Ben MeadorsandClaude Fable 5 252bdbc5df docs(usbnet): handoff update - crash root-caused, coredump recipe that works ...

The 4403-session crash is fixed and hardware-verified, so the handoff now
records the full causal chain (missing BOARD_HAS_PSRAM -> PSRAM never in
heap -> bad_alloc abort in the connect-time manifest walk) instead of an
open bug. The coredump section becomes the recipe that actually worked,
including the discovery that a panic reboot leaves the USB pads with the
OTG controller - there is no USB-Serial-JTAG window to race after a crash;
only a power-on reset or BOOT+RST brings the port back. New traps: the
Kconfig choice-collision that silently discards coredump config, Homebrew's
pio 6.1.18 uninstalling the platform, and BOARD_HAS_PSRAM being mandatory
on PSRAM boards. The cross-core tud_network_can_xmit() theory moves to
ruled-out-as-the-crash but stays in the backlog as a real threading-contract
violation with a proposed one-line fix.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

2026-08-06 15:24:10 -05:00

..

lora_region_preset_compatibility_client_spec.md

Regioninfo (#11056)

2026-07-19 06:22:39 -05:00

mesh_beacon_module.md

emdashes begone (#10847)

2026-07-01 19:01:27 -05:00

nexthop-routing-reliability.md

Extract mcp-server to its own repo (meshtastic/meshtastic-mcp) (#10861)

2026-07-02 14:33:04 -05:00

node_info_stores.md

TMM extend to ephemeral 3rd tier store (#11050)

2026-07-21 12:19:43 +02:00

traffic_management_module.md

TMM extend to ephemeral 3rd tier store (#11050)

2026-07-21 12:19:43 +02:00

usb-ncm-phone-api-handoff.md

docs(usbnet): handoff update - crash root-caused, coredump recipe that works

2026-08-06 15:24:10 -05:00