With my image, don't worry about this. ^^^ As long as there is no
"uEnv.txt" file on the microSD, u-boot will always use the factory one
i installed in the eMMC. And since it uses uuid's instead of the raw
partition name, it'll always find the "rootfs" partition no matter
what. So just blank/format your microSD as a simple ext4 partition.
Regards,
PS: if you want to get crazy, the same script that generated this
image, can also build a version of debian that'll fit in 64MB. But at
that point all you have is perl/apt-get/dpkg..
Regards,
Robert,
that was a fast one. Thanks again - discussions in other groups can
sometimes lead you to nowhere and you get lost. (There was written
that the uEnv.txt was mandatory) So I will not use your 64 MB image.
Sorry, I am happy that this one is running.
Have a nice day
and so long from Nowhere man
Hajo
Yeah, this requirement was a bug in the version of u-boot shipped with
the board.
You can see how i worked around the issue here:
Essentially the original factory u-boot, only checked for the presense
of the microSD, if found it would try to boot with it no matter what.
Instead, I set it up, to search for a uEnv.txt, try to load it and
test if "uenvcmd" was set. Thus a little more error proof..
Regards,
Robert,
Thanks for your response! I have 15 BBBs and one uSD card, so I am kinda leaning toward using the eMMC on each.
Robert said:
Correct, to meet everyone’s out of box pkg requirements, the eMMC is
mostly full. If you drop opencv/python/chromium you’ll gain a lot of
space back.
Pardon my ignorance, but is there a “magic scalpel” command to free up all space related to Opencv? …and then separately, Chromium? Something is not right with the approach I tried…
After this operation, 27.8 MB disk space will be freed.
Do you want to continue [Y/n]?
…maybe I just need to learn how to use “aptitude” ? I am guessing that an “uninstall” or “remove” or “autoremove” command is better than going through and deleting random directories full of opencv/chromium related things?
Thanks for any insights!!!
ps- I decided to remove all documentation on one of my BBBs, so I did this…
rm /usr/share/doc -R
…that seemed to free up 91MB, but it still wasnt enough!!!
Ran out of space again during “make”!!!
ola-rdm-discover.cpp:232:1: fatal error: closing dependency file .deps/ola-rdm-discover.Tpo: No space left on device
compilation terminated.
The bug is not reproducible, so it is likely a hardware or OS problem.
make[2]: *** [ola-rdm-discover.o] Error 1
make[2]: Leaving directory /usr/local/src/ola/examples' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory /usr/local/src/ola’
make: *** [all] Error 2
Thx!
-frenchy
Robert,
Thanks for your response! I have 15 BBBs and one uSD card, so I am kinda
leaning toward using the eMMC on each.Robert said:
Correct, to meet everyone's out of box pkg requirements, the eMMC is
mostly full. If you drop opencv/python/chromium you'll gain a lot of
space back.Pardon my ignorance, but is there a "magic scalpel" command to free up all
space related to Opencv? ...and then separately, Chromium? Something is
not right with the approach I tried...apt-get autoremove opencv*
After this operation, 27.8 MB disk space will be freed.
Do you want to continue [Y/n]?
apt-get remove libopencv-* --purge ; apt-get autoremove
apt-get autoremove chromium*
So due to the build requirements, chromium is currently not a *.deb
package. I'd like to change this. But to give you an idea, it takes a
Quad Core Cortex A9, running at 1.2Ghz with 2GB of ram and a 7200rpm
sata drive 8 hours to build..
SO just:
rm -rf /usr/lib/chromium/
rm -f /usr/bin/chromium
...maybe I just need to learn how to use "aptitude" ? I am guessing that an
"uninstall" or "remove" or "autoremove" command is better than going through
and deleting random directories full of opencv/chromium related things?Thanks for any insights!!!
ps- I decided to remove all documentation on one of my BBBs, so I did
this...
rm /usr/share/doc -R
...that seemed to free up 91MB, but it still wasnt enough!!!
you can also dump:
/usr/share/man/
Regards,
Robert,
How do you propose that users switch to v3.13.x (by which I assume you mean the Linux kernel version)?
