Find the answer to your Linux question:
Page 1 of 2 1 2 LastLast
Results 1 to 10 of 13
I upgraded from Lenny to Squeeze more than 3 weeks ago. It wouldn't boot. I fixed the problem of GRUB2 bootloader not loading. Then I started getting the message : ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
  1. #1
    Just Joined!
    Join Date
    Jun 2012
    Posts
    20

    Kernel panic ! Not able to sync...after upgrading to Squeeze


    I upgraded from Lenny to Squeeze more than 3 weeks ago. It wouldn't boot.

    I fixed the problem of GRUB2 bootloader not loading.

    Then I started getting the message : Alert! UUID xxx yyy does not exist. Dropping to a shell ! (posted in another thread earlier).

    On the good advice of jayd512 I carefully reinstalled the 2.6.32.5 kernel, and this time with make localmodconfig (I found this obscure command after a loooot of research, because I wasn't able to get the modules on the .config file correct).

    -----------------

    Now, I'm still unable to boot and the problem I face is as follows:

    Problem : On boot after selecting the newly installed kernel 2.6.32.5, the error message says, Kernel panic : Not able to sync VFS.

    Now I'm relying on a post by "whyzerman" in this forum [1]
    to solve my problem (the VFS syncing error during boot is fairly common).

    But I'm unable to :

    a) Install mdadm via dpkg -i mdadm
    The error I get is : FATAL : Coluld not load /lib/module 2.6.29.06-std122......(BTW this is the rescuecd kernel number, I'm using)

    b) As rightly said here [2] the mdadam.conf in my system doesn't show any partitions. But the thing is I can't run an mdadm -detail -scan

    HELP.

    References :

    [1] ht t p://forums.debian.net/viewtopic.php?f=10&t=74946#p416281

    [2] ht t p://ubuntuforums.org/showthread.php?t=1846484

  2. #2
    Just Joined!
    Join Date
    Mar 2011
    Posts
    68
    mdadm is for RAID systems. Are you using a RAID? If not you are barking up the wrong tree.

    Could you please post the output of /etc/fstab and your GRUB2 configuration (for the record I don't use GRUB2 because I find the configuration syntax unnecessarily complex but if you post it anyways someone else will probably find it useful.)

    Also, check your kernel configuration for the blocklayer being enabled and the driver for your hard drive is compiled into the kernel too.
    Make sure that the driver for the file system of your / partition is compiled into your kernel.
    When I say "compiled into" check that there is a[*] selecting it and not just a [M]. the [M] will make a modules only but the[*] will build right into the kernel binary, which is what your bootloader will push into memory.
    Last edited by bleedingsamurai; 06-24-2012 at 03:36 PM. Reason: Additional info

  3. #3
    Just Joined!
    Join Date
    Jun 2012
    Posts
    20
    The highly delayed response was unavoidable.

    Quote Originally Posted by bleedingsamurai View Post
    mdadm is for RAID systems. Are you using a RAID? If not you are barking up the wrong tree.

    Could you please post the output of /etc/fstab and your GRUB2 configuration (for the record I don't use GRUB2 because I find the configuration syntax unnecessarily complex but if you post it anyways someone else will probably find it useful.)

    Also, check your kernel configuration for the blocklayer being enabled and the driver for your hard drive is compiled into the kernel too.
    Make sure that the driver for the file system of your / partition is compiled into your kernel.
    When I say "compiled into" check that there is a[*] selecting it and not just a [M]. the [M] will make a modules only but the[*] will build right into the kernel binary, which is what your bootloader will push into memory.
    Thanks for pointing out that I don't have to use mdadm as I don't use RAID devices. I was needlessly worried over it.

    I will post my /etc/fstab once I get back home. However, it seemed fine from what I've seen else where.

    1) How do I check my kernel configuration for the block layer ? I mean which entry in the .config file should be checked for that ?

    2) To check the driver of the file system being compiled into the kernel, do I have to check the .config file or is there some entry in the interface of make menuconfig that I have to modify ?

    Thanks.

  4. $spacer_open
    $spacer_close
  5. #4
    Just Joined!
    Join Date
    Mar 2011
    Posts
    68
    Quote Originally Posted by Abhid View Post
    The highly delayed response was unavoidable.
    No problem, life catches up to you every now and then, I completely understand.
    1) How do I check my kernel configuration for the block layer ? I mean which entry in the .config file should be checked for that ?
    In your config file, check for:
    CONFIG_BLOCK=y
    CONFIG_BLK_DEV_BSG=y
    CONFIG_MSDOS_PARTITION=y

    There is actually a section in the root of menuconfig called:
    Code:
    Enable the block layer --->
    where you can double check over all this good stuff in more detail.

