Coredump to flash hanging at partition erase step

jsam589
Posts: 74
Joined: Sat Aug 17, 2019 9:31 pm

Coredump to flash hanging at partition erase step

Postby jsam589 » Thu Oct 24, 2019 5:38 pm

I cannot get coredump to a flash partition to work. I am using v4.0-beta1. I have 16 MByte flash WROOM device.
In a simple test using the blink example, I write to a null pointer to cause crash.

I found a few other reports of the same behavior:
https://github.com/espressif/esp-idf/issues/3857
https://github.com/espressif/esp-idf/issues/3954

The last line that appears on console is like this and comes from core_dump_common.c. Eventually resets due to WDT.
D (11868) esp_core_dump_common: Core dump len = 5596 (7 0)

The "prepare" function is in core_dump_flash.c and will call spi_flash_erase_range()

If this is a known issue, is it fixed in master/latest branch? I am using the tagged v4.0-beta1 right now.

My partitions file is
# Name, Type, SubType, Offset, Size, Flags
nvs, data, nvs, 0x9000, 16K,
otadata, data, ota, 0xd000, 8K,
phy_init, data, phy, 0xf000, 4K,
golden, app, factory, 0x10000, 2M,
ota_0, app, ota_0, 0x210000, 2M,
ota_1, app, ota_1, 0x410000, 2M,
log_0, data, spiffs, 0x610000, 1516K,
log_1, data, spiffs, 0x78B000, 1556K,
dump, data, coredump, 0x910000, 512K,
cert, 0x45, , 0x990000, 8K,
key, 0x45, , 0x992000, 4K,
storage, 0x44, , 0x993000, 2M,
# 0xB93000 - 0xFFFFFF (4532K): Remainder of 16 MByte is reserved for future use

jsam589
Posts: 74
Joined: Sat Aug 17, 2019 9:31 pm

Re: Coredump to flash hanging at partition erase step

Postby jsam589 » Fri Oct 25, 2019 3:26 am

I have tried again using IDF "latest" (v4.1-dev) and coredump works okay from that branch.

However, now I cannot seem to get MD5 to appear in my partitions table, regardless of how I set it in menuconfig. I can use gen_esp32part.py to create one with MD5 in it. Hopefully this will be fixed soon in IDF.

Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 111 guests