Solved: SimplyMEPIS v6.0 boots five or six time OK, then hangs on "Setting up LVM"

I have the same on my dell desktop dimension 3000.
The problem set in at beta 3 - beta 2 is fine.
haven't tried final yet but not expecting any progress.
After removing matrox vga card and going back to onboard
intel chip, it will fail twice on the trot each day and the third (ish) reboot goes thru ok...

you can check my earlier posts but no-one has come back with anything constructive - where do these numbers come from - kernel event manager? - so wher do you look for the codes - can we turn it off? etc.

Googling for "17179594.796000" reveals a page : http://mozy.org/vaio/ - which relates the number with some usb / touchpad stuff - not really any help, but quite an interesting page - see piccys at bottom page.

Perhaps drop an email to - as i don´t think there will be many people here who have any clue as to what is causing the boot to hang...

very interesting pics...
my desktop doesn't have a touchpad and nothing on usb...
emailed details to warren at dev@mepis but the only help was to try this or that beta to narrow the problem down

I'm having the same problem. I know it's not my hardware as it works fine in any other distro, besides, i've recently replaced my HDD. The system hangs when it wants to set up LVM volume management. I have to reboot a few times until Mepis finally boot up properly..

I can't seem to find any info anywhere on any other forum as to why this is happening...

My laptop has been working OK for nearly two weeks, now. I just don't do upgrades. I do install applications that I need or desire.

I learned that a full upgrade will kill me on the next boot-up. Also, the automatix script did the same thing. Automatix looks really cool. The next boot-up my laptop was dead at LVM.

One of these days I will start a binary search process on the updrades, attempting to isolate the culprit package.

OK -- this has nothing to do with LVM volumes and the numbers displayed are meaningless

Problem loading ndiswrapper and recognizing Broadcom card

[It's possible you could boot by manually turning off wireless -- I didn't try this]

add
blacklist ndiswrapper
to /etc/modprobe.d/blacklist [edit as root -- if you can't boot edit the blacklist file from the live CD as root]

remove
blacklist bcm43xx
{if you added it to blacklist}

Now you can boot -- but you will not have wireless --
you need to install
bcm43xx-fwcutter
and set up the Broadcom native driver
You can do this with wired LAN or by downloading the .deb on another PC and copying it and installing with kpackage

http://www.ubuntuforums.org/showthread.php?t=185174&highlight=bcm43xx-fwcutter

Mike

Post to http://www.mepislovers.com -- great forum
Use the wiki -- you can search it
http://www.mepislovers-wiki.org/
Download the user guide (link at left)

Thanks Mike. Your earlier posts were very helpfull to me to setup ndiswrapper.

The way i understand LVM is the way linux allocates and takes care of hard drive space etc... It would appear that over time your linux space is getting fuzzed up. Probably your partition scheme somewhere is not making linux happy. At least thats the way im reading it so maybe redefining your partitions or changing them a bit will fix it.

jim

light9, have you set up your HP for dual-boot, or is it only booting Mepis? I don't know specifics on the HP, but I am wondering if when you partitioned the drive, did there happen to be any "hidden" or "unusual" partitions that perhaps you wiped out? Do you know whether the HP has a hidden "diagnostic" or "recovery" partition? Some of the laptops do have that (I think some Dells and Toshibas are like that). If so, and if you wiped out that partition, your BIOS might be interfering, trying to "fix" things.

The only other thing I can think of, like Jim, is that your partition layout and/or file system(s) might be a little odd. Can you post your /etc/fstab file as well as the output of "mount" when your system is running well?

Did you have a chance to test mdmarmer's suggestion and what was the result?

Jon

I reported this problem to

Mike

First of all, thanks to Jim, Jon and Mike for your replies.

I have been trying to isolate the problem by reducing the variables. This simple situation will cause the hang on "Setting up LVM".

(1) Install SM v6.0 on two partitions: root (including /home) on an ext3 partition, and swap. Let MEPIS installer format the partition. Use defaults where possible.

(2) Configure ndiswrapper with the bcmwl5.inf driver as in Mike's prior comments. This is necessary to get the wi-fi onto wlan0.

(3) Execute apt-get update and apt-get upgrade.

(4) Reboot.

I now know that it does not matter how much time passes or what else is done between thest steps. I did a test of JUST these four steps over a two hour period and the results were:

Hangs on "Setting up LVM".

My /home was on the root partition for the simple test, but it is normally on a seperate partition.

I have preserved the WinXP partition on hda1. Originally, there was ONLY one partition observable on the laptop by gPartEd or QtPartEd: hda1 covering the entire hard disk. I looked for hidden partition as I am well aware of that practice. After shrinking hda1 (Windows), I created an extended partition over the free disk. Created three Linux partitions and three data partitions.