    2) To check the driver of the file system being compiled into the kernel, do I have to check the .config file or is there some entry in the interface of make menuconfig that I have to modify ?
    Personally I find that most of the time it is a lot easier to use menuconfig which uses the formatting I explained in my previous post. Manually editing the config file isn't fun, that is more for automating checks and changes with scripts.

    fire up menuconfig, in the root of the menu you should see something like:
    Code:
    File systems --->
    enter this sub-menu, and make sure whatever file system you formatted your (/)root partition with is selected with an (*) asterisk. For example, mine is ext4 so my configuration looks like this:
    Code:
    <*> The Extended 4 (ext4) filesystem

  6. #5
    Just Joined!
    Join Date
    Jun 2012
    Posts
    20
    Quote Originally Posted by bleedingsamurai View Post
    In your config file, check for:
    CONFIG_BLOCK=y
    CONFIG_BLK_DEV_BSG=y
    CONFIG_MSDOS_PARTITION=y
    These entries are proper.

    fire up menuconfig, in the root of the menu you should see something like:
    Code:
    File systems --->
    enter this sub-menu, and make sure whatever file system you formatted your (/)root partition with is selected with an (*) asterisk. For example, mine is ext4 so my configuration looks like this:
    Code:
    <*> The Extended 4 (ext4) filesystem
    I checked that entry.

    The result is the same. Here is my /etc/fstab :

    Code:
    proc               /proc       proc     defaults     0    0
    /dev/sda1        /           ext2       errors=remount-ro 0     1
    UUID=xxxx-yyyy-xzzzz
    
    /dev/sda5        none       swap     sw     0     0
    UUID=aaaaa-bbbbb
    
    /dev/sdc          /media/cdrom0      udf,iso9660   user,noauto     0     0
    
    /dev/cdrom        /media/cdrom0    udf,iso9660 user,noauto     0     0
    -----------

    I just re-compiled the kernel again and now I get that "Alert ! Dropping to a shell" thing once more (I had started a thread on this earlier). All the aforementioned entries in .config and menuconfig are unchanged as before.

    There is also one nagging question : Since the beginning, the GRUB bootmenu shows 2 kernels in (recovery mode). Why ?

  7. #6
    Just Joined!
    Join Date
    Mar 2011
    Posts
    68
    Are those really the values your UUIDs are set equal too? I think that is your problem, those aren't real UUIDs. Try commenting them out in fstab and formatting GRUB to boot without UUIDs

    Well, usually what that means is that instead of the kernel executing /sbin/init after it is loaded into memory, bootloader/kernel command line options are used to tell it to run a recovery shell, usually BusyBox. When distros push a kernel update they will usually edit the menu to list a regular kernel and then a boot configuration like I mentioned above labeled as recovery.

  8. #7
    Just Joined!
    Join Date
    Jun 2012
    Posts
    20

    Thumbs down

    Quote Originally Posted by bleedingsamurai View Post
    Are those really the values your UUIDs are set equal too? I think that is your problem, those aren't real UUIDs. Try commenting them out in fstab and formatting GRUB to boot without UUIDs
    Oh no, those aren't the real UUIDs. I just wrote a string blah--blah because it was a bit tedious to write. Its just a long alpha-numeric string.

    Quote Originally Posted by bleedingsamurai View Post
    Well, usually what that means is that instead of the kernel executing /sbin/init after it is loaded into memory, bootloader/kernel command line options are used to tell it to run a recovery shell, usually BusyBox. When distros push a kernel update they will usually edit the menu to list a regular kernel and then a boot configuration like I mentioned above labeled as recovery.
    Well, looks like I'm at a dead end then

  9. #8
    Just Joined!
    Join Date
    Mar 2011
    Posts
    68
    The last thing I can think of is an issue with your GRUB2 configuration. As I mentioned before, I tend to avoid GRUB2 but I'd be willing to try and sort through it to help you out.

    I'm thinking, in the configuration, GRUB2 is telling your kernel the wrong partition to look for / on.

    From what I gathered during my limited experience with GRUB2, they index partitions starting with 1 rather then with 0 like Legacy GRUB (but the hard drives are still indexed starting at 0!). So make sure your entry for root in /boot/grub/grub.cfg looks like: (hd0,1)
    Assuming the /etc/fstab you posted is correct.

  10. #9
    Just Joined!
    Join Date
    Jun 2012
    Posts
    20
    Quote Originally Posted by bleedingsamurai View Post
    I'm thinking, in the configuration, GRUB2 is telling your kernel the wrong partition to look for / on.

    From what I gathered during my limited experience with GRUB2, they index partitions starting with 1 rather then with 0 like Legacy GRUB (but the hard drives are still indexed starting at 0!). So make sure your entry for root in /boot/grub/grub.cfg looks like: (hd0,1)
    Assuming the /etc/fstab you posted is correct.
    If it helps, the root entry in my grub.cfg reads (hd0, msdos) instead of (hd0,1) or (hd0,2). Is this the reason for the "Alert! Falling to a shell" message ?

  11. #10
    Just Joined!
    Join Date
    Mar 2011
    Posts
    68
    the root entry in my grub.cfg reads (hd0, msdos)
    Well that just doesn't seem right. Try changing that to:
    Code:
    (hd0,1)
    and see how that works.
    Last edited by bleedingsamurai; 07-09-2012 at 03:14 AM. Reason: fixed spelling error

Page 1 of 2 1 2 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •