@@ -6,6 +6,79 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
6
6
and this project adheres to
7
7
[ Semantic Versioning] ( https://semver.org/spec/v2.0.0.html ) .
8
8
9
+ ## [ 1.27.0] - 2025-07-21
10
+
11
+ ### 📈 Added
12
+
13
+ - General:
14
+
15
+ - Add a reference software watchdog port for the STM32L4 series LPTIM
16
+ peripheral. Users of the STM32 HAL can now compile in the reference port and
17
+ the ` MemfaultWatchdog_Handler ` . The handler will save a coredump so the full
18
+ system state can be recovered when a watchdog takes place. More details can
19
+ be found in
20
+ [ ` ports/include/memfault/ports/watchdog.h ` ] ( ports/include/memfault/ports/watchdog.h ) .
21
+
22
+ - Add CLI commands ` wdg_enable ` , ` wdg_disable ` , and ` wdog_update <timeout_ms> `
23
+ for testing a software watchdog port. These commands are disabled by default
24
+ and can be enabled for platforms using the minimal shell/console with
25
+ ` MEMFAULT_DEMO_CLI_WATCHDOG ` .
26
+
27
+ - Zephyr:
28
+
29
+ - Add a new Kconfig option, ` CONFIG_MEMFAULT_HTTP_MAX_MESSAGES_TO_SEND ` , which
30
+ controls the max number of messages that will be sent in a single invocation
31
+ of ` memfault_zephyr_port_post_data() ` or similar APIs. The default is ` 100 ` ,
32
+ which is suitable for most applications. Before this change, the limit was
33
+ hard coded to 5 messages, which was too low for systems with infrequent
34
+ upload intervals.
35
+
36
+ - Support building for ` native_sim ` on arm64 hosts (specifically, the
37
+ ` native_sim/native/64 ` target), in addition to x86 hosts.
38
+
39
+ ### 🛠️ Changed
40
+
41
+ - Zephyr:
42
+
43
+ - Improve the default implementation of
44
+ ` memfault_platform_sanitize_address_range() ` to include all memory
45
+ accessible by the kernel. This enables collection of heap-allocated task
46
+ control blocks, which was previously unsupported. Users with discontiguous
47
+ memory regions should provide their own implementation, as before.
48
+
49
+ - Prioritize the thread bookkeeping array when collecting thread information
50
+ in a coredump (when ` CONFIG_MEMFAULT_COREDUMP_COLLECT_TASKS_REGIONS=y ` , the
51
+ default). This improves the quality of the processed coredump if the
52
+ coredump region is too small to collect all stacks for all threads in the
53
+ system (impacts systems with many threads and limited coredump storage
54
+ space).
55
+
56
+ - Add a new Kconfig option, ` CONFIG_MEMFAULT_HTTP_PACKETIZER_BUFFER_SIZE ` ,
57
+ which controls the size of the intermediate buffer used when reading data
58
+ from the underlying data source (coredump storage, log buffer, CDR, etc)
59
+ when uploading data to Memfault via HTTP. The default size is 128 bytes, and
60
+ 1024 bytes on nRF91x series SOCs to better support modem trace CDR upload.
61
+ Thanks to @DematteisGiacomo for submitting this in
62
+ [ #92 ] ( https://github.com/memfault/memfault-firmware-sdk/pull/92 ) .
63
+
64
+ - ESP-IDF:
65
+
66
+ - Handle deprecated Deep Sleep API calls for upcoming ESP-IDF v5.5 and v6.
67
+
68
+ - Supporting building with the Memfault CLI commands disabled,
69
+ ` CONFIG_MEMFAULT_CLI_ENABLED=n ` . Thanks to @finger563 for reporting this
70
+ issue and providing a fix in
71
+ [ #93 ] ( https://github.com/memfault/memfault-firmware-sdk/issues/93 ) 🎉!
72
+
73
+ ### 🐛 Fixed
74
+
75
+ - Zephyr:
76
+
77
+ - Fix an issue where the socket file descriptor can potentially be leaked when
78
+ the connection terminated unexpectedly during an HTTP chunk upload. Thanks
79
+ to @DematteisGiacomo for submitting this in
80
+ [ #92 ] ( https://github.com/memfault/memfault-firmware-sdk/pull/92 ) .
81
+
9
82
## [ 1.26.1] - 2025-06-30
10
83
11
84
This is a minor fix release, addressing one future compatibility issue with the
0 commit comments