I posted to on 7/25/06

Posted fstab and mount results follow. Thanks again,

Warren Light

/etc/fstab:

# Pluggable devices are handled by uDev, they are not in fstab
/dev/hda5 / ext3 defaults,noatime 1 1
/dev/hda7 swap swap sw,pri=1 0 0
none /proc proc defaults 0 0
none /proc/bus/usb usbfs devmode=0666 0 0
none /dev/pts devpts mode=0622 0 0
none /sys sysfs defaults 0 0
/dev/hda6 /home auto defaults,noatime 1 2
# Dynamic entries below, identified by 'users' option
/dev/hda1 /mnt/hda1 ntfs noauto,users,exec,ro,umask=0222 0 0
/dev/hda8 /mnt/hda8 ext3 noauto,users,exec 0 0
/dev/hda9 /mnt/hda9 vfat,ext3,ext2,reiserfs noauto,users,exec 0 0
/dev/hda10 /mnt/hda10 ext3 noauto,users,exec 0 0
/dev/cdrom /media/cdrom iso9660,udf noauto,users,exec,ro 0 0

mount results:

/dev/hda5 on / type ext3 (rw,noatime)
proc on /proc type proc (rw)
/sys on /sys type sysfs (rw)
varrun on /var/run type tmpfs (rw)
varlock on /var/lock type tmpfs (rw)
procbususb on /proc/bus/usb type usbfs (rw)
udev on /dev type tmpfs (rw)
devpts on /dev/pts type devpts (rw,gid=100,mode=0622)
devshm on /dev/shm type tmpfs (rw)
/dev/hda6 on /home type ext3 (rw,noatime)
capifs on /dev/capi type capifs (rw,mode=0666)
binfmt_misc on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)

My Hardware:

Product Name zv6201cl
US Product Number EC367UA#ABA
Microprocessor 2.0 GHz AMD Athlon™64 Processor 3200+
Microprocessor Cache 512KB L2 Cache
Memory 512MB 333MHz DDR System Memory (2 Dimm)
Video Graphics ATI RADEON® XPRESS 200M IGP
Video Memory 128MB DDR (dedicated)
Hard Drive 60GB (4200RPM) Hard Drive
Multimedia Drive DVD±R/RW and CD-RW Combo Drive with Double Layer Support
Display 15.4” WXGA High-Definition BrightView Widescreen (1280 x 800) Display
Network Card Integrated 10/100BASE-T Ethernet LAN (RJ-45 connector)
Broadcom BCM4318
Wireless Connectivity 54g™ 802.11b/g WLAN with 125HSM / SpeedBooster support
Sound Altec Lansing

Doesnt make a lot of sense so far but im wondering if you use grub for a boot loader? I suspect winders is playing with your drive somehow....

jim

Jim,

I did not boot or view Windows during the quick, two hour test described above. That is, from MEPIS install to failure without windows.

BTW, now that I have decided not to upgrade for a while, there has been no recurrance of the hang problem in over a week. In this period, I have booted Windows several times.

Yes, I use GRUB. See my menu.lst below. It is standard MEPIS install.

timeout 15
color cyan/blue white/blue
foreground ffffff
background 0639a1

gfxmenu /boot/grub/message

title MEPIS at hda5, kernel 2.6.15-26-386
root (hd0,4)
kernel /boot/vmlinuz-2.6.15-26-386 root=/dev/hda5 nomce quiet vga=791
boot

title Windows at hda1
rootnoverify (hd0,0)
chainloader +1

title MEMTEST
kernel /boot/memtest86+.bin

Huh well so mutch for that theory! Gonna have to think about this one.... Its so illogical , works till upgrade and then goes south. Cant think of a thing that would do that unless you have a bad disk thats muffin the same thing over and over.Very strange ....

jim

All,

Just to throw some of my observations into the works. Mine are less rigourous because I was unable to spend as much time with it, unlike light9's observatations.

I did this test on my Dell Inspiron 2600. I am currently running Mepis 3.3.1-1. I have run each version of 6.0 as it was released on a separate partition. So my "live" OS is on /dev/hda3, the "test" version is on /dev/hda2. I have a single swap partition on /dev/hda1, and "/home" (only used for "live") is mounted on /dev/hda6.

All versions of 6.0, up through RC3 worked just fine. However, with every version of Mepis, I have had to "tweak" the initial grub screen (both Live and installed) so that vga=773. The vga=791 is not liked by my system, whereas vga=773 sets a nice compressed, but readable text screen. When "splashy" came along with 3.4 it annoys me (I'll share more in a moment) so I normally remove it after installing each version of 6.0.

So, up to the present (yesterday). I have 6.0 rc3 loaded and it runs just fine. I started my system with 6.0 Release, as a Live CD. I set the grub line to vga=773, as usual, and let the system boot. When the "Starting splashy..." message came up, I pressed escape to return to a text boot screen. It booted through kdm, offered me the login screen, and I logged in as user root. KDE appeared to go through its boot sequence and froze on the folder icon on the boot screen. The kde boot screen has seven different icons IIRC while it's initializing things. It looks like it timed out during part of the file initialization. The screen cleared and I had to reboot my system. I could not simply restart X.

