@4: Don't know how to decode data: 64 0 2 3                                                                                                                                      
[2024-11-07 14:24:59+01] dht11 dht11@4: Don't know how to decode data: 160 0 1 1                                                                                                                                     
[2024-11-07 14:25:49+01] dht11 dht11@4: Don't know how to decode data: 69 0 1 4                                                                                                                                      
[2024-11-07 14:26:06+01] dht11 dht11@4: Don't know how to decode data: 70 0 1 1                                                                                                                                      
[2024-11-07 14:26:23+01] dht11 dht11@4: Don't know how to decode data: 72 0 0 7                                                                                                                                      
...
[2024-11-07 14:27:12+01] dht11 dht11@4: Don't know how to decode data: 75 0 0 4                                                                                                                                      
[2024-11-07 14:27:29+01] dht11 dht11@4: Don't know how to decode data: 75 0 0 3                                                                                                                                      
[2024-11-07 14:27:46+01] dht11 dht11@4: Don't know how to decode data: 76 0 0 2                                                                                                                                      
[2024-11-07 14:28:02+01] dht11 dht11@4: Don't know how to decode data: 77 0 0 1                                                                                                                                      
[2024-11-07 14:29:09+01] dht11 dht11@4: Don't know how to decode data: 79 0 1 129                                                                                                                                    
[2024-11-07 14:29:25+01] dht11 dht11@4: Don't know how to decode data: 80 0 1 129                                                                                                                                    
[2024-11-07 14:29:43+01] dht11 dht11@4: Don't know how to decode data: 81 0 1 130

It definitely looks like the subzero temps are counted from 1 129 (-0.1), 1 130 (-0.2) and up (I have returned to mainline kernel because the dev kernel have no headers available which I need to compile other things).
BTW: I compared it with values from BMP280 and with my meteo station and it seems that the readings from this DHT11 is more precise than from BMP because it was needed to correct by offset about 1.2˚C degrees, while DHT values was as nearly same as from the meteo station. Both sensors were near to each other at the same place. It was rather suprising...