Xiaomi WiFi Router 3G

Did you tried this method to flash LEDE?http://telegra.ph/Xiaomi-Router-3G-proshivka-LEDE-11-23
I mean, In breed you flash original image to both system0 and system1, then you should just follow the instructions in this topic to flash lede kernel1, rootfs0 flash by mtd. And of course the nvram is unnecessary because you can set breed to autoboot kernel1 by the GUI. 0x600000 as you wrote it previously.

I want to flash breed but i need to make sure i will have an option to failsafe if i flash a not working lede... it happens some time.. :smiley: Maybe unleash some more power from this CPU by overclock it..

i have a lot a message like this.

"Data buffer not 16 bytes aligned: 8eac4198"

i don't know who did that or if it's a problem but everything seems to work.

Are you using my builds? or something else? In my build yes there are a lot of it but i didn't figured it out what's causing this. But i don't see any bugs. It doesn't connected to optimalizations like O2, O3,Os might be connected with 4.4 kernel.

Others also seem to have it...http://lists.infradead.org/pipermail/lede-dev/2017-July/008358.html
https://forum.openwrt.org/t/data-buffer-not-16-bytes-aligned-what-does-it-mean/10829

no it's my own build from openwrt master (from 3 or 4 days)
and i'm on a 4.14 kernel

It seems this comes from target/linux/ramips/patches-4.14/0039-mtd-add-mt7621-nand-support.patch (IIRC I've seen the same message on 4.9 kernel too, so similar patch must have existed also for older kernels. I think it was introduced first in b4b0a55aabd2839c5c995bb226f528efca73eea5, for kernel 3.10, back in 2013). So I guess there is either really an unaligned write happening, or the check is somehow incorrect. However I lack the knowledge to analyse C code at this level.

1 Like

Interesting...
I had never seen such a message on my device...
Did you write much on internal flash?

Hi there. I installed LEDE one month ago on my Xiaomi router and everything was working fine.
Today, I did "opkg update" of my packages.
The upgrade of the "netifd" failed and I my putty exited during the upgrade.

Since this bad upgrade, I'm not able to SSH or even ping the router on all interfaces or wifi.
Can someone tell me what I can do ?
May be I could try to connect with a serial cable as mentionned before ?Any other solutions ?
Thanks

Use reset button to factory reset the device(long press 10 sec). If it's not working then use failsafe mode.
Power off the router press reset after first blink of the led then it will blinking very fast. Set a static ip to your pc then connect through telnet.
then issuethis commands

mount_root
firstboot

Thanks you very much HeadLess, I use the failsafe mode and I get SSH access.

Can I flash an upgraded firmware and keep the old configs and packages?

If you wanted to update my firmware yes, but i didn't uploaded yet, because there was a luci error in sysupgrade and i didn't build since i report the problem which is solved.

Most of the time you can restore your backup from different openwrt also. Just you have to make sure you know what you do.

For example i migrated my config backup from 1043nd v2 i just removed the network/wireless config/firewall konfig and i copied the required lines to the new config.

Packages are not recommended to keep. Because those may be depend on kernel and other library.

You can use the script OpenWrtScripts#opkgscriptsh from richb-hanover to save a list of installed packages, do a sysupgrade and restore packages with that script also.

Ensure that /etc/sysupgade.conf includes your "special" config or files for backup.

1 Like

HI! I flashed the latest snapshot and added a wireguard interface. whenever I start the interface the device reboots. I had to enter the failsafemode and remove all settings. After this I switched back to HeadLess Image. On his image wireguard is working fine.

Is anyone using a wireguard interface on the latest snapshot?

Ok, I've earned the basic trust level, so now I can reply. :wink:

Not much actually. I've seen it usually after first boot following sysupgrade. I supposed it had something to do with 'opkg update' and 'opkg install', because after each sysupgrade I restore ddns-scripts (with dependencies). Except those, I haven't seen any new "Data buffer not 16 bytes aligned" message for 4 days of uptime. So I've checked it specifically, with 'opkg update' (nothing in logs) and 'opkg install mtr' (several "Data buffer not 16 bytes aligned" entries). This looks like an 100% reproducible way for me.

Just did a build (OpenWrt SNAPSHOT r6745-b1f5dd3) with standard kernel (4.14.34) and wireguard seems to work just as expected. No reboots. The interface goes up automatically when the router boots and my Android device that is also running latest wireguard snapshot (0.0.20180420) can connect without issues.

Ok, i did never write anything other than configs because i build my firmware always with all needed packages...
So the unaligned access is only on writes.
Hopfully with no data loss or corruption...
If i had time, i will take a look in the mentioned mtd patch...

