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

Re: Partition Magic Chokes on my Partitions



On Thu, Jun 17, 2004 at 06:21:58PM +0200, Alexander Dalloz wrote:
> Am Do, den 17.06.2004 schrieb Per-Olof Litby - Reg'l Mgr Nordic/Baltic -
> Java System Software - Sun Microsystems um 17:52:
> 
> > Here's the problem: I have a triple boot system on this laptop - Win XP 
> > Pro, Fedora Core 2, and Sun's Java Desktop System (SuSE based). Here's 
> > the output of the fdisk -l command:
> > 
> >    Device Boot      Start         End      Blocks   Id  System
> > /dev/hda1               1        3335    26785048+   7  HPFS/NTFS      
> >          Windows XP, NTFS
> > /dev/hda2            3335        4864    12285000    f  W95 Ext'd 
> > (LBA)         Extended, containing hda5,6,7,8
> > /dev/hda5            3335        4100     6146248+  83  Linux            
> >        Fedora Core 2 / ext3
> > /dev/hda6            4100        4228     1028128+  82  Linux swap      
> >         Fedora Core 2 swap
> > /dev/hda7   *        4228        4789     4505728+  83  Linux            
> >        Sun JDS / reiserfs
> > /dev/hda8            4789        4864      604768+  82  Linux swap      
> >         Sun JDS swap
> 
> The partitions start on the same block with which the previous partition
> ends?

Good catch, but it might be OK. Those are cylinders, not blocks. It is
possible that a partition may end somewhere inside a cylinder (say, it
includes heads 0 and 1), and the next partition starts in that
cylinder (say, heads 2 and 3). I haven't seen such a beast, but with
LBA addressing or SCSI drives it is feasible. Such a scheme would work
only if one only accessed the drive with LBA. CHS addressing would
break.

The main advantage for ending partitions on cylinder boundaries was,
way back when, the laziness of programmers, or minimizing head
movement within a partition. With LBA and variable density zones on
modern drives, tracks, heads, sectors and "cylinder boundaries" are
illusory anyway.

I'd want to look at the partition table itself to be sure that my
guess is correct, and to be sure that the partitions don't
overlap. Does fdisk report such overlapping?

-- 

Charles Curley                  /"\    ASCII Ribbon Campaign
Looking for fine software       \ /    Respect for open standards
and/or writing?                  X     No HTML/RTF in email
http://www.charlescurley.com    / \    No M$ Word docs in email

Key fingerprint = CE5C 6645 A45A 64E4 94C0  809C FFF6 4C48 4ECD DFDB

Attachment: pgp00334.pgp
Description: PGP signature


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