No network after upgrade to 5.30


Recommended Posts

Armbian is a community driven open source project. Do you like to contribute your code?

I would like to describe more in detail.
The update from Armbian 5.25 to 5.30 / 4.9.12-sunxi does not work.
Using the u-boot-sunxi-with-spl.bin on / dev / mmcblk0 brings nothing.

Have some pictures made, maybe this will help?

 

 

20170614_234932.jpg

20170614_234959.jpg

Link to post
Share on other sites
4 hours ago, Igor said:

R1 with v5.31 legacy works. 

 

Will rebuild in the morning. 

 

Thank you for your quick response. It works perfectely for me.

 

-------------------------------------

 

[  134.321805] Unable to handle kernel NULL pointer dereference at virtual address 0000000c
[  134.330177] pgd = c0004000
[  134.333167] [0000000c] *pgd=00000000
[  134.336970] Internal error: Oops: 5 [#1] SMP THUMB2
[  134.342010] Modules linked in: aes_arm(+) ccm xt_nat xt_conntrack veth ir_xmp_decoder ir_mce_kbd_decoder ir_sharp_decoder ir_sanyo_decoder ir_sony_decoder ir_jvc_decoder ir_rc6_decoder ir_nec_decoder ir_rc5_decoder xt_CHECKSUM iptable_mangle ipt_MASQUERADE nf_nat_masquerade_ipv4 iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack xt_tcpudp bridge stp llc iptable_filter ip_tables x_tables rt2800usb rt2800lib rt2x00usb rt2x00lib evdev sun4i_codec mac80211 snd_soc_core snd_pcm_dmaengine snd_pcm nvmem_sunxi_sid nvmem_core snd_timer snd soundcore ir_lirc_codec lirc_dev sunxi_cir sun4i_ss fuse cpufreq_dt uio_pdrv_genirq uio bonding brcmfmac brcmutil cfg80211 rfkill realtek
[  134.405095] CPU: 0 PID: 1974 Comm: cryptomgr_test Not tainted 4.11.5-sunxi #1
[  134.412434] Hardware name: Allwinner sun7i (A20) Family
[  134.417808] task: ef13ea40 task.stack: ec560000
[  134.422479] PC is at crypto_remove_spawns+0x7c/0x180
[  134.427584] LR is at 0xec561f18
[  134.430820] pc : [<c04d513c>]    lr : [<ec561f18>]    psr: a00d0133
               sp : ec561f10  ip : ec561f20  fp : ec561f50
[  134.442608] r10: 00000401  r9 : c0c74108  r8 : ecc30688
[  134.447981] r7 : 00000000  r6 : ec561f10  r5 : ec561f18  r4 : 00000000
[  134.454689] r3 : ecc31888  r2 : bfbf5000  r1 : ecc307c0  r0 : c0c74100
[  134.461411] Flags: NzCv  IRQs on  FIQs on  Mode SVC_32  ISA Thumb  Segment none
[  134.468933] Control: 50c5387d  Table: 6c63806a  DAC: 00000051
[  134.474855] Process cryptomgr_test (pid: 1974, stack limit = 0xec560210)
[  134.481752] Stack: (0xec561f10 to 0xec562000)
[  134.486248] 1f00:                                     ee7dc9c0 ee7dc9c0 ecc319d8 ecc307c0
[  134.499289] 1f20: ec561f20 ec561f20 00000000 ecc20200 bfbf5000 c0c72e18 00000401 c0c74100
[  134.517233] 1f40: bfbf5028 bfbf5068 c0c74128 c04d5433 ec561f50 ec561f50 ec669640 edd6b300
[  134.535282] 1f60: ec669640 00000000 ec560000 edd6b300 c04d9e5d ec715d10 ec5e319c c04d9e6d
[  134.553678] 1f80: ec5e3180 c012ff49 ffffffff ec669640 c012fe4d 00000000 00000000 00000000
[  134.572384] 1fa0: 00000000 00000000 00000000 c0106351 00000000 00000000 00000000 00000000
[  134.591071] 1fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[  134.609729] 1fe0: 00000000 00000000 00000000 00000000 00000013 00000000 00180002 00004020
[  134.628386] [<c04d513c>] (crypto_remove_spawns) from [<c04d5433>] (crypto_alg_tested+0xfb/0x12c)
[  134.647670] [<c04d5433>] (crypto_alg_tested) from [<c04d9e6d>] (cryptomgr_test+0x11/0x20)
[  134.666368] [<c04d9e6d>] (cryptomgr_test) from [<c012ff49>] (kthread+0xfd/0x104)
[  134.684527] [<c012ff49>] (kthread) from [<c0106351>] (ret_from_fork+0x11/0x20)
[  134.697364] Code: 681c 42a3 d014 681c (68e3) 4283 
[  134.708311] ---[ end trace 34a6d32538f81975 ]---

 

Unfortunately I got the same issue as the Nanopi.  LXC containers are freezing the system with 4.11.x kernel.

