[Thermostat cluster] - OccupiedCoolingSetpoint writes outside limits do not return CONSTRAINT_ERROR (CON-1947)

[aggaddam]

Describe the bug
According to the Matter Thermostat cluster specification, writes to OccupiedCoolingSetpoint outside the range
MinCoolSetpointLimit ≤ OccupiedCoolingSetpoint ≤ MaxCoolSetpointLimit must be rejected with CONSTRAINT_ERROR.

AbsMinCoolSetpointLimit = 1600
AbsMaxCoolSetpointLimit = 3200
MinCoolSetpointLimit = 1600
MaxCoolSetpointLimit = 3000

However, on the ESP Matter implementation, writes outside this range return SUCCESS.

Observed behavior

./chip-tool thermostat write occupied-cooling-setpoint 3400 1 1  → SUCCESS
./chip-tool thermostat write occupied-cooling-setpoint 3200 1 1  → SUCCESS
./chip-tool thermostat write occupied-cooling-setpoint 1500 1 1  → SUCCESS
./chip-tool thermostat write occupied-cooling-setpoint 1400 1 1  → SUCCESS

Expected behavior:
Writes < MinCoolSetpointLimit (1600) or > MaxCoolSetpointLimit (3000) Should be rejected with:
IM Status = CONSTRAINT_ERROR (0x87), and the attribute value should remain unchanged

Environment

• ESP-Matter Commit Id: 75775c0 (release/v1.5)

• ESP-IDF Commit Id: 4c2820d377d1375e787bcef612f0c32c1427d183 (v5.4.1)

• SoC (eg: ESP32 or ESP32-C3): ESP32-C3

• Device Logs (Please attach the log file):

• Host Machine OS: Ubuntu

• Host Machine Python version:

• Commissioner app and versions if present: chip-tool

• Home hub app and versions if present: NA

• Commissioner's logs if present: NA

Any additional details
...

[jadhavrohit924]

@aggaddam I see that the commands getting success as a response, however the device does not allow to update value out of range. If you read the value of occupied-cooling-setpoint after setting out of bounds value, you will know. We are looking into why we get success as response.

[aggaddam]

@jadhavrohit924 Yes, Commands are getting success as a response, however attributes values won't get updated in case of out-of-bound values.

As per spec, the command response should be a constraint error instead of success response. Please refer to Test Case - TC-TSTAT-2.2

[jadhavrohit924]

Can you please try attached patch?

thermostat.patch

[aggaddam]

@jadhavrohit924 Can you share the full file for esp_matter_data_model_provider.cpp instead patch file ?

[jadhavrohit924]

@aggaddam Here you go!!

esp_matter_data_model_provider.cpp

[aggaddam]

@jadhavrohit924 Can you share esp_matter_data_model_provider.cpp file based on this commit d42e9d6 from release/v1.5 branch? Getting build errors with your file integrated

ributeRequest&, chip::app::AttributeValueDecoder&)':
/home/agaddam/MyFolder/esp_idf_v5_4_1/li_ac_matter_fw/targets/simplifi/AC/li_esp32matter_sdk/esp-matter/components/esp_matter/data_model_provider/esp_matter_data_model_provider.cpp:525:24: error: 'execute_callback' is not a member of 'esp_matter::attribute'; did you mean 'set_callback'?
525 | if (attribute::execute_callback(attribute::PRE_UPDATE, request.path.mEndpointId,
| ^~~~~~~~~~~~~~~~
| set_callback
/home/agaddam/MyFolder/esp_idf_v5_4_1/li_ac_matter_fw/targets/simplifi/AC/li_esp32matter_sdk/esp-matter/components/esp_matter/data_model_provider/esp_matter_data_model_provider.cpp: In lambda function:
/home/agaddam/MyFolder/esp_idf_v5_4_1/li_ac_matter_fw/targets/simplifi/AC/li_esp32matter_sdk/esp-matter/components/esp_matter/data_model_provider/esp_matter_data_model_provider.cpp:535:20: error: 'execute_callback' is not a member of 'esp_matter::attribute'; did you mean 'set_callback'?
535 | attribute::execute_callback(attribute::POST_UPDATE, request.path.mEndpointId,
| ^~~~~~~~~~~~~~~~
| set_callback
ninja: build stopped: subcommand failed.
ninja failed with exit code 1, output of the command is in the /home/agaddam/MyFolder/esp_idf_v5_4_1/li_ac_matter_fw/targets/simplifi/AC/build/log/idf_py_stderr_output_36526 and /home/agaddam/MyFolder/esp_idf_v5_4_1/li_ac_matter_fw/targets/simplifi/AC/build/log/idf_py_stdout_output_36526

[jadhavrohit924]

@aggaddam We have not created a version for release/v1.5 yet. We will first fix it on the main branch then backport it. That's why I shared you a patch at the first place, so that you can apply that and you will be unblocked.

[aggaddam]

@jadhavrohit924 Let me know If these changes are present in main branch, i will test out with main branch. Adding patch won't help directly as there will be dependencies...

[jadhavrohit924]

@aggaddam Fix is merged on main branch with 864b07f

[shubhamdp]

@aggaddam the fix was also cherry-picked to release/v1.5 in 53c1d87

[shubhamdp]

Closing since the fix is in the tree. Please re-open if required.

[aggaddam]

@shubhamdp Fix is working ok with the latest commit (e5257b6) from release/v1.5 branch