A question about disk partition layouts..

[Brynet-Inc]

I'm wondering, Why is their a 63 LBA gap between partitions?

Is there any historical significance to it? Why all the wasted space?

I haven't created many topics lately.. So I thought, "What the heck, I'll ask."

EDIT: 32256/512 = 63 - Apologies.

[elderK]

Aye, I wonder about this too.
On the bright side, it means I can install my Kernel and all of it's userspace onto a Filesystem, without actually taking any space up in the Filesystem.

(Yay for living in no-mans land! )

~Z

[Ready4Dis]

I'm wondering, Why is their a 64 LBA(32k) gap between partitions?

Is there any historical significance to it? Why all the wasted space?

I haven't created many topics lately.. So I thought, "What the heck, I'll ask."

It possibly just got aligned on a boundary of sorts. Or possibly the boot manager in use requires the space for booting, so it reserves it outside of a partition so that it won't be written to.

[Brynet-Inc]

Whoops, s/their/there - apologies for my illiteracy.

As for Ready4Dis, I wasn't asking about a specific instance, all disk partitioning utilities put a gap between partitions.
Layout of a DOS formatted disk is:

Code: Select all

ID type 0x06
Start CHS = 0/1/1
End CHS = 406/15/63

LBA Start 63, size 410193

ID type 0x05
Start CHS = 406/0/1
End CHS = 1013/15/63

LBA Start 410256, size 611856

As you can see, there is a gap between partitions.. and it's entirely full of zeros - is it a legacy CHS limit of some sort?

EDIT: It seems I was wrong, it's only a 32256 byte gap.. 63 LBA - Whoops * 2.

[bewing]

It all depends on your partitioning software. I have seen the gap you are talking about on m$ FDISK partitioned stuff. And yes, the software is trying to keep 1980's era disks partitioned on head/sector boundaries. The technique is utterly obsolete.

And I strongly suspect a typo in your CHS numbers. You are showing a 63 sector OVERLAP between the two partitions -- not a gap. I think you mean "End CHS = 405/15/63"

[Brynet-Inc]

And I strongly suspect a typo in your CHS numbers. You are showing a 63 sector OVERLAP between the two partitions -- not a gap. I think you mean "End CHS = 405/15/63"

Apologies:

ID type 0x05
Start CHS = 407/0/1
End CHS = 1013/15/63

The starting cylinder was 407, not 406 or 405.

So I believe the gap remains, and there is no partition overlapping bewing.

No big deal I guess, was simply curious.

[JAAman]

bewing is correct:

in the old days, it was common practice to start a partition on a cyl boundary (the first sector in the partition was always sector 1), some of the reasons, included easy calculating of disk offsets, and improved performance, partitions started simply aligning on a 63 sector boundary (the maximum sectors/cyl in the traditional CHS scheme)


however:
this space (at least the one at the beginning of the disk) should always be maintained for compatibility reasons, as a lot of disk managers and virus scanners use it (boot managers are often loaded from the 'in-between' space between the MBR and the first partition, some virus scanners also load themselves here, as do certain disk compression programs)

[bewing]

If the first partition ends at:

End CHS = 406/15/63

And the next partition starts at:

Start CHS = 407/0/1

Then they are contiguous. There is no gap. Those are sequential CHS addresses. (Sectors go from 1 to 63, Heads go from 0 to 15.)

[Brynet-Inc]

If the first partition ends at:

End CHS = 406/15/63

And the next partition starts at:

Start CHS = 407/0/1

Then they are contiguous. There is no gap. Those are sequential CHS addresses. (Sectors go from 1 to 63, Heads go from 0 to 15.)

Look, There is a gap between partitions..

Heck - check your own drive if you don't believe me..

[egos]

Look, There is a gap between partitions..

Heck - check your own drive if you don't believe me..

Where's gap? I don't see it!

..., 406/15(last logical surface number: 0...15)/63, 407/0/1(first logical sector number on the surface track: 1...63), ...

[Brynet-Inc]

It seems obvious the reason for the gap is legacy CHS related, regardless of that.. a physical gap exists between partitions if you'll take the time to look.

I believe JAAman's explanation clears things up, I was only curious about the gap - I'm not very used to MBR-style partitioning.