Bios freaks
Karl Larsen
k5di at zianet.com
Mon Aug 20 15:17:42 UTC 2007
Mikkel L. Ellertson wrote:
> Tim wrote:
>
>> On Sun, 2007-08-19 at 18:52 -0600, Karl Larsen wrote:
>>
>>> I have heard a thousand words a week about LVM and it never made the
>>> point that it moved /boot close to the near end of a hard drive.
>>>
>> I haven't heard about that happening. It's always been my experience
>> that if you used the install routines to prep your drive, that it made
>> the /boot partition the first one. I don't know if that's coincedental,
>> but it would be a good thing for it to deliberately do. Other
>> partitions get shuffled about, though. I don't know the reasoning
>> behind it (if there is any).
>>
>>
> I don't think Karl understood what people were trying to tell him.
> The problem is not LVM moving /boot close to the end of a hard
> drive, but that when you have /boot as part of the / partition,
> instead of giving its own partition, the files in /boot may end up
> towards the end of the drive.
>
>
>> e.g. If you had created /boot/, /home/, /tmp/, /usr/, /var/, in that
>> order, you might find that at the end of your manual intervention, it
>> actually created the partitions in another order, albeit with /boot/
>> being the first partition.
>>
>> Of course, thanks to how drives fake their number of heads and
>> cylinders, to accomodate how BIOSs (and IDE?) are still terribly poor at
>> handling large drives, there's still no guarentee that all of the first
>> partition is going to be where the BIOS can access it.
>>
>>
> The drives internal remapping should have no affect on what the BIOS
> can access. The BIOS just tells the drive I want this track, head,
> sector, and it could care less where it really is on the drive. If
> the partition is the first x blocks on the drive, the BIOS can
> access it, regardless of the physical location on the drive, as long
> as the logical h,t,s is in the range the BIOS can handle.
>
> Mikkel
>
You have not been following the other thread bash. I checked just
now and my /boot is at /dev/sdb6 and it works fine. Here is fdisk -l
Disk /dev/sdb: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/sdb1 * 1 1217 9775521 83 Linux
/dev/sdb2 1218 1945 5847660 83 Linux
/dev/sdb3 1946 1961 128520 82 Linux swap / Solaris
/dev/sdb4 1962 18534 133122622+ 5 Extended
/dev/sdb5 1962 7060 40957686 83 Linux
/dev/sdb6 7061 12159 40957686 83 Linux
/dev/sdb7 12160 18534 51207156 83 Linux
[root at k5di ~]#
Notice that sdb6 starts at 7061 cylinders! And the grub boot works
fine to a setup on the first hard drive (hd0) :-)
The 1994 BIOS is reading the grub message and going to another hard
drive finds the /boot at 7061 cylinder. 8-)
Puts a crimp in your theory.
--
Karl F. Larsen, AKA K5DI
Linux User
#450462 http://counter.li.org.
More information about the fedora-list
mailing list