dangernoodle-io / breadboard / 30418595123

97%

push

github

fix: retry reboot-record NVS save once before restarting (#1126)

The reboot-reason record persist ahead of a restart was best-effort with no
mitigation for a transient NVS failure -- the diagnostic was simply lost with
a log line as the only evidence (which may itself not survive esp_restart()).

Add bb_system_reboot_record_save_retry(): one retry on a non-encode failure
before giving up. bb_system_reboot_record_save() re-opens/re-writes/
re-commits the NVS handle from scratch on every call, so the retry is a
genuinely independent second attempt, not a no-op replay of the same commit.
Deliberately best-effort -- the restart proceeds regardless of either
attempt's outcome, since a stranded board is worse than a lost diagnostic.

Both call sites that persist-then-restart now share the retry:
bb_system_restart_reason_at() and factory_reset_reboot_work_fn().

Refs B1-1257.

7896 of 8500 branches covered (92.89%)

Branch coverage included in aggregate %.

11 of 11 new or added lines in 1 file covered. (100.0%)

13425 of 13580 relevant lines covered (98.86%)

2607.08 hits per line