Ubuntu 18.04 LTS for ODROID-C2 (64bits OS)

Post by **mdrjr** » Fri Feb 05, 2016 7:20 am

Hello,

We are releasing the Ubuntu 16.04 LTS for ODROID-C2.

Its the very first public image so you can expect some bugs.
Bugs can be reported on this thread and they will be addressed as possible for the next image.

Before posting any bug report make sure that you have your OS fully updated as many things will change very fast in the early stage.

Link: http://odroid.com/dokuwiki/doku.php?id= ... nux_ubuntu

Big note! (Updates to v2.0)

Some users had a issue after update that screen was with black borders.
A workaround for that is to simple run: "sudo apt install mali-x11"

Other issue fixed was "Slow console"
By updating your system this issue should be re-mediated now.
"sudo apt update && sudo apt dist-upgrade"

XeoSal · Post by **XeoSal** » Fri Feb 05, 2016 9:08 am

Cool, seems like everything is ready for a Debian/Ubuntu platform, first thing on my list is getting chromium to work here. Although, I haven't received the package yet, I am hopping it will be there by today or maybe tomorrow.

mlinuxguy · Post by **mlinuxguy** » Fri Feb 05, 2016 1:52 pm

Mine came in this morning. I've pulled down Ubuntu and booted the C2 up on it.
Currently I'm loading up the build environment with plans to pull down the source code from GIT and build it native on the C2

I had planned to test out a MPU-6050 which I wired to it but don't see any I2C module
I2cdetect fails for all the busses listed under /dev so I'm guessing there is no module yet

I'll know more after I get the source down and inspect it

Update
Ok I have got the source down and started inspecting it. From the file: odroidc2_defconfig
we can see:
CONFIG_I2C_AML=y
(so its built-in, assuming that is the correct driver)

Code: Select all

root@odroid64:/usr/src/linux# ls /dev/i2c*
/dev/i2c-0  /dev/i2c-1  /dev/i2c-2  /dev/i2c-3  /dev/i2c-4  /dev/i2c-5  /dev/i2c-6  /dev/i2c-7

But this repeats for each device:

Code: Select all

root@odroid64:/usr/src/linux# i2cdetect -y 0
Error: Could not open file `/dev/i2c-0': No such device

I'll have to continue digging through the source to see what is going on
Update #2

From the file: /usr/src/linux/arch/arm64/boot/dts/meson64_odroidc2.dts

Code: Select all

&i2c_a {
        status = "disabled";
};

Not sure why they have it disabled in the DTS file. Problems with it or they were rushed getting Ubuntu out

rooted · Post by **rooted** » Fri Feb 05, 2016 3:10 pm

I must find a eMMC not in use and get this installed, I have been giving Android a run through today.

Post by **mdrjr** » Fri Feb 05, 2016 3:46 pm

A kernel update will be available within 1 hour addressing the following issues:

i2c support is now fixed
i2s hifi dac board is supported.

venkatbo · Post by **venkatbo** » Fri Feb 05, 2016 3:59 pm

Overall a very spiffy experience - Nice !!! Good job, HK.

Observed a couple of low level issues on following ubuntu:

Code: Select all

$ uname -a
Linux odroid64 3.14.29-18 #1 SMP PREEMPT Thu Feb 4 14:45:38 BRST 2016 aarch64 aarch64 aarch64 GNU/Linux

Following set of update/upgrade commands (almost - see workaround below) work:

Code: Select all

$ sudo apt-get update && sudo apt-get upgrade 
$ sudo apt-get dist-upgrade && sudo apt-get install linux-image-c2
$ sudo apt-get autoremove

----

1) Possible LOCALE setup error:

Code: Select all

$ sudo apt-get dist-upgrade
...
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
	LANGUAGE = "en",
	LC_ALL = (unset),
	LC_TIME = "en",
	LC_MONETARY = "en",
	LC_ADDRESS = "en",
	LC_TELEPHONE = "en",
	LC_NAME = "en",
	LC_MEASUREMENT = "en",
	LC_IDENTIFICATION = "en",
	LC_NUMERIC = "en",
	LC_PAPER = "en",
	LANG = "en_US.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to a fallback locale ("en_US.UTF-8").
locale: Cannot set LC_ALL to default locale: No such file or directory
...

