Problem updating packages to 5.25

[man_of_war]

 

Hello everybody!
There is a big problem with the Banana PI Pro. The system is installed on the micro SD, when updating the packages I get the following result:

root@bananapipro:~# apt-get install mdadm
Reading package lists... Done
Building dependency tree       
Reading state information... Done
You might want to run 'apt-get -f install' to correct these:
The following packages have unmet dependencies:
 libc6-dev : Depends: libc6 (= 2.19-18+deb8u9) but 2.19-18+deb8u7 is to be installed
E: Unmet dependencies. Try 'apt-get -f install' with no packages (or specify a solution).
root@bananapipro:~# apt-get -f install
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Correcting dependencies... Done
The following extra packages will be installed:
  libc-dev-bin libc6
Suggested packages:
  glibc-doc
Recommended packages:
  manpages-dev
The following packages will be upgraded:
  libc-dev-bin libc6
2 upgraded, 0 newly installed, 0 to remove and 49 not upgraded.
2 not fully installed or removed.
Need to get 0 B/4,196 kB of archives.
After this operation, 0 B of additional disk space will be used.
Do you want to continue? [Y/n] Y
Preconfiguring packages ...
(Reading database ... 52471 files and directories currently installed.)
Preparing to unpack .../libc6_2.19-18+deb8u9_armhf.deb ...
Unpacking libc6:armhf (2.19-18+deb8u9) over (2.19-18+deb8u7) ...
dpkg: error processing archive /var/cache/apt/archives/libc6_2.19-18+deb8u9_armhf.deb (--unpack):
 unable to clean up mess surrounding `./usr/lib/arm-linux-gnueabihf/audit/sotruss-lib.so' before installing another version: Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/test-results-arm-linux-gnueabihf-libc': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/BUGS': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/copyright': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/README.Debian.gz': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/README.hesiod.gz': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/NEWS.gz': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/changelog.gz': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/NEWS.Debian.gz': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/doc/libc6/changelog.Debian.gz': Read-only file system
dpkg: error while cleaning up:
 unable to restore backup version of `/usr/share/lintian/overrides/libc6': Read-only file system
dpkg: error while cleaning up:
 unable to securely remove '/var/lib/dpkg/tmp.ci': Read-only file system
dpkg: error while cleaning up:
 unable to securely remove '/var/lib/dpkg/reassemble.deb': Read-only file system
dpkg-deb: error: subprocess paste was killed by signal (Broken pipe)
dpkg: error: unable to create new file '/var/lib/dpkg/status-new': Read-only file system
E: Sub-process /usr/bin/dpkg returned an error code (2)
E: Failed to write temporary StateFile /var/lib/apt/extended_states.tmp

dmesg:

[   21.114064] rc rc0: lirc_dev: driver ir-lirc-codec (sunxi-ir) registered at minor = 0
[   21.141516] sunxi-ir 1c21800.ir: initialized sunXi IR driver
[   21.167429] sun4i-codec 1c22c00.codec: Codec <-> 1c22c00.codec mapping ok
[   22.496820] EXT4-fs (mmcblk0p1): re-mounted. Opts: commit=600,errors=remount-ro
[   24.594576] Adding 131068k swap on /var/swap.  Priority:-1 extents:2 across:163836k SSFS
[   25.465944] systemd-journald[248]: Received request to flush runtime journal from PID 1
[   26.591321]  RX IPC Checksum Offload disabled
[   26.605358]  No MAC Management Counters available
[   26.605384] sun7i-dwmac 1c50000.ethernet eth0: fail to init PTP.
[   26.605736] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[   30.741639] sun7i-dwmac 1c50000.ethernet eth0: Link is Up - 1Gbps/Full - flow control rx/tx
[   30.746410] 8021q: adding VLAN 0 to HW filter on device bond0
[   30.865735] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[   31.637564] Link is Up - 1000/Full
[  403.827540] error 0 requesting status 0x80e00
[  403.935673] error 0 requesting status 0x80900
[  403.949068] blk_update_request: I/O error, dev mmcblk0, sector 2353400
[  403.965954] sunxi-mmc 1c0f000.mmc: smc 0 err, cmd 25, WR EBE !!
[  403.981194] sunxi-mmc 1c0f000.mmc: data error, sending stop command
[  403.996814] mmcblk0: mmc_blk_cmd_recovery: general error sending stop or status command, stop cmd response 0x80d00, card status 0x900
[  404.027099] mmcblk0: timed out sending r/w cmd command, card status 0x900
[  404.135650] sunxi-mmc 1c0f000.mmc: smc 0 err, cmd 25, WR EBE !!
[  404.150685] sunxi-mmc 1c0f000.mmc: data error, sending stop command
[  404.166049] mmcblk0: mmc_blk_cmd_recovery: general error sending stop or status command, stop cmd response 0x80d00, card status 0x900
[  404.196038] mmcblk0: timed out sending r/w cmd command, card status 0x900
[  404.211922] blk_update_request: I/O error, dev mmcblk0, sector 2353400
[  404.227523] blk_update_request: I/O error, dev mmcblk0, sector 2353408
[  404.242928] blk_update_request: I/O error, dev mmcblk0, sector 2353416
[  404.258004] blk_update_request: I/O error, dev mmcblk0, sector 2353424
[  404.272836] blk_update_request: I/O error, dev mmcblk0, sector 2353432
[  404.287332] blk_update_request: I/O error, dev mmcblk0, sector 2353440
[  404.301529] blk_update_request: I/O error, dev mmcblk0, sector 2353448
[  404.315695] EXT4-fs (mmcblk0p1): Delayed block allocation failed for inode 5550 at logical offset 0 with max blocks 7 with error 5
[  404.342500] EXT4-fs (mmcblk0p1): This should not happen!! Data will be lost

