- Zephyr port: added a fix for upcoming Zephyr 3.2 compatibility, thanks @nordicjm for the fix!
- Added default config header for PSoC 6 port ports/cypress/psoc6/psoc6_default_config.h so user doesn't have to create it
- ModusToolbox:tm: Software
port updates
- Added heartbeat metrics for heap and Wi-Fi performance tracking when using the default port for CAT1A (PSoC:tm: 6). See ports/cypress/psoc6/memfault_platform_core.c for more details
- Fixed reboot reason reported when PSoC 6 is fully reset to report "Power On Reset" instead of "Unknown"
- Zephyr port updates
- Memfault logs (eg
MEMFAULT_LOG_DEBUG()etc) are now routed to the Zephyr logging infrastructure. The typical set of Kconfig options for Memfault logs are available (CONFIG_MEMFAULT_LOG_LEVEL_WRNetc). See details in "Breaking Changes" below for enabling logs in your project. - Added a new Kconfig option,
MEMFAULT_ZEPHYR_FATAL_HANDLER, which can be used to disable the Zephyr fault handler print facilities.
- Memfault logs (eg
-
Users will no longer see internal Memfault log output by default, but will have to enable it explicitly to see the output:
# enable LOG CONFIG_LOG=y # not required- enabling the Memfault logging component enables including the # log buffer in coredumps CONFIG_MEMFAULT_LOGGING_ENABLE=y # if on pre-v3.1.0 zephyr, you can choose either the default LOG v1 # implementation, or select a LOG2 mode to enable LOG2. on zephyr 3.1.0+, LOG # v1 is removed and LOG v2 is now the only log implementation # CONFIG_LOG2_MODE_DEFERRED=y # make sure to select a log backend to see the output CONFIG_LOG_BACKEND_UART=y
The log statements affected by this change are likely only the internal Memfault SDK logs (
MEMFAULT_LOG_DEBUG()etc), unless those macros are used in the user application. -
Removed support for Zephyr LTS release 1.14 as it was superseded by LTS V2 almost a year ago now. A project using this release of Zephyr must target a memfault-firmware-sdk release less than 0.32.0.
-
More logically grouped Kconfig settings in Zephyr example app's prj.conf
-
Fixed a few typos in particle port documentation
-
Simplified compilation steps for the nRF91 sample test app when compiling with older releases of the nRF Connect SDK and refreshed the example to target the v2.0.2 release by default
-
Updated default demo CLI commands to better align with our suggested integration test commands. The default set now looks like this:
mflt> help clear_core: Clear an existing coredump drain_chunks: Flushes queued Memfault data. To upload data see https://mflt.io/posting-chunks-with-gdb export: Export base64-encoded chunks. To upload data see https://mflt.io/chunk-data-export get_core: Get coredump info get_device_info: Get device info test_assert: Trigger memfault assert test_busfault: Trigger a busfault test_hardfault: Trigger a hardfault test_memmanage: Trigger a memory management fault test_usagefault: Trigger a usage fault test_log: Writes test logs to log buffer test_log_capture: Trigger capture of current log buffer contents test_reboot: Force system reset and track it with a trace event test_trace: Capture an example trace event help: Lists all commands
-
Zephyr port: enable proper backtraces for Zephyr
__ASSERT()macro on aarch32/cortex_m. Prior to this fix, crashes from__ASSERT()triggering would show an incorrect PC/LR for the active thread. -
Support for pre-release nRF Connect SDK v2.0.99 and Zephyr > v3.1:
- Upcoming nRF Connect SDK and Zephyr releases removed logging v1. Add build support for these changes, and keep backwards compatibility for previous nRF Connect SDK/Zephyr releases
- Correct an issue in the Memfault logging v2 backend, when the system was
invoked from an ISR context. This could happen due to a recent change, in
Memfault SDK v0.31.1, where the Zephyr fatal informational logs were output
from
memfault_platform_reboot()by default. It did not impact the collected backtrace, but it would show a nuisance__ASSERT()in the console output, ifCONFIG_ASSERT=y.
- Fix a compilation issue in the Dialog example app from the removal of
memfault_demo_cli_cmd_print_chunk()in Memfault SDK release v0.31.4.
-
ESP32 port: add new Kconfig option,
CONFIG_MEMFAULT_AUTOMATIC_INIT, that can be explicitly set tonto skip automatically initializing the Memfault SDK on boot. This can be useful if Memfault SDK initialization needs to be deferred to application start. -
Zephyr port: add Kconfig options,
CONFIG_MEMFAULT_INIT_PRIORITY/CONFIG_MEMFAULT_INIT_LEVEL_POST_KERNELfor controlling the Memfault SDK initialization level and priority. This can be useful when needing Memfault to initialize earlier in the system startup sequence, for example for diagnosing crashes in an early driver initialization. -
Partial support, still in progress, for NRF Connect SDK + Zephyr v3.1:
- Remove reference to the now-removed Kconfig symbol,
NET_SOCKETS_OFFLOAD_TLSto enable building without warnings. NOTE: if mbedtls is enabled (CONFIG_MBEDTLS=y), but is not being used for HTTP transfers (eg, mbedtls is used for security functions, but the device does not use HTTP for transferring data), it may be necessary to explicitly setCONFIG_MEMFAULT_HTTP_USES_MBEDTLS=n.
- Remove reference to the now-removed Kconfig symbol,
-
Zephyr port: remove an unused header file,
ports/zephyr/common/memfault_zephyr_http.h -
Remove
memfault_demo_cli_cmd_print_chunk()demo function.memfault_data_export_dump_chunks()can be used instead, which is intended to be used with the "Chunks Debug" UI in the Memfault web application- see here for more details
- Support Zephyr v3.1+ by conditionally compiling out Logger v1 code, thanks to @tejlmand for the patch!
- Fixed a 🐛 in the Zephyr port HTTP implementation, where a socket file descriptor was leaked. This caused every HTTP operation after the first to fail on Zephyr platforms. Thanks to @rerickson1 for the fix!
- Added an update to improve the quality of stack traces when using
MEMFAULT_ASSERTwith the TI ARM Clang Compiler
- Enable the Zephyr fault handler (including console fault prints) after
Memfault handler runs. Can be disabled by implementing
memfault_platform_reboot(). See details in ports/zephyr/include/memfault/ports/zephyr/coredump.h
- Fixed compiler error in nRF91 sample test app when compiling with the nRF Connect SDK v2.0.0 release
- Added reference port for CAT1A (PSoC:tm: 6) based MCUs using the ModusToolbox:tm: Software stack. For more details see ports/cypress/psoc6 directory.
- Added a convenience utility function for posting chunks using the Memfault
http client. See
memfault_http_client_post_chunkfor more details!
- Fixed compiler error in nRF91 sample test app when compiling with the nRF Connect SDK 1.8 release
- ESP-IDF: add Memfault Compact Log example integration to the
examples/esp32project
- ESP-IDF: Fix backtraces when using ESP-IDF v4.4+
- nRF-Connect SDK: enable the Kconfig flag
MEMFAULT_NRF_CONNECT_SDKby default when targeting the nrf52 + nrf53 series SOCs (previously only enabled by default for nrf91 series)
Added clarifications around licensing in ports and examples folders. See README for more details.