Link to post
Share on other sites
2 hours ago, Charizard said:

Unfortunately I got the same issue as the Nanopi.  LXC containers are freezing the system with 4.11.x kernel.


You are using development kernel, which we don't support for many reasons.

 

6 hours ago, Aux said:

I would like to describe more in detail.
The update from Armbian 5.25 to 5.30 / 4.9.12-sunxi does not work.
Using the u-boot-sunxi-with-spl.bin on / dev / mmcblk0 brings nothing.

Have some pictures made, maybe this will help?

 


This indicate some problems with u-boot script. Check and change rootdev parameter in /boot/armbianEnv.txt

setenv rootdev "/dev/mmcblk0p1"

Link to post
Share on other sites
3 hours ago, Igor said:

This indicate some problems with u-boot script. Check and change rootdev parameter in /boot/armbianEnv.txt

setenv rootdev "/dev/mmcblk0p1"

I have changed it but no hope for improvement. Now comes "/dev/mmcblk0p1 not exist" ...

 

20 minutes ago, zador.blood.stained said:

This possibly destroyed the root partition, so you need to reinstall from scratch. It's not a boot script issue.

I know I have backup on sd card written and the update process from 5.25 to 5.30 again but without "linux-u-boot-next-bananapi_5.30_armhf.deb" because it was locked in the repository. Now I have the above error ....

Link to post
Share on other sites

Sorry guys but there is something completely wrong. Backup image recovered, update process from 5.25 to 5.30/31 started, reboot same errors.
This time the complete update with "linux-u-boot-next-bananapi_5.31_armhf.deb".

 

Dpkg.log - https://pastebin.com/SV22r1R3

 

There are no other usable logs or have I overlooked what?

On the system runs openmediavault, is it possible that has an influence?

 

before update

cat /etc/*-release
# PLEASE DO NOT EDIT THIS FILE
BOARD=bananapi
BOARD_NAME="Banana Pi"
VERSION=5.25
LINUXFAMILY=sunxi
BRANCH=next
ARCH=arm
IMAGE_TYPE=stable
PRETTY_NAME="Debian GNU/Linux 8 (jessie)"
NAME="Debian GNU/Linux"
VERSION_ID="8"
VERSION="8 (jessie)"
ID=debian
HOME_URL="http://www.debian.org/"
SUPPORT_URL="http://www.debian.org/support"
BUG_REPORT_URL="https://bugs.debian.org/

 

Link to post
Share on other sites

I wanted to know which package makes the system stupid and did the installation step by step look at the bottom

 

Install - sunxi-tools (1.4.2-1~armbian5.31+1) reboot status = all works
Install - linux-jessie-root-next-bananapi 5.31 reboot status = all works
Install - armbian-tools-jessie 5.30  reboot status = all works
Install - linux-dtb-next-sunxi 5.31  reboot status = system broken - This update caused the problems!!!

Link to post
Share on other sites
3 minutes ago, Aux said:

Install - linux-dtb-next-sunxi 5.31  reboot status = system broken - This update caused the problems!!!

Nope, but for whatever reasons dtb and kernel are separate packages and updating one without the other breaks stuff. Based on my support experiences not just with Armbian (but commercially in large installations) larger updates are also great to trigger storage errors (that existed already before but high IO activity during the update triggers then filesystem errors and things start failing). This applies especially to Armbian situation where most devices are running off unreliable SD card storage.

 

TL;DR: More than one reason why updates are failing exists, updating dtb without kernel or vice versa is known to fail and checking storage if updates fail is always a good idea (at least dmesg output or providing 'armbianmonitor -u' or even 'armbianmonitor -c $HOME' output)

Link to post
Share on other sites

Thank you for your answer @tkaiser!
But if I update all packages  have the same problem, what now?
Have the update at least 5 times in various combinations tasted, since is what is lazy. I have never had such problems with ARMBIAN before.