1 Like

Hi, can someone post guide or link for wireguard installation guide? I never use it before. Thanks :wink:

Also for me, I'm using Windows 10 with https://tunsafe.com/ client.

i tested it. here is the error from the crash log:

[   59.290912] CPU 3 Unable to handle kernel paging request at virtual address 00000010, epc == 8f122d44, ra == 8f122d1c
[   59.301505] Oops[#1]:
[   59.303770] CPU: 3 PID: 0 Comm: swapper/3 Not tainted 4.14.34 #0
[   59.309748] task: 8fc44b00 task.stack: 8fc70000
[   59.314252] $ 0   : 00000000 00000001 00000000 fffffffe
[   59.319465] $ 4   : 00000000 805c0000 8f243864 00000020
[   59.324689] $ 8   : 8f243874 c0000000 f0000000 ff000000
[   59.329906] $12   : ffff0000 00000007 00000018 c66b9a56
[   59.335116] $16   : 8e80c840 8f796000 8d810000 00000000
[   59.340325] $20   : 8e94e800 80550000 805568e8 ffffffff
[   59.345534] $24   : 00000000 8000cf54
[   59.350745] $28   : 8fc70000 8fc15ae8 8d810054 8f122d1c
[   59.355955] Hi    : 00002665
[   59.358815] Lo    : 94af5487
[   59.361721] epc   : 8f122d44 noise_handshake_begin_session+0xd04/0x12b0 [wireguard]
[   59.369348] ra    : 8f122d1c noise_handshake_begin_session+0xcdc/0x12b0 [wireguard]
[   59.376962] Status: 11007c03      KERNEL EXL IE
[   59.381136] Cause : 40800008 (ExcCode 02)
[   59.385133] BadVA : 00000010
[   59.388002] PrId  : 0001992f (MIPS 1004Kc)
[   59.392074] Modules linked in: pppoe ppp_async pppox ppp_generic nf_conntrack_ipv6 mt76x2e mt7603e mt76 mac80211 iptable_nat ipt_REJECT$
[   59.462829]  sch_htb sch_hfsc sch_ingress ledtrig_usbport ip6t_REJECT nf_reject_ipv6 nf_log_ipv6 nf_log_common ip6table_mangle ip6table$
[   59.488504] Process swapper/3 (pid: 0, threadinfo=8fc70000, task=8fc44b00, tls=00000000)
[   59.496553] Stack : 00000001 8e9e8958 8f11cc80 8e343400 8fd4c000 401d5829 00000040 8e80c840
[   59.504891]         805b1718 00000003 8e80c840 00000000 8d810000 00000000 8e94e800 80550000
[   59.513227]         805568e8 ffffffff 8d810054 8032ad50 8d810000 8032a918 00000000 00000000
[   59.521566]         00000000 80550000 8e80c840 805561b8 8d810000 805b1718 00000003 805788f8
[   59.529900]         8e94e800 00000001 8ead5d80 8032b5fc 8feffa08 8d810000 00000001 8fc15b94
[   59.538232]         ...
[   59.540682] Call Trace:
[   59.543144] [<8f122d44>] noise_handshake_begin_session+0xd04/0x12b0 [wireguard]
[   59.550437] [<8032b5fc>] __dev_queue_xmit+0x6f0/0x85c
[   59.555487] [<8033b6c0>] neigh_xmit+0x104/0x1ec
[   59.560011] [<8e9ca718>] nf_flow_offload_ip_hook+0x564/0x5ac [nf_flow_table]
[   59.567048] [<80366c64>] nf_hook_slow+0x58/0x104
[   59.571669] [<80325d34>] __netif_receive_skb_core+0x5f0/0xc5c
[   59.577387] [<8032c028>] netif_receive_skb_internal+0xd8/0xf0
[   59.583115] [<8041c824>] br_pass_frame_up+0xdc/0x144
[   59.588055] [<8041cdb8>] br_handle_frame_finish+0x52c/0x570
[   59.593603] [<8041d184>] br_handle_frame+0x330/0x3dc
[   59.598545] [<80325ecc>] __netif_receive_skb_core+0x788/0xc5c
[   59.604278] [<80328c90>] process_backlog+0x98/0x160
[   59.609135] [<8032c494>] net_rx_action+0x150/0x30c
[   59.613931] [<8044d610>] __do_softirq+0x128/0x2ec
[   59.618624] [<800328e0>] irq_exit+0x98/0xcc
[   59.622809] [<80239abc>] plat_irq_dispatch+0xfc/0x138
[   59.627844] [<8000b5a8>] except_vec_vi_end+0xb8/0xc4
[   59.632790] [<8000cf70>] r4k_wait_irqoff+0x1c/0x24
[   59.637581] [<8006620c>] do_idle+0xe4/0x168
[   59.641748] [<80066488>] cpu_startup_entry+0x24/0x2c
[   59.646687] Code: 8e020048  00431024  00402025 <8c420010> 8c420014  0040f809  27b30018  0040a825  8e02009c
[   59.656413]
[   59.658074] ---[ end trace 05987436b60c7b6c ]---

Did you change the kernel config?
Eg. preemption?