ESP32 Forum

When I run the simple_ota_example, I encountered the invalid segment length problem. The error message is "esp_image: invalid segment length 0xffffffff". Below is a capture from the development platform:

Toolchain path: /opt/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc
Toolchain version: crosstool-ng-1.22.0-80-g6c4433a5
Compiler version: 5.2.0
Python requirements from C:/msys32/home/Simon/esp/esp-idf/requirements.txt are s atisfied.
MONITOR
--- idf_monitor on com10 115200 ---
--- Quit: Ctrl+] | Menu: Ctrl+T | Help: Ctrl+T followed by Ctrl+H ---
ets Jun 8 2016 00:22:57

rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)
configsip: 0, SPIWP:0xee
clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00
mode:DIO, clock div:2
load:0x3fff0018,len:4
load:0x3fff001c,len:6184
load:0x40078000,len:10180
load:0x40080400,len:6660
entry 0x40080764
I (29) boot: ESP-IDF v3.2.2 2nd stage bootloader
I (29) boot: compile time 14:07:02
I (29) boot: Enabling RNG early entropy source...
I (33) boot: SPI Speed : 40MHz
I (38) boot: SPI Mode : DIO
I (42) boot: SPI Flash Size : 4MB
I (46) boot: Partition Table:
I (49) boot: ## Label Usage Type ST Offset Length
I (56) boot: 0 nvs WiFi data 01 02 00009000 00004000
I (64) boot: 1 otadata OTA data 01 00 0000d000 00002000
I (71) boot: 2 phy_init RF data 01 01 0000f000 00001000
I (79) boot: 3 factory factory app 00 00 00010000 00100000
I (86) boot: 4 ota_0 OTA app 00 10 00110000 00100000
I (94) boot: 5 ota_1 OTA app 00 11 00210000 00100000
I (101) boot: End of partition table
I (106) boot: Defaulting to factory image
I (110) esp_image: segment 0: paddr=0x00010020 vaddr=0x3f400020 size=0x1c878 (11
6856) map
I (160) esp_image: segment 1: paddr=0x0002c8a0 vaddr=0x3ffb0000 size=0x02ef4 ( 1
2020) load
I (165) esp_image: segment 2: paddr=0x0002f79c vaddr=0x40080000 size=0x00400 (
1024) load
0x40080000: _WindowOverflow4 at C:/msys32/home/Simon/esp/esp-idf/components/free
rtos/xtensa_vectors.S:1779

I (167) esp_image: segment 3: paddr=0x0002fba4 vaddr=0x40080400 size=0x0046c (
1132) load
I (176) esp_image: segment 4: paddr=0x00030018 vaddr=0x400d0018 size=0x857c0 (54
6752) map
0x400d0018: _stext at ??:?

I (2739) wifi: state: assoc -> run (10)
I (2759) wifi: connected with HELLO_INNOPLAY, channel 11
I (2759) wifi: pm start, type: 1

I (3759) event: sta ip: 10.0.1.67, mask: 255.255.255.0, gw: 10.0.1.1
I (3759) simple_ota_example: Connect to Wifi ! Start to Connect to Server....
I (5849) esp_https_ota: Starting OTA...
I (5849) esp_https_ota: Writing to partition subtype 16 at offset 0x110000
I (6049) esp_https_ota: esp_ota_begin succeeded
I (6049) esp_https_ota: Please Wait. This may take time
I (6059) esp_https_ota: Connection closed,all data received
I (6069) esp_image: segment 0: paddr=0x00110020 vaddr=0x3f400020 size=0x07518 ( 29976) map
E (6089) esp_image: invalid segment length 0xffffffff
E (6089) esp_https_ota: Error: esp_ota_end failed! err=0x5379. Image is invalid
E (6099) simple_ota_example: Firmware Upgrades Failed

Any idea what cause this problem?

I also tried the native_ota_example. The result is just the same as the simple_ota_example. Then I tried to print out the error code from esp_ota_end, the error code, update_handle is 1503HEX.

Is there anyone got any idea of the meaning of this error code? Is there any solution to this issue?

+1 on toolchain 8.2.0 and IDF v4.0 beta1
Edit: Also tried IDF v4.1-dev-369-g4dac7c7df-dirty
Failed exactly the same first go, succeeded second go.

Can you please try out suggestion from https://github.com/espressif/esp-idf/is ... -538271762? This will essentially try to download simple `hello-world` application hosted on github server instance, thus we can isolate if its server setup issue (if any) or device side issue. (Interim, we do have an internal CI test https://github.com/espressif/esp-idf/bl ... le_test.py, which ensures functionality of `simple_ota` example before every github repository sync).

Let us know update.

Thanks.

I have some trouble following the example due to our setup; only HTTP, but I can try putting the image on a known good server.

Tested again twice witout making any changes, updating locally. First time succeeded although a bit slow due to being in debug mode.

Second time:

D (53223) esp_https_ota: Written image length 5202
D (53223) HTTP_CLIENT: is_data_remain=1, is_chunked=1, content_length=-1
D (53223) HTTP_CLIENT: need_read=289, byte_to_read=289, rlen=289, ridx=0
D (53233) HTTP_CLIENT: http_on_body 289
D (53243) esp_https_ota: Written image length 5491
D (53243) HTTP_CLIENT: is_data_remain=1, is_chunked=1, content_length=-1
D (53253) HTTP_CLIENT: need_read=289, byte_to_read=289, rlen=177, ridx=0
D (53253) HTTP_CLIENT: http_on_body 177
D (53263) HTTP_CLIENT: is_data_remain=1, is_chunked=1, content_length=-1
D (58263) HTTP_CLIENT: need_read=112, byte_to_read=112, rlen=0, ridx=177
D (58263) esp_https_ota: Written image length 5668
D (58263) HTTP_CLIENT: is_data_remain=1, is_chunked=1, content_length=-1
D (63263) HTTP_CLIENT: need_read=289, byte_to_read=289, rlen=0, ridx=0
I (63263) esp_https_ota: Connection closed, all data received
D (63263) esp_image: reading image header @ 0x200000
D (63263) esp_image: image header: 0xe9 0x06 0x02 0x02 400813a4
I (63273) boot_comm: mismatch chip revision, expect 1, found 0
I (63283) esp_image: segment 0: paddr=0x00200020 vaddr=0x3f400020 size=0x4057c (263548) map
D (63283) esp_image: free data page_count 0x0000003b
I (63333) wifi.c: disconnect_socket(): Disconnected socket 56 (1). 0 client(s) left
E (63473) esp_image: invalid segment length 0xffffffff
E (63473) config.c: ota(): Error: config.c:135 ota() ret=esp_https_ota(&client) returned 5379
E (63483) config.c: ota(): OTA failed!

Connection is closed unexpectedly? I will try to connect it to the internet and OTA from there next.

Cannot reproduce in station mode on v4.0-beta1.

Problem is perhaps that the web server does not give a content-length so when the transfer is interrupted prematurely the OTA process does not know it does not have all data?

My problems seem to have been fixed by assuring a content-length is supplied in the HTTP header.

FYI, I had this, and it was caused by TCP closing the OTA early on and hence an invalid image, which in turn was caused by low dynamic memory. I only had 15k of free memory and that was not enough - but no clue or warning from debug logs to tell me why TCP had closed the connection!

How did you make sure that the content length is same as the one provided?