- minor updates to
scripts/eclipse_patch.pyto support NXP's MCUXpresso IDE
- Particle's Device OS port improvements:
- A user initiated reboot will now be recorded as a User Shutdown instead of a Low Power reset
- A custom hardware_version can now be specified using the
hardware_versionargument when initializing the Memfault library - Default hardware version now uses the
PLATFORM_NAMEmacro instead ofPRODUCT_SERIESmacro
- Zephyr port improvements
- Exposed lower level APIs to Memfault's HTTP post implementation to allow
easier custom handling. See
ports/zephyr/include/memfault/ports/zephyr/http.hfor more details
- Exposed lower level APIs to Memfault's HTTP post implementation to allow
easier custom handling. See
- Misc README documentation improvements
- Fix a build regression on nRF Connect SDK v1.2 caused by the new Kconfig flag
CONFIG_MEMFAULT_HTTP_USES_MBEDTLS
- Fix stack selection when in ISR context in some Zephyr versions
- Fix a build error when building Zephyr with
CONFIG_NORDIC_SECURITY_BACKENDenabled. New Kconfig flagCONFIG_MEMFAULT_HTTP_USES_MBEDTLScan be used to manually control this configuration if necessary (default should be set correctly in most cases)
- Fix CI unit test build error from older version of gcc
- Added a Task Watchdog optional module. This can be used to monitor and trigger a fault in the case of a task or thread that becomes stuck. See information in components/include/memfault/core/task_watchdog.h for how to configure and use the module
- Fix compilation when building for a Zephyr target that does not have the
CONFIG_ARM_MPUflag enabled - Fix compilation errors to enable compatibility with Zephyr v3.0.0
- Updated Memfault Diagnostic GATT Service (MDS) based on feedback. This service can be used to transparently forward data collected by the SDK to a Bluetooth Low Energy gateway and proxied to the cloud. See ports/include/memfault/ports/ble/mds.h
- Updated Mbed OS invoke commands to be more resilient against python package conflicts
- If your project is based on Zephyr < 2.6, you now need to explicitly set
CONFIG_OPENOCD_SUPPORT=yin yourprj.conf
- Added a port to Particle's Device OS. More details can be found in
ports/particle/README.md. - Added several more
reboot reason options:
kMfltRebootReason_KernelPanicfor explicitly tracking fatal resets from within a OS or RTOSkMfltRebootReason_FirmwareUpdateErrorfor explicitly tracking resets due to firmware update failures or rollbacks
- Added a convenience utility function for base64 encoding data in place. See
memfault_base64_encode_inplacefor more details! - Fixed compiler error in ESP-IDF port when compiling for ESP32-S2 targets
- Added configuration option,
MEMFAULT_COREDUMP_INCLUDE_BUILD_ID, which can be used to disable storing the Build Id in a coredump. - Fixed stale link in Mbed example app
README. - Added utility script that can be used to "arduino-ify" the code in this repo.
- Fixed linter errors in python scripts after addition of flake8-bugbear linter in CI.
- Fixed compiler error in nRF91 sample test app when compiling with the nRF Connect SDK 1.2 release
- Updated nRF91 sample test app to be compatible with nRF Connect SDK 1.9 release
- Updated python scripts to be compatible with new flake8 linter options
- Updated where
lcovis sourced from when running unit tests in CI
- Add an optional override flag to control the name used for Zephyr data
regions- see
ports/zephyr/common/memfault_zephyr_ram_regions.c. Only needed for unusual Zephyr + Memfault configurations prior to Zephyr v2.7 (for example, nRF Connect SDK v1.7.1 with Memfault SDK v0.27.3+) - Fix the STM32F7xx reboot reason port to correctly account for the internally wired Pin Reset
- Fix a function prototype mismatch in the STM32L4 flash port (thanks to @schultetwin for reporting this in #22!)
- Add support for setting string metrics, see
memfault_metrics_heartbeat_set_string()incomponents/include/memfault/metrics/metrics.hfor the new API. See the Memfault Docs for general information on using the metrics API. - Updated
memfault_metrics_heartbeat_debug_print()to also print current timer values, instead of showing0. Seecomponents/include/memfault/metrics/metrics.hfor details
- Update the STM32 QP/C example (
examples/qp) to compile and run correctly now - Add intructions for exercising Memfault OTA in the ESP32 example, see the
"Testing OTA" section in
examples/esp32/README.md - Update the Memfault HTTP client to URL-encode query params when checking for
OTA updates (in the case of Device properties containing reserved characters,
eg
+). Update the ESP port to check for reserved characters in query params and emit an error - Fix an outdated comment in
cmake/Memfault.cmake, as reported in issue #21 (thank you @C47D !)
- Update Python tests to use a mocked-out gdb instance
- Fix a build error for the Nordic Connect SDK v1.7.99 development version
- Correct an error in header file include order in the Mynewt port
- Update the esp32 and zephyr examples to use
1.0.0-devinstead of1.0.0+<6 digits of build id>for the version specifier. Build id is no longer required for symbol file reconciliation and the+character is a reserved character for URI schemes; this impacted OTA release requests. See this document for Memfault's recommended versioning strategy - Add a reboot reason port for the STM32F7xx family.
- Re-run python
blackandisortformatters on python code
- The Mynewt integration now supports the Memfault demo shell via
mflt_shell_init(), see the Mynewt port README.md for details. Huge thanks to @t3zeng for providing this implementation! - Add support for ESP-IDF v4.3.1. This update should fix the bootlooping issue seen when using the port with v4.3.1+ of ESP-IDF.
- Add support for
LOG2deferred mode on zephyr. This should fix bootloops when enablingLOG2. - Fix flipped args passed from
MEMFAULT_ASSERT_RECORD()toMEMFAULT_ASSERT_EXTRA_AND_REASON()(regression in v0.23.0). This affected the_extraadditional context value passed via this macro. - Fix a typo in
ports/esp_idf/memfault/common/memfault_platform_http_client.cwhich caused the OTA example to always return "OTA Update Available" when the current version is already the latest.
- Updated list of sample apps in (
examples/README.md)[examples/README.md]
- Extended mynewt RTOS port to capture coredumps via a
os_coredump_cbimplementation whenMEMFAULT_COREDUMP_CB=1
- Fixed a few compiler warnings emitted when compiling with clang
- Improved documentation for coredump config with nRF5 SDK
- Added support for using compact logs with the Memfault log subsystem.
- Added port for mynewt RTOS to Memfault SDK. (Huge thanks to @t3zeng for the help here!) See sdk/embedded/ports/mynewt for more details.
- Added support for Zephyr 2.7 release.
- Fixed a missing symbol linker error for
memfault_fault_handlerthat could arise when compiling with-flto. - Fixed a compiler error in
memfault_fault_handling_arm.cthat arose when using certain versions of the Clang compiler. - Cleaned up python scripts after enabling additional PEP8 naming convention linters.
- Updated
modem_key_mgmt_existsAPI usage to be compatible with changes upcomming in nRF Connect SDK 1.8.
-
Added preview of the Memfault Diagnostic GATT Service (MDS). This service can be used to transparently forward data collected by the SDK to a Bluetooth Low Energy gateway and proxied to the cloud. See ports/include/memfault/ports/ble/mds.h for more details!
- Added reference port of MDS for the DA1469x SDK to ports/dialog/da1469x/memfault_diagnostic_service.c
-
Implemented utility python script which can be used for quickly adding a
memfault-firmware-sdkport to an eclipse project. For more details, runpython scripts/eclipse_patch.py --help
-
Added example project demonstrating integration of Memfault on Cypress' CY8CKIT-064S0S2-4343W running Amazon-FreeRTOS publishing data using an AWS IoT MQTT broker. For more details about how to run the Memfault example, see examples/cypress/CY8CKIT-064S0S2-4343W/README.md.
-
Fixed a compiler warning emitted when using TI's GCC Compiler as reported by @albertskog in issue #18
- Apply some suggestions emitted by
flake8-pytest-stylelinter to scripts/tests/test_memfault_gdb.py.
- Added a workaround to
event_storage.hto prevent compilation errors when using the unity test framework to generate mocks. - Updated
makefiles/MemfaultWorker.mkto usesortto guarantee a deterministic file list order irrespestive of make version. A consistent order is useful for reproducible builds. - Make use of
__has_include()in Zephy port to remove the requirement of always needing to creatememfault_platform_config.h,memfault_metrics_heartbeat_config.def, &memfault_trace_reason_user_config.deffor a build to compile. To force a compile failure instead when any of these files do not exist, a user can setCONFIG_MEMFAULT_USER_CONFIG_SILENT_FAIL=n
- The current version of the Memfault Firmware SDK can now be accessed
programmatically from the
memfault/version.h. - Improved HTTP util parser when dealing with malformed status codes
- Updated nRF91 sample test app to be compatible with nRF Connect SDK 1.6
- Added a new utility API,
memfault_packetizer_set_active_sources, to the data_packetizer module, which let's one control the exact data sources that will get packetized. See description in the header for more details - Port Improvements:
- NXP i.MX RT102x
- Applied suggestions from @elliot-wdtl for the Zephyr ports
(#15):
- Updated software watchdog port to make use of
MEMFAULT_SOFTWARE_WATCHDOGmacro
- Updated software watchdog port to make use of
- Applied suggestions from @ioannisg & @mbolivar-nordic in
(#14) to change
the KConfig options used to select
CONFIG_MEMFAULT_HTTP_ENABLE&CONFIG_MEMFAULT_ROOT_CERT_STORAGE_NRF9160_MODEM&CONFIG_MEMFAULT_NRF_CONNECT_SDKwhen using nRF91 targets.
