Re: Linux and Windows partitioners fail to see opposite partitions
Elmer E. Dow wrote:
page says for the command W) with rewriting the mbr as you said. It's
sinking in now that evidently the mbr and the partition table are the
same thing. I need to do some reading to get a better understanding of
the workings of a computer system in general and the boot process in
particular.
In addition to Stephen's post, which covers most of what you need to
know in practical terms, there are plenty more gory details here:
http://en.wikipedia.org/wiki/Master_boot_record
To avoid confusion, I'd add that most (all?) MBR fixing utilities don't
actually touch the partition table entries, they write the first 446
bytes. Recreating a partition table by reading the raw drive data,
containing unknown partition types and filesystems, is not a trivial
job, and will not be attempted by a simple utility. So it's still
possible for an MBR fixing utility to replace 446 bytes of correct code
but leave a corrupted partition table in place.
It is also possible, but made clear not to be the case here by your
fdisk listing, for partition table entries to be placed out of order on
the physical drive i.e. fdisk will allow you to create hda6 with
higher-numbered cylinders than hda7. Most software which reads the
partition table can deal with this, some cannot and will fail to see all
partitions.
There was a Microsoft-approved method of installing NT4.0 to a hard
drive of greater than a certain size (I forget how big, but 8GB rings a
bell) which involved two NT4.0 installations, one to a primary partition
placed after an extended partition i.e. with higher cylinder numbers
than the extended and logical partitions. I recall at least one piece of
software, possibly Partition Magic or one of its competitors, freaking
out about that.
--
Joe
Reply to: