How to fix Gateway panic for ESP32-S3? (TZ-585)

[kivanychev]

Steps to reproduce.

1. clone [esp-zigbee-sdk] from github repository
2. set-target esp32s3
3. menuconfig -> set Wifi parameters
4. Build esp_zigbee_gateway with IDF 5.1
5. Flash the code into the ESP32-S3-DevkitC-1 connected to the prepared RCP (rx, tx, boot, rst)
6. Observe the panic log on gateway side:

I (6181) main_task: Returned from app_main()
I (6371) ESP_ZB_GATEWAY: ZDO signal: ZDO Config Ready (0x17), status: ESP_FAIL
I (6381) ESP_ZB_GATEWAY: Zigbee stack initialized
I (6561) ESP_ZB_GATEWAY: Zigbee rcp device booted
I (6561) ESP_ZB_GATEWAY: Running RCP Version: zigbee-e088c3766b-926a2fa-1.0.7-2024-01-13 13:32:23 UTC

abort() was called at PC 0x420131e6 on core 0
0x420131e6: zb_assert at ??:?

Backtrace: 0x40375a7a:0x3fcb9380 0x4037d76d:0x3fcb93a0 0x4038422a:0x3fcb93c0 0x420131e6:0x3fcb9430 0x4203d4cf:0x3fcb9450 0x4203d4f7:0x3fcb9470 0x4203928e:0x3fcb9490 0x42017502:0x3fcb9510 0x4202e6db:0x3fcb9540 0x4200ae63:0x3fcb9560 0x42009cd5:0x3fcb9580 0x40380181:0x3fcb95b0
0x40375a7a: panic_abort at F:/esp32/idf-511/components/esp_system/panic.c:452

0x4037d76d: esp_system_abort at F:/esp32/idf-511/components/esp_system/port/esp_system_chip.c:84

0x4038422a: abort at F:/esp32/idf-511/components/newlib/abort.c:38

0x420131e6: zb_assert at ??:?
0x4203d4cf: error_unhandled at zb_error_indication.c.obj:?
0x4203d4f7: zb_error_raise at ??:?
0x4203928e: zb_mlme_dev_reset_conf at ??:?
0x42017502: zb_sched_loop_iteration at ??:?
0x4202e6db: zboss_main_loop_iteration at ??:?
0x4200ae63: esp_zb_main_loop_iteration at ??:?
0x42009cd5: esp_zb_task at F:/esp32/esp-zigbee-sdk/examples/esp_zigbee_gateway/main/esp_zigbee_gateway.c:197 (discriminator 2)
0x40380181: vPortTaskWrapper at F:/esp32/idf-511/components/freertos/FreeRTOS-Kernel/portable/xtensa/port.c:162

ELF file SHA256: 77266215aa8a8fed

Rebooting...

PS: When the same code sample is configured and built for ESP32 it works fine on ESP32DevKit.

How can I fix such panic for S3?

[xieqinan]

Hello @kivanychev ,

This issue will be fixed in this week.