2) Package being held back:

For same dist-upgrade cmd above, the usb-modeswitch-data package is being held-back.
NOTE: Please do not upgrade this package per warning further in thread.

3) Unsure if the following is a real issue or not - could be wrong on it...
Time zone change is not taking effect immediately. Had to reboot for it to take effect. Normally, after a few moments of changes, if 'net connection exists, it should pick up from some NTP server somewhere, I thought...

Post by **mdrjr** » Fri Feb 05, 2016 4:06 pm

1. Set your correct locale on the UI on the first time usage. (System -> Control-center -> Language support options)
It will generate the locales automatically.

2. Don't touch usb-modeswitch-data package. Its a issue with Ubuntu ARM64 packages ATM.
If you force the installation of the older package many stuff will be broken.

venkatbo · Post by **venkatbo** » Fri Feb 05, 2016 4:24 pm

Another non-high-priority issue:

Though lsusb shows that the HK offered USB-based GPS dongle is detected, the data captured seems inaccurate

Code: Select all

odroid@odroid64:~$ sudo cat /dev/ttyACM0 | grep GPRMC
$GPRMC,,V,,,,,,,,,,N*53
$GPRMC,,V,,,,,,,,,,N*53
$GPRMC,,V,,,,,,,,,,N*53
...

Should be something like:[code]$GPRMC,161053.00,A,3719.54074,N,12201.49867,W,0.079,,110415,,,A*65

Post by **mdrjr** » Fri Feb 05, 2016 4:44 pm

That's correct.. you don't have a GPS fix .. that's why.

venkatbo · Post by **venkatbo** » Sun Feb 07, 2016 12:58 pm

Upgraded entire image just minutes ago (7:30pm US-PST).

Code: Select all

$ uname -a
Linux c2-1 3.14.29-19 #1 SMP PREEMPT Fri Feb 5 05:02:45 BRST 2016 aarch64 aarch64 aarch64 GNU/Linux.

Firefox 44.0.1 crashes often - at random events. Following are some of the activities, when firefox crashes:
. search from edit field on top bar (with Google set as the search engine)
. search from the google.com homepage
. when a download of file is initiated from a link
. in the middle of a download.

This version 44.0.1 is slightly better than the previous version that was available about 24hrs back. That version would crash much more often.

Following output may be helpful:

Code: Select all

$ strace firefox
...
gettimeofday({1454817159, 310451}, NULL) = 0
--- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x3985e82dcf2400} ---
unlinkat(AT_FDCWD, "/home/odroid/.mozilla/firefox/fyiaqdmx.default/lock", 0) = 0
close(7)                                = 0
rt_sigaction(SIGSEGV, {SIG_DFL, [], 0}, NULL, 8) = 0
rt_sigprocmask(SIG_UNBLOCK, [SEGV], NULL, 8) = 0
tgkill(2041, 2041, SIGSEGV)             = 0
--- SIGSEGV {si_signo=SIGSEGV, si_code=SI_TKILL, si_pid=2041, si_uid=1000} ---
+++ killed by SIGSEGV +++
Segmentation fault

The timezone is set at: US/Los Angeles. (PST)

Post by **odroid** » Sun Feb 07, 2016 1:13 pm

It is the main reason why I've used the Midori browser.

rooted · Post by **rooted** » Sun Feb 07, 2016 1:33 pm

Does chrome not work?

XeoSal · Post by **XeoSal** » Sun Feb 07, 2016 1:37 pm

rooted wrote:Does chrome not work?

I am trying to build it for arm64 right now, lots of difficulties because almost every variable is configured for armhf. ATM, there is no chromium-browser built for aarch64, I will keep trying though with different version, maybe v45 will be fine building instead of v48 I don't know.

Post by **odroid** » Sun Feb 07, 2016 1:44 pm

We also tried to build the Chromium. But it was almost impossible as XeoSal mentioned.

XeoSal · Post by **XeoSal** » Sun Feb 07, 2016 1:52 pm

odroid wrote:We also tried to build the Chromium. But it was almost impossible as XeoSal mentioned.