- Added
exportcommand to the demo cli to better mirror our suggested integration test commands
- If you are were using a custom nRF91 based board config (i.e neither
BOARD_NRF9160DK_NRF9160NSorBOARD_THINGY91_NRF9160NS), the following KConfig options will now be enabled by default. The following can be added to yourprj.confto restore the original behavior:CONFIG_MEMFAULT_HTTP_ENABLE=nCONFIG_MEMFAULT_NRF_CONNECT_SDK=nCONFIG_MEMFAULT_ROOT_CERT_STORAGE_TLS_CREDENTIAL_STORAGE=y
- Added "compact log" support to trace events. When enabled, the format string
will be removed at compile time from calls to
MEMFAULT_TRACE_EVENT_WITH_LOGand an integer along with arguments will be serialized instead. The actual string will recovered and formatted when it arrives in the Memfault cloud. This leads to a massive reduction in space & bandwidth needed to send trace events. For more details about how to set up, check out this guide! - Fixed a
-Wshadowcompiler error that would arise inmemfault_coredump_regions_armv7.cwhenMEMFAULT_COLLECT_MPU_STATEwas enabled - Updated debug print utility in
memfault_coredump_storage_debug.cto guard against potentially printing an uninitialized string. - Removed unnecessary extra argument from
MEMFAULT_SOFTWARE_WATCHDOG
- If you were already using
MEMFAULT_SOFTWARE_WATCHDOG, you will need to update your call site invocations to remove the argument being passed. i.e
- MEMFAULT_SOFTWARE_WATCHDOG(0);
+ MEMFAULT_SOFTWARE_WATCHDOG();- Support for Dialog DA1469x chip family (Huge thanks to @iandmorris for the
help here!)
- Example eclipse project and more details about how to add the port to any DA1469x based project can be found here.
- Added a simple utility to track heap allocations. This can be used to more
easily debug what allocations led to out of memory situations with the
addition of only several hundred bytes to a coredump capture. For more
details, see
memfault/core/heap_stats.h- For FreeRTOS users, automatic tracking of heap allocations can be enabled with the Memfault port. To enable, see the "Heap Tracking" section in the README at ports/freertos/.
- Added new convenience utilities for asserting when a watchdog event is
detected,
MEMFAULT_SOFTWARE_WATCHDOG. This will result in the issue in the Memfault UI being classified for as a "Software Watchdog" instead of an "Assert" for easier classification. - Fixed a 🐛 in Zephyr port where cpu runtime metrics were never getting reset after a heartbeat was collected leading to always increasing runtime values getting reported.
- Improved support for running tests against different versions of
clang and gcc and enabled more address sanitizers such as
-fsanitize=undefined - Misc documentation edits
- Reduced code space utilized by metric subsystem by transitioning from a string representation of metric names to an enum representation.
- Updated
memfault_gdb.pyhelper script to use latest Memfault API for uploading symbol files. - Removed "DST Root CA X3" from the required Memfault root certificate list as there is no infrastructure that relies on it anymore.
- Updated PEM representation of all root certificates to include newlines after 64-character intervals to improve portability with various TLS stacks.
- Updated
fw_build_id.pyscript. The same script can now also be installed viapypi:pip install mflt_build_id - Various improvements to example app documentation
- Zephyr / nRF Connect SDK port:
- Made periodic upload a named choice,
MEMFAULT_HTTP_PERIODIC_UPLOAD_CONTEXT, so the default can be overriden from other Kconfig files. - Added prompt text for
MEMFAULT_HTTP_DEDICATED_WORKQUEUE_STACK_SIZE
- Made periodic upload a named choice,
- Zephyr / nRF Connect SDK port:
CONFIG_MEMFAULT_NRF_CONNECT_SDKis now only enabled by default whenCONFIG_MEMFAULT=yCONFIG_DEBUG_THREAD_INFO=yis now selected by default (in prep for deprecation ofCONFIG_OPENOCD_SUPPORT)- Added new option (
CONFIG_MEMFAULT_USER_CONFIG_ENABLE=y) which can be used to remove requirement of providing any user specific configuration. - Added two choices for periodic posting of memfault data:
CONFIG_MEMFAULT_HTTP_PERIODIC_UPLOAD_USE_SYSTEM_WORKQUEUE. This is the default and matches previous release behavior of posting data to Memfault from the system work queueCONFIG_MEMFAULT_HTTP_PERIODIC_UPLOAD_USE_DEDICATED_WORKQUEUEThis is a new option which can be used to post data from a dedicated worker queue. This can be useful if the network stack may block for extended periods of time which would stall other system work queue jobs from getting processed.
- The Zephyr & nRF Connect SDK ports will now only enable the
MEMFAULT_SHELLby default when the Zephyr shell is enabled.CONFIG_SHELL=ymust now be enabled explicitly in yourprj.conffor the Memfault Shell to be enabled.
- Updated
sMemfaultBuildIdStoragestructure to track the build id length used for event serialization and updatedfw_build_id.pyscript to extract information.
- Zephyr / nRF Connect SDK port:
- Replaced
MEMFAULT_DEFAULT_REBOOT_REASON_IMPLKconfig option withMEMFAULT_REBOOT_REASON_GET_CUSTOMand updated default configuration for the nRF Connect SDK. This fixes an issue resulting in the generic memfault_reboot_reason_get getting linked rather than the nRF Connect SDK port.
- Replaced
- Updated memfault_fault_handling_arm.c to work around a compiler bug when using 6.x of the GNU ARM Toolchain
- Port Improvements:
- SAML10/SAML11
- Updated esp-idf port to streamline integrations making use of the amazon-freertos
- Zephyr
- Added several Kconfig options for better control over information
collected in a coredump:
CONFIG_MEMFAULT_COREDUMP_COLLECT_DATA_REGIONSto enable/disable collection of.dataregionCONFIG_MEMFAULT_COREDUMP_COLLECT_BSS_REGIONSto enable/disable collection of.bssregionCONFIG_MEMFAULT_COREDUMP_STORAGE_CUSTOM=ycan be used to opt out of the default RAM backed coredump implementation and provide a custom one in the port.
- Added several Kconfig options for better control over information
collected in a coredump:
- Reduced instruction cycles required to update a heartbeat metric
- Events will now store an abbreviated build id when serialized
-
Added support for collecting additional register information when a Hardfault takes place when using the Zephyr port. This information will be decoded and displayed in the Memfault UI in the "Exceptions" tab.
-
Updated
buffered_coredump_storage.hto usememmovinstead ofmemcpysincedstandsrcbuffers may overlap when all of.bssis saved in a coredump capture. -
Added a new Kconfig option to the Zephyr port,
CONFIG_MEMFAULT_METRICS_EXTRA_DEFS_FILE=y, which causesmemfault_metrics_heartbeat_extra.defto be included in the metric definitions. This can be utilized by a third party consumer of Zephyr to more easily extend the default heartbeat metrics collected when using memfault.
- Updated
memfault_gdb.pyhelper script to use latest Memfault API for uploading symbol files.
- If you are using nRF Connect SDK / Zephyr port, the SDK
will now automatically be picked up as a Zephyr Module! You will need to make
two changes:
- Remove the
ZEPHYR_EXTRA_MODULESaddition from your projects CMakeLists.txt, i.e
--- a/your_application/CMakeLists.txt +++ b/your_application/CMakeLists.txt @@ -3,7 +3,6 @@ - list(APPEND ZEPHYR_EXTRA_MODULES $ENV{ZEPHYR_BASE}/../modules/memfault-firmware-sdk/ports/nrf-connect-sdk)
- Add
CONFIG_MEMFAULT_NRF_CONNECT_SDK=yto your projectsprj.conf
- Remove the
- Support for Dialog DA145xx chip family (Huge thanks to @iandmorris for the
help here!)
- GCC & Keil based demo application for the DA14531 & DA14585/DA14586 can be found here.
- Ports for applications using the DA145xx SDK can be found here.
- ESP32 port improvements
- Added example of periodically posting data to memfault via a background task.