My BBB panics like this on every boot using the Debian 2014-03-04 image. I can’t even log in to the text console.
Dennis Cote
"panics" on every boot?
Do you have any error log? I can't really help with that limited info.
Easitest thing to do is, grab the non-flasher:
https://s3.amazonaws.com/beagle-debian/bone-debian-7.4-2014-03-04-2gb.img.xz
flash it to a microSD card..
mount the first fat partition, edit "uEnv.txt" remove the "quiet" from
optargs.. save unmount..
Next using a usb-serial convert log the full serial boot log for me.
Regards,
Dennis,
I am not sure what you mean by “BBB panics”, but I can tell you that I have been upgrading my kernel using the new Debian eMMC flasher image. I have done it on several BBBs so far with no problem…
Before the procedure:
uname -a
Linux beaglebone 3.8.13-bone41 #1 SMP Tue Mar 4 22:51:47 UTC 2014 armv7l GNU/Linux
Then Update Kernel:
cd /opt/scripts/
git pull
./tools/update_kernel.sh --beta-kernel
reboot
And after the procedure:
uname -a
Linux beaglebone 3.13.6-bone7 #1 SMP Sat Mar 8 01:11:45 UTC 2014 armv7l GNU/Linux
Hope this helps…
-frenchy
I'm not sure if you're talking about a non-patched 3.13.x, but I've recently tried (plain) 3.13.6 and received a panic (oops) straight on boot in musb (a bt-dongle was connected on boot). So I've just switched back to some heavy patched 3.11 I haven't many problems with.
Sorry, no log too and I'm currently too lazy to fiddle with the bone and produce one, but it might be a hint to try booting with some usb-device connected.
Regards,
Alexander Holler
He has done something wrong, or has missed a step in setting up his media.
Having a serial debug cable on the board and working would clear things up in a hurry.
“panics” on every boot?
Do you have any error log? I can’t really help with that limited info.
When I said “panics like this” I meant in the same way that as the user whose message you replied to. I have copied his log below. On my BBB it usually happens just after the 120 second mark, but I have also seen it happen at 180 seconds as below. The rest of the messages are identical (same addresses etc.).
[ 180.537526] INFO: task mmcqd/0:74 blocked for more than 60 seconds.
[ 180.544275] “echo 0 > /proc/sys/kernel/hung_task_timeout_secs” disables this message.
[ 180.552668] Kernel panic - not syncing: hung_task: blocked tasks
[ 180.559071] [] (unwind_backtrace+0x1/0x8a) from [] (panic+0x51/0x148)
[ 180.567727] [] (panic+0x51/0x148) from [] (watchdog+0x14f/0x194)
[ 180.575937] [] (watchdog+0x14f/0x194) from [] (kthread+0x67/0x74)
[ 180.584234] [] (kthread+0x67/0x74) from [] (ret_from_fork+0x11/0x34)
[ 180.592778] drm_kms_helper: panic occurred, switching back to text console
Easitest thing to do is, grab the non-flasher:
https://s3.amazonaws.com/beagle-debian/bone-debian-7.4-2014-03-04-2gb.img.xz
flash it to a microSD card…
That is what I did, and what I am running.
mount the first fat partition, edit “uEnv.txt” remove the “quiet” from
optargs… save unmount…Next using a usb-serial convert log the full serial boot log for me.
Will do.
Yeah this annoyance..
What brand of microSD cards are you using?
We've back ported a few mmc tweaks from later kernels, yet some cards
still show this issue..
Regards,
I’m using a 4GB Kingston Technology cards. Note, I have not expanded the partition to fill the card yet, I just to boot with it after copying the image.
I’ll get you the complete boot log if that is still of use.
Dennis Cote
Robert,
The uEnv.txt file from the FAT partition in this image (attached) does not have optargs defined. Both sections where optargs appears are commented out.
There is a line that says:
systemd=quiet init=/lib/systemd/systemd
I’m not sure if you mean to remove the quiet from this line, or perhaps something else.