These settings inside for GYP took me somewhere however, still some conflicts aborting compilation:

Code: Select all

ifeq (arm64,$(DEB_HOST_ARCH_CPU))
GYP_DEFINES += \
	arm_neon=1 \
	target_arch=arm64 \
	use_cups=1 \
	use_openssl=1 \
	v8_use_arm_eabi_hardfloat=true \
	arm_float_abi=hard \
	arm_thumb=1 \
	armv8=1 \
	arm_version=8 \
	$(NULL)
 endif

rooted · Post by **rooted** » Sun Feb 07, 2016 2:48 pm

They are building a chromium port on Tizen which is arm64, perhaps it would help to look at their flags

https://build.tizen.org/package/show?pa ... n%3AMobile

rooted · Post by **rooted** » Sun Feb 07, 2016 4:01 pm

I'm not sure if qupzilla is in the repo but it should be, it's quite a good browser and definitely supports arm64.

crashoverride · Post by **crashoverride** » Tue Feb 09, 2016 6:33 am

When booting this image for the first time, the device boots to the login screen and then suddenly reboots. I am guessing this is due to resizing partitions, but it makes it seem like the board crashed and rebooted.

The choice of default wallpaper needs to be re-thought. The current wallpaper features a star pattern. Unfortunately, this is reminiscent of display issues with bad memory and also similar to issues with bad HDMI cables.

Regarding SD card compatibility, I have tested with a Samsung 8GB MB-MS8GA. This card is known to fail in the original C1. It appears to operate correct and at full speed (class 6) with the C2.

meveric · Post by **meveric** » Tue Feb 09, 2016 7:16 am

I just got my C2 today (this evening to be exactly) I tried an 32GB Intenso micro SDHC card, which also is known to cause issues on the C1. Worked flaweless on the C2.

rooted · Post by **rooted** » Tue Feb 09, 2016 8:40 am

That's great news, improved SD compatibility.

crashoverride · Post by **crashoverride** » Tue Feb 09, 2016 8:46 am

The Mono (C#) runtime also appears to be a 64bit casualty. It does not appear for the arm64 architecture in the official mono repo, the ubuntu repo, or the debian repo.

[Edit]
I attempted to build it from source but it fails due to https://github.com/mono/mono/blob/maste ... -codegen.h referencing a file that does not exist. This led me to this link https://bugzilla.xamarin.com/show_bug.cgi?id=38223 which states that mono 64bit ARM support is not open source. This means it will not be available for any distro.

My long term solution will be to use CoreCLR (Microsoft open source) in its place once its available.

crashoverride · Post by **crashoverride** » Tue Feb 09, 2016 10:14 am

I also noticed my good friend "perl: warning: Setting locale failed." is back. This is again due to the /home/odroid/.pam_environment file.

Symptoms:

Code: Select all

perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
        LANGUAGE = "en",
        LC_ALL = (unset),
        LC_TIME = "en",
        LC_MONETARY = "en",
        LC_ADDRESS = "en",
        LC_TELEPHONE = "en",
        LC_NAME = "en",
        LC_MEASUREMENT = "en",
        LC_IDENTIFICATION = "en",
        LC_NUMERIC = "en",
        LC_PAPER = "en",
        LANG = "en_US.UTF-8"
    are supported and installed on your system.
perl: warning: Falling back to a fallback locale ("en_US.UTF-8").
locale: Cannot set LC_ALL to default locale: No such file or directory

Edit /home/odroid/.pam_environment as follows to resolve the errors:

Code: Select all

LC_ALL=en_US.UTF-8
PAPERSIZE=a4
LANGUAGE=en
LANG=en_US.UTF-8

[Edit]
Note that you must log out and back in for the changes to take effect.

wrxtasy · Post by **wrxtasy** » Tue Feb 09, 2016 6:06 pm

XeoSal wrote:
rooted wrote:Does chrome not work?
I am trying to build it for arm64 right now, lots of difficulties because almost every variable is configured for armhf. ATM, there is no chromium-browser built for aarch64, I will keep trying though with different version, maybe v45 will be fine building instead of v48 I don't know.

odroid wrote: 5. Ubuntu 16.04(LTS) ARM64 has some missing packages i.e., Chromium. It seems to have some compiling/linking issue with 64bit gcc/glibc.

EDIT: Tweaked the GCC options and progress !

XeoSal · Post by **XeoSal** » Tue Feb 09, 2016 9:22 pm

@mdrjr there is a problem that's very annoying caused by the "C2_init" fix script, it happens upon updating the initramfs either manually or automatically by an application requiring it for example, try this "sudo update-initramfs -c -t -k 3.14.29-18" then you should notice a small strange square showing up at the top left of your screen, it happens in both Debian/Ubuntu though.