- Added a new Kconfig option,
MEMFAULT_COREDUMP_USE_OTA_SLOT=ywhich can be used to save a coredump in an unused OTA slot rather than the default coredump partition. This can be useful in situations where Memfault is being integrated after a product has shipped and updating the partition table is no longer possible.
- Added
MEMFAULT_EVENT_STORAGE_NV_SUPPORT_ENABLED=0which can be used to disable dynamic configuration of non-volatile storage. Setting this flag when the non-volatile event storage API is not in use will save several hundred bytes of codespace. - Hardened memfault_http_parse_response() utility to parse HTTP responses with headers that exceed a length of 128 bytes
- Fixed a 🐛 in
memfault_log_save_preformatted()leading to invalid logs being reported when attempting to save log lines > 128 bytes. (thanks @alvarop for the report!) - Added a convenience API,
memfault_create_unique_version_string(), which can be used for easily appending a build id on the software version reported.
- Updates to demo cli:
MEMFAULT_DEMO_SHELL_RX_BUFFER_SIZEcan be used to shrink the maximum amount of bytes that can be buffered on a single line.- Made
memfault_demo_shell_commands.hpublic and moved it tomemfault/demo/shell_commands.hto facilitate easier overriding of the default set of commands used in a build.
- ESP32 Example App Updates:
- Added
exportcommand to demonstrate how data can be dumped via the console - Added Memfault Build ID to example app as a reference
- Added
- Fixed a 🐛 in
memfault_platform_sanitize_address_range()template example. - Refactored nRF5 example app to mirror integration steps listed in the latest integration guide
- Added a new configuration option,
MEMFAULT_PLATFORM_FAULT_HANDLER_CUSTOM, which can be used to explicitly disable the stubmemfault_platform_fault_handler()implementation - Improve the quality of backtrace recovery for asserts when using Arm Compiler
5 by removing use of noreturn function attribute for
memfault_fault_handling_assert()declaration.
- Added support for collecting ARMv7-M MPU regions as part of coredump
collection. To enable, set
MEMFAULT_COLLECT_MPU_STATE=1in yourmemfault_platform_config.h. Once enabled, the MPU will be automatically analyzed for configuration errors and results will be presented in the "MPU" tab in the Memfault UI for a coredump. - Added a new API,
memfault_log_trigger_collection, which can be used to "freeze" the current contents of the log buffer when unexpected behavior takes place on the device for upload to Memfault. The logs can then be uploaded to Memfault just like any other data and appear in the UI for a device. For more details about the Memfault log subsystem see https://mflt.io/logging
- Fixed compilation error when compiling the panics component against Cortex-M0+ with ARM Compiler 5
- Added several default heartbeat metrics to the Zephyr port around timer task stack usage and execution time.
- Refreshed esp32 example app README and updated instructions for the v3.3.5 esp-idf
- Added
test_log&trigger_logsCLI commands to nRF5 & Zephyr example applications to exercise new log collection functionality.
- Fixed a 🐛 in Zephyr port leading to a compilation error with nRF Connect
SDK when
CONFIG_DOWNLOAD_CLIENT=y&CONFIG_MEMFAULT_NRF_SHELL=y - Dialog DA1468x
QSPI coredump storage port
updates:
- default storage parition (
NVMS_LOG_PART) can be overriddenMEMFAULT_PLATFORM_COREDUMP_STORAGE_PARTITION - Max space used within partition can be limited using
MEMFAULT_PLATFORM_COREDUMP_STORAGE_MAX_SIZE_BYTES
- default storage parition (
- Updated zephyr example application and docs to be compatible with v2.5 release.
- Added
memfault_log_savestub to unit tests to facilitate easier testing of logging dependencies - Improved correctness of strategy used to capture
msp&pspregister in Cortex-M fault handler. - Added new convenience utility,
memfault_circular_buffer_read_with_callback(), to circular buffer api.
- Added new convenience APIs,
memfault_event_storage_bytes_used()&memfault_event_storage_bytes_free(), to event storage module. - Added a new configuration option,
MEMFAULT_ASSERT_HALT_IF_DEBUGGING_ENABLED. By default, it is off, but when enabled will causememfault_platform_halt_if_debugging()to be called prior to triggering the full coredump capture. - Fixed a type conversion compiler warning emitted by IAR and ARMCC in
resetreas_reboot_tracking.c. - Port Updates:
- Removed "Heartbeat triggered!" print when
memfault_metrics_heartbeat_debug_trigger()is called
- If you were using
ports/dialog/da1468x/reset_stat_reboot_tracking.c, thememfault_platform_reboot_tracking_boot()implementation from yourmemfault_platform_port.cfile can be removed and the one in the port can be picked up.
- Added a new convenience API,
memfault_device_info_dump()which can be used to pretty print the device information populated in thememfault_platform_get_device_info()dependency function. - Added
memfault_platform_sanitize_address_range(). This functrions is intended for use in yourmemfault_platform_coredump_get_regions()implementation when capturing regions that are not defined at compile time. - Fixed a 🐛 with
memfault_coredump_storage_debug_test_*which would generate a false positive test failure when the coredump storage area was not divisible by 16. - C++ header guards are now included in all headers in the ports/ directory for easier integration in mixed C/C++ environments
- Port Updates:
- Dialog DA1468x: Added patch which can be used to add GNU Build ID
- Added support for FreeRTOS 8 to the FreeRTOS port.
- STM32H7 family / STM32CubeH7:
- STM32WBxx family / STM32CubeWB:
- Improved configurability of RAM backed coredump storage port with new configuration options for to control where RAM is allocated, what memory regions are collected, and stack size to collect.
- Added additional comments to
ports/templatesdirectory to facilitate porting.
- Demo CLI Shell commands are now defined as weak symbols so they can be overriden with a custom set.
- If you were using the
ports/emlib/msc_coredump_storage.cport in your system, you must addMEMFAULT_PLATFORM_COREDUMP_STORAGE_USE_FLASH=1to yourmemfault_platform_config.h - If you were using
ports/panics/src/memfault_platform_ram_backed_coredump.c:memfault_platform_sanitize_address_range()must be implemented.- The default coredump storage RAM size was changed from 700 to 1024 bytes so
more information can be captured. The original size can be restored by
setting
MEMFAULT_PLATFORM_COREDUMP_STORAGE_RAM_SIZE 700in yourmemfault_platform_config.h
- Renamed
platformsfolder toexamplesto better capture that the folder contains "example" integrations of the Memfault SDK into hello world style apps from various SDKs - Port Updates:
- Zephyr: added a software watchdog port which can be used for capturing coredumps ahead of a hardware watchdog reset.
- EFR32: rich reboot reason info derived from EMU_RSTCAUSE register
- Updated nRF9160 Demo Application to be compatible with nRF Connect SDK 1.5
- Added support for capturing coredumps with GDB from GNU GCC 4.9 to
memfault_gdb.py
- Added support for automatically capturing logs with Zephyr when logging in
synchronous mode
CONFIG_LOG_IMMEDIATE=y - Unified nomenclature for references to the "Project Key" used to push data to Memfault
If you were linking any files from the platforms folder into your project, the
path needs to be updated to examples:
- ${MEMFAULT_FIRMWARE_SDK}/platforms/
+ ${MEMFAULT_FIRMWARE_SDK}/examples/- Reference platform API implementations for DA1468x:
- Fixed a 🐛 that led to a unit test failure in
test_coredump_storage_debugin some environments whereconstarrays were getting dynamically built on the stack at runtime. - Worked around a limitation in GNU GCC 4.9's extended ASM to fix a compiler bug
that would arise when compiling
memfault_fault_handling_arm.cfor Cortex-M0 targets. - Added a new
ports/templatesfolder that can be copy/pasted into a project and used as a starting point for a Memfault port!
- Improved documentation in README and components directories.
- Added coredump capture support to panics component for the AARCH64 architecture.
- Reference platform API implementations for the following MCUs/SDKs:
- STM32WBxx family / STM32CubeWB:
- nRF5 SDK
fw_build_id.pyscript improvements- script is now compatible with Python2 environments.
- Added new
--dump <chars>option to simplify extraction of build id in automation, i.e:python scripts/fw_build_id.py <ELF> --dump 7 3a3e81f
- Started work to enable automatic capture of logs in a coredump for the Zephyr & nRF Connect SDK.