Dennis Cote
uEnv.txt (1.36 KB)
OK, so I removed the quiet from the systemd definition and rebooted my BBB. It did log a lot more stuff. The complete boot log is attached.
Since you seem t think the problem is SD card specific, I will try again with a different card.
HTH
Dennis Cote
bootlog.txt (24.9 KB)
Using a different 8GB SD card my BBB boots this new image as expected. Hopefully the bootlog from the problem card will help identify the issue.
Dennis Cote
Please take the time to give a detailed look over this image and report any issues to the bug tracker on elinux.org:
On booting this new image I noticed a few issues immediately, but I’m not sure if one is by design. The ethernet interface is not setup in this image. Running ifconfig shows no IP address for the ethernet port.
Looking at /etc/network/interfaces I see that the eth0 section is commented out.
I tried adding the following to /etc/network/interfaces but I still have no IP address assigned (as if dhcp wasn’t aquirring an address). Commenting out the allow-hotplug line makes no difference.
auto eth0
allow-hotplug eth0
iface eth0 inet dhcp
Shouldn’t the ethernet interface be setup by default? How does one setup dhcp on Debian if not through the interfaces file?
Also, how is one supposed to use the Root Terminal in the LXDE environment? It won’t accept the admin password for the debian user, and I don’t know if there is a different root password?
Thanks.
Dennis Cote
Please take the time to give a detailed look over this image and report
any issues to the bug tracker on elinux.org:
http://bugs.elinux.org/projects/debian-image-releasesOn booting this new image I noticed a few issues immediately, but I'm not
sure if one is by design. The ethernet interface is not setup in this image.
Running ifconfig shows no IP address for the ethernet port.Looking at /etc/network/interfaces I see that the eth0 section is commented
out.I tried adding the following to /etc/network/interfaces but I still have no
IP address assigned (as if dhcp wasn't aquirring an address). Commenting out
the allow-hotplug line makes no difference.auto eth0
allow-hotplug eth0
iface eth0 inet dhcpShouldn't the ethernet interface be setup by default? How does one setup
dhcp on Debian if not through the interfaces file?
The wicd deamon should setup eth0 within 30 seconds on 2nd boot.
(first boot there is a slight delay as the ssh key's are generated).
If you uncomment out the eth0 interface in /etc/network/interfaces
boot time falls from 15seconds to 35ish..
Also, how is one supposed to use the Root Terminal in the LXDE environment?
It won't accept the admin password for the debian user, and I don't know if
there is a different root password?
Nice catch, i need to remove that application from the menu. It was
requested to blank out the root password for ease of use.
Regards,
The wicd deamon should setup eth0 within 30 seconds on 2nd boot.
(first boot there is a slight delay as the ssh key’s are generated).
I had rebooted many times. I hadn’t paid any attention to the wicd program since it seemed to be for WIFI which I am not using.
I had to open the preferences and check the option to always show wired interfaces (I also checked always switch to wired connection when available) before I saw the eth0 connection. I then clicked the connect button and waited until it finished. It still didn’t connect, but after I rebooted again, it did connect using dhcp. Since then it has been connecting on each boot.
If you uncomment out the eth0 interface in /etc/network/interfaces
boot time falls from 15seconds to 35ish…
I’m not sure what you mean by this. “unccomment out” is ambiguous. Did you mean “uncomment”, or did you mean “comment out”.
Is it faster to boot with the eth0 defined in /etc/network/interfaces, or is it faster with the eth0 section commented out? Why? Is it redundant to define eth0 here, and then have wicd also connect eth0; or does having defined in the interafces file cause wicd to skip its redundant setup later?
On an unrelated issue, how do you setup the timezone for the time display on the LXDE desktop? I have set /etc/timezone. I have also used tzselect and added the TZ environment variable to my .profile as suggested in the output of tzselect. It does not appear that my .profile file is being executed though. I have a user bin directory in my /home/debian directory, and it is not being added to the path as it seems it should be by reading the .profile file. I checked and I don’t have a .bash_profile or .bash_login file which would prevent .profile from executing. Any ideas?
Dennis Cote