OESF Portables Forum
Everything Else => Zaurus Distro Support and Discussion => Distros, Development, and Model Specific Forums => Archived Forums => Debian => Topic started by: ZDevil on November 09, 2007, 04:20:19 am
-
==================================
Quick notes on the configured debian eabi rootfs tarball
====================================
I. Introduction
II. Download
III. What is in this rootfs?
IV. Installation steps (brief)
V. Special notes
I. Introduction
This tarball is primarily for use with the yonggun (eabi) kernel. For users of other kernels (such as Angstrom and cortez's Poky), please refer to the special notes at the end.
This rootfs can run on the internal drive, a SD card and a CF card.
The yonggun kernel has been verified to be working on C1000, C3100 and C3200 (but not C3000 as it cannot use uboot due to its limited NAND space).
Feel free to post your questions here.
Update: It is now verified that the rootfs tarball also works fine on C3000 with the Angstrom kernel. You don't need uboot at all but just flash the Angstrom kernel in the usual Flash Menu (or the "OK mode"). So skip Steps 5 & 6, and extract jpmatrix's customization tarball (http://matrixmen.free.fr/zaurus/debian/custo_c3k.tar): tar xvf custo_c3k.tar -C /. , after extracting the rootfs and kernel modules.
II. Download
Download the 071201 rootfs tarball here (http://matrixmen.free.fr/zaurus/debian/zaurus_eabi_rootfs_071201.tar.bz2) OR the mirror here (http://debian.drun.net/rootfs/)
Changelogs
08-04-16- updated with updater.sh for C1000 (for sharp bootloader)
- updated cackohddfix to v0.2
08-03-21- updated the instructions to using yonggun kernel 2.6.24-3
07-12-01- removed leading path (/mnt/card) in the extracted rootfs
- fixed loadkeys in /etc/rc.local
- added yonggun kernel modules: 2.6.22.9, 2.6.23.1, 2.6.23.3 (so users just need to flash the kernel)
- new packages presinstalled: ssh, openssh-client, openssh-server
Note: Those who already got the previous rootfs running do not need this.
III. What is in this rootfs?
1) eabi rootfs tarball (Jan 2007)
2) apt-get dist-upgrade (Nov 28 2007) -- FULLY upgraded!
3) yonggun's (= 2or0's) deb packages & keymap (for details, see: http://yonggun.tistory.com/54 (http://yonggun.tistory.com/54) )
Highlights:
-- X-window fully set up (xfbdev)
-- Video and sound playback enabled (mplayer, alsa stuff, xmms)
-- Display auto-rotation between landscape/portrait modes
-- Wireless networking enabled
-- Keymap in pdaXrom's style
-- Mouse keybindings with xbindkeys: [Calendar] = left button; [Address] = middle button; [Mail] = right button
4) fine-tuning with my previous after installation scripts & settings: including hardware clock setting, sources.list with working repos, some basic network setup, usb modules loading, etc.
5) locales installed, generated, and set to en_US.UTF-8 (export added to ~/.profile)
6) additional packages installed:
-- file manager: pcmanfm
-- editor: leafpad
-- browsers: dillo, links2, epiphany-browser, epiphany-webkit
-- essential networking tools: ssh, cvs, subversion
-- hardware management tools: pcmciautils, hdparm
-- archiving tools: bzip2, unrar-free, unzip, xarchiver (tar and zip are already presinstalled in the rootfs)
7) touchscreen calibrated (/dev/input/ts0)
8) executable paths, compiler architecture optimization options and QT3/QT4 paths (commented) set in ~/.bashrc
9) gtk-icon-theme set to "gnome" (in ~/.gtkrc-2.0)
10) default window manager = icewm + some basic fine-tuning (e.g. the [Menu] key call the wm menu); custom settings files in ~/.icewm
11) backlight adjustment enabled ([Fn]+3, [Fn]+4)
12) sound playback (tuned in alsamixer) with headphone/speaker hot-swapping functionality
IV. Installation steps (last modified: 080321)
Step 1:
Prepare an ext2/ext3 formatted media (internal drive/SD/CF) and at least two partitions (root, swap). A media of at least 1GB of memory is required.
Step 2:
Extract the rootfs tarball to the target media: tar -xjpf zaurus_eabi_rootfs_071201.tar.bz2 -C to the root level of a CF/SD(depending on which media you want to load Debian)
(NOTE: This can take *quite* a while because the extracted rootfs is almost 700MB. More patience, more joy.)
Also download and extract "modules-2.6.24-yonggun.tar.gz" from http://yonggun.tistory.com/68 (http://yonggun.tistory.com/68) to the same location.
Step 3:
Edit /etc/fstab and assign the correct mount points to the root partition and the swap partition. The default is set to boot from the internal drive. Just comment the lines for /dev/hdaX and uncomment the lines for /dev/mmcblk0pX if debian boots from SD.
Step 4:
Flash a new kernel
There are two sets of kernels to work with Debian EABI, one using the default Sharp bootloader and the other using uboot. Both work fine.
================================================================================
Kernel installation for Sharp bootloader
I. Get the latest yonggun kernel here: http://yonggun.tistory.com/68 (http://yonggun.tistory.com/68) (under "For Sharp Bootloader"), rename the kernel image to "zImage.bin", put it together with this at the root level of a CF/SD. Remove the ".txt" file extension before use.
updater.sh (for C3000/3100/3200) (https://www.oesf.org/forum/index.php?act=attach&type=post&id=5291) or this updater.sh (for C1000) (https://www.oesf.org/forum/index.php?act=attach&type=post&id=5403)
II. Reboot the machine into the normal Flash Menu (press [OK] before powering up/rebooting)
III. In the Japanese menu, choose "4", then choose the media, then choose "Y" to confirm.
================================================================================
Kernel installation for Uboot
I. Install the pdaXrom uboot (http://www.pdaxrom.org/?q=node/106 (http://www.pdaxrom.org/?q=node/106))
II. Press [OK] to reboot into the uboot emergency mode.
III. Install the yonggun kernel
For all Cxx00 series: 2.6.24 (http://yonggun.tistory.com/68)
by running this autoboot.sh (remove the ending .txt extension) on a SD/CF.
[ You are not allowed to view attachments ]
IV: As soon as booting starts,
press and hold [1] to boot from SD
press and hold [2] to boot from the internal drive
press and hold [3] to boot from CF
================================================================================
Step 5:
First boot
Only the root account is set up: login = root; password = zaurus
Step 6:
Set the correct date and time:
date MMDDHHMMYYYY e.g. 112814002007 (= Nov 28, 14:00, 2007)
hwclock -w
Step 7:
Make the kernel modules work
depmod -a
Step 8:
Set up network connection
Set up your own network profile in /etc/network/interfaces . A sample file has been provided. After that run /etc/init.d/networking restart
Then plug in your wifi CF card if not inserted yet. If the card fails to work, either take out the card and insert in again, or run pccardctl eject 1 && pccardctl insert 1
That's it. Debian EABI is ready to run.
Run startx to launch the window manager (icewm).
Run apt-get install PACKAGE_NAME to install your desired packages.
Run apt-get update to update the package lists.
Run apt-get dist-upgrade to upgrade all the upgradable packages.
V. Special notes:
Although this is a tailor-made rootfs for running with the yonggun kernels, you may still use it with other eabi kernels, such as the Angstrom kernels and cortez's Poky kernels. YMMV.
Touchscreen issue:
This tarball is primarily tailor made for the yonggun kernel. The touchscreen device is /dev/input/ts0. If you use other kernels, such as Angstrom's, you may have to change to another device, such as /dev/input/event1 (to be verified). Specify the setting in ~/.profile.
Or if you find that the touchscreen pointer goes wild, exit X and run ts_calibrate in the login shell.
-
That's good idea, I think.
BTW, we can also simplify the UBOOT INSTALLATION and KERNEL INSTALLATION steps by uploading internal flash backups for all Z models. For example, I can upload my internal flash for C1000 with already installed Uboot and Kernel kernel 2.6.22.9 yonggun. So, it will make installation a little bit simpler (via NAND flash restore DM feature).
-
Sounds good too. But I wonder how to install the NAND backup image.
-
I am very interested in this! I would really like to see how debian runs on my SL-C1000, but I have not had much luck. But thats mostly user error
-
Good. Will start working on this tonight....
-
Indeed, sounds very interesting for me too!!
-
Give it a try. It would be nice to play a bit with yet another distro until I finally settle down on a final decision .
-
Sounds good too. But I wonder how to install the NAND backup image.
It's described here, for example: https://www.oesf.org/forum/index.php?showtopic=24199 (https://www.oesf.org/forum/index.php?showtopic=24199)
Issue: My Z won't boot after trying to install pdaXrom. I want to go back to ROM xyz.
Solution: Do a NAND restore. Download the correct file from here (select your Z model on the right):
http://www.trisoft.de/en_downloads.htm (http://www.trisoft.de/en_downloads.htm)
and put the SYSTCXXX.DBK file on an SD/CF card (FAT16 formatted). remove all power from the Z (battery and AC), then hold Fn+D+M and insert the adapter. YOu should get a diagnostics menu. Go to the third page and select NAND restore. Note that you should only have one card (either SD or CF in the Z at this point).
-
I'll be eagerly waiting for this! Thanks for your hard work!
-
Sounds good too. But I wonder how to install the NAND backup image.
It's described here, for example: https://www.oesf.org/forum/index.php?showtopic=24199 (https://www.oesf.org/forum/index.php?showtopic=24199)
Issue: My Z won't boot after trying to install pdaXrom. I want to go back to ROM xyz.
Solution: Do a NAND restore. Download the correct file from here (select your Z model on the right):
http://www.trisoft.de/en_downloads.htm (http://www.trisoft.de/en_downloads.htm)
and put the SYSTCXXX.DBK file on an SD/CF card (FAT16 formatted). remove all power from the Z (battery and AC), then hold Fn+D+M and insert the adapter. YOu should get a diagnostics menu. Go to the third page and select NAND restore. Note that you should only have one card (either SD or CF in the Z at this point).
You can't do a NAND restore of a backup that has uBoot installed. uBoot messes up something so the restore tool won't restore it correctly, or so I gather.
-
Would be awesome I think to have a pre-packaged rootfs.
-
You can't do a NAND restore of a backup that has uBoot installed. uBoot messes up something so the restore tool won't restore it correctly, or so I gather.
Thanks for clarification, Capn_Fish.
-
I personally think the nand part of the process is the easiest, include the exact kernel.img and files along with the image. That way its two relatively clean steps. Step one setup u-boot and step two untar the rootfs. Would be awesome to have a setup rootfs right off ... no fiddling with ts this and that
-
hoy m8s !
too late and too much pain for my poor eyes, i need too sleep
so i neither release anything nor write some docs or explanations now...
just wanna say that i could post a working rootfs of Debian Sid running e16 with right click support (well, seems to be, as i get menu/settings then i tap on the desktop with fn/ctrl keys )
i'm very surprise about anything quite important: the speed.
Last time, xfce4 was so slow that i tried to run matchbox, blackbox, and so on. I though that was acceptable, but that EABI wasn't as fast as i had heard.
now, i can tell you that that's uncomparatively faster than pdaxrom or old OABI builds of OE
especielly E16 is even faster than openbox
for now vim is broken, and i need some time to polish that rootfs. but if anyone is interested, i would post it
if anyone knows how to get rid of the "help board" of e16 (explaining what would happen if i do this or that), that'll be very nice to tell me
i also can set up a feed for new compiled eabi packages on my server, which as a low bandwith, but it can be convenient ?
by for now, hope i'll could get up in 5 hours ]
-
just wanna say that i could post a working rootfs of Debian Sid running e16 with right click support (well, seems to be, as i get menu/settings then i tap on the desktop with fn/ctrl keys )
cool
why e16 and not e17 ? e17 works great under angstrom
i'm very surprise about anything quite important: the speed.
Last time, xfce4 was so slow that i tried to run matchbox, blackbox, and so on. I though that was acceptable, but that EABI wasn't as fast as i had heard.
did you try icewm too ? i wonder if e16/e17 is faster than icewm...
i also can set up a feed for new compiled eabi packages on my server, which as a low bandwith, but it can be convenient ?
by for now, hope i'll could get up in 5 hours ]
wake up ! it's 14h07 in France !!!
-
cool
why e16 and not e17 ? e17 works great under angstrom
cuz i'm using E17 on my laptop and i think that it is a little too heavy for the Z (unless you choose a color only wallpaper, no window effects and so on...)
well, what's new on E ? new libraries: don't think it uses less memory or cpu (on my laptop it takes more than XFCE), wm overall look improved, new applications.
And that's buggy. (at least much more than e16 which is on a stable state) where are the advantages ? i don't care about using bleeding edge apps if i have to spend 3 times more to open a terminal
but i'll certainly make a try
did you try icewm too ? i wonder if e16/e17 is faster than icewm...
that's fast too. that's ahrd to compare the both... MHO, that's just a matter of choices
wake up ! it's 14h07 in France !!! wink.gif
hmmm... i got up à 7h to go to school
well, have a good day, you guys, i'm goin to bed for now sorry for my english, i'm really exhausted
-
As my Debian EABI on my 3200 has been set up successfully and started making packages missing in the armel feeds, I am thinking of rolling up a new, fresh, and fully configured tarball, specially tailor-made for the yonggun kernels with the following ingredients:
1) Base = The January 07 EABI rootfs
+
2) apt-get dist-upgrade
+
3) 2or0's finetuning packages
+
4) The other configuration stuff in my after installation tarball
There are two big advantages AFAIK:
-- It can save new users a lot of time and frustrations setting up the system; no post installation step is needed except setting up network connection.
-- It will certainly work because it must be working on my machine in the first place.
If someone is interested (even if there is just one or two users) I will do it before the weekend.
So does anyone want it?
Is the EABI rootfs available for download?
-
Sorry for keep you waiting.
At first I was trying to create the tarball on my C860 from the ground up. But now it does not have a working (yonggun) kernel yet, and I was terribly busy over this couple of weeks. May try to do that on my 3200 this weekend.
-
Sorry for keep you waiting.
At first I was trying to create the tarball on my C860 from the ground up. But now it does not have a working (yonggun) kernel yet, and I was terribly busy over this couple of weeks. May try to do that on my 3200 this weekend.
No problem. Anyway i got my EABI debian installed but with Angstrom 2.6.23 kernel instead of yonggun kernel.
-
No problem. Anyway i got my EABI debian installed but with Angstrom 2.6.23 kernel instead of yonggun kernel.
+1
can younggun make a special kernel for the c3000, without uboot or altboot or i don't know what-boot ?
-
Sorry for keep you waiting.
At first I was trying to create the tarball on my C860 from the ground up. But now it does not have a working (yonggun) kernel yet, and I was terribly busy over this couple of weeks. May try to do that on my 3200 this weekend.
Hi ZDevil, any progress on this?
-
Hey, Fatty! Glad to see you here again!
Okay, i'll get the (tar)ball rolling tonight.
-
Hey, Fatty! Glad to see you here again!
Okay, i'll get the (tar)ball rolling tonight.
Longing for playing with the Z w/ Debian! You'll save my life!
I did buy a new Sandisk 8G and tried installing again, but still got the same kernel panic issue. But I finally figured out what had been happening: I didn't press [2] while booting with uboot! So the issue is not CF related
[OT] not back to the place-you-know-where for some time, and only recently noticed you won't be back there. So see you here . My wife is getting better now but the health issue is on me though ...
-
The tarball is ready! It's 195MB big (close to 700MB when extracted, because it's fully configured and installed with many goodies). jpmatrix, may I upload it to your ftp?
-
The tarball is ready! It's 195MB big (close to 700MB when extracted, because it's fully configured and installed with many goodies). jpmatrix, may I upload it to your ftp?
go go go !!!!
-
Finished uploading (in /uploads ). How can we access it?
Once the file is open for d/l I will post some instructions for using the tarball.
BTW, the other stuff in /uploads are not useful any more. So feel free to clean theme up.
-
cool
why e16 and not e17 ? e17 works great under angstrom
cuz i'm using E17 on my laptop and i think that it is a little too heavy for the Z (unless you choose a color only wallpaper, no window effects and so on...)
well, what's new on E ? new libraries: don't think it uses less memory or cpu (on my laptop it takes more than XFCE), wm overall look improved, new applications.
And that's buggy. (at least much more than e16 which is on a stable state) where are the advantages ? i don't care about using bleeding edge apps if i have to spend 3 times more to open a terminal
but i'll certainly make a try
did you try icewm too ? i wonder if e16/e17 is faster than icewm...
that's fast too. that's ahrd to compare the both... MHO, that's just a matter of choices
wake up ! it's 14h07 in France !!! wink.gif
hmmm... i got up à 7h to go to school
well, have a good day, you guys, i'm goin to bed for now sorry for my english, i'm really exhausted
Hello, what repository do you use, because i've only E16 available.
-
Finished uploading (in /uploads ). How can we access it?
Once the file is open for d/l I will post some instructions for using the tarball.
BTW, the other stuff in /uploads are not useful any more. So feel free to clean theme up.
ok!
it will be available within 1 hour on:
http://matrixmen.free.fr/zaurus/debian/ (http://matrixmen.free.fr/zaurus/debian/)
as http://matrixmen.free.fr/zaurus/debian/zau..._071128.tar.bz2 (http://matrixmen.free.fr/zaurus/debian/zaurus_eabi_rootfs_071128.tar.bz2)
-
This sounds interesting. You probably remember I had a kernel panic issue with Titchy due to the kernel not recognizing my A-Data 16GB internal. I'm guessing that this kernel won't either (though it would be handy). This raises 2 possibilities-- will u-boot handle my internal cf issue, and if not is there a reasonable way to do a 1000 style sd installation on a 3100 (and then use the internal cf as /home, or something like, ultimately)?
Also, how is the sound?
-
Thanks, jpmatrix!
Added a short instruction in the first post for easy reference.
@adf: the A-Data card issue all depends on the kernel. But the rootfs tarball can be boot on the internal drive, SD and CF. I created this tarball on an SD and everything works. Sound plays well.
-
Hey, Fatty! Glad to see you here again!
Okay, i'll get the (tar)ball rolling tonight.
Longing for playing with the Z w/ Debian! You'll save my life!
I did buy a new Sandisk 8G and tried installing again, but still got the same kernel panic issue. But I finally figured out what had been happening: I didn't press [2] while booting with uboot! So the issue is not CF related
[OT] not back to the place-you-know-where for some time, and only recently noticed you won't be back there. So see you here . My wife is getting better now but the health issue is on me though ...
Fatty, the tarball is ready now. Have fun!
Just forget about the incidence in the place-you-know-where.... I can't think of any word to describe it except "ridiculous".
Take good care of yourself and your wife!
-
This is working good on my C1000 with kernel 2.6.22.9 yonggun on a 2GB SD card (Sandisk 133X)
Thank you very much for sharing your work !
-
Dutifully downloading this one, great job guys
-
perhaps i'll backup my existing eabi and try the zdevil rootfs on my C3000 with angstrom 2.6.23
that way we'll know if works well with this configuration
-
Fatty, the tarball is ready now. Have fun!
Tks. Downloading.
If you don't mind, I'll direct those interested from you-know-where to this post.
-
This sounds interesting. You probably remember I had a kernel panic issue with Titchy due to the kernel not recognizing my A-Data 16GB internal. I'm guessing that this kernel won't either (though it would be handy). This raises 2 possibilities-- will u-boot handle my internal cf issue, and if not is there a reasonable way to do a 1000 style sd installation on a 3100 (and then use the internal cf as /home, or something like, ultimately)?
Also, how is the sound?
I used to think I have the same kernel panic issue with my swapped CF (TwinMOS 4GB). Recently I have swapped it to another Sandisk ultraII 8G, and got the same result. Till yesterday I finally figured out that it's all because I didn't press [2] (boot from Internal CF) while booting uboot. Now my Z can boot in Debian with the original TwinMOS CF. Stupid me
-
Fatty, the tarball is ready now. Have fun!
Tks. Downloading.
If you don't mind, I'll direct those interested from you-know-where to this post.
No problem. Just let me know if there is any issue of the tarball.
Feel free to post it *anywhere*.
BTW i've already posted it in Dinorex's forum.
-
This sounds interesting. You probably remember I had a kernel panic issue with Titchy due to the kernel not recognizing my A-Data 16GB internal. I'm guessing that this kernel won't either (though it would be handy). This raises 2 possibilities-- will u-boot handle my internal cf issue, and if not is there a reasonable way to do a 1000 style sd installation on a 3100 (and then use the internal cf as /home, or something like, ultimately)?
Also, how is the sound?
I used to think I have the same kernel panic issue with my swapped CF (TwinMOS 4GB). Recently I have swapped it to another Sandisk ultraII 8G, and got the same result. Till yesterday I finally figured out that it's all because I didn't press [2] (boot from Internal CF) while booting uboot. Now my Z can boot in Debian with the original TwinMOS CF. Stupid me
..and me maybe, it would appear
-
zdevil,
your rootfs seems to work on my c3000 !
mainly i had to :
-unable to open initial console on 1st boot : solved by rebooting
-modules.dep not found on boot : solved by depmod -a on first root login
-startx : icewm launch is ok ! but stylus doesn't go where i click !
apparently /dev/input/ts0 does not exist
and ts.conf is misconfigured (module_raw corgi is not commented)
lets's nano-ing a little
EDITED:
stylus troubles solved by :
rmmod mousedev (for the stylus jittering)
cd /dev/input
ln -s ts0 event1
then startx, open a console and ts_calibrate to recalibrate the screen
thanks for your work Zdevil ! it looks it works on a c3000 too !
-
I have no idea what causes the issues of initial console and depmod. They played well in my case. I wonder if it has to do with something one has to do with the freshly installed kernel (I did this rootfs on my already working 3200 setup).
The stylus thing is tricky. As I suspect at the end of the instructions, you need to play with other options if you use non-yonggun kernels.
Thanks, jpmatrix for the feedback. Perhap it's not such a bad idea to keep the tarball polished and updated every now and then?
-
from where you make the step 3 to format internal drive? from uboot?
-
I have no idea what causes the issues of initial console and depmod. They played well in my case. I wonder if it has to do with something one has to do with the freshly installed kernel (I did this rootfs on my already working 3200 setup).
The stylus thing is tricky. As I suspect at the end of the instructions, you need to play with other options if you use non-yonggun kernels.
Thanks, jpmatrix for the feedback. Perhap it's not such a bad idea to keep the tarball polished and updated every now and then?
yesssss
what do you think is better : i make another thread (or a special entry in the wiki for the c3000) or you modify your initial first thread to add the c3000 ?
-
from where you make the step 3 to format internal drive? from uboot?
Hmm. Here is a thing about uboot: it seems not being able to handle ext3 partition properly (but correct me if I am wrong).
So perhaps it is safer to do the extraction before installing uboot (if it has not been installed yet). Let me swap steps 1, 2, 3. That doesn't hurt anything at all.
Thanks for reminding me of this.
-
How would I go about installing this on my SL-C3200's internal drive? I'm wanting to do a complete clean install of the yongun kernel and use this rootfs. What would be the easiest way to do this coming from Angstrom?
-
from where you make the step 3 to format internal drive? from uboot?
Hmm. Here is a thing about uboot: it seems not being able to handle ext3 partition properly (but correct me if I am wrong).
So perhaps it is safer to do the extraction before installing uboot (if it has not been installed yet). Let me swap steps 1, 2, 3. That doesn't hurt anything at all.
Thanks for reminding me of this.
True. I didn't find mkfs tool within uboot. So I have to nand back to Sharp rom, do format and reinstall uboot. It doesn't cost much time though.
BTW, how can I re-calibrate the pointer? It isn't quite accurate on top and bottom region.
Lastly, as I'm not quite familiar with uboot, after uboot/kernel being flashed, where is the remaining space of the flash memory going?
-
BTW, how can I re-calibrate the pointer? It isn't quite accurate on top and bottom region.
just run in a console :
ts_calibrate
-
Lastly, as I'm not quite familiar with uboot, after uboot/kernel being flashed, where is the remaining space of the flash memory going?
The remaining NAND space is for: 1) a eabi-kernel friendly pdaXrom build (if there is one); 2) an little more empty flash space for storage (if there is mtd support)
-
I have no idea what causes the issues of initial console and depmod. They played well in my case. I wonder if it has to do with something one has to do with the freshly installed kernel (I did this rootfs on my already working 3200 setup).
The stylus thing is tricky. As I suspect at the end of the instructions, you need to play with other options if you use non-yonggun kernels.
Thanks, jpmatrix for the feedback. Perhap it's not such a bad idea to keep the tarball polished and updated every now and then?
yesssss
what do you think is better : i make another thread (or a special entry in the wiki for the c3000) or you modify your initial first thread to add the c3000 ?
I added a note with link to this post in the first post.
It's great to know the rootfs also works for C3000.
So I suppose it can also get along with Cxx0, if the kernel is up to the job?
-
I added a note with link to this post in the first post.
It's great to know the rootfs also works for C3000.
So I suppose it can also get along with Cxx0, if the kernel is up to the job?
great!
here are my customization files after installing your rootfs :
http://matrixmen.free.fr/zaurus/debian/custo_c3k.tar (http://matrixmen.free.fr/zaurus/debian/custo_c3k.tar)
just
cd /
tar xvf custo_c3k.tar
-
Can someone help me ...
I got the error mentioned here:
https://www.oesf.org/index.php?title=Q18:_I...ternal_drive%3F (https://www.oesf.org/index.php?title=Q18:_I_am_using_a_C3x00%2C_and_I_want_to_use_kernel_2.6.23.1_yonggun_with_uboot._How_can_I_install_DebiaBI_on_the_internal_drive%3F)
about the initial console error.
When I booted into emergency system, I wanted to do this:
mount /dev/hda1 /mnt/cf
cd /mnt/cf/dev
MAKEDEV console
but I could not:
first - it's already mounted as /mnt/cf as ext3 - that's ok
when I try to do ls -al /mnt/cf - it freezes without doing anything.
after this I tried
umount /mnt/cf
but I got "Device or resource busy"
Looks there isn't even /dev/hda1 ! only /dev/hda.
I'm getting this behaviour for the second time it's happening only under emergency system. Under the initial B+D - it works ok.
Does anyone knows some way to fix this? Seems I cannot even install debian I'm not talking about testing it ...
-
Step 2: Extract the rootfs tarball to the target media: tar -xjpf zaurus_eabi_rootfs_071128.tar.bz2 -C /path/to/the/target/media
I did:
cd /mnt/card
tar -xjpf zaurus_eabi_rootfs_071128.tar.bz2 -C /mnt/hd/
This made me /mnt/hd/mnt/card/ directories - which I had to move to /mnt/hd/ manually - verified on desktop - the archive contains /mnt/card/ path.
-
Step 2: Extract the rootfs tarball to the target media: tar -xjpf zaurus_eabi_rootfs_071128.tar.bz2 -C /path/to/the/target/media
I did:
cd /mnt/card
tar -xjpf zaurus_eabi_rootfs_071128.tar.bz2 -C /mnt/hd/
This made me /mnt/hd/mnt/card/ directories - which I had to move to /mnt/hd/ manually - verified on desktop - the archive contains /mnt/card/ path.
yes i forgot that one too i had to move them too
about your console problem, just try to wait a little after the message. if it doesn't work, just reboot removing the battery and try again. it worked for me.
-
Can someone help me ...
I got the error mentioned here:
https://www.oesf.org/index.php?title=Q18:_I...ternal_drive%3F (https://www.oesf.org/index.php?title=Q18:_I_am_using_a_C3x00%2C_and_I_want_to_use_kernel_2.6.23.1_yonggun_with_uboot._How_can_I_install_DebiaI_on_the_internal_drive%3F)
about the initial console error.
When I booted into emergency system, I wanted to do this:
mount /dev/hda1 /mnt/cf
cd /mnt/cf/dev
MAKEDEV console
but I could not:
first - it's already mounted as /mnt/cf as ext3 - that's ok
when I try to do ls -al /mnt/cf - it freezes without doing anything.
after this I tried
umount /mnt/cf
but I got "Device or resource busy"
Looks there isn't even /dev/hda1 ! only /dev/hda.
I'm getting this behaviour for the second time it's happening only under emergency system. Under the initial B+D - it works ok.
Does anyone knows some way to fix this? Seems I cannot even install debian I'm not talking about testing it ...
UBOOT emergency has a bug that you mentioned. For C3x00, the internel disk (hda1) seems to be recognized as cf card, and the disk partition information disappears. Whenever I entered into the emergency mode, I repartitioned the internel disk to recover the partition information. Fortunately, all data are not destroyed.
-
Can someone help me ...
I got the error mentioned here:
https://www.oesf.org/index.php?title=Q18:_I...ternal_drive%3F (https://www.oesf.org/index.php?title=Q18:_I_am_using_a_C3x00%2C_and_I_want_to_use_kernel_2.6.23.1_yonggun_with_uboot._How_can_I_install_Debia_on_the_internal_drive%3F)
about the initial console error.
When I booted into emergency system, I wanted to do this:
mount /dev/hda1 /mnt/cf
cd /mnt/cf/dev
MAKEDEV console
but I could not:
first - it's already mounted as /mnt/cf as ext3 - that's ok
when I try to do ls -al /mnt/cf - it freezes without doing anything.
after this I tried
umount /mnt/cf
but I got "Device or resource busy"
Looks there isn't even /dev/hda1 ! only /dev/hda.
I'm getting this behaviour for the second time it's happening only under emergency system. Under the initial B+D - it works ok.
Does anyone knows some way to fix this? Seems I cannot even install debian I'm not talking about testing it ...
UBOOT emergency has a bug that you mentioned. For C3x00, the internel disk (hda1) seems to be recognized as cf card, and the disk partition information disappears. Whenever I entered into the emergency mode, I repartitioned the internel disk to recover the partition information. Fortunately, all data are not destroyed.
I also got this problem.. Could you give me hint/command how to repartition the internal disk under uboot and make it works?
When I do fdisk /dev/hda
I got "Segmentation fault"
-
Exact same problem here
I extracted the rootfs and moved it to /mnt/cf but I get a kernel panic when I try to boot.
Please append a correct "root=" boot option; here are the available partitions:
b300 922000 mmcblk0
0300 6000278 hda driver: ide-disk
0301 5371600 hda1
0302 628488 hda2
Kernel Panic - not syncing: VFS: unable to mount root fs on unknown-block(0,0)
Any suggestions?
EDIT: Ok, I was confused about which reboot I was on. It is throwing this kernel panic during the reboot immediately after selecting yes to install Debian in the UBOOT emergency loader.
-
Step 2: Extract the rootfs tarball to the target media: tar -xjpf zaurus_eabi_rootfs_071128.tar.bz2 -C /path/to/the/target/media
I did:
cd /mnt/card
tar -xjpf zaurus_eabi_rootfs_071128.tar.bz2 -C /mnt/hd/
This made me /mnt/hd/mnt/card/ directories - which I had to move to /mnt/hd/ manually - verified on desktop - the archive contains /mnt/card/ path.
yes i forgot that one too i had to move them too
about your console problem, just try to wait a little after the message. if it doesn't work, just reboot removing the battery and try again. it worked for me.
I had this too - you can use the '--strip-components 2' option. This way tar will remove two top directory levels leaving just the root layout in the target extraction directory. That is you won't need to move everything two levels up afterwards.
-
So there are three issues to deal with:
1. Failure to open initial console
-- The solution is just a reboot ...
2. Failure to do things on the internal disk/CF in the uboot emergency system
-- I've changed the order of the installation steps: prepare the rootfs disk first, and then install uboot & the kernel. This is the buggy uboot. Unless someone can make a new one....
3. rootfs tarball contains /mnt/card
-- I will make a new one right away. Also I will add a couple essential programs to it, such as ssh.
-
whoooowww! using zdevil rootfs, i've sound now working out of the box, either via headphone and zaurus speaker !!!! that's great step
-
whoooowww! using zdevil rootfs, i've sound now working out of the box, either via headphone and zaurus speaker !!!! that's great step
No, kudos go to 2or0. It's his package.
BTW I've finished uploaded a newer rootfs tarball with a few minor fixes. Those who got the previous one working don't need this. Once the link is released (thanks jpmatrix) I'll add a note to the first post.
-
whoooowww! using zdevil rootfs, i've sound now working out of the box, either via headphone and zaurus speaker !!!! that's great step
No, kudos go to 2or0. It's his package.
BTW I've finished uploaded a newer rootfs tarball with a few minor fixes. Those who got the previous one working don't need this. Once the link is released (thanks jpmatrix) I'll add a note to the first post.
the rootfs is now uploaded to matrixmen.free.fr/zaurus/debian
about younggun debs i've some error messages when reprepro-ing them i'll send them to you
-
here are the error messages about younggun debs :
/mnt/lan/zaurus/debian/uploads/yonggun_deb/console-tools_0.2.3dbs-65_armel.deb: component guessed as 'main'
Skipping inclusion of 'console-tools' '1:0.2.3dbs-65' in 'sid|main|armel', as it has already '1:0.2.3dbs-65'.
Character 0x3a not allowed in sourcename: 'http://www.pdaXrom.org/src/cpuoverclock-0.0.3.tar.bz2'!
To ignore this use --ignore=forbiddenchar.
Character 0x3a not allowed in sourcename: 'http://gpe.linuxtogo.org/download/source/gpe-conf-0.2.6.tar.bz2'!
To ignore this use --ignore=forbiddenchar.
Character 0x3a not allowed in sourcename: 'http://handhelds.org/pub/projects/gpe/source/gpe-icons-0.25.tar.gz'!
To ignore this use --ignore=forbiddenchar.
Error: Field 'Source' contains unexpected character 'f'!
/mnt/lan/zaurus/debian/uploads/yonggun_deb/libconsole_0.2.3dbs-65_armel.deb: component guessed as 'main'
Skipping inclusion of 'libconsole' '1:0.2.3dbs-65' in 'sid|main|armel', as it has already '1:0.2.3dbs-65'.
Error: Field 'Source' contains unexpected character 'f'!
/mnt/lan/zaurus/debian/uploads/yonggun_deb/libiw28_28-1_armel.deb: component guessed as 'main'
Skipping inclusion of 'libiw28' '28-1' in 'sid|main|armel', as it has already '28-1'.
Error: Field 'Source' contains unexpected character 'f'!
Error: Field 'Source' contains unexpected character 'f'!
Error: Field 'Source' contains unexpected character 'f'!
Error: Field 'Source' contains unexpected character 'f'!
Character 0x3a not allowed in sourcename: 'file://mplayer.conf'!
To ignore this use --ignore=forbiddenchar.
Error: Field 'Source' contains unexpected character 'f'!
Error: Field 'Source' contains unexpected character 'f'!
Error: Field 'Source' contains unexpected character 'f'!
/mnt/lan/zaurus/debian/uploads/yonggun_deb/wireless-tools_28-1_armel.deb: component guessed as 'main'
Skipping inclusion of 'wireless-tools' '28-1' in 'sid|main|armel', as it has already '28-1'.
Error: Field 'Source' contains unexpected character 'f'!
Error: Field 'Source' contains unexpected character 'f'!
Character 0x3a not allowed in sourcename: 'http://gpe.linuxtogo.org/download/source/xst-0.15.tar.gz'!
To ignore this use --ignore=forbiddenchar.
/mnt/lan/zaurus/debian/uploads/yonggun_deb/x-yonggun_0.0_armel.deb: component guessed as 'main'
Skipping inclusion of 'x-yonggun' '0.0' in 'sid|main|armel', as it has already '0.0'.
Error: Field 'Source' contains unexpected character 'f'!
There have been errors!
edited: i realize most errors are warnings or duplicate entries so they should be harmless...
matrixmen repositories updated!
-
Can someone help me ...
I got the error mentioned here:
https://www.oesf.org/index.php?title=Q18:_I...ternal_drive%3F (https://www.oesf.org/index.php?title=Q18:_I_am_using_a_C3x00%2C_and_I_want_to_use_kernel_2.6.23.1_yonggun_with_uboot._How_can_I_install_Debiaon_the_internal_drive%3F)
about the initial console error.
When I booted into emergency system, I wanted to do this:
mount /dev/hda1 /mnt/cf
cd /mnt/cf/dev
MAKEDEV console
but I could not:
first - it's already mounted as /mnt/cf as ext3 - that's ok
when I try to do ls -al /mnt/cf - it freezes without doing anything.
after this I tried
umount /mnt/cf
but I got "Device or resource busy"
Looks there isn't even /dev/hda1 ! only /dev/hda.
I'm getting this behaviour for the second time it's happening only under emergency system. Under the initial B+D - it works ok.
Does anyone knows some way to fix this? Seems I cannot even install debian I'm not talking about testing it ...
UBOOT emergency has a bug that you mentioned. For C3x00, the internel disk (hda1) seems to be recognized as cf card, and the disk partition information disappears. Whenever I entered into the emergency mode, I repartitioned the internel disk to recover the partition information. Fortunately, all data are not destroyed.
I also got this problem.. Could you give me hint/command how to repartition the internal disk under uboot and make it works?
When I do fdisk /dev/hda
I got "Segmentation fault"
For my case, I first unmount /mnt/cf when I entered into emergency mode. You can use fdisk to repartition the /dev/hda. If you know the partition information, you can use that to recover the data in your internel disk.
-
edited: i realize most errors are warnings or duplicate entries so they should be harmless...
matrixmen repositories updated!
Thanks
-
For my case, I first unmount /mnt/cf when I entered into emergency mode. You can use fdisk to repartition the /dev/hda. If you know the partition information, you can use that to recover the data in your internel disk.
well, after umount /mnt/cf - i cannot see any hda1/hda2 partitions. So this means the "emergency" mode is not useful
-
For my case, I first unmount /mnt/cf when I entered into emergency mode. You can use fdisk to repartition the /dev/hda. If you know the partition information, you can use that to recover the data in your internel disk.
well, after umount /mnt/cf - i cannot see any hda1/hda2 partitions. So this means the "emergency" mode is not useful
You cannot see hda1/hda2 partitions before you make the repartition to see hda1/hda2.
-
For my case, I first unmount /mnt/cf when I entered into emergency mode. You can use fdisk to repartition the /dev/hda. If you know the partition information, you can use that to recover the data in your internel disk.
well, after umount /mnt/cf - i cannot see any hda1/hda2 partitions. So this means the "emergency" mode is not useful
You cannot see hda1/hda2 partitions before you make the repartition to see hda1/hda2.
apparentlu you have to partition the drive and untar the tarball BEFORE installing the OS. I'm in the midst of this myself (and have only manged 3 nand restores to show for it)
Is there a step missing in the installation directions that must be obvious to those who have done it and less so to those of us who haven't?
Anyway, I've had: uboot go wrong in surprising ways, and my shiny new taerball extract with an unexpected eof error. I'm sure I'll be happy I did this--but atm...ugh.
-
I can't get /mnt/cf to unmount
umount /mnt/cf
umount: /mnt/cf: Device or resource busy
umount -f /mnt/cf
umount: forced umount of /dev/hda1 failed!
is there some magic I'm missing?
-
No joy with the tarball. I've used wget twice and opera once (On Ubuntu dapper, where /media/sdb1 is a 4G hitachi MD on a card reader I use regularly) --and I always get an eof error. Any suggestions? I've included an excerpt below:
tar -xjpf /home/adf/zeabi/zaurus_eabi_rootfs_071201.tar.bz2 -C /media/sdb1
gets me
bzip2: Data integrity error when decompressing.
Input file = (stdin), output file = (stdout)
It is possible that the compressed file(s) have become corrupted.
You can use the -tvv option to test integrity of such files.
You can use the `bzip2recover' program to attempt to recover
data from undamaged sections of corrupted files.
tar: Unexpected EOF in archive
tar: Unexpected EOF in archive
tar: Error is not recoverable: exiting now
-
I can't get /mnt/cf to unmount
umount /mnt/cf
umount: /mnt/cf: Device or resource busy
umount -f /mnt/cf
umount: forced umount of /dev/hda1 failed!
is there some magic I'm missing?
It seems some people got the same issue.
I want to know what step I missed or the guild not covered..
I follow the wiki guild Q18 to install Debian Eabi on my 3200 internal drive. All steps run smooth, just stop after the line of "unable to open an initial console"
-
No joy with the tarball. I've used wget twice and opera once (On Ubuntu dapper, where /media/sdb1 is a 4G hitachi MD on a card reader I use regularly) --and I always get an eof error. Any suggestions? I've included an excerpt below:
you're right, the file is corrupted.
zdevil, please upload it again
-
Uploading. 25 minutes left.
I also found it strange: my original zaurus_eabi_rootfs_071201.tar.bz2 is 200.6MB, but I see there is a zaurus_eabi_rootfs_071201.tar which is just 204.2MB. The tar should be around 600.MB.
Anyway, here is my md5sum. jpmatrix, could you also help me check it?
[size=]b89abbed88dbe5671112c33703b92668[/size] zaurus_eabi_rootfs_071201.tar.bz2
-
Uploading. 25 minutes left.
I also found it strange: my original zaurus_eabi_rootfs_071201.tar.bz2 is 200.6MB, but I see there is a zaurus_eabi_rootfs_071201.tar which is just 204.2MB. The tar should be around 600.MB.
Anyway, here is my md5sum. jpmatrix, could you also help me check it?
[size=]b89abbed88dbe5671112c33703b92668[/size] zaurus_eabi_rootfs_071201.tar.bz2
that's strange, the md5sum is correct (even for the older file) but i'm unable to extract the tar with 7-zip....
you can join me on irc on #debian-zaurus
-
Okay. Which irc server?
Uploading finished. In 071201.
-
Okay. Which irc server?
Uploading finished. In 071201.
irc.freenode.net
as seen on https://www.oesf.org/forum/index.php?showto...5000&hl=irc (https://www.oesf.org/forum/index.php?showtopic=25000&hl=irc)
-
Would it be possible to create a debian install distro that works like a regular Zaurus rom?
Start with a striped Angstrom image. Have it automatically try to set up a network connection, download a debian root filesystem, install it to a location of the user's preference and set up altboot to automatically boot the new system.
Basically using altboot as a boot loader and Angstom as a installation tool.
As far as the end user would be concerned installing debian would be like installing any other rom, it would just take longer.
-
Both jpmatrix and I found that the rootfs tarball is corrupted. It seems something went wrong went i run "tar -cpjf" (but it worked for the very first tarball). Now I am making a new tarball. Will upload it in a few hours (if things go well). Sorry for this.
-
Update:
A new tarball is ready. It is tested (twice actually) and it is fine. Now waiting for jpmatrix's ftp to be up again.
-
Update:
A new tarball is ready. It is tested (twice actually) and it is fine. Now waiting for jpmatrix's ftp to be up again.
you're in a lucky day : ftp is up !
-
Update:
A new tarball is ready. It is tested (twice actually) and it is fine. Now waiting for jpmatrix's ftp to be up again.
you're in a lucky day : ftp is up !
Hope the 'new' tarball and scripts are thoroughly debugged. I also hope I won't be trying to use it any time soon. The EABI install I did using cortez's method, in combination with the ZDevil scripts and your tips at the wiki, is still OK! Thanks also to dlj0 for his X tips.
Obligatory stupid question:- Is it possible to do this install without u-boot?
My reluctance to use u-boot is from the same cause I'm reluctant to use angstrom or any of the earlier OPIE stuff. When I scan the relevant forums: so many recurring problems....
My only real complaint with my EABI system is the inability to use X as non-root. But I'll figure something out soon I hope. (Before someone asks what I'm worrying about::- My fingers sometimes outrun the brain and root is very dangerous on those occasions! )
Cheers
-
new rootfs is available :
http://matrixmen.free.fr/zaurus/debian/zau..._071201.tar.bz2 (http://matrixmen.free.fr/zaurus/debian/zaurus_eabi_rootfs_071201.tar.bz2)
enjoy!
-
Thanks, jpmatrix. Just updated the link in the first post too.
-
Hope the 'new' tarball and scripts are thoroughly debugged. I also hope I won't be trying to use it any time soon. The EABI install I did using cortez's method, in combination with the ZDevil scripts and your tips at the wiki, is still OK! Thanks also to dlj0 for his X tips.
Obligatory stupid question:- Is it possible to do this install without u-boot?
My reluctance to use u-boot is from the same cause I'm reluctant to use angstrom or any of the earlier OPIE stuff. When I scan the relevant forums: so many recurring problems....
My only real complaint with my EABI system is the inability to use X as non-root. But I'll figure something out soon I hope. (Before someone asks what I'm worrying about::- My fingers sometimes outrun the brain and root is very dangerous on those occasions! )
Cheers
Of course if you already have a working setup running, you don't need this rootfs. It is for fresh and new install. Yet it is also a better backup rootfs than the unconfigured ones in case things got messed up.
A few users and I have verified that the rootfs is working, and it is working without using any more after installation scripts.
To be fair, u-boot and a distro are two totally different things. Indeed so many users (I for one) have qualms about the un-userfriendlieness of uboot. But talking about the distros themselves is a separate issue here.
As for running X as non-root user, flux has already come up with a good solution, which has already been wikified iirc.
-
thx mate...
i've just kicked off the dl rite away just now....
-
Good luck, man! Tell me how things go here.
-
Good luck, man! Tell me how things go here.
Tks ZDevil!
Your 1st tar ball did help me bringing up real Debian on my dog! Unluckily, as I messed up with wifi setting, I rebooted it and only found that /dev/hda1 failed to be mounted again. Not sure what caused that, so it's been put aside for a while (well, also because I'd been busy for some days).
I'll try the new tar ball again and report.
OT: say hi to dinorex also
-
hi all zdevil / fatty,
i am too bc in sydney for my new home (gardening etc), therefore it's hard to get me in working around with the doggie debian...
hope i can get that around asap.... (after grabbing that tarball when i back home tonite)..... cheers.../dinorex
-
Downloaded twice, the archive is broken
wasp@home:/storage/dl/zaurus-debian$ bzip2 -d zaurus_eabi_rootfs_071201.tar.bz2
bzip2: Data integrity error when decompressing.
Input file = zaurus_eabi_rootfs_071201.tar.bz2, output file = zaurus_eabi_rootfs_071201.tar
It is possible that the compressed file(s) have become corrupted.
You can use the -tvv option to test integrity of such files.
You can use the `bzip2recover' program to attempt to recover
data from undamaged sections of corrupted files.
bzip2: Deleting output file zaurus_eabi_rootfs_071201.tar, if it exists.
wasp@home:/storage/dl/zaurus-debian$
-
Downloaded twice, the archive is broken
wasp@home:/storage/dl/zaurus-debian$ bzip2 -d zaurus_eabi_rootfs_071201.tar.bz2
bzip2: Data integrity error when decompressing.
Input file = zaurus_eabi_rootfs_071201.tar.bz2, output file = zaurus_eabi_rootfs_071201.tar
It is possible that the compressed file(s) have become corrupted.
You can use the -tvv option to test integrity of such files.
You can use the `bzip2recover' program to attempt to recover
data from undamaged sections of corrupted files.
bzip2: Deleting output file zaurus_eabi_rootfs_071201.tar, if it exists.
wasp@home:/storage/dl/zaurus-debian$
don't tell me that!!!!
could you check the md5 checksum of the file you downloaded?
-
don't tell me that!!!!
could you check the md5 checksum of the file you downloaded?
Sorry to say this but ... (after editing the md5sum file to remove some white spaces) :
wasp@home:/storage/dl/zaurus-debian$ md5sum -c zaurus_eabi_rootfs_071201.tar.bz2.md5sum
zaurus_eabi_rootfs_071201.tar.bz2: OK
-
don't tell me that!!!!
could you check the md5 checksum of the file you downloaded?
Sorry to say this but ... (after editing the md5sum file to remove some white spaces) :
wasp@home:/storage/dl/zaurus-debian$ md5sum -c zaurus_eabi_rootfs_071201.tar.bz2.md5sum
zaurus_eabi_rootfs_071201.tar.bz2: OK
so it is corrupted from zdevil
zdeviiiiil !!!
we're going to make it ! ... one day ))
-
That's strange. I have already extracted it twice on my Z and don't get this. Would there be something going on during the ftp transmission?
-
That's strange. I have already extracted it twice on my Z and don't get this. Would there be something going on during the ftp transmission?
Sorry to tell but... yes, I've the same result
I d/l it twice, with Firefox in XP, double checked the checksum which is okey, but got file corruption when unzip with 7z midway.
-
Well, if that's the case,
Two solutions:
-- I upload it again ...
-- Someone offers a space so that I can upload it there as well
-
That's strange. I have already extracted it twice on my Z and don't get this. Would there be something going on during the ftp transmission?
that's strange, i cannot extract it .... moreover it worked the first time....
very very strange.... there's a new version of filezilla server... i try to upgrade....
-
That's strange. I have already extracted it twice on my Z and don't get this. Would there be something going on during the ftp transmission?
wait wait
if the md5sum is ok, i wonder how the file can be corrupted????
-
That's strange. I have already extracted it twice on my Z and don't get this. Would there be something going on during the ftp transmission?
wait wait
if the md5sum is ok, i wonder how the file can be corrupted????
the md5 sum was created _after_ the file got corrupted?
-
That's strange. I have already extracted it twice on my Z and don't get this. Would there be something going on during the ftp transmission?
wait wait
if the md5sum is ok, i wonder how the file can be corrupted????
the md5 sum was created _after_ the file got corrupted?
Not quite. I did the md5sum before uploading.
-
Well, if that's the case,
Two solutions:
-- I upload it again ...
-- Someone offers a space so that I can upload it there as well
Check your PM, you've got the space needed
Not quite. I did the md5sum before uploading.
I know this. But this doesn't means the tar.bz2 was not corrupted before you make the md5 sum ... anyway ... I've no idea but it's corrupted.
-
Not quite. I did the md5sum before uploading.
I know this. But this doesn't means the tar.bz2 was not corrupted before you make the md5 sum ... anyway ... I've no idea but it's corrupted.
The thing is, I extracted the tarball twice (for testing) on my Zaurus after creating the checksum...
-
The thing is, I extracted the tarball twice (for testing) on my Zaurus after creating the checksum...
Then md5sum is telling wrong please check on your zaurus and check against md5sum ...
-
The thing is, I extracted the tarball twice (for testing) on my Zaurus after creating the checksum...
Then md5sum is telling wrong please check on your zaurus and check against md5sum ...
in the same time, i'm trying to extract it on my zaurus too... i'll let you know
-
The thing is, I extracted the tarball twice (for testing) on my Zaurus after creating the checksum...
Then md5sum is telling wrong please check on your zaurus and check against md5sum ...
I've just done a md5sum with my local copy (not the one in the ftp). Same as what I did on my Mac. And I got it extracted and booted this morning on my C860 ...
-
The thing is, I extracted the tarball twice (for testing) on my Zaurus after creating the checksum...
Then md5sum is telling wrong please check on your zaurus and check against md5sum ...
I've just done a md5sum with my local copy (not the one in the ftp). Same as what I did on my Mac. And I got it extracted and booted this morning on my C860 ...
men, this is a total mystery ! or a time shifting failure
i get this only error when untaring :
tar: Error is not recoverable: exiting now
-
here is my checksum on my window$ machine :
D:\warez\zaurus\debian\uploads>c:\Win32App\fciv.exe zaurus_eabi_rootfs_071201.ta
r.bz2
//
// File Checksum Integrity Verifier version 2.05.
//
b89abbed88dbe5671112c33703b92668 zaurus_eabi_rootfs_071201.tar.bz2
D:\warez\zaurus\debian\uploads>more zaurus_eabi_rootfs_071201.tar.bz2.md5sum
b89abbed88dbe5671112c33703b92668 zaurus_eabi_rootfs_071201.tar.bz2
D:\warez\zaurus\debian\uploads>
-
Lol.
Here is mine: e3133206a7846f585bd41c8dfb58f0b2
Same on my Z and my Mac ...
-
zdevil,
what is your tar version and what do you type to extract it ?
-
Lol.
Here is mine: e3133206a7846f585bd41c8dfb58f0b2
Same on my Z and my Mac ...
oh!! it is not what you put in the md5sum file on my ftp !
ps: i'm on irc too
-
Lol.
Here is mine: e3133206a7846f585bd41c8dfb58f0b2
Same on my Z and my Mac ...
But they are TOTALLY different from yours!
wasp@home:/storage/dl/zaurus-debian$ md5sum zaurus_eabi_rootfs_071201.tar.bz2
b89abbed88dbe5671112c33703b92668 zaurus_eabi_rootfs_071201.tar.bz2
wasp@home:/storage/dl/zaurus-debian$ cat zaurus_eabi_rootfs_071201.tar.bz2.md5sum
b89abbed88dbe5671112c33703b92668 zaurus_eabi_rootfs_071201.tar.bz2
wasp@home:/storage/dl/zaurus-debian$
-
WHY?
I also uploaded my checksum to jpmatrix's ftp, together with the tarball. So I suppose things would be fine if people compare the downloaded copy's checksum with mine.
Would there be something wrong when moving the upload around in the ftp?
-
WHY?
I also uploaded my checksum to jpmatrix's ftp, together with the tarball. So I suppose things would be fine if people compare the downloaded copy's checksum with mine.
Would there be something wrong when moving the upload around in the ftp?
You tell me maybe you've got intelligent virus on your mac
Please please re-upload them - I've got a brick right now instead of Z
-
WHY?
I also uploaded my checksum to jpmatrix's ftp, together with the tarball. So I suppose things would be fine if people compare the downloaded copy's checksum with mine.
Would there be something wrong when moving the upload around in the ftp?
You tell me maybe you've got intelligent virus on your mac
Please please re-upload them - I've got a brick right now instead of Z
Ok... just to play safe. Let me boot this rootfs up on my C860 first. Then I will upload the tarball again with the checksum to both yours and jpmatrix's servers.
-
Test done! The rootfs does boot on SD on my C860 (Angstrom kernel 2.6.21), via altboot.
But I discovered one curious thing.
Boot stops when loading sshd. After removing ssh, openssh-client and openssh-server, boot goes through!
This happens to the C860 only; the same rootfs boot on my 3200 without any issue (with ssh installed).
Will proceed to (re)uploading right away.
Now also i am thinking of making another tarball for 860 (and perhaps also suits C1000 if using angstrom kernel).
-
As my 3200 is now running two gftp side by side (one uploading to jpmatrix's, the other kkazakov13's ), I found one more thing interesting about the different behaviors of the same rootfs on 3200 and 860.
The landscape/portrait function of switchevd in 2or0's zaurusd package behave in opposite ways!
On 3200, it works just fine.
On 860, X will start in the portrait mode. Turning the screen into portrait makes the screen switch in landscape mode.
So I need to adjust the values in /etc/zaurud/hinge-landscape.d/20-matchbox-landscape (normal -> left) and /etc/zaurusd/hinge-portrait.d/20-matchbox-portrait (right -> normal) . Then it works as expected.
Now go dealing with the touchscreen...
jpmatrix, I have a question about your trick in post #39:
You said you did this:
rmmod mousedev (for the stylus jittering)
cd /dev/input
ln -s ts0 event1
then startx, open a console and ts_calibrate to recalibrate the screen
But,
"rmmod mousedev" doesn't work, (ERROR: Module is in use)
And in /dev/input, event1 is already there, but no ts0, so why "ln -s ts0 event1"?
I will try to create a tarball for 860 if the touchscreen is also fixed.
[EDIT]
touchscreen fixed.
Just comment out the line "module_raw corgi" in /etc/ts.conf, and then
Edit ~/.profile. Change /dev/input/ts0 to /dev/input/event1. No symlink is necessary.
One more different behaviour using 2or0's yonggun-specific packages with Angstrom.
The ipaq-sleep package does not seem to work. Pressing power button will not power off the machine. It stays on and returns to the previous display in a few seconds. Removing the package suspend/resume work as expected.
-
Lol.
Here is mine: e3133206a7846f585bd41c8dfb58f0b2
Same on my Z and my Mac ...
oh!! it is not what you put in the md5sum file on my ftp !
ps: i'm on irc too
Yes, you are right. My apology, sir!
Anyway, now uploading again. Please check again. Thanks.
-
jpmatrix, I have a question about your trick in post #39:
You said you did this:
rmmod mousedev (for the stylus jittering)
cd /dev/input
ln -s ts0 event1
then startx, open a console and ts_calibrate to recalibrate the screen
But,
"rmmod mousedev" doesn't work, (ERROR: Module is in use)
And in /dev/input, event1 is already there, but no ts0, so why "ln -s ts0 event1"?
you've to exit X to rmmod mousedev
and apology too, it was : ln -s event1 ts0
yes you can also set tslib_tsdevice=/dev/input/event1
ps: uploading new rootfs (checked!) to matrixmen.free.fr !
-
Just d/l the newest tarball and the checksum is good, able to be unzipped via 7zip.
Tks everyone!
-
jpmatrix, I have a question about your trick in post #39:
You said you did this:
rmmod mousedev (for the stylus jittering)
cd /dev/input
ln -s ts0 event1
then startx, open a console and ts_calibrate to recalibrate the screen
But,
"rmmod mousedev" doesn't work, (ERROR: Module is in use)
And in /dev/input, event1 is already there, but no ts0, so why "ln -s ts0 event1"?
you've to exit X to rmmod mousedev
and apology too, it was : ln -s event1 ts0
yes you can also set tslib_tsdevice=/dev/input/event1
ps: uploading new rootfs (checked!) to matrixmen.free.fr !
Wait, I got this error outside X. So ...?
And in my case there is no need to do the symlinking. Just do the changes I decribed above and that's it. I do this for my C860 using Angstrom kernel 2.6.21.
-
Just d/l the newest tarball and the checksum is good, able to be unzipped via 7zip.
Tks everyone!
Great. Do let me know how it goes.
I will not be online till tomorrow night. So good luck, man!
-
ZDevil: you can post mine mirror too in the first post - so everyone can download from there too. It's not only for me
but meanwhile - here it is http://debian.drun.net/rootfs/ (http://debian.drun.net/rootfs/)
P.P. ZDevil: tnx, I'm downloading it when back from work, will check and report
-
Done. Thanks for hosting it.
It's good to have a mirror of jpmatrix's server.
So kkazakov13, do you think you can create a mirror repo of jpmatrix's in your server (of course only after you get debian set up and start playing)?
-
"rmmod mousedev" doesn't work, (ERROR: Module is in use)
And in /dev/input, event1 is already there, but no ts0, so why "ln -s ts0 event1"?
If you install udev, please check /etc/init.d/udev and then remove the word "mousedev" in this file not to autoload mousedev module at booting.
-
Done. Thanks for hosting it.
It's good to have a mirror of jpmatrix's server.
So kkazakov13, do you think you can create a mirror repo of jpmatrix's in your server (of course only after you get debian set up and start playing)?
These days are very intense for me ... I will create a mirror saturday or sunday (hopefully!)
-
Just curious: How is the new tarball now? Does it work properly?
-
Just curious: How is the new tarball now? Does it work properly?
will install asap--but that might be a day or two. thanks for doing it in advance, though.
I plan to install to cf, then format and istall to the internal A-data from there. hopefully that works. otherwise i will install to sd
-
Just curious: How is the new tarball now? Does it work properly?
Just installed, w/ kernel 2.6.23.9 yonggun, and played for awhile. Everything works so far
-
Just curious: How is the new tarball now? Does it work properly?
Just installed, w/ kernel 2.6.23.9 yonggun, and played for awhile. Everything works so far
Good. Is there any other action that you have to take apart/different from those in the instructions?
Enjoy the debian experience.
-
with 2.6.23.1 i could not calibrate, some ts_open errors, with 2.6.23.9 is ok.
mplayer with pxa from 2or0's site does not work on both kernels. Reinstalled everything fresh on my C3200, no problems whatsoever except that weird shadows that look like something desynchronized (it flickers a bit too), on the light windows over dark windows or background or sth like this ... it's really annoying are there any settings to tweak about this desynchronization or flickering or whatever it is ...
-
Can you post a screenshot of the window shadowing thing you've got? I still haven't seen this on both my 3200 and 860.
For the ts_open error, make sure that:
1) There is a line "export TSLIB_CONFFILE="/etc/ts.conf" in ~/.profile, besides export TSLIB_TSDEVICE.
2) The directory /usr/lib/ts is present.
Also I've tried mplayer -vo pxa to play .mpg files and it works great.
-
Also I've tried mplayer -vo pxa to play .mpg files and it works great.
I know it's working with you ... it's working with everyone else than me
could you give me more info from your system ... trying to find out where the problem might be. Like:
ls -al /dev/fb*
cat /proc/fb
would be nice
P.P. Another question - when I hold down the stylus while in X it's moving around and I often cannot click a button. Is it like this with you ?
-
P.P. Another question - when I hold down the stylus while in X it's moving around and I often cannot click a button. Is it like this with you ?
it's the actual mouse jittering trouble we're having
the solution is to exit X, rmmod mousedev and restart X
that worked for me, although there's still some jittering but less than with mousedev loaded
-
... except that weird shadows that look like something desynchronized (it flickers a bit too), on the light windows over dark windows or background or sth like this ... it's really annoying are there any settings to tweak about this desynchronization or flickering or whatever it is ...
Okey, I think I understand you. When look closer, I notice a bit flickering on the background. When the menu pops up, the whole background directly above appears to be darker. Well, a bit annoying too, but not as in pdaxrom overclocked and w/o cf.
I don't think we can take a photo to show that.
Good. Is there any other action that you have to take apart/different from those in the instructions?
Just a minor thing: The emergency mode in Sharp ROM is not bz2 capable. So I have to unzip the rootfs file on PC first. Also, I do have 'unable to open console' error, but I always fail to MAKEDEV within uboot emergency mode. So my install sequence is: at Sharp ROM emergency mode, prepare the internal CF, untar the rootfs, do MAKEDEV, and then install uboot blah blah blah.
-
... except that weird shadows that look like something desynchronized (it flickers a bit too), on the light windows over dark windows or background or sth like this ... it's really annoying are there any settings to tweak about this desynchronization or flickering or whatever it is ...
Okey, I think I understand you. When look closer, I notice a bit flickering on the background. When the menu pops up, the whole background directly above appears to be darker. Well, a bit annoying too, but not as in pdaxrom overclocked and w/o cf.
Sorry, I still don't understand this, because I don't see what you describe ...
Good. Is there any other action that you have to take apart/different from those in the instructions?
Just a minor thing: The emergency mode in Sharp ROM is not bz2 capable. So I have to unzip the rootfs file on PC first. Also, I do have 'unable to open console' error, but I always fail to MAKEDEV within uboot emergency mode. So my install sequence is: at Sharp ROM emergency mode, prepare the internal CF, untar the rootfs, do MAKEDEV, and then install uboot blah blah blah.
The initial console error: instead of MAKEDEV, a simple reboot seems to fix this as well.
I see. I extracted debian eabi under debian oabi and angstrom, so I was not aware of this uncompression issue. Thanks. Will add this to the instructions in the first post.
-
Also I've tried mplayer -vo pxa to play .mpg files and it works great.
I know it's working with you ... it's working with everyone else than me
could you give me more info from your system ... trying to find out where the problem might be. Like:
ls -al /dev/fb*
cat /proc/fb
would be nice
P.P. Another question - when I hold down the stylus while in X it's moving around and I often cannot click a button. Is it like this with you ?
Okay. Here is the output of ls -al /dev/fb*
crw-rw---- 1 root video 29, 0 2007-12-07 20:39 /dev/fb0
crw-rw---- 1 root video 29, 1 2007-12-07 20:39 /dev/fb1
crw-rw---- 1 root video 29, 2 2007-12-07 20:39 /dev/fb2
crw-rw---- 1 root video 29, 3 2007-12-07 20:39 /dev/fb3And cat /proc/fb :
0 PXA
1 overlay1
2 overlay2
3 cursorDoes this help?
Sidenote: apt-get dist-upgrade will replace the preinstalled mplayer (a debianized angstrom package) with the official one, which seems to fail to play with pxa option. Just dpkg -i install the yonggun mplayer deb will get it back.
The jittering issue: how about simply do a ts_calibrate or reboot?
(Posting using epiphany-browser on my 3200 )
-
Well ... it's just simply not working - nothing from things you guys mention that work with you.
jittering is not removed with rmmod mousedev, slightly less, but still annoying and hard to click sometimes. mplayer DOES NOT work, no matter what i try, what i install - it just does NOT.
and the calibrate - well, i NEVER got it to work like - wherever I click - to be exactly there the pointer. I tried calibrating it about 10 times - there's always some difference.
The most annoying thing still remains these artifacts under X - they are hardly visible sometimes ... mostly are visible on single color darker background - but are visible.
The problem is that I need to work with zaurus sometimes ... and I still cannot without any problems.
-
Umm... suppose you just did exactly what other people (including me) did, and you got all these mplayer, touchscreen and shadowing issues, then I don't really have anything to say on top of my head... and I can't really explain how these things work for me either.
If i were you, I would just do it all over again if I still want to run Debian. Otherwise, just blame on the bad luck...
-
Umm... suppose you just did exactly what other people (including me) did, and you got all these mplayer, touchscreen and shadowing issues, then I don't really have anything to say on top of my head... and I can't really explain how these things work for me either.
If i were you, I would just do it all over again if I still want to run Debian. Otherwise, just blame on the bad luck...
I followed everything exactly - except that i bzip2 -d 'ed the file on my machine before unpacking to hdd. Did reinstallations about 4 or 5 times ...
Anyway - I guess it's just a bad luck
-
Then, kkazakov13, if you don't mind me asking again:
-- What do you have in /dev/fb* and /proc/fb ? Does it help if you (re)install the three mplayer-related debs here: http://yonggun.tistory.com/54 (http://yonggun.tistory.com/54) (under "Video" at the very bottom)?
-- How is your touchscreen settings (e.g. ~/.profile, /etc/ts.conf )
-- How to reproduce the window shadowing, so that I can try and see if I can also see that?
-
Then, kkazakov13, if you don't mind me asking again:
-- What do you have in /dev/fb* and /proc/fb ? Does it help if you (re)install the three mplayer-related debs here: http://yonggun.tistory.com/54 (http://yonggun.tistory.com/54) (under "Video" at the very bottom)?
I have the same things in /dev/fb* and /proc/fb as yours. The mplayer included in the rootfs had not pxa option, the one you mention have pxa but does not work. I installed the 3 files (removing first the original ones)
-- How is your touchscreen settings (e.g. ~/.profile, /etc/ts.conf )
Haven't touched them - whatever is in the rootfs
-- How to reproduce the window shadowing, so that I can try and see if I can also see that?
Can't tell you - it's just there! the first time I boot IceWM from the rootfs. Did _nothing_.
P.P. Anyway - I'm stopping to experiment with Debian for now as I need _working_ machine ( going back to my heavily customized beta 3 ... seems this is the most stable for now). I really wanted to watch 640x480 movies without overclocking ...
After some time maybe I will try again. Will keep monitoring the Debian section closely
-
Then, kkazakov13, if you don't mind me asking again:
-- What do you have in /dev/fb* and /proc/fb ? Does it help if you (re)install the three mplayer-related debs here: http://yonggun.tistory.com/54 (http://yonggun.tistory.com/54) (under "Video" at the very bottom)?
I have the same things in /dev/fb* and /proc/fb as yours. The mplayer included in the rootfs had not pxa option, the one you mention have pxa but does not work. I installed the 3 files (removing first the original ones)
-- How is your touchscreen settings (e.g. ~/.profile, /etc/ts.conf )
Haven't touched them - whatever is in the rootfs
-- How to reproduce the window shadowing, so that I can try and see if I can also see that?
Can't tell you - it's just there! the first time I boot IceWM from the rootfs. Did _nothing_.
P.P. Anyway - I'm stopping to experiment with Debian for now as I need _working_ machine ( going back to my heavily customized beta 3 ... seems this is the most stable for now). I really wanted to watch 640x480 movies without overclocking ...
After some time maybe I will try again. Will keep monitoring the Debian section closely
i don't want to be sad but could all your problems come from an actual hardware problem ? perhaps oriented toward the touchscreen ? and which appear in certain situations only.... i don't know....
before going back to pdaxrom can you try latest angstrom to see if you have same problems?
-
i don't want to be sad but could all your problems come from an actual hardware problem ? perhaps oriented toward the touchscreen ? and which appear in certain situations only.... i don't know....
before going back to pdaxrom can you try latest angstrom to see if you have same problems?
Definately _not_ a hardware problem. I'm already back to beta3 ... needed to write some docs later in the bed, while listening some music ... anyway ... i have no shadow issues under X, and no jittering under beta 3. Also I did tried angstrom about 2 weeks ago - I did not had these problems there either. Don't remember the kernel version there. BVDD driver also works under pdaXrom beta3, so I suppose my pxa layer is not broken either
-
Hello Zdevil,
Thanks for your excellent work. I've installed it on my C3200 and it's working very well (the wifi card is connecting very very fast and flawlessly).
The only remark I have to do is that, in the new tarball, the mplayer installed does not have the pxa driver option. I've had to install the younggun ones to make it work with pxa (I'm having troubles with aspect correction in pxa but I'll open another topic to deal with it).
I have the 2.6.23.9 kernel installed, are the modules of that kernel already in the tarball (if so, the first topic must be coreected).
Thanks,
glac
-
Just got my 3100, copied the tarball over etc ...
stuck at Warning: unable to open an initial console.
Kernel panic - not syncing: Attempted to kill init!
I get this regaurdless of the kernel or image ... even get it when I copy the working C1000's debian off SD to the HDD.
Very frustrating thing is the emergency console has nothing that enable's me to access the hdd. And rebooting does not fix the console error for me.
-
Just got my 3100, copied the tarball over etc ...
stuck at Warning: unable to open an initial console.
Kernel panic - not syncing: Attempted to kill init!
I get this regaurdless of the kernel or image ... even get it when I copy the working C1000's debian off SD to the HDD.
Very frustrating thing is the emergency console has nothing that enable's me to access the hdd. And rebooting does not fix the console error for me.
just reboot and try to wait a little at this error message
to have a beautiful emergency console, boot pdaxii13 install and choose "go to console" (something like this)
-
How did you do that? VERY interested.
Yeah the kernel panic is a stopper ... what am I doing wrong? PdaXii13 is looking better and better ...
-
How did you do that? VERY interested.
Yeah the kernel panic is a stopper ... what am I doing wrong? PdaXii13 is looking better and better ...
go get the pdaxii13 install files, or get them here : http://matrixmen.free.fr/zaurus/pdaxii13/ (http://matrixmen.free.fr/zaurus/pdaxii13/)
then proceed to flash process, with battery off, reset, then battery on, then OK-ON, choose 4, SD or CF, then Y
then at the pdaxii13 install menu, just choose to go to console, there you can mount hdd to do the MAKEDEV procedure
of course don't flash anything and just reboot
-
I gave it a whirl-- used up a few hours, never got anything working right- saw kkazakov's posts and headed back to pdaxii13.
I think this project, if any, has a future. The feature set in 2or0's kernels alone makes it worth watching. I'll try again soon (I have a vacation coming). In the meantime any idiot-proofing of the installation and guide would be much appreciated. Also, is there some way to backport the goodies, like webkit, from sid to lenny (an keep the project on testing)? Sid breaks things, and it is tough enough to manage a shoe-horned port alone, without managing a "port" (I know, not technically what it is, but you get the idea) of an already unstable system.
I'm back to pdaxii13v2 with a touch of etch for the time being. Hopefully that isn't seen as criticism or condemnation. A "big" distro, like debian, on the Z was what I'd been hoping for since I bought my 5600 almost 4 years ago. You guys seem to have achieved that, though not quite in a way I'm comfortable using, yet. Keep up the good work, please
-
Just spent four days strait ticking myself off. The kernel panic has stumped me. For now I will head back to PdaXii13. I think Cortez has something, a simple boot flash ... just my network card isn't supported. I will keep an eye out.
-
it's "funny" to see how people can have lots of difficulties to install their debian on zaurus and how it works without any major problems for other people........ i try to make a poll about that right now
-
Indeed, its strange ... had no issues getting it onto the C1K, went easily. But the HDD on the C3.1K seems to want to fight, I will in the end return (can't help it).
-
Remember the "old" days when lilo was "it" and systems had a /boot partition....? would this not be much easier (on all but 3000..and maybe even that) if we did some thing like Nand=/boot? Probably there is some obvious difficulty, but if there isn't it seems like this would be a great way to handle the various media issues.
As someone with a comparatively limited skillset and non-standard hardware, this sort of solution has obvious personal appeal.
Jpmatrix--sign me up as 1 count of "struggled to no avail."
-
Back on Debian EABI -lol- Could not stay away, so twitchy
So now with all the tweaking again ...
-
I'm using Debian EABI 2.26.23.9 Y on mine SL-C3200 with the first Zdevil's tarball. I have one question:
How do I overclock my system? When I try to use cpuoverclock.py in X, everytime I change the clock and I hit ok, nothing happens. Some messages appear in my console complaining of a config file not being found. Is there any other way to overclock in console mode? Or is there any fix of this script? Maybe I could try reinstalling younggun's cpuoverclock and see what happens.
glac.
-
I'm using Debian EABI 2.26.23.9 Y on mine SL-C3200 with the first Zdevil's tarball. I have one question:
How do I overclock my system? When I try to use cpuoverclock.py in X, everytime I change the clock and I hit ok, nothing happens. Some messages appear in my console complaining of a config file not being found. Is there any other way to overclock in console mode? Or is there any fix of this script? Maybe I could try reinstalling younggun's cpuoverclock and see what happens.
glac.
In console, "dvfm 16 6 1 2 1" is for 624 and "dvfm 16 4 1 2 1" for 416.
-
tarball downloaded, installing now on C1k, keep your fingers crossed
-
Fu**ing u-boot! Nothing for the fainted heart!
Ended here: https://www.oesf.org/forum/index.php?showtopic=24199 (https://www.oesf.org/forum/index.php?showtopic=24199) , doing a NAND-restore now.
Maybe I'll try it again tomorrow.
tarball downloaded, installing now on C1k, keep your fingers crossed
-
I find u-boots not so scary.
If you do not have a linux system to host your efforts, flash first to pdaxii13 ... do all the file work and such from it. Once your ready, flash u-boot, upgrade the kernel too 2.6.22 (for the c1k) and your good to go.
-
Hi,
that's exactly what i've done. Prepared the sd-card with the rootfs on my debian (sidux) box. But maybe i used the wrong mkfs for the cf-card flashing uboot, don't know.
No charging light after flashing u-boot, having troubles entering Fn-D-M. Thanks god i managed to enter the diagnosis, did a NAND-restore and reflashed pdaxii3.
Anyway, i am keen enough on apt-get to give u-boot another try this evening
So please, folks: ext2 or vfat for uboot? And which version of u-boot? The one for pdaxrom r198 http://distro.ibiblio.org/pub/linux/distri...akita-uboot.zip (http://distro.ibiblio.org/pub/linux/distributions/pdaxrom/download/r198/c1000-c3x00/pdaXrom-1.1.0r198-Kathrin-akita-uboot.zip) ?