- If you were linking
memfault_nrf5_coredump.cin your project:- the file has been split into
nrf5_coredump_regions.c(which defines the regions to collect in a coredump) &nrf5_coredump_storage.c(which implements the port for saving coredumps to internal flash). Both of these files must be added to your build system. - Linker script names for region to save coredump to and regions to collect
need to be updated:
- Update
__CoreStart&__MemfaultCoreStorageEndto__MemfaultCoreStorageStart&__MemfaultCoreStorageEndin linker script. - Update
__MfltCoredumpRamStart&__MfltCoredumpRamEndto__MemfaultCoredumpRamStart&__MemfaultCoredumpRamEnd
- Update
- the file has been split into
- The SDK now includes a central configuration header at
components/include/memfault/config.h. Platform overrides can be defined inmemfault_platform_config.h - Reference platform API implementations for the nRF5 SDK
- All headers for Memfault "components" can now be found under
components/include/memfault/. This simplifies adding components to build systems as only a single path is now needed!
- You must create the file
memfault_platform_config.hand add it to your include path. This file can be used in place of compiler defines to tune the SDK configurations settings. - If you are not using a Memfault
build system helper
- Remove all include paths at
${MEMFAULT_FIRMWARE_SDK}/components/${COMPONENT}/include - Add include path
${MEMFAULT_FIRMWARE_SDK}/components/includeto your build system
- Remove all include paths at
- If you were linking any of the nRF5 example app files in your project, the directory has changed:
- ${MEMFAULT_FIRMWARE_SDK}/nrf5/libraries/memfault/platform_reference_impl/memfault_platform_reboot_tracking.c
+ ${MEMFAULT_FIRMWARE_SDK}/ports/nrf5_sdk/resetreas_reboot_tracking.c
- ${MEMFAULT_FIRMWARE_SDK}/nrf5/libraries/memfault/platform_reference_impl/memfault_platform_coredump.c
+ ${MEMFAULT_FIRMWARE_SDK}/ports/nrf5_sdk/memfault_nrf5_coredump.c- ESP8266 Updates
- Added new Kconfig option which can be set via
make menuconfigand be used to disable the Memfault integration,CONFIG_MEMFAULT=n. - Fixed a 🐛 leading to a compilation error when
both
CONFIG_USING_ESP_CONSOLE=nandCONFIG_MEMFAULT_CLI_ENABLED=n
- Added new Kconfig option which can be set via
- Added default implementations for
MEMFAULT_GET_LR(),MEMFAULT_GET_PC(), andMEMFAULT_BREAKPOINT()tocompiler_gcc.hto facilitate compilations of the SDK against other architectures.
- Reference platform API implementations for the following MCUs/SDKs:
- nRF Connect SDK
- Added support for nRF Connect SDK v1.3.x
- Added support for FOTA with Memfault. See
memfault_fota_start()for more details
- Zephyr
- Added implementations for
memfault_platform_metrics_timer_boot()&memfault_platform_get_time_since_boot_ms()memfault dependencies to Zephyr port. A custom implementation can still be provided by settingMEMFAULT_METRICS_TIMER_CUSTOM=y - Metrics support is now enabled by default when
CONFIG_MEMFAULT=yand can be disabled by settingCONFIG_MEMFAULT_METRICS=n - Added support for periodically uploading Memfault data in the background.
This is off by default and can be enabled with the
CONFIG_MEMFAULT_HTTP_PERIODIC_UPLOAD=yoption
- Added implementations for
- nRF Connect SDK
- Fixed a 🐛 that could lead to an invalid coredump being sent to Memfault
when
memfault_packetizer_abort()was called after a coredump was partially sent.
- Added support for nRF Connect SDK v1.2.x and updated the integration guide accordingly.
- Added full support for the ESP8266 (Tensilica Xtensa LX106) MCU architecture! A step-by-step integration guide can be found here and the port to the ESP8266 RTOS SDK can be found here.
- Added a convenience header for picking up includes for all Memfault components. If you are using the build system helpers, this path will be picked up automatically.
#include "memfault/components.h"
// call to any Memfault API in the components folder- Fixed a 🐛 leading to Root CAs not get loaded correctly when using the nRF
Connect SDK port and the
MEMFAULT_ROOT_CERT_STORAGE_TLS_CREDENTIAL_STORAGE=yKconfig option. - Applied suggestions from @rerickson1 for the Zephyr and nRF Connect SDK ports:
CONFIG_MEMFAULT_METRICS=ycan now be used to compile the metrics component into the Zephyr and nRF Connect SDK ports.CONFIG_MEMFAULT=ymust now be specified to enable the Memfault integration.
- Refactored the nRF Connect SDK port to build directly on top of the zephyr port reducing code duplication and facilitate the rollout of additional features to both SDKs at the same time moving forward.
- If you are using a Memfault
build system helper and not using
the trace or metrics functionality, you will now need to create a
memfault_trace_reason_user_config.deformemfault_metrics_heartbeat_config.deffile, respectively, and add it to your include path. - When using the Zephyr port, the memfault integration must be enabled
explicitly by adding
CONFIG_MEMFAULT=yto yourprj.conf
- Reference platform API implementations for the following MCUs/SDKs:
- STM32F4 family / STM32CubeF4:
- STM32L4 family / STM32CubeL4
- nRF Connect SDK
- rich reboot reason info derived from PMU RESETREAS register
- refactored HTTP port to support multiple backing storage strategies for
root certificates. See
MEMFAULT_ROOT_CERT_STORAGE_*Kconfig options for more details - Added support for Memfault OTA downloads. See memfault_nrfconnect_port_ota_update() for more details
- Updated
memfault_freertos_ram_regions.cport to collect all TCBs and then stacks. This way, the state of all tasks can be recovered even if the coredump storage regin is filled while writing all the task stacks. - Reference platform API implementations for the following MCUs/SDKs:
- STM32F4 family / STM32CubeF4
- STM32L4 family / STM32CubeL4
- NXP's S32K1xx family / S32K1 SDL
- Silicon Lab's EFM/EFR family / v3.0 Gecko SDK
- Added several more
reboot reason options:
kMfltRebootReason_PinResetfor explicitly tracking external pin resets.kMfltRebootReason_SoftwareWatchdog&kMfltRebootReason_HardwareWatchdogfor easier disambiguation between watchdog resets where a coredump was captured versus ones where no software handler ran and hardware reset the device.kMfltRebootReason_ClockFailurefor explicit tracking of resets due to loss of a clock signal or PLL lock.kMfltRebootReason_Lockupfor explicit tracking of faults from within the Hardfault or NMI exceptions on ARM Cortex-M MCUs.
- Added a utility which can be used to verify a platform coredump storage implementation is working as corrected. For more details about how to use, see memfault_coredump_storage_debug.c.
- Added infrastructure to coredump collection in
panicscomponent to support ESP8266 (Tensilica Xtensa LX106) MCU architecture.
-
nRF Connect Updates:
- Updated port to support nRF Connect SDK v1.4.0
- Added port of HTTP client to post data to Memfault
- Added support for capturing state of all Zephyr tasks during a crash. This way the state of all threads can be seen in the Memfault UI when a crash is uploaded.
- Updated memfault_demo_app to use the nRF9160-DK
- Added notes to the step-by-step integration guide for the nRF9160.
- Added Memfault OTA support to esp-idf port. Updates can now be performed by
calling
memfault_esp_port_ota_update(). More details can be found inports/esp_idf/memfault/include/memfault/esp_port/http_client.h - The esp-idf port debug CLI can now easily be disabled by using the
MEMFAULT_CLI_ENABLED=nKconfig option. - Added FreeRTOS utility to facilitate collecting minimal set of RAM in a
coredump necessary to recover backtraces for all tasks. More details can be
found in
ports/freertos/include/memfault/ports/freertos_coredump.h - Previously, if the Memfault event storage buffer was out of space, a "storage out of space" error would be printed every time. Now, an error message is printed when the issue first happend and an info message is printed when space is free again.
- Added a reference software watchdog port for the STM32H7 series LPTIM
peripheral. Users of the STM32 HAL can now compile in the reference port and
the
MemfaultWatchdog_Handler. The handler will save a coredump so the full system state can be recovered when a watchdog takes place. More details can be found inports/include/memfault/ports/watchdog.h.
