Sonoff POW R2 crashes after changing initial admin password

[ruimarinho]

After the initial flash (using latest commit 36f6d39 from dev branch), the Sonoff POW R2 boots correctly with the default ESPURNA SSID.

As usual, I enter the default password to connect to it and then navigate to the device's IP where I'm presented with a password change page. After changing the default password, the device disconnects and the wifi responds to correctly to the new password but the web server does not start — the device always enter safe mode.

My device has board 2017-11-30 V1.0 with CSE77598-SG (1741D1A).

[xoseperez]

I cannot reproduce the issue. How do you know the device enters safe mode? Can you still connect to it via AP? Can you type crash to dump the latest crash stack?
BTW, the latest commit changes the way the authentication works, it could affect this issue too.

[ruimarinho]

The web server does not start but the telnet server does. The console output says the device is in safe mode.

Regarding the crash command, please see the output at #761 (comment)

[ruimarinho]

@xoseperez is there any way I can help debug this issue?

[ServerCare]

Just wanted to add that I seem to have the same or similar issue. Possibly my description of the issue may assist. I actually have this now happening on three new R2's that were flashed with the stable build. In my case after changing the name things seem to be fine still, but after I try to add them to my wifi (using a fixed IP) things go wrong. From that point on the devices can only be seen as wifi AP's (curiously with the changed name in two devices). Connecting to the AP works, I can ping to 192.168.4.1 but web- browsing to it does not work. In addition: all three devices report ' -ERROR buffer overflow' when you telnet to them.

The first few seconds after a 'hard reset' (pulling the power) I can still switch the relais manually, but that stops quickly and the device is unresponsive.

What can I do to 'reset' the devices?
I have a few more devices, let me know if there is anything I can test

[ServerCare]

Did some further testing with a 4rd R2 (yes I have plenty of them ;-)
This time I was more careful with the setup procedure and I disabled 'scan networks' (didn't work anyway) did not change the name and left the IP up to DHCP.

This time I can connect & things seem to work, however; very soon after booting the device no longer accepts commands in telnet or debug window, simply responding -error unknown command (multiple times, at random) and error- buffer overflow (in the telnet window). The debug window typicall responds with '[WEBSOCKET] Requested action: dbgcmd' when I type in (for example) 'info'

[mcspr]

@ServerCare Would you mind trying this build?
https://github.com/mcspr/espurna-travis-test/releases/download/20180619/espurna-1.13.0c.gitb7915c3-itead-sonoff-pow-r2.bin
https://github.com/mcspr/espurna-travis-test/releases/tag/20180619

By the looks of it serial connection is active. And that was solved by this commit in the dev: 40b9fdf

[ServerCare]

I have done an OTA update on the one device that I could still access over the web interface (which worked like a charm!) and can confirm that the telnet issues are now over. Output looks fine after the update.

Would this also fix the issues I have on my other 3 devices that I cannot reach anymore? I'll see if I can update them using the flash procedure tomorrow.

Some more good news; after the upgrade I have changed the name of the device as well as changed to a static IP without any issues. AND I was able to scan the network without any issues...

[ServerCare]

Further good news: I have re-flashed my three unresponsive R2's, all have come back to life & are now acting as expected. Thanks!

@xoseperez just to let you know this fix is of a huge impact on all R2's I have implemented this on (4 in total) In essense: without this fix the current stable is not reliable for the R2.

[xoseperez]

@ruimarinho Can you confirm if @ServerCare is having the same issue and if dev branch is working for you?

[ruimarinho]

Will try to erase flash and reflash with master this weekend. I have tried multiple times to run master after I had confirmed 1.12.6 was unstable for me, but never got to a working point. Perhaps I was simply unlucky with the commits I was testing with :)

On Wed, 20 Jun 2018 at 21:32 Xose Pérez ***@***.***> wrote: @ruimarinho <https://github.com/ruimarinho> Can you confirm if @ServerCare <https://github.com/ServerCare> is having the same issue and if de branch is working for you? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#901 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AARnxYNPMn5xttbk_4U9oyFjPGmcj3E4ks5t-rFEgaJpZM4USVxO> .

-- -- Rui Marinho

[ruimarinho]

@xoseperez tried with 1.13.1a (before I also erased flash using esptool.py) but still not working. Now I cannot even get an IP via DHCP. Using a static IP of 192.168.4.1, I was able to telnet into the device (the web server was down) and also connected to serial.

INFO
[189695]

[189695] [INIT] ESPURNA 1.13.1a
[189696] [INIT] xose.perez@gmail.com
[189696] [INIT] http://tinkerman.cat

