[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Solution: ext2/boot & reserfs/root + yaboot



Hi,

>The current version of yaboot with ReiserFS patch (avavilable on SuSE ftp) 
>does NOT supports yet separate /boot ext2 and root ReiserFS partitions (it 
>throws in open firmware). However, it works without any problem if /boot 
>and /root are located on single ReiserFS partition. Also, the system is 
>bootable from the CD.

Olaf Hering and Jeff Mahoney from SuSE have identified problem in yaboot 
which prevents to use ext2/boot & reserfs/root with yaboot under certain 
circumistances. Fix will be available very soon.


This is reply from SuSE:

>    Yaboot doesn't know, or care, what filesystem root is. The extent of its
>    knowledge is simply that root=/dev/path is another parameter to be passed
>    to the kernel. It may be having issues loading a kernel from your ext2
>    filesystem, but it shouldn't - my dual g4 loads its kernels from /boot
>    (part of the ext2 root filesystem).
>
>    One of the changes in my Yaboot patches is that it removes checking for
>    mac-style partition "naming", since it's not really valid. However, 
>the old
>    Yaboot will still check for these. My development system was 
>misconfigured
>    - I just used the last partition without changing the name from
>    "apple_free" as my root filesystem. Linux could handle it just fine, 
>since
>    it appeared as just another partition, but Yaboot would ignore the
>    "apple_free" partition type, and refuse to load the image (or even 
>look at
>    the filesystem at all). You may want to verify that this isn't the case.
>
>    The default catch bug I think you're refering to is one that I've got
>    tracked down and fixed. It's related to a failure while attempting to 
>open
>    a file. Later on in the code, it tries to close the file - but since it
>    hasn't been opened, it tries to follow a null pointer. I haven't released
>    a new patch for this yet, since there is another bug that I need to track
>    down that is a little more elusive. Since this bug is affecting you,
>    I'll post that patch later today.
>
>    You also may want to try to compile yaboot with DEBUG=1 (in the 
>Makefile).
>    It will tell you a little bit about _why_ it's failing to load a kernel
>    from an ext2 fs.
>
>    -Jeff






********************************************************
*   Best Regards   ---   Andrei Verovski               *
*   e-mail         ---   andrei.verovski@bigfoot.com   *
*                                                      *
*   Personal Home Page                                 *
*   http://homepage.mac.com/macgurutemple/             *
********************************************************




Reply to: