Text 14062, 110 rader
Skriven 2005-12-30 02:36:00 av MARTIN ATKINS (1:123/140)
     Kommentar till en text av ALAN ZISMAN
Ärende: Disc size mystery
=========================
-=> ALAN ZISMAN wrote to ALL <=-


 AZ> I'm suprised at both of these-- that this relatively recent BIOS didn't
 AZ> recognize the full capacity of the drive, and that WINXP didn't seem to
 AZ> use the (incorrect) BIOS settings-- but instead is using its own, much
 AZ> better (though still incorrect) settings.

 AZ> It would be nice to be able to get the full capacity of the drive
 AZ> recognized... but I can probably live with pretending it's a 130 GB
 AZ> drive.

Using the old BIOS limitation it should be able to get 128 real GB.
1048576 = 1 gig (notional) so 128 * 1048576 = 134,217,720 or 134 GB/dec
probably reporting as 130 Gig/decimal. I don't know much about NTFS or 
XP but it depends on whether it's geared up as a 8, 16, 32, 64, or 128 
bit operating system. If your BIOS sees 8 gig on an 8 bit system then it 
will be 16 gig on 16 bit system and all the way up to 128 gig on a 
128 bit sys. So the theory goes. :-} 

However Disk capacity can be expanded by doubling the cylinders from the 
old BIOS limitation of 1024 cylinders to 2048 using a small amount of each 
of the first 1024 cylinders to address the next 1024. So in theory there 
are at least two ways to expand away from the BIOS limitations. 

There are other ways of expanding things but to be honest I've found that 
these two rules give me a good guide to the disks potential capacity.

NTFS may be restricted to the 1024, 2048, 4096, rule but that does
not restrict you from using the spare space because the other 30 gig
or so can be another partition.  

<cut paste>

AZ> The BIOS 'recognizes' the drive as 8 GB; it won't let me manually 
AZ> enter Samsung's recommended manual settings: 312, 581, 808...

I'm not sure what these numbers represent but i think you only need
know HEADS SECTORS and CYLINDERS. I can't drop to BIOS at the moment
but just because the BIOS can't see beyond 8 gig it doesn't mean your
operating system can't (as is obvious but i had to point it out).

The only way i can explain things to myself is that the first 
1024 cls are addressed by the BIOS and after that the operating
system takes over. Try using what ever your latest version of WIN 
"fdisk" is and read the the way it sees the disk and that is what
you will get. 

I know you are familiar with Linux so also use Linux "fdisk" and
see how it interprets the same disk. There may be discrepancies
between the two because the software takes over after the BIOS
has interpreted the disk. I'm trusting your in a good mood so i
will bore you with some Linux stuff. :) So i "fdisk" my hard
disk:-

fdisk /dev/hda

The number of cylinders for this disk is set to 9729.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
   (e.g., DOS FDISK, OS/2 FDISK)
  
STOP

Notice it warns me that my disk is outside the 1024 range. ;)
But using the the double cylinder rule 1024 becomes 2048 and
then 4096 and then 8192 and this is the true bootable disk
range by the BIOS. So mechanically the BIOS, depending on the
operating system will see the first 1024 cls as the first
gig addressing the next 1024 cls with heaps to spare (you only
need a few bits or bytes to address each new track on the next
bank of clys).

Since the disk on my system has 9729 cls then we have:-

9729 - 8192 = 1537

The BIOS can't address the next 1537 cylinders because it can
only deal with complete disks of 1024 cls or powers of 2 and the
next power of 2 is 16384.

Disk /dev/hda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hda1   *           1        2096    16836088+   b  W95 FAT32
/dev/hda2            2097        9730    61314624    5  Extended
/dev/hda5            2097        2684     4723078+  83  Linux
/dev/hda6            2685        3579     7189056   83  Linux
/dev/hda7            3580        3968     3124611   83  Linux
/dev/hda8            3969        4535     4554396   83  Linux
/dev/hda9            4536        5350     6546456   83  Linux
/dev/hda10           5351        5381      248976   82  Linux swap
/dev/hda11           5382        9730    34927830   83  Linux
       
Although in this example both WIN and Linux have what appear to
be odd numbers both fall bellow powers of 2 of the 1024 rule
if you remember this disk has 9729 cylinders then only
/dev/hda11 falls outside what i concider the safe bootable range. 

It is also is sometimes an operating systems limitation when 
interpretation disk capacity
   
I really must get a life. ;)

--- MultiMail/Linux v0.46
 * Origin: Try Our Web Based QWK: DOCSPLACE.ORG (1:123/140)