Forum

irexec is working fine at boot.

(1st Nov, 2013 10:44 PM)zilexa Wrote: [ -> ]Is this the link to the latest img?
http://xbian.brantje.com/devel/others/xbian.beta2.7z

Since I can't use my RPi for playback atm I might as well use it to test beta2.
Any special setup required after writing img to sd?

@zilexa

I wrongly labeled the filesystem, so have to edit the image and reupload. so either wait 2-3-4 hours or grab it and update cmdline.txt in /boot by changing root=LABEL=xbian-copy to root=LABEL=xbian-beta2

---

(2nd Nov, 2013 12:42 AM)CurlyMo Wrote: [ -> ]Root partition LABEL=xbian-copy missing
mount: mounting LABEL=xbian-copy on /rootfs failed: No such file or directory
mount: mount point /rootfs/proc does not exist
mount: mount point /rootfs/boot does not exist
mount: mount point /rootfs/dev does not exist
mount: mount point /rootfs/sys does not exist
mount: mount point /rootfs/lib/modules does not exist

introduced --fs-label to allow custom label for clone, changed on all places where used but not when creating cmdline.txt . already corrected the tool, will correct img and reupload. correct LABEL is xbian-beta2.

---

(1st Nov, 2013 10:13 PM)lordgandhi Wrote: [ -> ]@mk01
I just wanted to bring up the issue with inputlirc again.
The autostart of the xbian lirc is disabled.
I attached the log of strace here: http://pastebin.com/dL0ZhHFv

@lordgandhi

inputlirc is lirc daemon provided by Debian upstream repo? Do you have experience with other RPI XBMC installs providing info that it is behaving differently?

what happens if you create secondary char device inside /dev with the same major/minor numbers as input1, input2? would this be a solution?

(2nd Nov, 2013 01:49 AM)mk01 Wrote: [ -> ]

(1st Nov, 2013 10:44 PM)zilexa Wrote: [ -> ]Is this the link to the latest img?
http://xbian.brantje.com/devel/others/xbian.beta2.7z

Since I can't use my RPi for playback atm I might as well use it to test beta2.
Any special setup required after writing img to sd?

@zilexa

I wrongly labeled the filesystem, so have to edit the image and reupload. so either wait 2-3-4 hours or grab it and update cmdline.txt in /boot by changing root=LABEL=xbian-copy to root=LABEL=xbian-beta2

---

(2nd Nov, 2013 12:42 AM)CurlyMo Wrote: [ -> ]Root partition LABEL=xbian-copy missing
mount: mounting LABEL=xbian-copy on /rootfs failed: No such file or directory
mount: mount point /rootfs/proc does not exist
mount: mount point /rootfs/boot does not exist
mount: mount point /rootfs/dev does not exist
mount: mount point /rootfs/sys does not exist
mount: mount point /rootfs/lib/modules does not exist

introduced --fs-label to allow custom label for clone, changed on all places where used but not when creating cmdline.txt . already corrected the tool, will correct img and reupload. correct LABEL is xbian-beta2.

---

(1st Nov, 2013 10:13 PM)lordgandhi Wrote: [ -> ]@mk01
I just wanted to bring up the issue with inputlirc again.
The autostart of the xbian lirc is disabled.
I attached the log of strace here: http://pastebin.com/dL0ZhHFv

@lordgandhi

inputlirc is lirc daemon provided by Debian upstream repo? Do you have experience with other RPI XBMC installs providing info that it is behaving differently?

what happens if you create secondary char device inside /dev with the same major/minor numbers as input1, input2? would this be a solution?

Yep,
it's the inputlirc daemon package provided by the debian wheezy repo.
I didn't test it with other RPI distros, but until beta2 now; from the alphas to beta1.1 it worked.

I'm sorry, but I'm not sure what you mean by your second part, since input1 and 2 correspond to the mouse respective keybooard parts of my MCE remote. I'm not surce if you want me to create ohter devices in /dev and if so I don't know how to do that.

(2nd Nov, 2013 03:02 AM)lordgandhi Wrote: [ -> ]I'm sorry, but I'm not sure what you mean by your second part, since input1 and 2 correspond to the mouse respective keybooard parts of my MCE remote. I'm not surce if you want me to create ohter devices in /dev and if so I don't know how to do that.

ok, that looks then as xbmc is doing this differently now. to be sure you can grab any of those http://xbian.brantje.com/pool/main/x/xbian-package-xbmc/ and test with beta2.

apt will not allow you to downgrade to those versions, but with following few steps you can test & revert back in few minutes:

the btrfs-auto-snapshot command will take just few seconds for creating the point and rollback as well.