- A log can now be captured alongside a trace event by using a new API:
MEMFAULT_TRACE_EVENT_WITH_LOG(reason, ...). This can be useful to capture arbitrary diagnostic data with an error event or to capture critical error logs that you would like to be alerted on when they happen. For example:
// @file memfault_trace_reason_user_config.def
MEMFAULT_TRACE_REASON_DEFINE(Critical_Log)// @file your_platform_log_implementation.h
#include "memfault/core/trace_event.h"
#define YOUR_PLATFORM_LOG_CRITICAL(fmt, ....) \
MEMFAULT_TRACE_EVENT_WITH_LOG(Critical_Log, fmt, __VA_ARGS__)// @file your_platform_temperature_driver.c
void record_temperature(void) {
// ...
// erase flash to free up space
int rv = spi_flash_erase(...);
if (rv != 0) {
YOUR_PLATFORM_LOG_CRITICAL("Flash Erase Failure: rv=%d, spi_err=0x%x", spi_bus_get_status());
}
}- The error tracing facilities are now initialized automatically for the esp-idf
- Fixed a 🐛 where an erroneous size was reported from
memfault_coredump_storage_check_size()ifMEMFAULT_COREDUMP_COLLECT_LOG_REGIONS=1andmemfault_log_boot()had not yet been called
- ESP32 port improvements:
- The Memfault
metricscomponent is now included by default in the ESP32 port MEMFAULT_LOG_DEBUGmessages now print by default- Added a
heartbeat_dumpCLI command for easy viewing of current heartbeat metrics - Custom handling of collecting Memfault data can now easily be implemented in
the ESP32 port using the new
memfault_esp_port_data_available()&memfault_esp_port_get_chunk()APIS. This can be useful in scenarios where there are external MCUs forwarding Memfault chunks to the ESP32.
- The Memfault
- The platform port for the memfault log dependency can now be implemented by
macros (rather than the
memfault_platform_logdependency). Seecomponents/include/memfault/core/debug_log.hfor more details.
- If you were using the ESP32 port:
- Call to
memfault_metrics_boot()can now be removed - Custom implementations for
memfault_platform_metrics_timer_boot()&memfault_platform_get_time_since_boot_ms()can be removed as they are now provided as part of the port.
- Call to
- Coredumps will now be truncated (instead of failing to save completely) when
the memory regions requested take up more space than the platform storage
allocated for saving. A warning will also be displayed in the Memfault UI when
this happens. Regions are always read in the order returned from
memfault_platform_coredump_get_regions()so it is recommended to order this list from the most to least important regions to capture. - Updated FreeRTOS port to use static allocation APIs by default when the
configSUPPORT_STATIC_ALLOCATION=1configuration is used.
- Added several more
reboot reason options:
kMfltRebootReason_SoftwareReset&kMfltRebootReason_DeepSleep. - Extended ESP32 port to include integrations for reboot reason tracking and log collection.
- Apply missing check to unit test reported on Github
- Added a new convenience API,
memfault_coredump_storage_check_size(), to check that coredump storage is appropriately sized. - Fixed a 🐛 with heartbeat timers that would lead to an incorrect duration being reported if the timer was started and stopped within the same millisecond.
- Fixed an issue when using TI's compiler that could lead to the incorrect register state being captured during a fault.
- If you were not using the
error tracing functionality, you will need to
create the configuration file "memfault_trace_reason_user_config.def" and add
it to your include path. This removes the requirement to manually define
MEMFAULT_TRACE_REASON_USER_DEFS_FILEas part of the compiler flags.
- Add support for sending multiple events in a single chunk. This can be useful
for optimizing throughput or packing more data into a single transmission
unit. The behavior is disabled by default but can be enabled with the
MEMFAULT_EVENT_STORAGE_READ_BATCHING_ENABLEDcompiler flag. More details can be found in memfault_event_storage.c - Added convenience API,
memfault_build_id_get_string, for populating a buffer with a portion of the Memfault Build ID as a string. - Added default implementations of several Memfault SDK dependency functions when using FreeRTOS to ports/freertos.
- A status or error code (i.e bluetooth disconnect reason, errno value, etc) can
now be logged alongside a trace event by using a new API:
MEMFAULT_TRACE_EVENT_WITH_STATUS(reason, status_code).
- Added support for TI's ARM-CGT Compiler
- Removed dependency on NMI Exception Handler for
MEMFAULT_ASSERTs. Instead of pending an NMI exception, the assert path will now "trap" into the fault handler by executing audfinstruction. This unifies the fault handling paths within the SDK and leaves the NMI Handler free for other uses within the user's environment. - Added several more
reboot reason options:
kMfltRebootReason_PowerOnReset,kMfltRebootReason_BrownOutReset, &kMfltRebootReason_Nmi.
- Added utility to facilitate collection of the memory regions used by the
logging module as part of a
coredump. With this change, when the SDK is compiled with
MEMFAULT_COREDUMP_COLLECT_LOG_REGIONS=1, the logging region will automatically be collected as part of a coredump. Step-by-step details can also be found in the logging integration guide. - Added
MEMFAULT_METRICS_KEY_DEFINE_WITH_RANGE()which can be used for defining the minimum and maximum expected range for a heartbeat metric. This information is used by the Memfault cloud to better normalize the data when it is presented in the UI.
- If you are not using our CMake or Make
build system helpers and are using
the
panicscomponent, you will need to manually add the following file to your build system:$(MEMFAULT_SDK_ROOT)/components/panics/src/memfault_coredump_sdk_regions.c
- Added a port for projects using the nRF Connect SDK along with a step-by-step integration guide.
- Disabled optimizations for
memfault_data_export_chunk()to guarantee the GDB chunk test utility can always be used to post chunks using the data export API.
- Added memfault/core/data_export.h API to facilitate production and evaluation use cases where Memfault data is extracted over a log interface (i.e shell, uart console, log file, etc). See the header linked above or the integration guide for more details.
- Fixed a 🐛 that would cause the demo shell to get stuck if backspace chracters were entered while no other characters had been entered.
- Updated the GDB chunk test utility to automatically detect when the data export API is integrated and post-chunks to the cloud directly from GDB when the function is invoked.
- If you are not using our CMake or Make
build system helpers and want to make
use of the new data export API, you will need to manually add the following
files to your build system:
- Add:
$(MEMFAULT_SDK_ROOT)/components/core/src/memfault_data_export.c - Add:
$(MEMFAULT_SDK_ROOT)/components/util/src/memfault_base64.c
- Add:
- Updated code to support compilations with
-Wunused-paramater, GNU GCC's-Wformat-signedness, and Clang's-Wno-missing-prototypes&-Wno-missing-variable-declarations. - Updated unit test setup to compile with newly supported warnings treated as errors
- Misc utility additions including support for encoding floats and int64_t's in the cbor utility
- Add additional utilities to the http component to facilitate easier release management integration in environments with no pre-existing http stack.
- Add new cli command,
mflt get_latest_release, to the Zephyr demo application (tested on the STM32L4) to demonstrate querying the Memfault cloud for new firmware updates.
- Refactored
democomponent to make it easier to integrate an individual CLI commands into a project since some of the commands can be helpful for validating integrations. More details can be found in the README at components/demo/README.md.
- If you are using the "demo" component and are not making use our CMake or
Make build system helpers, you will
need to make the following changes:
- Update:
$(MEMFAULT_SDK_ROOT)/components/demo/src/{memfault_demo_cli.c => panics/memfault_demo_panics.c} - Update:
$(MEMFAULT_SDK_ROOT)/components/demo/src/{ => panics}/memfault_demo_cli_aux.c - Add:
$(MEMFAULT_SDK_ROOT)/components/demo/src/memfault_demo_core.c - Add:
$(MEMFAULT_SDK_ROOT)/components/demo/src/http/memfault_demo_http.c
- Update:
- If you are using the
httpcomponent, the following macro names changed:
-#define MEMFAULT_HTTP_GET_API_PORT()
-#define MEMFAULT_HTTP_GET_API_HOST()
+#define MEMFAULT_HTTP_GET_CHUNKS_API_PORT()
+#define MEMFAULT_HTTP_GET_CHUNKS_API_HOST()- Moved
reboot_tracking.htocorecomponent since it has no dependencies on anything from thepanicscomponent. This allows the reboot tracking to be more easily integrated in a standalone fashion. - Published a new guide detailing how to manage firmware updates using Memfault.