[189696] [INIT] CPU chip ID: 0x4EB844
[189697] [INIT] CPU frequency: 80 MHz
[189697] [INIT] SDK version: 1.5.3(aec24ac9)
[189697] [INIT] Core version: 2.3.0
[189698] [INIT] Core revision:
[189698]
[189698] [INIT] Flash chip ID: 0x1640EF
[189700] [INIT] Flash speed: 40000000 Hz
[189700] [INIT] Flash mode: QIO
[189700]
[189700] [INIT] Flash size (CHIP)   :  4194304 bytes / 1024 sectors (   0 to 1023)
[189702] [INIT] Flash size (SDK)    :  1048576 bytes /  256 sectors (   0 to  255)
[189702] [INIT] Reserved            :     4096 bytes /    1 sectors (   0 to    0)
[189703] [INIT] Firmware size       :   476928 bytes /  117 sectors (   1 to  117)
[189703] [INIT] Max OTA size        :   544768 bytes /  133 sectors ( 118 to  250)
[189704] [INIT] EEPROM size         :     4096 bytes /    1 sectors ( 251 to  251)
[189704] [INIT] Reserved            :    16384 bytes /    4 sectors ( 252 to  255)
[189705]
[189705] [INIT] EEPROM sectors: 251, 250
[189705]
[189706] [INIT] BOARD: ITEAD_SONOFF_POW_R2
[189707] [INIT] SUPPORT: ALEXA BROKER DEBUG_TELNET DEBUG_WEB DOMOTICZ HOMEASSISTANT MDNS_SERVER MQTT NTP SCHEDULER SENSOR TELNET TERMINAL THINGSPEAK WEB
[189707] [INIT] SENSORS: CSE7766
[189707]
[189708] [INIT] Last reset reason: Power on
[189708] [INIT] Settings size: 266 bytes
[189709] [INIT] Free heap: 13864 bytes
[189710] [INIT] Power: 3007 mV
[189710] [INIT] Power saving delay value: 10 ms
[189710]
[189710] [WIFI] ------------------------------------- MODE STA
[189710] [WIFI] SSID
[189710] [WIFI] IP    0.0.0.0
[189711] [WIFI] MAC   2C:3A:E8:4E:B8:44
[189711] [WIFI] GW    0.0.0.0
[189711] [WIFI] DNS   208.67.222.222
[189711] [WIFI] MASK  0.0.0.0
[189711] [WIFI] HOST  http://ESP_4EB844.local
[189711] [WIFI] BSSID FF:FF:FF:FF:FF:FF
[189712] [WIFI] CH    1
[189712] [WIFI] RSSI  31
[189712] [WIFI] -------------------------------------- MODE AP
[189712] [WIFI] SSID  ESPURNA-4EB844
[189712] [WIFI] PASS  fibonacci
[189712] [WIFI] IP    192.168.4.1
[189713] [WIFI] MAC   2E:3A:E8:4E:B8:44
[189713] [WIFI] ----------------------------------------------
[189713] +OK

KEYS
[139737] Current settings:
[139738] > board => "71"
[139738] > boardName => "ITEAD_SONOFF_POW_R2"
[139738] > btnGPIO0 => "0"
[139739] > btnRelay0 => "0"
[139739] > cf1GPIO => "13"
[139739] > cfGPIO => "14"
[139739] > cfg => "3"
[139739] > hostname => "ESPURNA-4EB844"
[139740] > ledGPIO0 => "15"
[139740] > ledLogic0 => "1"
[139740] > mqttEnabled => "0"
[139741] > relayBoot0 => "0"
[139741] > relayGPIO0 => "12"
[139741] > relayPulse0 => "0"
[139741] > relayTime0 => "0.00"
[139741] > relayType0 => "0"
[139741] > selGPIO => "5"
[139742] Number of keys: 17
[139742] Current EEPROM sector: 250
[139743] Free EEPROM: 3830 bytes (93%)
[139743] +OK

CRASH
[210885] [DEBUG] No crash info
[210927] +OK
UPTIME
[222667] Uptime: 222 seconds
[222668] +OK
ERASE.CONFIG

Then reconnecting via telnet shows:

CRASH
[076058] [DEBUG] Latest crash was at 226863 ms after boot
[076059] [DEBUG] Reason of restart: 2
[076059] [DEBUG] Exception cause: 29
[076059] [DEBUG] epc1=0x4020cf07 epc2=0x00000000 epc3=0x00000000
[076059] [DEBUG] excvaddr=0x00000000 depc=0x00000000
[076060] [DEBUG] >>>stack>>>
[DEBUG] 3fff1f80: 00000000 00000001 ffffffe6 4020cf03
[DEBUG] 3fff1f90: 40104ac0 0432c3cf 3fff2060 4021b895
[DEBUG] 3fff1fa0: 3fff07f0 3fff2060 3fff0ff0 3fff2060
[DEBUG] 3fff1fb0: 3fffdad0 00000001 3fff07f0 4021c39d
[DEBUG] 3fff1fc0: 402017e6 0000000a 0000000a 3fff0fe4
[DEBUG] 3fff1fd0: 3fffdad0 3fff0bd8 3fff07ec 40202bef
[DEBUG] 3fff1fe0: 00000000 3fff0bd8 00000002 40204120
[DEBUG] 3fff1ff0: 3fffdad0 00000000 3fff0fde 40229c08
[DEBUG] 3fff2000: feefeffe feefeffe 3fff0ff0 401006fc
[DEBUG] <<<stack<<<
[076106] +OK

and on serial monitor just:

 ets Jan  8 2013,rst cause:1, boot mode:(3,7)

load 0x4010f000, len 1384, room 16 
tail 8
chksum 0x2d
csum 0x2d
v09826c6d
~ld

 ets Jan  8 2013,rst cause:1, boot mode:(3,7)

load 0x4010f000, len 1384, room 16 
tail 8
chksum 0x2d
csum 0x2d
v09826c6d
~ld

[xoseperez]

I see that the board is using QIO flash mode, but it should be DOUT....