I was able to duplicate the above sequence about four times. Then, I rebooted and did one thing differently: I let "splashy" run. Normally, "splashy" does not render properly on my laptop; it's a big vertical blur with a too-wide status bar; perfectly worthless. That is why I kill it. This time, I let it run (and annoy me). When I did that, I was able to log into kde as root

I thought I had a clue as to why this thing did not work, so I rebooted from power off and tried the same sequence (letting "splashy" run), and then interrupting "splashy". Alas, I was unable to duplicate the failure

I will try this sequence of tests again today to see if I can prove/disprove my results.

So there is definitely something ever-so-slightly screwball about 6.0 Release that doesn't like some computers. I also don't know if this helps identify the problem better (I doubt it). Is there some aspect of xorg configuration/initialization that is being set very early on, that is then getting lost just as the system is being booted? Perspiring minds want to know!

Jon

Jon,

I saw something similar yesterday trying to boot a five year old Compac laptop with Mepis 6.0 live CD. He said it worked the night before, but I could not get past the setup icons after the login screen. As I think about it now, all our attempts yesterday (with various options) were following a restart.

Is restart == cold boot ? That scares me.

The suggestions I gave in my previous post were not guesses -- they actually worked to get a good bootup on a non-booting HP L2000 laptop (Turion 64 ML-37, 1GB RAM, Broadcom wireless builtin)

OK -- my HP L2000 would not reboot after I installed 6.0, set up ndiswrapper (with blacklist of bcm43xx) and installed quite a few packages with Automatix ... I had only installed, rebooted once, got wireless working and run Automatix ... then I shutdown and tried to reboot the second time and it hung at LVM

I tried several things from the live CD, first removing LVM startup and shutdown scripts -- this did not help -- boot failed with a different message -- something about RAID -- as I don't have RAID (or LVM volumes) I removed RAID startup and shutdown scripts (startup and shutdown scripts are in /etc/rcx.d where x is 0-6 or S for the different runlevels 1-5 {0,6 and S are extra ones for shutdown and other special cases} -- it's pretty easy to see which ones start and stop which process) -- none of this helped ...

So I thought -- the install worked -- it rebooted once -- what had I done to mess it up? Seemed like it must have been using ndiswrapper to get wireless working and/or blacklist of bcm43xx -- so I blacklist ndiswrapper and remove the blacklist of bcm43xx
-- and it boots up fine !!!

Of course then I have no wireless connection, but wired LAN is working and bcm43xx is loaded -- I read up on how to get bcm43xx working -- I installed bcm43xx-fwcutter using the wired LAN ... it's very easy to get wireless working with the bcm43xx driver and I now have a good, bootable install

Mike

I've been working with 6 for a couple weeks now on a few machines and have a similar problem with this laptop.

After finally figuring out the windoze drivers for the bcm43xx that came with the laptop don't work my son found one that does.

I wrote a script fix_my_wireless.sh that nukes all the drivers from the system, grabs the good ones off one of my sites, installs and fires off the wireless. Seems to work great, but as we tear into this little box we get to a point where we think its best to start over. Scripting is handy even if bash is warty.

Then I have another script that does all kinds of stuff my son wants/needs for his dual boot linux box. Things like add skype, debian-multimedia for libdvdcss to the sources.list, do an update, upgrade, install mozplugger, kmplayer, ipod stuff, vim, update /etc/profile yada yada.

Well, the wonderful lock up described here started after one or two things that I'm still digging into. Ran the script and did an update and upgrade, then stopped the script and the problem was there - reinstalled the OS and remarked the upgrade from the shell script. Everything seemed fine.

Got dvds playing, ipod working, playing mp4's, movies off the ipod playing - seemed like a done deal. Then rebooted, it reared it's ugly head again.

One thing - at times the usb wireless mouse would freak, the load would goto about 10 and nothing would happen. Seems like hal-device-manager fixed this issue - not for sure tho, doesn't make sense and not sure I care unless it starts acting up again.

I'm going to leave the ipod out of the picture, don't upgrade and see how it goes. If problems persist I'll nuke the wireless as suggested and see if this helps.

I'll keep the progress posted and monitor this thread.

So far Mepis has been the most enjoyable Linux desktop I've tried, I just don't have a couple hundred hours for configuration (although I'm 1/2 way there).

I looked through the logs and when it started freezing, a new device/hardware id showed that was not in the logs.

My wireless card had a different ID from what was mentioned above as I can remember also.

Food for thought.

I re-installed SimplyMEPIS v6.0 on the same hardware/same partition and then did "Mark All Upgrades" after "Reload". I used the default repository set. No mods or tricks. Some 50+ packages were updated. Whoever corrected the software glitch in one or more packages, somewhere, I thank you. All works well.