Results 1 to 9 of 9

Thread: large block device support missing in 3.7.1?

  1. #1
    Join Date
    Nov 2007
    Posts
    1,353

    Default large block device support missing in 3.7.1?

    Under the menu heading "Enable the block layer" there should be an option called "Support for large (2TB+) block devices and files" which is required for just about every modern filesystem, but it seems to be missing. Without it, I cant get my / partition to mount r/w....

    Am I missing something obvious? I looked all through the menuconfig and couldnt find it. Without that option ext4 is worthless. and I dont know what happened to it.

    Any help would be awesome. if you guys need more information I'll gladly get it for you.

    EDIT: I know that there are some kernel options that only become visible if some prerequisite is checked, does this option also have a prereq that needs checked for it to be visible?
    Last edited by duby229; 12-18-2012 at 06:36 PM.

  2. #2
    Join Date
    Jul 2009
    Location
    Germany
    Posts
    523

    Default

    Are you running a 32bit kernel?
    The option is only visible with 32bit kernels:
    Code:
    Symbol: LBDAF [=n]
    Type : boolean
    Prompt: Support for large (2TB+) block devices and files
    Defined at block/Kconfig:26
    Depends on: BLOCK [=y] && !64BIT [=y]
    Location:
    -> Enable the block layer (BLOCK [=y])

  3. #3
    Join Date
    Nov 2007
    Posts
    1,353

    Default

    Quote Originally Posted by droste View Post
    Are you running a 32bit kernel?
    The option is only visible with 32bit kernels:
    Code:
    Symbol: LBDAF [=n]
    Type : boolean
    Prompt: Support for large (2TB+) block devices and files
    Defined at block/Kconfig:26
    Depends on: BLOCK [=y] && !64BIT [=y]
    Location:
    -> Enable the block layer (BLOCK [=y])
    ok yeah that is interesting.... Yeah I'm running a 64bit kernel....

    So if that option is not required for 64bit, then my problem is something entirely different than what I thought. This is a typed out as closely as I could remember of the exact error message dmesg gave me. I cant get a kernel log because the / fs is mounting ro and so it isnt being written to disk.

    Code:
    EXT4-fs (sdd3): mounted filesystem with ordered data mode
    VFS: Mounted root (ext4 filesystem) readonly on device

  4. #4
    Join Date
    Nov 2007
    Posts
    1,353

    Default

    here is some info if it helps at all....

    /dev/sdd1 /boot ext2 defaults 1 2
    /dev/sdd3 / ext4 defaults 0 1
    /dev/sdd2 none swap sw 0 0


    default 0
    timeout 0

    title Gentoo Linux
    root (hd0,0)
    kernel /boot/vmlinuz root=/dev/sdd3 rootfstype=ext4

    Really this should be all thats required to boot. I dont understand why its mounting / ro

  5. #5
    Join Date
    Jul 2009
    Location
    Germany
    Posts
    523

    Default

    I'm not sure what exactly is needed to boot from grub1 with an ext4 root without initrd/initramfs, but a quick google search brings up this:
    http://forums.gentoo.org/viewtopic-t-880313.html

    It states that you are probably missing devtmpfs in your kernel config:
    Code:
    CONFIG_DEVTMPFS=y
    CONFIG_DEVTMPFS_MOUNT=y

  6. #6
    Join Date
    Nov 2007
    Posts
    1,353

    Default

    well, i did not have those options in my kernel, so i added them and rebuilt it, but i am still getting exactly the same problem.... however according to dmesg devtmpfs is being loaded after the kernel mounts the fs ro..

  7. #7
    Join Date
    Nov 2007
    Posts
    1,353

    Default

    I couldnt figure this problem out, so I decided to re-install on the tried and true ext3...... exactly the same problem..... I'm completely stumped. I dont have a useable os, and I have no idea what is wrong so I can fix it.

  8. #8
    Join Date
    Apr 2010
    Posts
    1,946

    Default

    I have "rw,errors=remount-ro" for my / disk in fstab, instead of "defaults".
    There is also nothing wrong mount it as ro, so long you make relevant directories rw. For example server hardening use that.
    But assigning ro to whole / is stupid. Try change this in your fstab.

    I don't think tmpfs has anything to do with it.

  9. #9
    Join Date
    Apr 2010
    Posts
    1,946

    Default

    There could also be other causes. If you post dmesg somewhere, it could really allow to help you.
    Also, you could try compile and boot older kernel with exactly same config and look if it works, just to narrow the issue down.

Posting Permissions

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