I find u-boots not so scary.
If you do not have a linux system to host your efforts, flash first to pdaxii13 ... do all the file work and such from it. Once your ready, flash u-boot, upgrade the kernel too 2.6.22 (for the c1k) and your good to go.
-
FAT16 CF, r198 uboot works for me.
-
First: Thanks a lot for your great work ZDevil. I got Debian running on my C1000 and it "feels" good and fast. My only problem (so far): Is there a chance to get a WPA connection working with your tarball-installation? If yes how? (I didn't find wpa-supplicant).
Thanks in advance
zaurier
-
First: Thanks a lot for your great work ZDevil. I got Debian running on my C1000 and it "feels" good and fast. My only problem (so far): Is there a chance to get a WPA connection working with your tarball-installation? If yes how? (I didn't find wpa-supplicant).
Thanks in advance
zaurier
I found wpasupplicant package in the feed.
-
zaurus_eabi_rootfs_071201.tar.bz2 don't work *.bz2 with emergency sharp-menu [D+B] trick
zaurus_eabi_rootfs_071201.tar is BIG - 605M
solution with my SD 512M
zogs@linux:~/wrk> bunzip2 zaurus_eabi_rootfs_071201.tar.bz2
zogs@linux:~/wrk> gzip -c zaurus_eabi_rootfs_071201.tar > zaurus_eabi_rootfs_071201.tar.gz
zogs@linux:~/wrk> ls -alh zaurus_eabi_rootfs_071201.tar.gz
-rw-r--r-- 1 zogs users 226M 2007-12-11 16:05 zaurus_eabi_rootfs_071201.tar.gz
now with emergency sharp-menu [D+B] trick and my SD-512M...I can [y0y0]
# mount /dev/hda1 /hdd1
# cd /mnt/card
# cp zaurus_eabi_rootfs_071201.tar.gz /hdd1
# cd /hdd1
# tar -xvpzf zaurus_eabi_rootfs_071201.tar.gz
-
Thanks, i'm on the road again...
FAT16 CF, r198 uboot works for me.
-
Hm, seems no luck for me with u-boot.
Installed u-boot, pressed ok when rebooting started. Entered the emergency system. Fine so long.
Inserted the cf with the kernel, mount to see if cf exists and is mounted, cd /mnt/cf. Still fine.
But now: ls made me wait for several minutes, nothing happend. Turned the Z around and pressed the reset switch.
Now nothing happens, just like yesterday. Doing a NAND restore. But: i will try it again. Before flashing u-boot, i will prepare another cf with the kernel.
Just to proof i'm doing right: Entering the emergeny console, cd /mnt/cf and ./autoboot.sh will install the kernel ?
Man, i can't wait for apt-get
-
autoboot.sh (posted in this topic ) don't work..
i have write in u-boot emergency shell:
nandlogical /dev/mtd1 WRITE 0x5a0000 0x160000 /mnt/card/kernel.img
-
Awesome! Great!
Debian booting on my C1k.
Just changed the command for /mnt/cf, it flashed so fast...
My spectrum24 network-cf didn't work (missing firmware ?), so i configured usb-networking. Runnning apt-get update now
autoboot.sh (posted in this topic ) don't work..
i have write in u-boot emergency shell:
nandlogical /dev/mtd1 WRITE 0x5a0000 0x160000 /mnt/card/kernel.img
-
I have still problems connecting wireless via WPA-TKIP. I installed wpa_supplicant and got it running but no success (wap_cli shows a pingpong between "accociated with..." and "CTRL-EVENT-DISCONNECTED - Disconnect event - remove keys")... Before i go in details my first question: Is anybody out there, who get this tarball installation running with wireless WPA connection?
thanks zuaurier
-
autoboot.sh (posted in this topic ) don't work..
i have write in u-boot emergency shell:
nandlogical /dev/mtd1 WRITE 0x5a0000 0x160000 /mnt/card/kernel.img
I didn't have any problem using the attached script. The command you used is already in the autoboot.sh (notice the autoboot.sh for debian is different from the one for pdaXrom) Or is it only me?
-
Hoy all
As I explained on my blog: http://www.zecrazytux.net/GNU-Linux/Zaurus...x-on-the-Zaurus (http://www.zecrazytux.net/GNU-Linux/Zaurus/13-Debian-GNU-Linux-on-the-Zaurus) I got busy for a time... And i did'nt succeed to get clicks working.
I did'nt noticed that a younggun kernel existed for c1000 zauriis.
Lots of thanks for your work zdevil ! it works perfectly !
-
Hoy all
As I explained on my blog: http://www.zecrazytux.net/GNU-Linux/Zaurus...x-on-the-Zaurus (http://www.zecrazytux.net/GNU-Linux/Zaurus/13-Debian-GNU-Linux-on-the-Zaurus) I got busy for a time... And i did'nt succeed to get clicks working.
I did'nt noticed that a younggun kernel existed for c1000 zauriis.
Lots of thanks for your work zdevil ! it works perfectly !
blog bookmarked
il fait beau à Troyes ?
-
autoboot.sh (posted in this topic ) don't work..
i have write in u-boot emergency shell:
nandlogical /dev/mtd1 WRITE 0x5a0000 0x160000 /mnt/card/kernel.img
I didn't have any problem using the attached script. The command you used is already in the autoboot.sh (notice the autoboot.sh for debian is different from the one for pdaXrom) Or is it only me?
scenario:
pdaXrom emergency system
==================
pdaXrom [Borzoi]
Loading /etc/sysconfig/keyboard/Borzoi/keymap.map
Bootscript /mnt/card/autoboot.sh, go...
: command not foundsh : line 2:
: command not foundsh : line 4:
: command not foundsh : line 6:
Proceed with debian installation or boot to emergency system? [y/n]
: command not foundsh : line 8:
y
' : not a valid identfierne 9: read: 'ans
: command not foundsh : line 10:
'mnt/card/autoboot.sh: line 19: syntax error near unexpected token 'do
'mnt/card/autoboot.sh: line 19: 'for file in $LOC/kernel.img $LOC/KERNEL.IMG; do
Processing /etc/profile... Done
-sh-3.08# cd /mnt/card
-sh-3.08# sh autoboot.sh
[command not foundsh..etc etc repeat]
-sh-3.08# ./autoboot.sh
-s: ./autoboot.sh: /bin/sh^M: bad interpreter: No such file or directory
-sh-3.08# vi autoboot.sh
o my god !! ^M on end of every line
delete all ^M from autoboot.sh and :wq
restart with u-boot emrgency for kernel installation and it work fine
the problem is the format of autoboot.sh sorry
-
WPA connection still not working (I use the Ambicom WL1100C-CF card with my c1000). Before I try another 1000 configuration examples with wpa_supplicant.conf and network configuration files my question: Is anybody out there, who actually succeeded in WPA-connecting with this eabi-tarball-installation?
zaurier
-
I have found that things that aren't Z specific can be found in alot of the debian forums if that helps. Its where I have had to go to get a number of configuration options.
-
I have found that things that aren't Z specific can be found in alot of the debian forums if that helps. Its where I have had to go to get a number of configuration options.
I know that and of course I searched the (debian) forums for configuration problems/options and tried REALLY a lot... But none of the tips and configuration-samples helped/worked. I ask in this forum because my AMBICOMcard worked with all other ROMs I have installed so far. I just want to know whether somebody succeeded with THIS eabi-tarball-installation to make sure its "only" (my) configuration problem and it is possible to get the card working with THIS INSTALLATION and WPA encryption...
thanks zaurier
-
Before i go in details my first question: Is anybody out there, who get this tarball installation running with wireless WPA connection?
thanks zuaurier
Ok, maybe it's about time to go in details?
-
hello,
can someone please help:
* i want to use zdevils instructions to install debian on my z and went
into problems with uboot
* i read a thread of capn_fish what to do if something goes wrong while
installing uboot, but nothing helped.
* meanwhile i was able to install uboot using
pdaXrom-1.1.0r198-Kathrin-akita-uboot.zip coming from a fresh sharprom
(trisoft nand)/(fresh)cackorom installation, but after that i cant reboot
the zaurus using the on/off button and i cant get into the desired emergency
mode. nothing happens, cant get into the japanese mode or do c+d+reset to
install the kernel. sometimes even (fn+)d+m didnt worked, no headaches yet:-)
* i guess the installation order is alright. do i need to prepare the rootfs
on sd as the first step? i guess no...
my hope was to keep the cackorom in the flash and to use it (with kexec
from angstrom).
-
hello,
can someone please help:
* i want to use zdevils instructions to install debian on my z and went
into problems with uboot
* i read a thread of capn_fish what to do if something goes wrong while
installing uboot, but nothing helped.
* meanwhile i was able to install uboot using
pdaXrom-1.1.0r198-Kathrin-akita-uboot.zip coming from a fresh sharprom
(trisoft nand)/(fresh)cackorom installation, but after that i cant reboot
the zaurus using the on/off button and i cant get into the desired emergency
mode. nothing happens, cant get into the japanese mode or do c+d+reset to
install the kernel. sometimes even (fn+)d+m didnt worked, no headaches yet:-)
* i guess the installation order is alright. do i need to prepare the rootfs
on sd as the first step? i guess no...
my hope was to keep the cackorom in the flash and to use it (with kexec
from angstrom).
Forget about Cacko on flash when you want to use uboot and debian : they use different kernels.
(To make sure uboot is working : flash pdaXrom r198 following the instructions.
If you can boot it, uboot is ok.)
Do you need to prepare the rootfs : yes.
What else do you need to to : flash the kernel (actually, if you can flash the kernel uboot is ok)
Then : just reboot.
About installing uboot, flashing the kernel :
uboot is very picky about sd-cards and how they are formatted/partitioned. I always use a card with 2 partitions where the first one is fat-formatted and I always succeed. Before I used this card I used a card with 1 partition (a new one !) and never got it working.
Good luck,
Chero.
-
I agree with Chero, if you have PdaXrom 198 running uboot is fine, flashing it is easy from there. I am thinking that it won't be a happy marriage if you try to run cacko and u-boot. So far I have only successfully got a dual boot setup out of Angstrom & Debian.
I too use a dual partitioned card, fat on first and ext3 on the larger second partition (lesson learned while messing with the C1k and debian).
-
thanx for the quick answers. it is possible to run angstrom from hd and cacko
from nand using kexec. does the necessity of uboot kill this kind of scenario for
yongun/debian and cacko? i mean booting to a debian console, executing kexec
from the angstrom feed... maybe uboot changes the nand flash in such a way that
this becomes impossible.
-
partitioning the cards and flashing r198 is not needed. bad news for me: kexec executes
the cacko kernel ripped from tools.tar, but the kernel hangs while booting. the preinstalled
nand image seems to be disturbed by uboot and/or the yonggun kernel. maybe i will try
it from a clean cacko or sharprom install...
-
1) uboot installation from cf card and japanese menu: works now for me
but only when the sd card isnt inserted (not uboot specific)
2) rebooting works fine after installing the yongun kernel, i.e. using
reset or a reboot from console. before its a really headache, the zaurus
seems not to react on anything or arbitrarily concerning wakeups/reboots or
d+m stuff. can someone confirm this behaviour?
3) script didnt worked for me, executed the nandlogical command by hand
4) some progress with kexec and a fresh cacko installation, getting a lot
of error messages and warnings, but now i can see the cacko screen, starting
linux and qtopia... the story ends here:-)
-
i am very lost, when i boot into emergeny mode how do I repartation the drive with commands and put the file systems on them. The instructions looks like they assume I know what im doing. I am using a C3000 and trying to get EABI to work with the armstrong kernal. The kernal on Titchys homepage is the correct kernal right?
-
i am very lost, when i boot into emergeny mode how do I repartation the drive with commands and put the file systems on them. The instructions looks like they assume I know what im doing. I am using a C3000 and trying to get EABI to work with the armstrong kernal. The kernal on Titchys homepage is the correct kernal right?
hey a C3000 friend ! i thought i was the only one
well, i usually use the pdaxii13 flash kit to deal with my C3000 partitions. you'll find the files there :
http://matrixmen.free.fr/zaurus/pdaxii13/ (http://matrixmen.free.fr/zaurus/pdaxii13/)
put them on a card, apply the C3000 flash procedure (battery off, reset, battery on, on-ok power up, then 4, sd, y) and go to pdaxii13 console menu
then format your hdd, and un-tar there the debian rootfs
next prepare a card with angstrom kernel (all files execpt the hdimage1.tgz) then flash angstrom kernel
example of angstrom files needed : http://matrixmen.free.fr/zaurus/angstrom/ (http://matrixmen.free.fr/zaurus/angstrom/)
that should leave you a working eabi on your c3000
ps: once into debian you should untar modules2.6.23.tgz, depmod -a and reboot
-
Before i go in details my first question: Is anybody out there, who get this tarball installation running with wireless WPA connection?
thanks zuaurier
Ok, maybe it's about time to go in details?
I think I'm now a bit closer to a solution with my wireless WPA-connection/encryption problem. After investigating the modules I found out that the module "ieee80211_crypt_tkip.ko" is missing in /lib/modules/2.6.22.9-yonggun/kernel/net/ieee80211 but I found this module in the other two module-directories (2.6.23.1 and 2.6.23.9). Maybe it was just "forgotten" in the 2.6.22.9-yonggun kernel. Unfortunately does the c1000 only work with the 2.6.22.9-kernel. I think this module is essential for wpa-encryption with the hostap-driver. How can I get/compile/install this module also for "my" 2.6.22.9-yonggun kernel? Sorry for this stupid question, but I'm only a "user" and not so familiar with compiling and "those things"...
zaurier
-
2or0 Has an experiemental 2.6.23 kernel that works on the C1000, personally used it so I know it works.
-
2or0 Has an experiemental 2.6.23 kernel that works on the C1000, personally used it so I know it works.
Thanks for this information. Where can I get this "experimental kernel"? I tried to install 2.6.23.1-yonggun and 2.6.23.9-yonggun kernel. After getting into uboot emergency boot I succeeded in flashing the kernel like described in the first post of this threat but unfortunately no rebooting after flashing (tried with both 2.6.23 yonggun kernels). Reinstallation of my "old" 2.6.22.9-yonggun kernel worked out of the box. Some things I should know?
zaurier
-
Hello!
Thank you very much for this detailed instructions and the rootfs
After using it, I have a few suggestions that would make the installation IMHO easier and better:
-Please, include pre-installed wpa_supplicant. It is a really small package that will bring no more dependencies, and it gives the possibility to join our wifi networks at home when installing ( because all of you have WPA, right ? )
-Please, include dropbear instead of openssh-server. Its memory footprint is really smaller and the funcionality is almost the same.
-
Love ssh
Its all getting there, I am just in awe it runs at all on this size device!
-
After using it, I have a few suggestions that would make the installation IMHO easier and better:
-Please, include pre-installed wpa_supplicant. It is a really small package that will bring no more dependencies, and it gives the possibility to join our wifi networks at home when installing ( because all of you have WPA, right ? )
-Please, include dropbear instead of openssh-server. Its memory footprint is really smaller and the funcionality is almost the same.
Glad that you like it and thanks for the suggestions.
I am planning to make an updated tarball (before christmas? ) to include some minor fixes:
1. a working mplayer (the one packaged by 2or0 and not the one from the debian armel repos)
2. a revised repo list (/etc/apt/sources.list) to catch up with the change of the feed servers
3. a preinstalled application menu (for icewm and others), which is missing in the current tarball
4. (I remember I forgot a couple of other things, but they don't pop up on top of my head. so any other suggestion?)
Surely if i manage to make it happen i will attach a brief changelog so that existing users will know what to add (i guess many have already figured it out), instead of reinventing the wheel and flash everything afresh.
And regarding your suggestions:
wpa_supplicant: I am not sure how to get this work and where to get what because my cf wifi cards don't seem to support wpa. Any hint?
dropbear: will test it out. in fact my C860 halts during boot (loading ssh server) with the latest rootfs with openssh-server installed. Still don't know what happens. Removing it boot goes fine.
-
3. a preinstalled application menu (for icewm and others), which is missing in the current tarball
by the way, is there a simple way to update icewm menus as you install more and more apps ?
-
Is there a way to dual boot cacko and debian EABI using uboot?
-
Is there a way to dual boot cacko and debian EABI using uboot?
good question, see also my postings above. i used kexec from the angstroem
feed, had no luck with this, but try it yourself. it boots the cacko kernel. maybe
this is possible with special nand tools.
-
After using it, I have a few suggestions that would make the installation IMHO easier and better:
-Please, include pre-installed wpa_supplicant. It is a really small package that will bring no more dependencies, and it gives the possibility to join our wifi networks at home when installing ( because all of you have WPA, right ? )
-Please, include dropbear instead of openssh-server. Its memory footprint is really smaller and the funcionality is almost the same.
Glad that you like it and thanks for the suggestions.
I am planning to make an updated tarball (before christmas? ) to include some minor fixes:
1. a working mplayer (the one packaged by 2or0 and not the one from the debian armel repos)
2. a revised repo list (/etc/apt/sources.list) to catch up with the change of the feed servers
3. a preinstalled application menu (for icewm and others), which is missing in the current tarball
4. (I remember I forgot a couple of other things, but they don't pop up on top of my head. so any other suggestion?)
Surely if i manage to make it happen i will attach a brief changelog so that existing users will know what to add (i guess many have already figured it out), instead of reinventing the wheel and flash everything afresh.
And regarding your suggestions:
wpa_supplicant: I am not sure how to get this work and where to get what because my cf wifi cards don't seem to support wpa. Any hint?
dropbear: will test it out. in fact my C860 halts during boot (loading ssh server) with the latest rootfs with openssh-server installed. Still don't know what happens. Removing it boot goes fine.
I'm waiting this new tarball while trying to take some time to install it. Thanks for your work.
-
Well, I've finally installed this rootfs. Everything was ok during install but I get an annoying surprise when starting X. I get a weird "shadow effect" in every window, bars and pointer. I've seen kkazakov has experienced the same with no luck in solving it so I tried on another machine. Both my 3100 and 3200 had the same issue. Any ideas on what could be the cause??
-
Seen it it both OABI and EABI versions of Debian. Figure its a kdrive issue on Debian specifically. Not sure if theres a fix.
-
Zdevil said in this thread that he didn't appreciate it. In titchy linux (OABI) I don't have any issue. Everything looks great on both machines. It would be interesting to solve it. I'll try asap video performance (after network card is configured) and share my exprience. Thanks.
-
Zdevil said in this thread that he didn't appreciate it. In titchy linux (OABI) I don't have any issue. Everything looks great on both machines. It would be interesting to solve it. I'll try asap video performance (after network card is configured) and share my exprience. Thanks.
Not exactly. It's just that I don't see window shadowing on both my C860 and C3200, so there is nothing I can say about that.
BTW the r3 of the EABI rootfs is underway. Now I am waiting for 2or0's packages to be added to jpmatrix's feed, so that apt-get and apt-pinning can work smoothly.
wpa-supplicant will also be added to the rootfs, but I can't test it because I am not using it now.