(2nd Nov, 2013 03:27 AM)mk01 Wrote: [ -> ]

(2nd Nov, 2013 03:02 AM)lordgandhi Wrote: [ -> ]I'm sorry, but I'm not sure what you mean by your second part, since input1 and 2 correspond to the mouse respective keybooard parts of my MCE remote. I'm not surce if you want me to create ohter devices in /dev and if so I don't know how to do that.

ok, that looks then as xbmc is doing this differently now. to be sure you can grab any of those http://xbian.brantje.com/pool/main/x/xbian-package-xbmc/ and test with beta2.

apt will not allow you to downgrade to those versions, but with following few steps you can test & revert back in few minutes:

Code:

cd /tmp
sudo -i
wget http://xbian.brantje.com/pool/main/x/xbian-package-xbmc/xbian-package-xbmc_2.3-0.1_armhf.deb
btrfs-auto-snapshot snapshot --name xbmctest root
stop xbmc
rm -fr /usr/local/lib/xbmc
rm -fr /usr/local/share/xbmc
dpkg -x xbian-package-xbmc_2.3-0.1_armhf.deb /
start xbmc
.
.
.
btrfs-auto-snapshot rollback root/@xbmctest
reboot

the btrfs-auto-snapshot command will take just few seconds for creating the point and rollback as well.

Ok, thanks will test that tomorrow when I've got the time.

Changing the label from xbian-copy to xbian-beta2 worked.

One small suggestion. When you load the settings of a specific tab in xbian-config-xbmc and try to load the settings from another tab. The second tab waits for the first tab to complete. It would be great to have some sort of threading so you can load various tabs at the same time.

Small issues
- xbian-config doesn't autostart when logging in through ssh.
- when i stop xbmc through service xbmc stop and then start it again i get splash message network started
- when i reboot, it get confronted with the init process and fs resize splashes.
- the default timezone is Europe/Bratislava

That was it till so far.

(1st Nov, 2013 06:58 AM)mk01 Wrote: [ -> ]can you check md5sum img.file?

this is as it should be 7560f1d9b7df6ef780d50d0f07d339cb xbian.beta2.resized.img

MD5 checked and was 7560f1d9b7df6ef780d50d0f07d339cb

(2nd Nov, 2013 03:50 AM)CurlyMo Wrote: [ -> ]Small issues
- xbian-config doesn't autostart when logging in through ssh.
- when i stop xbmc through service xbmc stop and then start it again i get splash message network started
- when i reboot, it get confronted with the init process and fs resize splashes.
- the default timezone is Europe/Bratislava

autostart is fixed, system cleanup tool will revert this always to autorun
i was aware of the "network started" - but until now nobody realized ;-) lol. fixed
the reboot and again resize is triggered by creating swap partition. normally it should be created during first boot if parameter present in cmdline.txt, but wasn't working like that. always after second reboot. as I was changing the label I found why (couldn't find it before for almost two months). should be fixed
heh. what it is by default? system timezone should then be reverted as well.

testing now all the fixes, if ok will re-upload image. removed it for the time being.

Matus

---

(2nd Nov, 2013 04:06 AM)Killerbee Wrote: [ -> ]

(1st Nov, 2013 06:58 AM)mk01 Wrote: [ -> ]can you check md5sum img.file?

this is as it should be 7560f1d9b7df6ef780d50d0f07d339cb xbian.beta2.resized.img

MD5 checked and was 7560f1d9b7df6ef780d50d0f07d339cb

ok, the same. no idea.

wait for the new image and please report back.

Also, consider changing the systemtotaluptime in guisettimgs.xml to zero on first boot:

Quote:One small suggestion. When you load the settings of a specific tab in xbian-config-xbmc and try to load the settings from another tab. The second tab waits for the first tab to complete. It would be great to have some sort of threading so you can load various tabs at the same time.

it's in a thread (one thread by loading cat), first version wasn't, but if i remember, i didn't have pushed it, so your version should be threaded.
Normally it should be possible to load multiple cat
but i develop under vnc, and sometimes it freeze, don't know if it's the case without vnc.
you can activate Debug, and check log, you should see if loading is started.