- Disabled optimizations for
memfault_fault_handling_assert(). This improves the recovery of local variables of frames in the backtrace when certain optimization levels are used. - Updated
memfault_sdk_assert.cto address a GCC warning (-Wpointer-to-int-cast) emitted when compiling the file for 64 bit architectures. - Misc README improvements.
- If you are already using reboot tracking in your system, you will need to update the following includes in your code:
-#include "memfault/panics/reboot_tracking.h"
-#include "memfault/panics/reboot_reason_types.h"
+#include "memfault/core/reboot_tracking.h"
+#include "memfault/panics/reboot_reason_types.h"-
If you are not using our CMake or Make build system helpers, you will need to update the path for the following files:
$(MEMFAULT_SDK_ROOT)/components/{panics => core}/src/memfault_ram_reboot_info_tracking.c$(MEMFAULT_SDK_ROOT)/components/{panics => core}/src/memfault_reboot_tracking_serializer.c
-
eMfltResetReasonwas renamed toeMemfaultRebootReason.
- Added native SDK support for tracking and generating a unique firmware build
id with any compiler! Quick integration steps can be found in
memfault/core/build_info.h.
It is very common, especially during development, to not change the firmware
version between editing and compiling the code. This will lead to issues when
recovering backtraces or symbol information because the debug information in
the symbol file may be out of sync with the actual binary. Tracking a build id
enables the Memfault cloud to identify and surface when this happens! You can
also make use of two new APIs:
memfault_build_info_dump()can be called on boot to display the build that is running. This can be a useful way to sanity check that your debugger succesfully flashed a new image.memfault_build_info_read()can be used to read the build id for your own use cases. For example you could append a portion of it to a debug version to make it unique.
- The CMake build system helper is now compatible with any 3.x version of CMake (previously required 3.6 or newer).
- The unique firmware build id is stored automatically as part of coredump collection.
- If you are not using our CMake or Make
build system helpers, you will need
to add the following files to your build:
$(MEMFAULT_SDK_ROOT)/components/core/src/memfault_build_id.c$(MEMFAULT_SDK_ROOT)/components/core/src/memfault_core_utils.c
- We also encourage you to add a unique build id to your build (several line code change).
- Added support for (optionally) storing events collected by the SDK to non-volatile storage mediums. This can be useful for devices which experience prolonged periods of no connectivity. To leverage the feature, an end user must implement the nonvolatile_event_storage platform API.
- Added an assert used internally by the SDK which makes it easier to debug API
misuse during bringup. The assert is enabled by default but can easily be
disabled or overriden. For more details see
memfault/core/sdk_assert.h. - Added a default implementation of
memfault_platform_halt_if_debugging()for Cortex-M targets. The function is defined as weak so a user can still define the function to override the default behavior.
- Updated
memfault install_chunk_handlerto work with older versions of the GNU Arm Toolchain.
- If you are not using our CMake or Make
build system helpers, you will need
to add
$(MEMFAULT_SDK_ROOT)/components/core/src/memfault_sdk_assert.cto your project.
- Moved
trace_event.htocorecomponent since it has no dependencies on anything from thepanicscomponent. This allows the trace event feature to be more easily integrated in a standalone fashion.
- If you are already using
MEMFAULT_TRACE_EVENT()in your project, you will need to update the include as follows:
-#include "memfault/panics/trace_event.h"
+#include "memfault/core/trace_event.h"- If you are not using our CMake or Make
build system helpers, you will need
to update the source path for
components/panics/src/memfault_trace_event.ctocomponents/core/src/memfault_trace_event.c
- Added a new GDB command which can be used to post packetized Memfault data directly from GDB to the Memfault cloud. This can be helpful as a way to quickly test data collection functionality while working on an integration of the SDK.
- The
captured_datefor an event can now be set by implementingmemfault_platform_time_get_current(). If the API is not implemented, thecaptured_datewill continue to be set based on the time the event was received by the memfault cloud.
- Added a reference implementation of
reboot reason tracking to the
NRF52 demo app
and a new
rebootCLI command to easily exercise it. - A
reset_reasoncan now optionally be provided as part ofsResetBootupInfo. This can be useful for scenarios where the reboot reason is known on bootup but could not be set prior to the device crashing. - A reboot reason event will now always be generated when
memfault_reboot_tracking_boot()is called even if no information about the reboot has been provided. In this scenario, the reset reason will bekMfltRebootReason_Unknown
- Updated a few Kconfig options in the Zephyr demo app to improve the ability to
compute stack high water marks (
CONFIG_INIT_STACKS=y) and determine if stacks has overflowed (CONFIG_MPU_STACK_GUARD=y).
device_serialis no longer encoded by default as part of events. Instead, thedevice_serialin an event is populated from the the unique device identifier used when posting the data to the chunks REST endpoint. This leads to ~20% reduction in the size of a typical event. Encodingdevice_serialas part of the event itself can still be enabled by adding-DMEMFAULT_EVENT_INCLUDE_DEVICE_SERIAL=1as a compilation flag but should not be necessary for a typical integration.
- Added
memfault_log_read()API to make it possible to use the module to cache logs in RAM and then flush them out to slower mediums, such as a UART console or Flash, from a background task.
- A pointer to the stack frame upon exception entry is now included in
sCoredumpCrashInfowhen memfault_platform_coredump_get_regions is invoked. This can (optionally) be used to configure regions collected based on the state or run platform specific cleanup based on the state. - Added Root Certificates needed for release downloads to
MEMFAULT_ROOT_CERTS_PEM.
- All sources that generate events now use the same utility function,
memfault_serializer_helper_encode_metadata()to encode common event data.
- Introduced a lightweight logging module. When used, the logs leading up to a crash will now be decoded and displayed from the Memfault Issue Details Web UI. For instructions about how to use the module in your project, check out log.h.
- The metrics component will now automatically collect the elapsed time and the number of unexpected reboots during a heartbeat interval. The Memfault cloud now uses this information to automatically compute and display the overall uptime of your fleet.
- Added a new document walking through the design of Memfault event serialization.
- Cleaned up
test_memfault_root_cert.cppand moved thebase64implementation within the unit test to theutilcomponent so it is easier to share the code elsewhere in the future. - Added
constto a few circular_buffer.h API signatures. - Misc code comment improvements.
- The function signature for
memfault_metrics_boot()changed as part of this update. If you are already using themetricscomponent, you will need to update the call accordingly. See the notes in metrics.h for more details. - If you are not using our CMake or Make
build system helpers, you will need
to add
$(MEMFAULT_SDK_ROOT)/components/core/src/memfault_log.cto your project. (Note that untilmemfault_log_boot()is called, all calls made to the logging module will be a no-op).
- Improve the
memfault_platform_coredump_storage_clear()NRF52 reference implementation for situations where the SoftDevice is enabled and there is a lot of Bluetooth activity. (In this scenario, NRF52 flash operations may need retries or take a while to complete). - Fixed compiler error that could arise with the metrics component when using Arm Compiler 5 due to multiply defined weak symbols.
- Add support for ESP32 (Tensilica Xtensa LX6 MCU) to the panics component.
- A step-by-step integration guide can be found here.
- A drop-in port for an existing v3.x or v4.x based ESP-IDF project can be found at ports/esp_idf.
- An example application exercising the memfault-firmware-sdk can be found here.
- If a software watchdog has been
implemented on a Cortex-M device,
MemfaultWatchdog_Handlercan now be registered as the Exception Handler to automatically collect a coredump. - For heartbeat metrics, instead of serializing the name of each metric, we now recover it from the debug information in the symbol file in the Memfault cloud. For a typical heartbeat this reduces the serialization size by more than 50% and results in a smaller footprint than other structured serialization alternatives such as Protobuf.
- Remove usage of
__has_includemacro for IAR compiler since not all versions fully support it.
- Add support for calling
MEMFAULT_TRACE_EVENT()from interrupts. Note: If you are not using our CMake or Make build system helpers, this change requires that you add$(MEMFAULT_SDK_ROOT)/components/core/src/arch_arm_cortex_m.cto your project. - Misc documentation improvements.
- Add support for compressing coredumps as they are sent using Run Length Encoding (RLE). More details can be found in memfault/core/data_source_rle.h.
- Update metrics component to support compilation with the IAR ARM C/C++ Compiler.