The only way to get rid of the annoying square is by rebooting the device, unfortunately. I think we really need to solve this before you guys decide to ship the product on sales by March. By the way, the mali-fbdev package on your repository needs "Replaces: mesa-foo, etc" so it doesn't conflict with the default mesa packages.

Regarding FireFox on Ubuntu, currently it's considered not even useable we can't use this as it crashes every second, although Iceweasel running on Debian Jessie is much better and it's very useable, this was the only way I could enjoy web surfing on my ODROID-C2.

P.S. Can you provide us with the Mali-X11 binary blobs? I want to experiment witt them, I think we can get some limited 3d acceleration for "xorg" through the fbturbo DDX driver with GLESv2.

venkatbo · Post by **venkatbo** » Tue Feb 16, 2016 2:41 pm

With the latest image:

Code: Select all

odroid@odroid64:~$ uname -a
Linux odroid64 3.14.29-21 #1 SMP PREEMPT Sun Feb 14 23:04:01 BRST 2016 aarch64 aarch64 aarch64 GNU/Linux

The clock on the desktop (top right corner on system menu bar) updates to the correct date/time ~50 seconds after boot/login.

So I added the following to /etc/rc.local (before exit line):

Code: Select all

( /etc/init.d/ntp stop
until ping -nq -c3 8.8.8.8; do
   echo "Waiting to setup ntp based clock..."
done
ntpd -gq
/etc/init.d/ntp start )&

Now, the clock gets updated ~5 secs after boot/login (checked by running date cmd from linux prompt), but the clock on the desktop takes 20+ seconds to update.

venkatbo · Post by **venkatbo** » Wed Feb 17, 2016 1:05 pm

I updated the system just now (02/16/16 8PM PST). While the version info remains unchanged, some components did get updated/upgraded.

Noticed something unusual - I've lost the ability to add shortcuts to the desktop, via "Add this launcher to desktop" menu item. However, I can launch the app and also add it to the panel on top.

Post by **mdrjr** » Fri Feb 19, 2016 12:36 am

venkatbo wrote:I updated the system just now (02/16/16 8PM PST). While the version info remains unchanged, some components did get updated/upgraded.

Noticed something unusual - I've lost the ability to add shortcuts to the desktop, via "Add this launcher to desktop" menu item. However, I can launch the app and also add it to the panel on top.

Nothing ODROID specific.. Ubuntu may have changed this

venkatbo · Post by **venkatbo** » Fri Feb 19, 2016 3:29 am

I did not mention clearly... actually prior to updating, I was able to add the shortcuts... and lost the ability after updating/upgrading. So presumed something from our side caused this... If ubuntu changed between these two C2 images, then that could be the cause.

Will reflash the original images, and retry updating/upgrading just to make sure the update/upgrade process did not ruin things...

thanks.

Post by **mdrjr** » Fri Feb 19, 2016 3:39 am

venkatbo wrote:I did not mention clearly... actually prior to updating, I was able to add the shortcuts... and lost the ability after updating/upgrading. So presumed something from our side caused this... If ubuntu changed between these two C2 images, then that could be the cause.

Will reflash the original images, and retry updating/upgrading just to make sure the update/upgrade process did not ruin things...

thanks.

That's correct. As I said.. Nothing from our side.. We didn't pushed any new packages or updates.

Those changes your are complaining came directly from Canonical Repos.

Remember that here, we use Canonical repos..
We add hardkernel repo only containing the board specific stuff. (Kernel, bootloader, GPU,...).
There's no change to any other part of the system.

Post by **odroid** » Sun Feb 28, 2016 5:44 pm

We've released a new version of Ubuntu image (v1.1)
http://odroid.com/dokuwiki/doku.php?id= ... _note_v1.1

chlorisdroid · Post by **chlorisdroid** » Tue Mar 01, 2016 5:34 pm

Tested. Works fine. But to get MIDI running a lot of tweaking is necessary.
-recompile the kernel
-strip off aufs (don't need docker)
-activate a lot of kernel modules for audio support
-remove all blacklisted modules from Ubuntu file system.
-install Midi-Software
-configure audio system properly to get everything running.
But... it works - with external Midi-Instruments, too

otec · Post by **otec** » Wed Mar 09, 2016 5:43 am

How can I disable sit0 and ip6tnl0 interfaces ?

Code: Select all

ip6tnl0   Link encap:UNSPEC  HWaddr 00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00
          UP RUNNING NOARP  MTU:1452  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

sit0      Link encap:IPv6-in-IPv4
          UP RUNNING NOARP  MTU:1480  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

Blacklisting sit module in blacklist.conf, doesn't work as it seems that sit isnt a module.

I disabled ipv6 in sysctl.d by these:

Code: Select all

net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
net.ipv6.conf.lo.disable_ipv6 = 1

but those interfaces still show up.

Post by **mdrjr** » Wed Mar 09, 2016 6:01 am

Here:

Edit ~ line 73 as on the example file below
https://github.com/mdrjr/c2_bootini/blo ... ot.ini#L73

and on the end add ipv6.disable=1

otec · Post by **otec** » Wed Mar 09, 2016 6:37 am

mdrjr wrote:Here:

Edit ~ line 73 as on the example file below
https://github.com/mdrjr/c2_bootini/blo ... ot.ini#L73

and on the end add ipv6.disable=1

Forgot to mention that I have this. Interfaces still show up in ifconfig.

otec · Post by **otec** » Thu Mar 10, 2016 2:29 am

Any other ideas how to disable those interfaces?

Sent from my Nexus 6P using Tapatalk

Post by **mdrjr** » Thu Mar 10, 2016 2:32 am

Nope, sorry.

Obiwantje · Post by **Obiwantje** » Thu Mar 10, 2016 4:19 am

Is there any way I can disable overscan? My TV does not support this.

Hdlife · Post by **Hdlife** » Thu Mar 10, 2016 5:05 am

hi,

i'm not sure, if it's the good topic to talk about my issues (only 2)

.
I begin :
After update and upgrade by sudo apt update and sudo apt upgrade. (the installation was finish correctly)

My mouse didn't work, i tried with a another mouse it's the same case. i need to reinstall the os on the sd card.

I'm french, i would like to use ubuntu on this language, but i installed all french language package, remove another language, and english stay by fault with french's selected.

Maybe can you help me ?

thks a lot

Post by **mdrjr** » Thu Mar 10, 2016 5:46 am

Hdlife wrote:hi,

i'm not sure, if it's the good topic to talk about my issues (only 2) .
I begin :
After update and upgrade by sudo apt update and sudo apt upgrade. (the installation was finish correctly)

My mouse didn't work, i tried with a another mouse it's the same case. i need to reinstall the os on the sd card.

I'm french, i would like to use ubuntu on this language, but i installed all french language package, remove another language, and english stay by fault with french's selected.

Maybe can you help me ?

thks a lot

Please create your own thread and we'll take from there ok?

mlinuxguy · Post by **mlinuxguy** » Thu Mar 10, 2016 8:00 am

otec wrote:Any other ideas how to disable those interfaces?

Sent from my Nexus 6P using Tapatalk

You will have rebuild the kernel and disable sit
# grep -i SIT /root/linux/.config
CONFIG_IPV6_SIT=y

If they hadn't built that module into the kernel you could have just blacklisted the sit module

otec · Post by **otec** » Thu Mar 10, 2016 5:59 pm

mlinuxguy wrote:
otec wrote:Any other ideas how to disable those interfaces?

Sent from my Nexus 6P using Tapatalk
You will have rebuild the kernel and disable sit
# grep -i SIT /root/linux/.config
CONFIG_IPV6_SIT=y

If they hadn't built that module into the kernel you could have just blacklisted the sit module

That's what I suspected.

Wonder why sit isn't build as a module ....

Sent from my Nexus 6P using Tapatalk

otec · Post by **otec** » Thu Mar 10, 2016 6:02 pm

Is there a guide somewhere how to rebuild the kernel?

Sent from my Nexus 6P using Tapatalk

elatllat · Post by **elatllat** » Fri Mar 11, 2016 1:42 am

otec wrote:Is there a guide somewhere how to rebuild the kernel?

Sent from my Nexus 6P using Tapatalk

http://odroid.com/dokuwiki/doku.php?id= ... ing_kernel

otec · Post by **otec** » Fri Mar 11, 2016 2:14 am

Thanks!

Sent from my Nexus 6P using Tapatalk

dkfuel · Post by **dkfuel** » Tue Mar 15, 2016 5:44 am

I just want to make sure I'm not spinning my wheels here, as I've reformatted and re-imaged my SD Card 3 times now. When I run apt-get upgrade I get the following issue:

Code: Select all

The following packages have unmet dependencies:
 libperl5.22 : Depends: perl-modules-5.22 (>= 5.22.1-9) but 5.22.1-7 is installed
 perl : Depends: perl-base (= 5.22.1-9) but 5.22.1-7 is installed
        Depends: perl-modules-5.22 (>= 5.22.1-9) but 5.22.1-7 is installed
        Recommends: rename but it is not installed
E: Unmet dependencies. Try using -f.

and sudo apt-get -f install results in:

Code: Select all

The following additional packages will be installed:
  perl-base perl-modules-5.22
The following packages will be upgraded:
  perl-base perl-modules-5.22
2 upgraded, 0 newly installed, 0 to remove and 279 not upgraded.
3 not fully installed or removed.
Need to get 0 B/3,781 kB of archives.
After this operation, 8,192 B disk space will be freed.
Do you want to continue? [Y/n] 
(Reading database ... 179339 files and directories currently installed.)
Preparing to unpack .../perl-base_5.22.1-9_arm64.deb ...
Unpacking perl-base (5.22.1-9) over (5.22.1-7) ...
dpkg: error processing archive /var/cache/apt/archives/perl-base_5.22.1-9_arm64.deb (--unpack):
 unable to stat './usr/lib/aarch64-linux-gnu/perl-base/auto/POSIX/POSIX.so' (which I was about to install): Input/output error
dpkg-deb: error: subprocess paste was killed by signal (Broken pipe)
Processing triggers for man-db (2.7.5-1) ...
Errors were encountered while processing:
 /var/cache/apt/archives/perl-base_5.22.1-9_arm64.deb
E: Sub-process /usr/bin/dpkg returned an error code (1)

Any effort to remedy these issues prove futile. Should I just ignore for now or is running apt-get upgrade just a bad idea overall for now while all of the bugs get worked out? Can I function like this or should I reformat and re-image again and then not run apt-get update/upgrade at all? I have people waiting on me to deploy this as a kiosk for a 4K monitor here in our ad agency. Do I need to tell them that this device is not ready for prime time yet?

Better yet, is there a 32-bit build that runs smoothly on this (for instance, you can install the 32-bit Raspbian on the 64-bit Raspberry Pi 3 and have it run with no issues)? I just don't know if you can do 4K on a 32-bit build and that's essential for the project.

Thanks in advance...

Post by **mdrjr** » Tue Mar 15, 2016 9:32 am

What kind of sdcard you are using..

What I'm seeing here is IO error.
Looks to me like a microsd issue.

dkfuel · Post by **dkfuel** » Tue Mar 15, 2016 11:22 am

I *was* using a Silicon Power 32GB up to 85MB/s MicroSDHC UHS-1 Class10 (SP032GBSTHBU1V20SP) purchased from Amazon. I've switched to a Sandisk 32Gig card (don't have the model number at the moment) and it works flawlessly. I wish I had thought to try another card before wasting my time on the Silicon Power card -- but thanks for pointing me in the right direction. I'm up and running now and apt-get upgrades are working perfectly!

I'll need to figure out if it is an incompatiblity with the card or if that particular card is defective. "The game's afoot!"

*EDIT* I have two of the Silicon Power cards and both were horrifically unstable in the C-2. I even wrote zeros to one and retried it and it was still bad. So avoid the Silicon Power SP032GBSTHBU1V20SP in this device, unless you enjoy headaches.

I appreciate your time and help.

William

sboyce · Post by **sboyce** » Fri Mar 18, 2016 12:32 am

** If my Ubuntu 16.04 problems are not experienced by anyone else, the UK suppliers are willing to replace my -C2. ***
** When I tried Android, it locks up with the blue LED solid on.

Using the same +5V power and peripherals move over from the -C1.

Booted up with a new Sandisk 64 GB Ultra HC-1 class 10. The -C1 also used the same type of 64GB SDHC card.
Did the update and during the dist-upgrade the -C2 froze not allowing normal mouse and keyboard operations in X or in VC though the mouse pointer was operating and the blue LED was flashing.