Any more suggestion?
-
Not exactly. It's just that I don't see window shadowing on both my C860 and C3200, so there is nothing I can say about that.
Wow ... are there different C3200 hardware configurations that I'm not aware of ? That's really strange some to expirience the problem, others not.
But on second thought, it's not hardware related, because this does not happen on any other distro
-
After testing a little harder with both machines I've seen something weird. It's subtle but I see the same effect on Pdaxii13v2 with openbox and icewm. So it should be a cross-rom problem. Maybe X related?
-
Tried running full X instead of Xfbdev? Would that have an effect on it? Seeings as almost all of them run Xfbdev, including most installs of debian on the Z.
-
Great Work on this.
Any chance of a new version with Cortez's Libertas Kernel and the Libertas drivers available during the install?
I am asking this as my D-Link CF WiFi card died and the most readily available CD card these days in the UK seems to be the Sparklan 802.11g card.
Or am I being dim and will putting the module on the Zaurus and replacing the Kernel image in the package in the 1st post of this thread be enough?
Thanks to everyone who has put work into this endeavour.
David
-
Zdevil said in this thread that he didn't appreciate it. In titchy linux (OABI) I don't have any issue. Everything looks great on both machines. It would be interesting to solve it. I'll try asap video performance (after network card is configured) and share my exprience. Thanks.
Not exactly. It's just that I don't see window shadowing on both my C860 and C3200, so there is nothing I can say about that.
BTW the r3 of the EABI rootfs is underway. Now I am waiting for 2or0's packages to be added to jpmatrix's feed, so that apt-get and apt-pinning can work smoothly.
wpa-supplicant will also be added to the rootfs, but I can't test it because I am not using it now.
Any more suggestion?
Try to add http://www.angstrom-distribution.org/unsta..._1.0-r3_all.ipk (http://www.angstrom-distribution.org/unstable/feed/all/prism-firmware_1.0-r3_all.ipk)
I used this package (with some tweaks, something in rc.d if I recall correctly) in order to load a WPA enabled firmware in my Ambicom WL1100C-CF at boot time
-
hey everybody
does anyone know if there's a kernel with ieee80211 tkip support for C1000 ?
youggun's 2.6.22 does not :-(
-
I think there's a yonggun 2.6.23.x out now, at least I saw it mentioned somewhere.
-
Indeed there is.
-
is it possible to install a new kernel without flashing (and maybe to be able to choose which kernel to boot) ?
-
hey everybody
does anyone know if there's a kernel with ieee80211 tkip support for C1000 ?
youggun's 2.6.22 does not :-(
Here is the ieee80211 tkip module for 2.6.22.9. Just remove 'tar.gz' in the file name.
[ You are not allowed to view attachments ]
-
2or2,
as you're in a compilation rush, would you compile kernel 2.6.23 without uboot or altboot or i-don't-know-boot, AND/OR compile the fuse modules for it please?
moreover, how do you compile your kernel ? i tried to compile the fuse module for my angstrom 2.6.23 kernel but it doesn't want to insmod. i need the fuse module for gmailfs and flickrfs
thanks in advance!
-
hey everybody
does anyone know if there's a kernel with ieee80211 tkip support for C1000 ?
youggun's 2.6.22 does not :-(
Here is the ieee80211 tkip module for 2.6.22.9. Just remove 'tar.gz' in the file name.
[ You are not allowed to view attachments ]
lots of thanks 2or0 ! i tried to compile my own kernel, with the gnu eabi toochain which is packaged for debian
i got a kernel.img of about 1.5mo but unfortunatly it doesn't work with my c1000 :-(
Now that's broken :s but thanks to your help i'll repair it and make wifi to work
-
hey everybody
does anyone know if there's a kernel with ieee80211 tkip support for C1000 ?
youggun's 2.6.22 does not :-(
Here is the ieee80211 tkip module for 2.6.22.9. Just remove 'tar.gz' in the file name.
[ You are not allowed to view attachments ]
well, it does'nt work :-(
vermagic is different, as the module is for 2.6.22.9 and the kernel is 2.6.22.9-yonggun... I don't know if that's the reason.
I get an error inserting the module: invalid format :-(
can you share kernel sources, pathces and the .config file ?
lots of thanks !
Edit: in fact i think i could build most of my own modules if you give me the module/build directory
-
BTW the r3 of the EABI rootfs is underway. Now I am waiting for 2or0's packages to be added to jpmatrix's feed, so that apt-get and apt-pinning can work smoothly.
wpa-supplicant will also be added to the rootfs, but I can't test it because I am not using it now.
Any more suggestion?
Try to add http://www.angstrom-distribution.org/unsta..._1.0-r3_all.ipk (http://www.angstrom-distribution.org/unstable/feed/all/prism-firmware_1.0-r3_all.ipk)
I used this package (with some tweaks, something in rc.d if I recall correctly) in order to load a WPA enabled firmware in my Ambicom WL1100C-CF at boot time
Thanks! I am also thinking of upgrading my home network security to WPA. For now the Z is the roadblock.
So to add to the booting script?
If it works for me I will add these stuff to the updated rootfs.
-
Thanks! I am also thinking of upgrading my home network security to WPA. For now the Z is the roadblock.
So to add to the booting script?
Yes.
If I recall correctly (It's been months since I booted my zaurus last time) the ipk install script tries to add something to rcN.d's scripts but it fails because of an updtate-rc.d's wrong syntax. If you examine the ipk contents, just check the install script, and correct the update.rc.d command.
It would be useful to build a correct deb package and to add it to the repositories.
If it works for me I will add these stuff to the updated rootfs.
Good.
I want to refresh my Debian install, but I'm too lazy to redo my complicate setup, so I'll wait for your ready rootfs
I think that all your work could lead to another installation method.
We could use your rootfs to build an installer using this following steps:
1) Install u-boot
2) Prepare an ext-2 sd with your rootfs (or a minimal rootfs, without X, but with your zaurus customizations).
3) Boot from the sd with your fs. The fs should be slightly modified in order to have some installation scripts. It should have some sort of script able to:
a ) Setup the network
b ) Partition and format the internal microdrive
c ) Install a base debian rootfs using debootstrap (I love debootstrap because It prepares an uptodate minimal debian system)
d ) Run a script that applies all post-install steps (install zaurus specifics debs and customizations, save the network config and the fstab prepared during installation); this script could be downloaded online in order to be updated without touching the installer.
3b) Alternatively to 3) user can choose an option that clones the SD rootfs using rsync on the internal microdrives and then modifies just the network config and the fstab.
4) Reboot from internal microdrive and all eventually run some post install script (one could prepare something in order to interactively change the timezone and the locales)
This could lead to a definitive, and neutral, Debian installer, that installs an uptodate minimal sid filesystem, with the minimal customizations needed to run on the Zaurus.
P.S.: how do you manage the user issue in your rootfs? I mean, I had to tweak startx scripts in order to run X server as root through sudo and all the other stuff as a normal user.
-
Here is the ieee80211 tkip module for 2.6.22.9. Just remove 'tar.gz' in the file name.
[ You are not allowed to view attachments ]
It's doesn't work :'(
-
If I recall correctly (It's been months since I booted my zaurus last time) the ipk install script tries to add something to rcN.d's scripts but it fails because of an updtate-rc.d's wrong syntax. If you examine the ipk contents, just check the install script, and correct the update.rc.d command.
I checked the zaurus and there's no rcN.d init script. The whole process is managed by udev/hostap, so you have just to install the ipk
-
3. a preinstalled application menu (for icewm and others), which is missing in the current tarball
This should be done automatically with the debian update-menus command.
Any supposed release date for the updated rootfs?
-
Okay, now I have a little time slipping back to this project. (My Eee PC is just keeping me occupied all the time ... it's my primary computer day in and day out now )
3. a preinstalled application menu (for icewm and others), which is missing in the current tarball
This should be done automatically with the debian update-menus command.
Any supposed release date for the updated rootfs?
Exactly. But I am still thinking about using this because what the update-menus generates is not entirely convenient. Just too many submenus. And its syntax is a bit clumsy to edit.
Can someone confirm that wpa_supplicant works in EABI? It would be super nice if someone can post a concise how-to.
Besides the program menu, I am testing a nice app called Mercury: http://www.marengo-ltd.com/mercury/ (http://www.marengo-ltd.com/mercury/) It works very well on my Eee PC 701 (xubuntu 7.10 Gutsy).
It's practically the linux version of Quicksilver for Mac.
Any thought here?
-
Thanks for this great image it works perfectly for me, with sharp-bootloader-yonggun kernel.
Where can I find kexec? Its not in debian rep for arm...
EDIT: OK, I used the one in Angstrom feed
-
I am overwhelmingly by so many interesting updates and solutions popping up over these weeks!
ATM I am still very busy with work(, apart from helping in debugging a few things on the eee pc... ) Will get back on track asap.
To catch up with the fast development, here is a brief summary of the new cool stuff for the next rootfs tarball. Feel free to suggest what you want to add and what you think may be useful to general users)
The goal is very simple: to create a constantly updated, kernel-neutral (as possible) rootfs integrating the fruits scattering in the threads, bootable with uboot AND sharp bootloader and with one or more than one systems (sharp/cacko/pdaX/Angstrom).
0) Kernel
2or0's new kernel:
https://www.oesf.org/forum/index.php?showtopic=25351 (https://www.oesf.org/forum/index.php?showtopic=25351)
http://yonggun.tistory.com/66 (http://yonggun.tistory.com/66)
Dualboot without uboot:
http://yonggun.tistory.com/67 (http://yonggun.tistory.com/67)
(Thanks to 2or0's great work!!).
Solutions/Patches under consideration:
1) Jumping stylus pointer:
rmmod mousedev
source: https://www.oesf.org/forum/index.php?showto...st&p=174536 (https://www.oesf.org/forum/index.php?showtopic=25433&view=findpost&p=174536)
2) mplayer:
https://www.oesf.org/forum/index.php?showtopic=25456 (https://www.oesf.org/forum/index.php?showtopic=25456), esp Post#4
3) browser (icehamster):
https://www.oesf.org/forum/index.php?showtopic=25435 (https://www.oesf.org/forum/index.php?showtopic=25435)
4) internet phone:
Kphone/pi, wengophone
https://www.oesf.org/forum/index.php?showtopic=25326 (https://www.oesf.org/forum/index.php?showtopic=25326)
5) WPA-supplicant
https://www.oesf.org/forum/index.php?showtopic=25325 (https://www.oesf.org/forum/index.php?showtopic=25325)
6) Simple right-click (1-step tap-plus-keypress)
https://www.oesf.org/forum/index.php?showtopic=25459 (https://www.oesf.org/forum/index.php?showtopic=25459)
7) fbiterm
https://www.oesf.org/forum/index.php?showtopic=25463 (https://www.oesf.org/forum/index.php?showtopic=25463)
8) Additional packages/programs to include:
-- localepurge
-- Xfce4 (or any other lightweight WM that can generate and handle program menus easily)
-- Iceweasel (to be tested)
-- OpenOffice (to be aliened and tested)
9) Some tweaks to reduce writes on flash drives
From Eee PC sources:
http://wiki.debian.org/DebianEeePC/TipsAnd...40c3a4e9b00ac1f (http://wiki.debian.org/DebianEeePC/TipsAndTricks#head-634f28cee3f156fd4ce96013640c3a4e9b00ac1f)
https://help.ubuntu.com/community/EeePC#hea...598caa70481700b (https://help.ubuntu.com/community/EeePC#head-94f55ed3ed5ccd1078c593314598caa70481700b)
http://ubuntu-eee.tuxfamily.org/index.php5...our_Flash_Drive (http://ubuntu-eee.tuxfamily.org/index.php5?title=How_to:_Reduce_Disk_Writes_to_Prolong_the_Life_of_your_Flash_Drive)
-
nice! How is hardware support--bluetooth/GPRS dialup, wifi, screen rotatation..that sort of thing?
...and then all that is need is a glitch free android running in debian on this setup
-
8) Mercury (to replace the system menu; update-menus is okay but has rooms for improvement)
https://www.oesf.org/forum/index.php?showto...st&p=174312 (https://www.oesf.org/forum/index.php?showtopic=25029&view=findpost&p=174312)
I recommend grun which is in the feed.
-
8) Mercury (to replace the system menu; update-menus is okay but has rooms for improvement)
https://www.oesf.org/forum/index.php?showto...st&p=174312 (https://www.oesf.org/forum/index.php?showtopic=25029&view=findpost&p=174312)
I recommend grun which is in the feed.
Actually gmrun is already included in the present rootfs tarball posted here and is part of the system menu (if you pick "Run"). The difference between gmrun and mercury is that the latter does not only launch programs but also features incremental searching and lists most frequently used programs on the top, which is pretty close to what QuickSilver does on Mac OS: http://noypeee.blogspot.com/2007/12/mercur...uicksilver.html (http://noypeee.blogspot.com/2007/12/mercury-eee-answer-to-mac-quicksilver.html)
-
8) Mercury (to replace the system menu; update-menus is okay but has rooms for improvement)
https://www.oesf.org/forum/index.php?showto...st&p=174312 (https://www.oesf.org/forum/index.php?showtopic=25029&view=findpost&p=174312)
I recommend grun which is in the feed.
Actually gmrun is already included in the present rootfs tarball posted here and is part of the system menu (if you pick "Run"). The difference between gmrun and mercury is that the latter does not only launch programs but also features incremental searching and lists most frequently used programs on the top, which is pretty close to what QuickSilver does on Mac OS: http://noypeee.blogspot.com/2007/12/mercur...uicksilver.html (http://noypeee.blogspot.com/2007/12/mercury-eee-answer-to-mac-quicksilver.html)
Not gmrun but grun.
Description: gRun includes features such as a history, command completion from the history and from PATH, recognition of console mode applications and launching a terminal for them, file extension associations and a dual fork()/execvp() application launcher.
-
Hmm. I will look into grun again. IIRC, grun has a problem with window positioning and size, and gmrun is an improved version of grun. But 2or0 is right, there are far more functionality in grun (and gmrun also) than I was aware of.
Two more things to add to the list:
WM:
I will pick up xfce4 again and see how it fares in Debian. Instead of running menu-update, a better solution to the system menu in IceWM is to make use of the menus generated in Xfce4 (the gnome menus), which are much more tidy and organized. Having two WMs in a preinstalled system to choose from and play with is not such a bad idea. If there is xfce4 I will also do some basic setup on the desktop for OOTB use.
Locales:
Will preinstall and include the localepurge package, such that users can run it once upon installation to reclaim tens of MBs.
But first of all, I will have to do a NAND flash restore to uboot my setup on 3200 again, change it back to ext3 from ext2, and turn to 2or0's sharp bootloader way again... when I get extremely bored.
-
Oh yea, I almost forgot two very important things!
Iceweasel: it seems to be back again/soon! The debian-armel mailing list says the segfault error that everyone is experiencing may be due to libgtk2 version problem. I've libgtk2 has been updated and iceweasel in some unofficial feed.
http://groups.google.com/group/linux.debia...4f87cdfd206636e (http://groups.google.com/group/linux.debian.ports.arm/browse_thread/thread/44f87cdfd206636e)
OpenOffice!! I bump into a Fedora armel repo and find armel ports of OOos there!
http://ooo.services.openoffice.org/pub/Ope.../armeabiport01/ (http://ooo.services.openoffice.org/pub/OpenOffice.org/cws/upload/armeabiport01/)
These are armel ports compiled with the Java options. It's interesting to see how alien can manage that. It will be cool if we can finally manage to put OOo back in Debian/Zaurus again after the good old Pocketworkstation.
-
OpenOffice!! I bump into a Fedora armel repo and find armel ports of OOos there!
http://ooo.services.openoffice.org/pub/Ope.../armeabiport01/ (http://ooo.services.openoffice.org/pub/OpenOffice.org/cws/upload/armeabiport01/)
These are armel ports compiled with the Java options. It's interesting to see how alien can manage that. It will be cool if we can finally manage to put OOo back in Debian/Zaurus again after the good old Pocketworkstation.
i've yet tried this one some times ago :
https://www.oesf.org/forum/index.php?showtopic=24789 (https://www.oesf.org/forum/index.php?showtopic=24789)
i had no luck... but perhaps it will work now...
-
I came across several cool tips to reduce writing to the flash disks in several pointers for Eee PC. I am testing them in me Debian/Zaurus set up and they seem to work. Will include them also.
-
so what about the new rootfs ? i'd like to try before leaving the zaurus scene
-
My spectrum24 network-cf didn't work (missing firmware ?), so i configured usb-networking. Runnning apt-get update now
I am trying to get my usb networking going also, but am not as successful as you were. I have tried several different /etc/network/interfaces options but do not know what is going on. I have a usb network adapter and a basic home network. My network automatically hands out ip addresses so I thought
auto eth0
iface eth0 inet dhcp
would do it. No luck. How did you configure your usb-networking? Any pointers will be appreciated.
Edit: fixed. Newbie at the controls.
allow-hotplug eth0
iface ith0 inet dhcp is what worked in /etc/network/interfaces.
I also have to do echo -n 1 >/sys/bus/usb/devices/1-2/bConfigurationValue to get the usb-ethernet adapter working because debian thinks there is not enough power to run the device.
https://www.oesf.org/forum/index.php?showto...amp;hl=pegasus# (https://www.oesf.org/forum/index.php?showtopic=20739&hl=pegasus#)
I also had eth4 showing up for some reason when doing ifconfig -a. This seemed to be causing some MAC address collision or some such, according to this post over at ubuntu:
http://ubuntuforums.org/showthread.php?t=221768 (http://ubuntuforums.org/showthread.php?t=221768)
I just rebooted and things seemed to clear up.
-
so what about the new rootfs ? i'd like to try before leaving the zaurus scene
Just leave ... why do you care? Why is your effort about Z still high after you know at 100% that you sell it and don't use it ?
-
Is there anyway to make this a hdImage.tgz so I can actually use this image? I can't seem to unzip the bz2 stuff for some reason to the microdrive of my 3200.
-
Is there anyway to make this a hdImage.tgz so I can actually use this image? I can't seem to unzip the bz2 stuff for some reason to the microdrive of my 3200.
What is the error? Are you trying to extract the tarball in uboot emergency mode?
The easier way is to use the Sharp bootloader kernels (new yonggun/poky/angstrom).
Or try extracting the rootfs tarball in another ROM/on another computer first.
-
regarding the updated install desription - i'm still in the cacko 1.23 "user group" but very eager to try debian.
i don't understand:
-- added yonggun kernel modules: 2.6.22.9, 2.6.23.1, 2.6.23.3 (so users just need to flash the kernel)
....
I. Get the latest yonggun kernel here: http://yonggun.tistory.com/66 (http://yonggun.tistory.com/66)
how these two fit together, because the newest kernel is a 2.6.24 kernel whose modules are not included.
Do i understand something wong?
George
-
regarding the updated install desription - i'm still in the cacko 1.23 "user group" but very eager to try debian.
i don't understand:
-- added yonggun kernel modules: 2.6.22.9, 2.6.23.1, 2.6.23.3 (so users just need to flash the kernel)
....
I. Get the latest yonggun kernel here: http://yonggun.tistory.com/66 (http://yonggun.tistory.com/66)
how these two fit together, because the newest kernel is a 2.6.24 kernel whose modules are not included.
Do i understand something wong?
George
It's because the rootfs tarball was made before the release of the yonggun kernel 2.6.24. That's why i added in Step 2 to extract the 2.6.24 kernel modules on top of the rootfs tarball.
The 2.6.24 kernel modules will be included in the next rootfs tarball, while I would keep the old ones there for those who want to experiment things.
There is a list of things to fine-tune/improve/modify in the current rootfs. But it is still working.
-
Hi,
I have just installed Yonggun's kernel.img.2.6.24-yonggun-hda1 kernel and modules-2.6.24-yonggun.tar.gz modules....aswell as Zdevil's zaurus_eabi_rootfs_071201.tar.bz2 Rootfs onto my internal CF drive.
Everything seems to work ok so far, I just have a couple of questions:
1) Should such things as my SD card "automount" when inserted? With my installation, when I insert an SD card it is not automatically detected. So, is this normal for this installation?
2) Also, a few applications, such as XMMS and Dillio, etc are nowhere to be seen in the program launcher/taskbar. I have to execute these apps from the command line instead. Infact, very few apps (system file browsers, word processors, etc) are visible from the application launch menu - which is surprising considering how large the Rootfs is. I guess they are all hidden somewhere on the drive? Again, is this standard for this distro?
3) There are no app links setup on the desktop, no access to any GUI Filebrowser on desktop (or even in the app launcher) , etc unlike distros such as PdaXRom where everything is immediately accessible...
So, I just want to know if these things are normal or have I screwed up the installation somehow causing the configuration of these things not to work properly...
thanks for any info...
-
Is there anyway to make this a hdImage.tgz so I can actually use this image? I can't seem to unzip the bz2 stuff for some reason to the microdrive of my 3200.
What is the error? Are you trying to extract the tarball in uboot emergency mode?
The easier way is to use the Sharp bootloader kernels (new yonggun/poky/angstrom).
Or try extracting the rootfs tarball in another ROM/on another computer first.
ZDevil,
I had the same problem today trying to decompress the bz2 file. When you boot in the emergency mode on the 3200 and login and then try to decompress with the 'tar -xjpf' option it returns an error message saying that it can't use or find the 'j' option...in other words, it doesn't have any bz2 decompress capability available.
Don't understand what you mean exactly by 'use the Sharp bootloader kernels (new yonggun/poky/angstrom)' ?
In the meantime, I am going to try and extract tarball on another computer now....
UPDATE: Here's what I'm getting so far during extraction...do these messages mean that I have problems with extracted files???
marks$ tar -xjpf zaurus.tar.bz2
tar: ./bin/uncompress: Cannot hard link to `./bin/gunzip': Operation not supported
tar: ./bin/bunzip2: Cannot hard link to `./bin/bzip2': Operation not supported
tar: ./bin/bzcat: Cannot hard link to `./bin/bzip2': Operation not supported
tar: ./bin/bzegrep: Cannot hard link to `./bin/bzgrep': Operation not supported
tar: ./bin/bzfgrep: Cannot hard link to `./bin/bzgrep': Operation not supported
tar: ./bin/bzless: Cannot hard link to `./bin/bzmore': Operation not supported
tar: ./bin/bzcmp: Cannot hard link to `./bin/bzdiff': Operation not supported
tar: ./dev/tty4: Cannot mknod: Operation not permitted
tar: ./dev/sndstat: Cannot mknod: Operation not permitted
tar: ./dev/ram14: Cannot mknod: Operation not permitted
tar: ./dev/loop5: Cannot mknod: Operation not permitted
tar: ./dev/tty2: Cannot mknod: Operation not permitted
tar: ./dev/full: Cannot mknod: Operation not permitted
tar: ./dev/midi02: Cannot mknod: Operation not permitted
tar: ./dev/audio1: Cannot mknod: Operation not permitted
tar: ./dev/ram7: Cannot mknod: Operation not permitted
tar: ./dev/ram6: Cannot mknod: Operation not permitted
tar: ./dev/rmidi0: Cannot mknod: Operation not permitted
tar: ./dev/smpte2: Cannot mknod: Operation not permitted
tar: ./dev/dsp2: Cannot mknod: Operation not permitted
tar: ./dev/urandom: Cannot mknod: Operation not permitted
tar: ./dev/midi00: Cannot mknod: Operation not permitted
tar: ./dev/mixer3: Cannot mknod: Operation not permitted
tar: ./dev/initctl: Cannot mkfifo: Invalid argument
tar: ./dev/midi01: Cannot mknod: Operation not permitted
tar: ./dev/tty9: Cannot mknod: Operation not permitted
tar: ./dev/agpgart: Cannot mknod: Operation not permitted
tar: ./dev/ram5: Cannot mknod: Operation not permitted
tar: ./dev/ptmx: Cannot mknod: Operation not permitted
tar: ./dev/midi03: Cannot mknod: Operation not permitted
tar: ./dev/ram16: Cannot mknod: Operation not permitted
tar: ./dev/dsp3: Cannot mknod: Operation not permitted
tar: ./dev/ram3: Cannot mknod: Operation not permitted
tar: ./dev/midi0: Cannot mknod: Operation not permitted
tar: ./dev/audio: Cannot mknod: Operation not permitted
tar: ./dev/mixer: Cannot mknod: Operation not permitted
tar: ./dev/tty8: Cannot mknod: Operation not permitted
tar: ./dev/ram8: Cannot mknod: Operation not permitted
tar: ./dev/loop3: Cannot mknod: Operation not permitted
tar: ./dev/audio3: Cannot mknod: Operation not permitted
tar: ./dev/ram0: Cannot mknod: Operation not permitted
tar: ./dev/ram15: Cannot mknod: Operation not permitted
tar: ./dev/loop0: Cannot mknod: Operation not permitted
tar: ./dev/mixer1: Cannot mknod: Operation not permitted
tar: ./dev/loop4: Cannot mknod: Operation not permitted
tar: ./dev/ram4: Cannot mknod: Operation not permitted
tar: ./dev/zero: Cannot mknod: Operation not permitted
tar: ./dev/mem: Cannot mknod: Operation not permitted
tar: ./dev/sequencer: Cannot mknod: Operation not permitted
tar: ./dev/smpte3: Cannot mknod: Operation not permitted
tar: ./dev/loop2: Cannot mknod: Operation not permitted
tar: ./dev/midi1: Cannot mknod: Operation not permitted
tar: ./dev/mpu401stat: Cannot mknod: Operation not permitted
tar: ./dev/tty1: Cannot mknod: Operation not permitted
tar: ./dev/ram12: Cannot mknod: Operation not permitted
tar: ./dev/tty: Cannot mknod: Operation not permitted
tar: ./dev/tty5: Cannot mknod: Operation not permitted
tar: ./dev/smpte0: Cannot mknod: Operation not permitted
tar: ./dev/tty6: Cannot mknod: Operation not permitted
tar: ./dev/mixer2: Cannot mknod: Operation not permitted
tar: ./dev/midi3: Cannot mknod: Operation not permitted
tar: ./dev/audioctl: Cannot mknod: Operation not permitted
tar: ./dev/ram1: Cannot mknod: Operation not permitted
tar: ./dev/null: Cannot mknod: Operation not permitted
tar: ./dev/loop6: Cannot mknod: Operation not permitted
tar: ./dev/dsp: Cannot mknod: Operation not permitted
tar: ./dev/ram11: Cannot mknod: Operation not permitted
tar: ./dev/loop7: Cannot mknod: Operation not permitted
tar: ./dev/rmidi1: Cannot mknod: Operation not permitted
tar: ./dev/console: Cannot mknod: Operation not permitted
tar: ./dev/dsp1: Cannot mknod: Operation not permitted
tar: ./dev/midi2: Cannot mknod: Operation not permitted
tar: ./dev/rmidi3: Cannot mknod: Operation not permitted
tar: ./dev/ram13: Cannot mknod: Operation not permitted
tar: ./dev/rmidi2: Cannot mknod: Operation not permitted
tar: ./dev/tty3: Cannot mknod: Operation not permitted
tar: ./dev/ram10: Cannot mknod: Operation not permitted
tar: ./dev/port: Cannot mknod: Operation not permitted
tar: ./dev/mpu401data: Cannot mknod: Operation not permitted
tar: ./dev/random: Cannot mknod: Operation not permitted
tar: ./dev/smpte1: Cannot mknod: Operation not permitted
tar: ./dev/audio2: Cannot mknod: Operation not permitted
tar: ./dev/kmem: Cannot mknod: Operation not permitted
tar: ./dev/ram2: Cannot mknod: Operation not permitted
tar: ./dev/loop1: Cannot mknod: Operation not permitted
tar: ./dev/ram9: Cannot mknod: Operation not permitted
tar: ./dev/tty7: Cannot mknod: Operation not permitted
tar: ./dev/tty0: Cannot mknod: Operation not permitted
tar: ./dev/xconsole: Cannot mkfifo: Invalid argument
tar: ./dev/fb0: Cannot mknod: Operation not permitted
tar: ./dev/fb1: Cannot mknod: Operation not permitted
tar: ./dev/fb2: Cannot mknod: Operation not permitted
tar: ./dev/fb3: Cannot mknod: Operation not permitted
tar: ./dev/fb4: Cannot mknod: Operation not permitted
tar: ./dev/fb5: Cannot mknod: Operation not permitted
tar: ./dev/fb6: Cannot mknod: Operation not permitted
tar: ./dev/fb7: Cannot mknod: Operation not permitted
tar: ./dev/apm_bios: Cannot mknod: Operation not permitted
tar: ./etc/X11/kb/Corgi/xmodmap_ru-phonetic: implausibly old time stamp 1969-12-31 16:00:00
tar: ./etc/X11/kb/Corgi/xmodmap_ru: implausibly old time stamp 1969-12-31 16:00:00
tar: ./etc/X11/kb/Corgi/xmodmap_us: implausibly old time stamp 1969-12-31 16:00:00
tar: ./etc/X11/kb/Tosa/xmodmap_ru: implausibly old time stamp 1969-12-31 16:00:00
tar: ./etc/X11/kb/Tosa/xmodmap_us: implausibly old time stamp 1969-12-31 16:00:00
tar: ./etc/X11/kb/Akita/xmodmap_ru-phonetic: implausibly old time stamp 1969-12-31 16:00:00
tar: ./etc/X11/kb/Akita/xmodmap_ru: implausibly old time stamp 1969-12-31 16:00:00
-
@ArchiMark:
Make sure that the partition containing the debian rootfs (the destination of the extraction) is ext3 formatted.
Extract the rootfs tarball in another environment/on another computer (as long as it can read ext3). The emergency linux mode has rather limited functionality.
Also do not extract the rootfs to a FAT partition.
By kernels using the Sharp bootloader, I mean those kernels which are flashed and boot using the original bootloader used in Sharp rom, Cacko rom, pdaXrom (up to beta3), pdaXii, and Angstrom, i.e. no uboot is installed.
-
I have just installed Yonggun's kernel.img.2.6.24-yonggun-hda1 kernel and modules-2.6.24-yonggun.tar.gz modules....aswell as Zdevil's zaurus_eabi_rootfs_071201.tar.bz2 Rootfs onto my internal CF drive.
Everything seems to work ok so far,
Glad to know that. I know Debian works because it just works on both my C860 and C3200.
So, I just want to know if these things are normal or have I screwed up the installation somehow causing the configuration of these things not to work properly...
It seems to me that there is nothing messed up in your setup. What you need is just a few configurations. Here they are:
1) Should such things as my SD card "automount" when inserted? With my installation, when I insert an SD card it is not automatically detected. So, is this normal for this installation?
Just edit /etc/fstab and add an entry for mount SD card (/dev/mmcblk0p1) there.
2) Also, a few applications, such as XMMS and Dillio, etc are nowhere to be seen in the program launcher/taskbar. I have to execute these apps from the command line instead. Infact, very few apps (system file browsers, word processors, etc) are visible from the application launch menu - which is surprising considering how large the Rootfs is. I guess they are all hidden somewhere on the drive? Again, is this standard for this distro?
There are several solutions:
Solution 1: Edit "~/.icewm/menu" and add whatever you want there. A tedious way.
Solution 2: Install a program call "menu" and run "update-menus". But the result is less than satifactory as there are too many trivial folders and categories and duplicated entries.
Solution 3: Add "icewm-menu-gnome2" in ~/.icewm/menu . See http://myy.helia.fi/~karte/icewm.html (http://myy.helia.fi/~karte/icewm.html) for instructions. But again I find the menu generated not so optimal.
Solution 4: Use my "icewm_meuntools" scriptpack which is a slightly modified and enhanced version of Meanie's tools in pdaXii. Extract the tarball somewhere, then enter the extracted "icewm_menutools_0.1a" directory, and run "./install.sh" there. A few seconds later you'll see a new IceWM menu with program icons sorted, plus a few goodies. [ You are not allowed to view attachments ]
3) There are no app links setup on the desktop, no access to any GUI Filebrowser on desktop (or even in the app launcher) , etc unlike distros such as PdaXRom where everything is immediately accessible...
Again there are several solutions.
Solution 1: Use Rox desktop, just like in pdaXrom. See here for how to install it: http://www.debianhelp.co.uk/rox.htm (http://www.debianhelp.co.uk/rox.htm)
Solution 2: Use idesk, and it requires manual editing of the config file. Just "apt-get install idesk". See http://idesk.sourceforge.net/wiki/index.php/Main_Page (http://idesk.sourceforge.net/wiki/index.php/Main_Page) for how to set up the menu file.
Solution 3: Use pcmanfm. Go to "Preferences > Desktop", then check "Show file icons on desktop". You can also set desktop wallpaper there (but I find it slower than using icewmbg directly). Also add "pcmanfm &" in .xinitrc, above the line "icewm-session".
-
Just to add:
From now on before the release of the next OOTB rootfs tarball, I will put things up in scriptpacks and post them in the forum every now and then.
The scriptpacks address (some of) the issues and questions raised about the present rootfs tarball and provide easy, and ideally painless and foolproof, solutions.
I think doing things this way is both easier for me to refine the present rootfs tarball, and existing users can keep track of the changes and choose to install them to their liking.
How's that?
-
@ArchiMark:
Make sure that the partition containing the debian rootfs (the destination of the extraction) is ext3 formatted.
Extract the rootfs tarball in another environment/on another computer (as long as it can read ext3). The emergency linux mode has rather limited functionality.
Also do not extract the rootfs to a FAT partition.
By kernels using the Sharp bootloader, I mean those kernels which are flashed and boot using the original bootloader used in Sharp rom, Cacko rom, pdaXrom (up to beta3), pdaXii, and Angstrom, i.e. no uboot is installed.
As always, thanks for your help, ZDevil!
I just realized that I had wanted to extract the files to my internal drive on my 3200, which I formatted as instructed. However, when I had problem extracting it there from the SD card since the emergency mode doesn't seem to be able to handle bz2 files, only tar files, I decided to try extracting it onto the card itself using my MacBook/OSX/Terminal....but you're right the SD card is formatted as FAT...hence the problems....
So, I guess I need to reformat the SD card and re-copy the files to it....
Would be a lot simpler if we could either
A) get the rootfs tarball in tar format, not bz2 so that we could extract it from card over to internal drive on 3200 or
find out if there is a way to extract bz2 file on card over to internal drive....
Thanks.
-
That would be nice!
This way it should be easier to concentrate on one problem - solve it - and move on.
Although a bit busy at the moment, i'll be definitely be a testing those scriptpacks.
Just to add:
From now on before the release of the next OOTB rootfs tarball, I will put things up in scriptpacks and post them in the forum every now and then.
...
How's that?
-
Thanks! I will also add uninstaller scripts if possible, so users can revert the changes and start afresh.
I am doing the scriptpacks just because I am too lazy to make a .deb.
Perhaps I can put all the scriptpacks together into an new customization .deb (with choices for installing /not installing each module) and post it together with the new customized rootfs.
Things already done:
1) IceWM: new root program menu, desktop icons, wallpaper rotation (landscape <-> portrait), more menu entries (rotation, screenshot, gksu, etc.)
2) Mouse right click: new slim and fast binaries
3) Dualboot option/desktop icon (may add it to IceWM menu)
4) New packages (browser, localepurge, ssh, etc)
5) gnome-helper (yelp) hack: to fix key [1] acting as F1 in gnome apps and calling gnome-helper in e.g. gnumeric and epiphany
6) New working repos list in /etc/apt/sources.list
7) Some tweaks to reduce disk writing (esp. for flash disks)
Things to be done (any help greatly appreciated!):
1) bluetooth
2) WPA support
3) automount SD/CF cards
-
...
Things already done:
...
7) Some tweaks to reduce disk writing (esp. for flash disks)
That's really great for my C1k, maybe it speeds things up a little bit?
...
Things to be done (any help greatly appreciated!):
1) bluetooth
2) WPA support
3) automount SD/CF cards
So what's up with bluetooth? Some time ago i managed to sync my phone contacts with opensync and evolution by bluetooth. Maybe i can help?
-
So what's up with bluetooth? Some time ago i managed to sync my phone contacts with opensync and evolution by bluetooth. Maybe i can help?
Yes, Please do! Thanks!
-
Want to install Debian to my internal 6GB microdrive.
So, extracted the tarball and modules file on ext3 formatted SD card using another computer. Seems OK,
Meanwhile, using fdisk I formatted my internal microdrive into 2 partitions; large ext3 partition (hda1) and a small swap partition (hda2).
Then I formatted the hda1 partition using mke2fs -j /dev/hda1.
So, now some questions please to help minimize my making mistakes....
1) I did not format the hda2 swap partition. Do I need to do that?
2) Do I need to mount the hda1/hda2 partitions before doing rest of install? If so, mount to?
Instructions say Step 3:
Edit /etc/fstab and assign the correct mount points to the root partition and the swap partition.
The default is set to boot from the internal drive.
Just comment the lines for /dev/hdaX and uncomment the lines for /dev/mmcblk0pX if debian boots from SD.
However, I just looked at '/etc/fstab' file on SD card. It shows the following:
/dev/root / ext2 rw 1 1
/dev/shm /tmp tmpfs size=16M 1 2
/dev/mmcda1 /mnt/card auto noauto, owner 0 0
none /dev/pts devpts gid=5, modes=620 0 0
So, this is a lot different than what instructions say I'll see.....
3) Then do I need to copy over all the files from SD card to internal microdrive in order to install it there or can I install to microdrive using files on SD? What would be best way?
Appreciate all the help!