Link to post
Share on other sites
1 hour ago, Aux said:

Sorry guys but there is something completely wrong. Backup image recovered, update process from 5.25 to 5.30/31 started, reboot same errors.


I just test the upgrade once again. Starting with this image: https://dl.armbian.com/bananapi/archive/Armbian_5.25_Bananapi_Debian_jessie_next_4.9.7.7z

 

apt update

apt upgrade

 

Zero problems. 

Spoiler

 ____                                  ____  _ 
| __ )  __ _ _ __   __ _ _ __   __ _  |  _ \(_)
|  _ \ / _` | '_ \ / _` | '_ \ / _` | | |_) | |
| |_) | (_| | | | | (_| | | | | (_| | |  __/| |
|____/ \__,_|_| |_|\__,_|_| |_|\__,_| |_|   |_|
                                               

Welcome to ARMBIAN 5.31 stable Debian GNU/Linux 8 (jessie) 4.11.5-sunxi   
System load:   0.51 0.14 0.05   Up time:       0 min
Memory usage:  5 % of 996MB     IP:            xxxxxxxx
CPU temp:      41°C           
Usage of /:    8% of 15G    

[ General system configuration: armbian-config ]
New to Armbian? Check the documentation first: https://docs.armbian.com

 

 

Logs: http://sprunge.us/QjUT

Edited by Igor
link to logs
Link to post
Share on other sites
6 minutes ago, Aux said:

I have never had such problems with ARMBIAN before.

Care to accept that you might not run into an Armbian problem here? The 5.30 upgrade f*cked up networking on some boards (fixed with 5.31) while you reported boot loops. Those are either a sign of underpowering or corrupt SD card. What about giving that a try?

 

BTW: All storage will die eventually, all PSUs will die eventually. If you're lucky it's a 'Boom! Dead!' situation but if not and things start to die slowly it always looks like software misbehaving. Not saying that this is the case for you just that you should also look into these possibilities.

Link to post
Share on other sites
11 minutes ago, tkaiser said:

Not saying that this is the case for you just that you should also look into these possibilities.

But I have already tried other SD card, now will still apply another power supply, when it does not work then I probably switch to other disto. Since there is apparently no solution for it, much more - I am apparently not the only one @ Wolf2000 wrote me by e-mail that he exactly the same problem has.

Except for that, I will now not open a new thread to just have to re-write and write the same ...

Link to post
Share on other sites

Just to waste a bit more of my own time too... 

 

Testing 5.25 Debian Jessie next from archive on a Banana Pi with sufficient powering (the USB cable is the problem most of the times) and reliable SD card (genuine Samsung EVO+, the one in the upper right corner of the picture here -- there are also a few links where anyone interested can learn what problems are related to SD cards in general!)

  • armbianmonitor -u prior to upgrade: http://sprunge.us/EaHX (Armbian 5.25 with kernel 4.9.7)
  • armbianmonitor -u after 5.31 upgrade: http://sprunge.us/PfGa (Armbian 5.31 with kernel 4.11.5 and obvisouly network connectivity)

Upgrade from 5.25 to 5.31 on this platform took 12:20 (dpkg log) but with SD cards showing low random performance it might take a lot longer (especially linux-header compilation might take ages if using a crappy SD card).

 

Obviously 5.31 fixed the network issues affecting a few boards after 5.30 upgrade but as usual SD card crappiness is a problem (and problems can be triggered with high IO activity during upgrades) and with boards that use the crappy Micro USB connector for power you always have to take care about insufficient powering. Again see the USB cable is the problem most of the times and my personal experiences with Banana Pi powering crappiness.

 

TL;DR: Unfortunately Armbian 5.30 upgrade introduced a NETWORK problem on some boards that got fixed with 5.31 in the meantime. Upgrading now is safe. SD cards as well as power supplies are a well known problem that always need some investigations if things start to fail for obscure reasons. If you run into boot loops or your board seems to be bricked use latest backup to recover or follow general guidelines (see our documentation) and always keep in mind that you might just experience your installation media dying.

Link to post
Share on other sites
Just now, Wolf2000 said:

Kommt der kram mit der SD-Karte und Stromversorgung das ist es sicher nicht!!!!!!

Und das löst das Problem nicht

 

Stimme ich vollkommen zu !

 

@tkaiser

Oh come the junk with the SD card and power supply it is certainly not !!!!!
And that does not solve the problem.  It ran for over a year and now the cable or SD card is defective and in two cases at the same time? This is a no go

 

Link to post
Share on other sites
Guest
This topic is now closed to further replies.