[  404.368220] error 0 requesting status 0x80900
[  404.380212] blk_update_request: I/O error, dev mmcblk0, sector 2353456
[  404.394670] sunxi-mmc 1c0f000.mmc: smc 0 err, cmd 25, WR EBE !!
[  404.408187] sunxi-mmc 1c0f000.mmc: data error, sending stop command
[  404.422062] mmcblk0: mmc_blk_cmd_recovery: general error sending stop or status command, stop cmd response 0x80d00, card status 0x900
[  404.449171] mmcblk0: timed out sending r/w cmd command, card status 0x900

SD card, new. There are no bad blocks, it is checked. This also arises on the new pure Armbian_5.25_Bananapipro_Debian_jessie_next_4.9.7.img and on my main distribution, also updated to 5.25

How to solve this problem?

[man_of_war]

I will answer myself, there was an incompatibility of the SD card. I used another card, also class 10, everything works well. The reason for this remains only unclear.

[tkaiser]

On 31.5.2017 at 8:43 PM, man_of_war said:

SD card, new. There are no bad blocks, it is checked.

You can't check for bad blocks on SD cards or any other flash based media. This is just fooling yourself. Explanation here (look also few posts above): https://forum.armbian.com/index.php?/topic/3988-sd-card-from-ancient-times/&do=findComment&comment=29226

 

 

4 hours ago, man_of_war said:

The reason for this remains only unclear.

Nope, just look into my signature. You can't build houses on quicksand and you can't run SBC reliably off broken SD cards.

[man_of_war]

The problem is clearly not in the SD card. On the same SD card with the kernel 3.4.113 everything is fine, but with the kernel 4.9.12 there is an error. Perhaps, some settings are slightly different or reversal at the hardware level. Additionally, I checked the same SD card on Orange PI Zero, everything works with the kernel 3.4.113.

[tkaiser]

Can you please provide the output of the following running with 4.9 kernel:

cat /sys/power/axp_pmu/{ac,vbus}/voltage

And then also from

stress -c 1 & sleep 1 && cat /sys/power/axp_pmu/{ac,vbus}/voltage && pkill stress

 

[man_of_war]

root@bananapi:/# cat /sys/power/axp_pmu/{ac,vbus}/voltage
4909600
0
root@bananapi:/# stress -c 1 & sleep 1 && cat /sys/power/axp_pmu/{ac,vbus}/voltage && pkill stress
[1] 31311
stress: info: [31311] dispatching hogs: 1 cpu, 0 io, 0 vm, 0 hdd
4846700
0
[1]+  Terminated              stress -c 1
root@bananapi:/# 

The board on this power supply has been working for a long time.

[tkaiser]

16 hours ago, man_of_war said:

The board on this power supply has been working for a long time.

 

What's the purpose of this statement? Every electronic device will die sooner or later. When PSUs start to die slowly symptoms are 'funny' since stuff happens that looks like broken software, the problems are not 100% reproducable and you waste a few days or even weeks of your life hunting non existing bugs.

 

You already suffer from undervoltage (4.85V with a rather light load as 'stress -c 1') so I would fix this first. The problem with this shitty Micro USB connector those Banana boards are using is not only cable resistance usually way too high but also contact resistance (touch the cable slightly and voltage drops another 0.3V easily). In the comments section of this blog post some more reliable Micro USB cables are listed http://tech.scargill.net/a-question-of-lifespan/ but the best you could do on a Banana Pi or Pro is to use the SATA power connector to reliably power these boards.

 

Then you already discovered an 'incompatibility' of your SD card since with another card 'it works'. As usual it's either powering or storage or both. In case you want to save time and frustration in the future simply follow Armbian's 'Getting Started Guide' (which is there for exactly this reason: Save you some time and frustration): https://docs.armbian.com/User-Guide_Getting-Started/

 

So test your SD cards first (badblocks is USELESS, use appropriate tools), only use Etcher to write images to SD cards (since Etcher does a verify, almost all other tools allow to write garbage to cards since they don't compare written result with original image -- it's as stupid as it sounds but still reality in 2017), check for powering problems and when you run into instabilities related to storage (that's what dmesg above is telling you), check your card (that's 'armbianmonitor -d $HOME') and think about running a somewhat heavy benchmark including 'random IO':

iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2

If the card dies or random IO performance is low then immediately throw it away (or ask for a refund since you should do this directly after purchase). For some numbers and good cards please look here: https://forum.armbian.com/index.php?/topic/954-sd-card-performance/