---

Also only somes are freezing (Service,Packages and preference),
loading is same code for all, so i suspect a 'race condition' bug,
i was first believe that processor was full, but it's happens also in preference where code is really simple (read small data on a small file).

i have looking a bit before, but didn't find what's the problem. i've found a bug with progress dialog and xbmcbultin call, and it take me very long time to find. if i've found time, will look again, but it's not a priority for me

and there is a lock around xbian-config call, i use now only one bash session, and communicate via stdin/stdout. So only one call is allowed at the same time.
pipe & co is not my cup of tea, do you think it's possible that each call has stdout redirect to a different pipe?
but this is not the main problem i think as preference has no xbian-config call

The reboots after the latest updates don't seem to reboot but merely shut down. I need to unplug and replug the power cable to get XBian booting again.

@Smultie so you are on the happy users having shutdown sequence longer than 30s ??? I will recheck whether I didn't switch the run levels.

just out of curiosity. Can you post your ls -la /etc/rc0.d and ls -la /etc/rc6.d ?

---

@all

image uploaded at http://xbian.brantje.com/devel/others/xbian.beta2.2013-01-11.7z

(2nd Nov, 2013 07:18 AM)mk01 Wrote: [ -> ]@Smultie so you are on the happy users having shutdown sequence longer than 30s ??? I will recheck whether I didn't switch the run levels.

just out of curiosity. Can you post your ls -la /etc/rc0.d and ls -la /etc/rc6.d ?

Sure:

xbian@xbian ~ $ ls -la /etc/rc0.d
total 32
drwxr-xr-x 1 root root 186 Nov 1 21:59 .
drwxr-xr-x 1 root root 3142 Nov 1 22:04 ..
lrwxrwxrwx 1 root root 15 Nov 1 21:59 K01samba -> ../init.d/samba
lrwxrwxrwx 1 root root 29 Nov 1 21:59 K02unattended-upgrades -> ../init.d/unattended-upgrades
lrwxrwxrwx 1 root root 18 Nov 1 08:00 K03sendsigs -> ../init.d/sendsigs
lrwxrwxrwx 1 root root 22 Nov 1 08:00 K04umountnfs.sh -> ../init.d/umountnfs.sh
lrwxrwxrwx 1 root root 18 Nov 1 08:00 K05umountfs -> ../init.d/umountfs
lrwxrwxrwx 1 root root 20 Nov 1 08:00 K06umountroot -> ../init.d/umountroot
lrwxrwxrwx 1 root root 14 Nov 1 08:00 K07halt -> ../init.d/halt
-rw-r--r-- 1 root root 353 Oct 15 2012 README

xbian@xbian ~ $ ls -la /etc/rc6.d
total 32
drwxr-xr-x 1 root root 190 Nov 1 21:59 .
drwxr-xr-x 1 root root 3142 Nov 1 22:04 ..
lrwxrwxrwx 1 root root 15 Nov 1 21:59 K01samba -> ../init.d/samba
lrwxrwxrwx 1 root root 29 Nov 1 21:59 K02unattended-upgrades -> ../init.d/unattended-upgrades
lrwxrwxrwx 1 root root 18 Nov 1 08:00 K03sendsigs -> ../init.d/sendsigs
lrwxrwxrwx 1 root root 22 Nov 1 08:00 K04umountnfs.sh -> ../init.d/umountnfs.sh
lrwxrwxrwx 1 root root 18 Nov 1 08:00 K05umountfs -> ../init.d/umountfs
lrwxrwxrwx 1 root root 20 Nov 1 08:00 K06umountroot -> ../init.d/umountroot
lrwxrwxrwx 1 root root 16 Nov 1 08:00 K07reboot -> ../init.d/reboot
-rw-r--r-- 1 root root 351 Oct 15 2012 README

@Smultie are you running any transmissiond, or SickBears or anything? or you have over 200 shares mounted on RPI ?

None of those
I just used "check deb upgrades" after which XBian told me it needed a reboot.
I pressed "Yes" to let it reboot, the reboot sequence started with the XBian-logo.
Normally it would reboot after that, but now: only a black screen.
I also couldn't SSH in without powercycling it by removing the power cable.