- Update Mbed OS 5 port to use
memfault_demo_shellinsteadmbed-client-cli, sincembed-client-cliis not part of the main Mbed OS 5 distribution. - Update nrf52 example application to only collect the active parts of the stack to reduce the overall size of the example coredump.
- Add a new API ("Trace Event") for tracking errors. This allows for tracing issues that are unexpected but are non-fatal (don't result in a device reboot). A step-by-step guide detailing how to use the feature can be found at: https://mflt.io/error-tracing
- Update GDB coredump collection script to work with the ESP32 (Tensilica Xtensa LX6 Architecture)
- Remove
__taskfrom IAR Cortex-M function handler declarations since it's not explicitly required and can lead to a compiler issue if the function prototype does not also use it. - Misc documentation and comment tweaks to make nomenclature more consistent
- Update panics component to support compilation with IAR ARM C/C++
Compiler. More details about integrating IAR can be found at
https://mflt.io/iar-tutorial. As part of the change
MEMFAULT_GET_LR()now takes an argument which is the location to store the LR to (void *lr = MEMFAULT_GET_LR()->void *lr;MEMFAULT_GET_LR(lr))
- Update the chunk protocol to encode CRC in last chunk of message instead of first. This allows the CRC to be computed incrementally and the underlying message to be read once (instead of twice). It also makes it easier to use the data packetizer in environments where reads from data sources need to be performed asynchronously. More details can be found at https://mflt.io/data-to-cloud
- Fixed a couple documentation links
- Guarantee that all accesses to the platform coredump storage region route
through
memfault_coredump_readwhile the system is running. - Scrub unused portion of out buffer provided to packetizer with a known pattern for easier debugging
- Add convenience API,
memfault_packetizer_get_chunk(), to data_packetizer module. - Add a new eMfltCoredumpRegionType,
MemoryWordAccessOnlywhich can be used to force the region to be read 32 bits at a time. This can be useful for accessing certain peripheral register ranges which are not byte addressable. - Automatically collect Exception / ISR state for Cortex-M based targets. NOTE
that the default config requires an additional ~150 bytes of coredump storage.
The feature can be disabled completely by adding
-DMEMFAULT_COLLECT_INTERRUPT_STATE=0to your compiler flags. More configuration options can be found in memfault_coredump_regions_armv7.c. - Improve documentation about integrating the SDK within a project in README
- Update Release note summary to use markdown headings for easier referencing.
- Update try script used to collect coredumps via GDB to also collect Exception/ISR register information.
- Add ARMv6-M fault handling port to panics component for MCUs such as the ARM Cortex-M0 and Cortex-M0+.
- Update panics component to support compilation with Arm Compiler 5.
- Updated Cortex-M fault handler in panics component to also collect
pspandmspregisters when the system crashes. This allows for the running thread backtrace to be more reliably recovered when a crash occurs from an ISR. - Added optional
MEMFAULT_EXC_HANDLER_...preprocessor defines to enable custom naming of exception handlers in the panics component. - Add port for Cortex-M based targets using Quantum Leaps' QP™/C & QP™/C++ real-time embedded framework. See ports/qp/README.md for more details.
- Add demo application running Quantum Leaps' QP™/C running on the STM32F407 discovery board. See examples/qp/README.md for more details.
- Added demo application and port for Arm Mbed OS 5 running on the STM32F429I-DISC1 evaluation board. See examples/mbed/README.md for more details.
- Changed
print_coretoprint_chunkin demo applications to better align with the Memfault nomenclature for data transfer.
- Expose root certificates used by Memfault CI in DER format for easier integration with TLS libraries which do not parse PEM formatted certificates.
- Add utilities to the http component for constructing Memfault cloud chunk endpoint POST requests to facilitate easier integration in environments with no pre-existing http stack.
- Add port for Cortex-M based targets in the Zephyr RTOS. Ports are available for the 1.14 Long Term Support Release as well as the 2.0 Release. See ports/zephyr/README.md for more details
- Add Zephyr demo application (tested on the STM32L4). See zephyr demo app directory for more details
- Release of metrics component. This API can easily be used to monitor device health over time (i.e connectivity, battery life, MCU resource utilization, hardware degradation, etc) and configure Alerts with the Memfault backend when things go astray. To get started, see this document
- Updated
memfault_platform_coredump_get_regions()to take an additional argument,crash_infowhich conveys information about the crash taking place (trace reason & stack pointer at time of crash). This allows platform ports to dynamically change the regions collected based on the crash if so desired. This will require an update that looks like the following to your port:
-const sMfltCoredumpRegion *memfault_platform_coredump_get_regions(size_t *num_regions) {
+const sMfltCoredumpRegion *memfault_platform_coredump_get_regions(
+ const sCoredumpCrashInfo *crash_info, size_t *num_regions) {- Added a new API,
memfault_coredump_storage_compute_size_required()which can be called on boot to sanity check that platform coredump storage is large enough to hold a coredump. For example:
sMfltCoredumpStorageInfo storage_info = { 0 };
memfault_platform_coredump_storage_get_info(&storage_info);
const size_t size_needed = memfault_coredump_storage_compute_size_required();
if (size_needed > storage_info.size) {
MEMFAULT_LOG_ERROR("Coredump storage too small. Got %d B, need %d B",
storage_info.size, size_needed);
}
MEMFAULT_ASSERT(size_needed <= storage_info.size);
- Added a convenience RAM backed reference port for coredump platform APIs. This can be used for persisting a coredump in RAM across a reset.
- Enhanced Reboot Tracking module within the panics component. Reboots that
don't result in coredumps can now be easily tracked (i.e brown outs, system
watchdogs, user initiated resets, resets due to firmware updates, etc). See
memfault/core/reboot_tracking.hfor more details and https://mflt.io/2QlOlgH for a step-by-step setup tutorial. - Added Event Storage module within the core component. This is a small RAM backed data store that queues up traces to be published to the Memfault cloud. To minimize the space needed and transport overhead, all events collected within the SDK are stored using the Concise Binary Object Representation (CBOR)
- Added helper makefile (
makefiles/MemfaultWorker.mk). A user of the SDK can include this makefile when using a make as their build system to easily collect the sources and include paths needed to build Memfault SDK components - Make unit tests public & add CI to external SDK
- Added firmware side support for the Memfault cloud chunk endpoint. This is
a sessionless endpoint that allows chunks of arbitrary size to be sent and
reassembled in the Memfault cloud. This transport can be used to publish any
data collected by the Memfault SDK. The data is read out on the SDK side by
calling the
memfault_packetizer_get_next()API indata_packetizer.h. More details here - Updated demo apps to use the new chunk endpoint
- Added a new weak function,
memfault_coredump_read()so platform ports can easily add locking on reads to coredump storage if necessary for transmission. - Updates to this version of the sdk will require the util component get compiled when using the panics API
- Added example port and demo for the STM32H743I-NUCLEO144 evaluation board
running ChibiOS. See
examples/stm32/stm32h743i/README.mdfor more details. - general README documentation improvements
- Fixed compilation error that could arise in
memfault_fault_handling_arm.cwhen using old versions of GCC
- Updated panics SDK to support complex hardware toplogies such as systems
with multiple MCUs or systems running multiple binaries on a single MCU. More
details here. Users of the SDK will need to update
the implementation for
memfault_platform_get_device_info() - Updated all ports to be in sync with SDK updates
- Updated panics core code to always collect fault registers for ARMv6-M & ARMv7-M architectures. The Memfault cloud will auto-analyze these and present an analysis.
- Updated https://try.memfault.com gdb script to collect Cortex-M MPU region information for auto-analysis
- general README documentation improvements
- improved error reporting strategy and documentation in
memfault/core/errors.h
- added example port and demo of panics SDK for the Nordic nRF52840
(PCA10056) development kit. See
examples/nrf5/README.mdfor more details. - Made SDK headers suitable for includion in C++ files
First Public Release of Memfault Firmware SDK 0.0.2 - June 26, 2019
- published initial Memfault SDK, see
README.mdin root for summary - published panics API which is a C SDK that can be integrated into any
Cortex-M device to save a "core" (crash state) on faults and system asserts.
See
components/panics/README.mdfor more details - Added example port and demo of panics SDK for the BCM943364WCD1 evaluation
board running the WICED Wifi stack. More details in
examples/wiced/README.md - add python invoke based CLI wrapper for demo ports