Tried Sandisk Ultra HC-1 16GB and 32GB cards with the same problem encountered.

"sudo apt-get upgrade" ran for quite a while until the -C2 froze with the blue LED still flashing.
The only way to recover is "ALT-SYSRQ-s", "ALT-SYSRQ-u" followed by "ALT-SYSRQ-B" and when it comes up it doesn't boot fully and shortly after it freezes.

Moved the SD card to na x86_64 Ubuntu PC.
root@sdrbox:~# fsck -a /dev/sdd1
fsck from util-linux 2.26.2
fsck.fat 3.0.28 (2015-05-16)
0x25: Dirty bit is set. Fs was not properly unmounted and some data may be corrupt.
Automatically removing dirty bit.
Performing changes.
/dev/sdd1: 5 files, 10304/65399 clusters
root@sdrbox:~# fsck -a /dev/sdd1
fsck from util-linux 2.26.2
fsck.fat 3.0.28 (2015-05-16)
/dev/sdd1: 5 files, 10304/65399 clusters
root@sdrbox:~# fsck -a /dev/sdd2
fsck from util-linux 2.26.2
rootfs was not cleanly unmounted, check forced.
rootfs: Deleted inode 958 has zero dtime. FIXED.
rootfs: Deleted inode 982 has zero dtime. FIXED.
rootfs: Deleted inode 992 has zero dtime. FIXED.
rootfs: Deleted inode 1050 has zero dtime. FIXED.
etc., etc.
rootfs: Deleted inode 128886 has zero dtime. FIXED.
rootfs: Entry 'console_setup' in /usr/share/initramfs-tools/scripts/panic (20300) has deleted/unused inode 20302. CLEARED.
rootfs: Entry 'xorg.lst' in /usr/share/doc/keyboard-configuration (23337) has an incorrect filetype (was 7, should be 1).

rootfs: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
(i.e., without -a or -p options)
root@sdrbox:~#

root@sdrbox:~# fsck /dev/sdd2
fsck from util-linux 2.26.2
e2fsck 1.42.12 (29-Aug-2014)
rootfs contains a file system with errors, check forced.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Entry 'xorg.lst' in /usr/share/doc/keyboard-configuration (23337) has an incorrect filetype (was 7, should be 1).
Fix<y>? yes
Entry 'cat1' in /var/cache/man (130211) has an incorrect filetype (was 2, should be 1).
Fix<y>? yes
Entry 'cat3' in /var/cache/man (130211) has an incorrect filetype (was 2, should be 1).
Fix<y>? yes
Entry 'cat4' in /var/cache/man (130211) has an incorrect filetype (was 2, should be 1).
Fix<y>? yes
Entry 'cat5' in /var/cache/man (130211) has deleted/unused inode 1675. Clear<y>? yes
Entry 'cat6' in /var/cache/man (130211) has deleted/unused inode 1678. Clear<y>? yes
Entry 'index.db' in /var/cache/man/bo (2081) has deleted/unused inode 30879. Clear<y>? yes
Entry 'preinst' in /var/lib/dpkg/tmp.ci (393239) references inode 393367 found in group 48's unused inodes area.
etc., etc.
rootfs: ***** FILE SYSTEM WAS MODIFIED *****
rootfs: 196600/4087808 files (0.4% non-contiguous), 1480280/16350848 blocks
yroot@sdrbox:~# fsck /dev/sdd2
fsck from util-linux 2.26.2
e2fsck 1.42.12 (29-Aug-2014)
rootfs: clean, 196600/4087808 files, 1480280/16350848 blocks

"sudo dpkg --configure -a" recovers it.
"sudo apt-get dist-upgrade" successful.
Commands "ls" and "uname -r" OK, "ifconfig" and it freezes.

castillan · Post by **castillan** » Fri Mar 18, 2016 2:17 am

I got the same issue : "sudo apt-get upgrade" ran for quite a while until the -C2 froze with the blue LED still flashing ". - black screen, cannot recover....will probably reload the sd card tomorrow...

sboyce · Post by **sboyce** » Fri Mar 18, 2016 3:00 am

I have just ordered a pre-loaded eMMC to try.

I had one other report that the same problems occur on this guy's.