Thanks,
Mark
-
A suggestion of partitioning of microdrive: /dev/hda1 = swap partition; /dev/hda2 = root partition
Reason: swap partition is accessed faster in the beginning of the drive. (Flash disks don't seem to have this issue)
Swap partition: you have to do: mkswap /dev/hda1 (or /dev/hda2)
"Installation" of the rootfs: Just copy/extract the content of the rootfs tarball to the microdrive partition. Make sure that your kernel matches the partition. So if your rootfs is in /dev/hda1, use yonggun's hda1 kernel, and so is hda2.
-
About /etc/fstab:
Change to this:
/dev/hda2 / ext3 rw,noatime 1 1
/dev/hda1 /none swap sw 0 0
proc /proc proc defaults 0 0
sysfs /sys sysfs defaults 0 0
/dev/shm /tmp tmpfs size=16M 1 2
-
A suggestion of partitioning of microdrive: /dev/hda1 = swap partition; /dev/hda2 = root partition
Reason: swap partition is accessed faster in the beginning of the drive. (Flash disks don't seem to have this issue)
Swap partition: you have to do: mkswap /dev/hda1 (or /dev/hda2)
"Installation" of the rootfs: Just copy/extract the content of the rootfs tarball to the microdrive partition. Make sure that your kernel matches the partition. So if your rootfs is in /dev/hda1, use yonggun's hda1 kernel, and so is hda2.
Thanks, ZDevil!
OK, re-did my microdrive as you suggested.... think I'm getting close now to getting install working...
Edited fstab as you suggested... (however, when I reopened fstab in vi, this time it looked like the way it was mentioned in instructions, so was OK, weird....).
When I tried to copy over tarball contents to microdrive, it couldn't do it, said there wasn't room....
Now, I realize that I made a mistake by trying to mount hda2 per criteria in fstab, as it was trying to copy file to root, but this is the small area in the linux emergency mode....
So, now, I need to mount hda2 properly to copy files over to it using cp -R /mnt/card /mnt/???
What is the right mount point for hda2 in order to be able to copy the files over it?
Do I need to do the same thing now for hda1, the swap partition??
Also, regarding yonggun's kernel, I downloaded the hda1 one and renamed it to zImage.bin as in instructions...so, the only other Sharp bootloader files on that webpage were for mmcdablk0p1 and mmcdablk0p2....none for hda2. So, is this OK to use zImage.bin that I downloaded?
Thanks,
-
Follow the first post of this thread: https://www.oesf.org/forum/index.php?showtopic=25029 (https://www.oesf.org/forum/index.php?showtopic=25029)
I made a mistake about the /etc/fstab. You don't need the line for tmpfs (i.e. the last line in my previous post).
-
Follow the first post of this thread: https://www.oesf.org/forum/index.php?showtopic=25029 (https://www.oesf.org/forum/index.php?showtopic=25029)
I'm trying, ZDevil, I'm trying....
But, don't I need to copy over the files from my SD card to my internal drive before flashing kernel?? Or are saying that I can install those files from SD card to microdrive during install or later??
If I need to copy files over now, then I need to know what is the proper way to mount hda2 from commant prompt...
I made a mistake about the /etc/fstab. You don't need the line for tmpfs (i.e. the last line in my previous post).
Oh, OK....
Thanks!
-
mkdir /mnt/cf (the name of the mount point doesn't matter)
mount -t ext3 /dev/hda2 /mnt/cf
Ready your root partition and rootfs before flashing the kernel.
-
mkdir /mnt/cf (the name of the mount point doesn't matter)
mount -t ext3 /dev/hda2 /mnt/cf
Right, ZDevil....mounted OK now....
Ready your root partition and rootfs before flashing the kernel.
If by this you mean copy over the extracted tarball and module files to root level of hda2 then that is what I'm doing now on Z..... given the size of the files, I think it will take some time to complete copying all the files...
THANKS again for all your help. Truly appreciate it!!!!
UPDATE: Just tried flashing and it says error message something like :
Zaurus SD Updater
/mnt/card/$UPDATE_SHELL not found
So, something is not ready yet....
-
Which partition is your rootfs in?
If it is /dev/hda1, flash the hda1 yonggun kernel.
If it is /dev/hda2, flash the hda2 yonggun kernel.
Use the "updater.sh" attached in the first post.
-
Which partition is your rootfs in?
If it is /dev/hda1, flash the hda1 yonggun kernel.
If it is /dev/hda2, flash the hda2 yonggun kernel.
Use the "updater.sh" attached in the first post.
rootfs is in hda2....but as mentioned before, I didn't see any hda2 kernel on yonggun's webpage, just hda1 and the others for mmcda....
-
My bad!! Sorry for that. Here it is: http://yonggun.tistory.com/68 (http://yonggun.tistory.com/68)
Now go correct the post ...
-
My bad!! Sorry for that. Here it is: http://yonggun.tistory.com/68 (http://yonggun.tistory.com/68)
Now go correct the post ...
OK, no problem.... except how do I now find and erase all the modules that I already extracted so that I can replace them with the ones from the new link???
-
You don't have to remove the older kernel modules. Just leave them there and extract the new one to the root of the root partition.
-
You don't have to remove the older kernel modules. Just leave them there and extract the new one to the root of the root partition.
OK, ZDevil, I'VE GOT DEBIAN EABI ! ! ! ! !
So far, working great, IceWM, ethernet card, apt-get..........
Could not have done it without, ZDevil !!!!
T H A N K Y O U ! ! ! ! ! [/b]
-
You don't have to remove the older kernel modules. Just leave them there and extract the new one to the root of the root partition.
OK, ZDevil, I'VE GOT DEBIAN EABI ! ! ! ! !
So far, working great, IceWM, ethernet card, apt-get..........
Could not have done it without, ZDevil !!!!
T H A N K Y O U ! ! ! ! ! [/b]
is blutooth/gprs working? tell us more
-
OK, ZDevil, I'VE GOT DEBIAN EABI ! ! ! ! !
So far, working great, IceWM, ethernet card, apt-get..........
is blutooth/gprs working? tell us more
Dunno yet....
I have an old Belkin USB BT adaptor F8T001 that maybe I can try out....
-
Congrats.
Try this scriptpack here: https://www.oesf.org/forum/index.php?showtopic=25644 (https://www.oesf.org/forum/index.php?showtopic=25644) , which gives a more user-friendly IceWM setting.
-
Congrats.
Try this scriptpack here: https://www.oesf.org/forum/index.php?showtopic=25644 (https://www.oesf.org/forum/index.php?showtopic=25644) , which gives a more user-friendly IceWM setting.
Thanks, ZDevil....will try that scriptpack...
Meanwhile, I installed something that said I should reboot....when I do that I get various error messages mainly about my HD; hda2.....
Checking root file system...fsck 1.40.8 (13-Mar-2008)
/dev/hda2 - contains a file system with errors. Check forced.
/dev/hda2: Inodes part of a corrupted orphan linked list found.
/dev/hda2: UNEXPECTED INCONSISTENCY: RUN fsck MANUALLY.
(ie. without -a or -p options)
fsck died with exit status 4
[color="#FF0000"]failed (code 4)[/color]
[color="#FF0000"]**[/color]An automatic file system check (fsck) of the root file system failed.
A manual fsck must be performed, then the system restarted.
The fsck should be performed in maintenance mode with the root filesystem mounted in read-only mode.
[color="#FF0000"]*[/color] The root filesystem is currently mounted in read-only mode.
A maintenance shell will now be started.
After performing system maintenance, press CONTROL-D
to terminate the maintenance shell and restart the system.
Give root password for maintenance (or type CONTROL-D to continue):
Zaurus: ~#
One other problem now is that I can't type a '/' character in this shell, when I press that key it gives a ',' or a '<' if I use the shift key...
What fsck command should I use or do something else ???
and how can I type a '/' so that I can type '/dev/hda2' ????
Funny thing is that debian/IceWM seemed to be working OK...
Thanks,
Mark
-
Easy:
In this case you SHOULD run a fsck to check and repair the filesystem.
To do so, reboot and you would get to that warning message again.
Just type your root password (default is "zaurus" if you haven't changed it).
Then run fsck .. [tab] dev/hda1 -- just run fsck without any option (as already suggested by the warning message -- a very cool and helpful feature in such a complete system)
(The [tab] after ".." does the trick to get you the "/" )
When prompted to repair something, just answer yes by "y" then [Enter].
This way fsck will get through and things will be fine again.
-
Easy:
In this case you SHOULD run a fsck to check and repair the filesystem.
To do so, reboot and you would get to that warning message again.
Just type your root password (default is "zaurus" if you haven't changed it).
Then run fsck .. [tab] dev/hda1 -- just run fsck without any option (as already suggested by the warning message -- a very cool and helpful feature in such a complete system)
(The [tab] after ".." does the trick to get you the "/" )
When prompted to repair something, just answer yes by "y" then [Enter].
This way fsck will get through and things will be fine again.
Thanks, ZDevil !
Your suggestions worked and fsck is fscking away right now....says FIXED after each error found.....
-
Many thanks for this, all working (but needing a ton more tweaking) here. Took about an hour and a half to untar the root on the Zaurus
My notes are here - http://rogalian.blogspot.com/2008/03/debia...s-sl-c3200.html (http://rogalian.blogspot.com/2008/03/debian-eabi-on-sharp-zaurus-sl-c3200.html)
Out of interest, why the huge amount of (non-debian) sources list?
-
Hi Guys,
Please help, I have 2 partitions, hda1 ext3 and hda2 swap. I untared the rootfs to hda1, flashed the kernel and get :
kernel panic, no init found.
Please help
-
Which kernel are you using? Which model is your Z?
-
Which kernel are you using? Which model is your Z?
Hi ZDevil, Thanks for the quick reply.
I have a 3200 and am using kernel 2.6.24.3, hda1, what does this error mean and how can I solve it? Thanks
Oh, the error message also say :
Try passing init= option to kernel
-
Checking root file system...fsck 1.40.8 (13-Mar-2008)
/dev/hda2 - contains a file system with errors. Check forced.
/dev/hda2: Inodes part of a corrupted orphan linked list found.
I've been getting exactly the same inode errors quite regularly, even when shutting down the machine properly. Not sure what's causing them yet.
In other news, I've had some success with the rather more mature evtouch driver in Debian - http://rogalian.blogspot.com/2008/04/zauru...erver-xorg.html (http://rogalian.blogspot.com/2008/04/zaurus-touch-screen-with-xserver-xorg.html) Rotating works with the touch screen driver, but sadly it breaks the framebuffer driver...
The same bug previously occured with a non-rotated screen and evtouch, so I'm hoping the fix is similar.
-
Which kernel are you using? Which model is your Z?
Hi ZDevil, Thanks for the quick reply.
I have a 3200 and am using kernel 2.6.24.3, hda1, what does this error mean and how can I solve it? Thanks
Oh, the error message also say :
Try passing init= option to kernel
Can you check/try these:
-- Which brand/model is your internal disk?
-- The kernel download is intact.
-- The updater.sh is the correct one (use the one attached in the first post)
-- The root partition is ext3 formatted. (you have to do "mkfs.ext3 /dev/hda1" after using fdisk).
-
Which kernel are you using? Which model is your Z?
Hi ZDevil, Thanks for the quick reply.
I have a 3200 and am using kernel 2.6.24.3, hda1, what does this error mean and how can I solve it? Thanks
Oh, the error message also say :
Try passing init= option to kernel
Can you check/try these:
-- Which brand/model is your internal disk?
-- The kernel download is intact.
-- The updater.sh is the correct one (use the one attached in the first post)
-- The root partition is ext3 formatted. (you have to do "mkfs.ext3 /dev/hda1" after using fdisk).
Hi ZDevil,
-- Hitachi 6GB HD
-- The kernel is fine, downloaded from http://yonggun.tistory.com/68 (http://yonggun.tistory.com/68)
-- update.sh comes from the 1st post
-- I did the following to the root partition : mke2fs -j /dev/hda1
I untared the rootfs on the SD then copied it across to the hda1.
Thanks for the support....
-
How did you uncompress the rootfs tarball?
Can you find /etc/inittab in the rootfs (you can check in the linux emergency cnosole)?
-
How did you uncompress the rootfs tarball?
Can you find /etc/inittab in the rootfs (you can check in the linux emergency cnosole)?
Hi ZDevil,
I have that file. I unzipped it then coppied it accross to hda1. I experienced problems trying to execute tar -xjpf zaurus_eabi_rootfs_071201.tar.bz2 -C to hda1
-
How did you uncompress the rootfs tarball?
Can you find /etc/inittab in the rootfs (you can check in the linux emergency cnosole)?
I have repartitioned the hard drive. Swap in the first partition and rootfs in the second.
When I execute tar -xjpf zaurus_eabi_rootfs_071201.tar.bz2 -C /mnt/ide2
I get the following error :
tar : bzip2: Cannot exec: No such file or directory
tar : Child returned status 2
-
How did you uncompress the rootfs tarball?
Can you find /etc/inittab in the rootfs (you can check in the linux emergency cnosole)?
I have repartitioned the hard drive. Swap in the first partition and rootfs in the second.
When I execute tar -xjpf zaurus_eabi_rootfs_071201.tar.bz2 -C /mnt/ide2
I get the following error :
tar : bzip2: Cannot exec: No such file or directory
tar : Child returned status 2
In your case you may try extracting the tarball before flashing the kernel. Don't do it in the linux emergency console as it doesn't seem to be able to handle the uncompression. The best way is to do it on another *nix box.
And now since you have your rootfs on hda2, you should use the yonggun kernel for hda2, not the one for hda1.
-
In your case you may try extracting the tarball before flashing the kernel. Don't do it in the linux emergency console as it doesn't seem to be able to handle the uncompression. The best way is to do it on another *nix box.
I copy tar et bzip from a Cacko rootfs (same kernel as emergency boot 2.4.20), to my sdcard, boot in emergency mode, then unlink tar et use bin from my sD.
It works.
-
In your case you may try extracting the tarball before flashing the kernel. Don't do it in the linux emergency console as it doesn't seem to be able to handle the uncompression. The best way is to do it on another *nix box.
I copy tar et bzip from a Cacko rootfs (same kernel as emergency boot 2.4.20), to my sdcard, boot in emergency mode, then unlink tar et use bin from my sD.
It works.
Or... you can "unbzip" the tar in any machine (*nix or windows with, for example, 7z) and just run "-xpf" command instead of "-xjpf"
P.D.: just waiting the new tarball to see all the improvements ZDevil has done.
-
In your case you may try extracting the tarball before flashing the kernel. Don't do it in the linux emergency console as it doesn't seem to be able to handle the uncompression. The best way is to do it on another *nix box.
I copy tar et bzip from a Cacko rootfs (same kernel as emergency boot 2.4.20), to my sdcard, boot in emergency mode, then unlink tar et use bin from my sD.
It works.
Or... you can "unbzip" the tar in any machine (*nix or windows with, for example, 7z) and just run "-xpf" command instead of "-xjpf"
P.D.: just waiting the new tarball to see all the improvements ZDevil has done.
I got it working yesterday. Thanks for all the support. I'm impressed it's fast Well done All.
One more question, I don't have a wifi card, how can I setup usb networking? Thanks.
-
I got it working yesterday. Thanks for all the support. I'm impressed it's fast Well done All.
One more question, I don't have a wifi card, how can I setup usb networking? Thanks.
Here it is: Usb Networking Post (https://www.oesf.org/forum/index.php?showtopic=25025)
-
hi zdevil and all,
i just installed debian on my z - and it's just wonderful! thanks so much for this post and tarball.
-
In watching my system boot, I noticed the following line & wondered if it was something to be concerned about:
Mount failed or selinuxfs on /seliunx: no such file of directory
-
Fixed, had to comment out icewm and insert startxfce4
-
I didn't find this mentioned anywhere...
How big do you recomend to make the swap? 256M? 512M?
And, the root should be aroung 1 gig right? Should it be bigger?
-
Mines 256 MB Swap
Never found a solution too the /selinux message in boot.
-
Wow! After using angstrom for some months, I've just installed the tarball and yonggun 2.6.24 kernel on my sl-c3200 internal drive. I'd like to say thanks for this tremendous piece of work. It's great to have a comprehensive system with online man pages and all the command line tools. The install was straightforward.
In case anyone is thinking of doing the same, here is a summary of the only things that have bothered me at all:
(1) The "/," key when unshifted defaults to "," in the console but "/" in the x terminal.
(2) I can't find a way to type "`" (backquote) character on the keyboard. Some shell scripts need it.
(3) The g_file_storage module doesn't work, which means that the Z can't look like a mass storage device to a PC. DMA errors are reported and bus resets. It has never worked under angstrom either, so I think sadly the pxa2xx_udc driver has a long-term (permanent?) unfixed bug.
(4) The most serious thing is that the cursor is very jittery when the touchscreen is touched making it quite hard to hit the small buttons on the title bars. It is useable, but a bit frustrating and possibly a show-stopper. I know this has been discussed in the forums but I don't think any solution has been found and it seems not everyone experiences it. Inspecting the output of /dev/input/ts0 with no xserver indicates that the driver is producing noisy X coordinate values (even going below 0 if on left edge of screen). It looks as if the tslib filters are doing their best to smooth things out but the noise is too great. Angstrom works fine in this area. Could it be a subtle kernel timing issue, related to some other configuration issue (root device?
Rod
-
This looks awesome! Great work!
I've noticed a couple of strange problems though (btw. I'm using the yonggun kernel and on a 3200):
1./ there is some sort of extremely strict password policy enforced which means I have to change roots password on login.
2./ im getting an error about ifup and ifdown not being able to read /etc/network/interfaces.
Could this be an issue with the install?
-
1./ there is some sort of extremely strict password policy enforced which means I have to change roots password on login.
Creating a non root user has fixed this!
The network problem was a mistake in the syntax of /etc/network/interfaces.
-
Does this pack includes yonggun's rootfs goods?
Thnx
Tasos
-
Every time I try to use an interface, ethernet(I have a buffalo 10base card) or g_ether I get kernel panic and errors... ifup eth3 for example is an instant kernel panic....
I have 2.6.24.4 kernel I think... the latest
thnx
Tasos
-
Every time I try to use an interface, ethernet(I have a buffalo 10base card) or g_ether I get kernel panic and errors... ifup eth3 for example is an instant kernel panic....
I have 2.6.24.4 kernel I think... the latest
thnx
Tasos
I had the same issue! I switched back to kernel 2.6.24.3 and got USB networking and wlan working.
I think one of the modules is broken in 2.6.24.4. Probably g_ether.
Achim
-
I had the same issue! I switched back to kernel 2.6.24.3 and got USB networking and wlan working.
I think one of the modules is broken in 2.6.24.4. Probably g_ether.
Achim
Same deal for me with 2.6.24.4 I got errors about af_packet and symbols, also udev was renaming eth0 to eth3. Went to the 2.6.24.3 kernel and everything works great.
TronDD.
-
The links to download this tarball are broken.
Anyone got a working link?