Using ESP32 Diagnostic Logs Provider

This document outlines how to integrate the ESP32 Diagnostic Logs Provider into your Matter application.

Configuration Options

Enable the following configuration options to use the ESP32 Diagnostic Logs Provider:

CONFIG_ENABLE_ESP_DIAGNOSTICS=y

This option enables the diagnostic logs storage functionality. By default, this setting enables both traces and metrics.

If you want to enable only metrics or only traces, configure the following options accordingly:

To enable only metrics:

CONFIG_ENABLE_ESP_DIAGNOSTIC_TRACES=n
CONFIG_ENABLE_ESP_DIAGNOSTIC_METRICS=y

To enable only traces:

CONFIG_ENABLE_ESP_DIAGNOSTIC_TRACES=y
CONFIG_ENABLE_ESP_DIAGNOSTIC_METRICS=n

After modifying the configuration options, make sure to perform a clean build to ensure the changes are applied correctly:

idf.py fullclean
idf.py build

Implementation Reference

The ESP32 diagnostic logs provider delegate implementation is available at:

examples/platform/esp32/diagnostics/diagnostic-logs-provider-delegate-impl.h
examples/platform/esp32/diagnostics/diagnostic-logs-provider-delegate-impl.cpp

These files contain the implementation of the LogProvider class which implements the DiagnosticLogsProviderDelegate interface.

Integration Steps

1. Include the Header File

Add the diagnostic logs provider delegate header to your application:

#include <diagnostic-logs-provider-delegate-impl.h>

2. Define Buffer Storage

Define buffers to store and retrieve diagnostic data:

#ifdef CONFIG_ENABLE_ESP_DIAGNOSTICS
static uint8_t retrievalBuffer[CONFIG_RETRIEVAL_BUFFER_SIZE]; // Buffer for retrieving diagnostics
static uint8_t endUserBuffer[CONFIG_END_USER_BUFFER_SIZE];    // Buffer for storing diagnostics

using namespace chip::app::Clusters::DiagnosticLogs;
#endif // CONFIG_ENABLE_ESP_DIAGNOSTICS

The buffer sizes can be configured through Kconfig options:

• CONFIG_RETRIEVAL_BUFFER_SIZE: Size of the buffer used for retrieving diagnostic data inside the diagnostic-logs-provider-delegate.
• CONFIG_END_USER_BUFFER_SIZE: Size of the buffer used to store diagnostic data in esp32_diagnostics backend.

3. Initialize the Log Provider

Implement the diagnostic logs cluster initialization callback:

#ifdef CONFIG_ENABLE_ESP_DIAGNOSTICS
void emberAfDiagnosticLogsClusterInitCallback(chip::EndpointId endpoint)
{
    auto & logProvider = LogProvider::GetInstance();
    logProvider.Init(endUserBuffer, CONFIG_END_USER_BUFFER_SIZE, retrievalBuffer, CONFIG_RETRIEVAL_BUFFER_SIZE);
    DiagnosticLogsServer::Instance().SetDiagnosticLogsProviderDelegate(endpoint, &logProvider);
}
#endif // CONFIG_ENABLE_ESP_DIAGNOSTICS

This callback initializes the log provider with the configured buffers and sets it as the delegate for the DiagnosticLogs cluster.

4. Include Diagnostic Logs Server

Make sure to include the diagnostic logs server in your application:

#include <app/clusters/diagnostic-logs-server/diagnostic-logs-server.h>

Additional Configuration

Crash Logs Support

To enable crash logs retrieval, configure the ESP32 core dump functionality:

CONFIG_ESP_COREDUMP_ENABLE_TO_FLASH=y
CONFIG_ESP_COREDUMP_DATA_FORMAT_ELF=y

Example Integration

The examples/temperature-measurement-app/esp32/ demonstrates proper integration: For more information related to the retrieval of diagnostics through diagnosticlogs cluster refer readme.md in app folder.

Important Notes

• The diagnostic logs provider must be explicitly enabled through the CONFIG_ENABLE_ESP_DIAGNOSTICS option
• Buffer sizes should be adjusted based on your application's needs
• The provider supports end-user support logs and crash logs (when configured)