How do I set the physical geometry?

How do I set the physical geometry?

Post by K. Poser » Sun, 31 Dec 1899 09:00:00



Hi.

I've got a IBM 45GB Deskstar DTLA 307045 with CHS=16383/16/63 (printed
on the label of the harddisk).
I suppose these CHS-values are the values of the physical geometry?!
But I don't get linux to show me these values as "physical geometry".
When I type "cat /dev/ide/hdc/geometry" the physical geometry is always:

89355/16/63
The logical geometry vary according to what I passed to the kernel at
boot-time (with "append=" in lilo.conf):
1027/255/63 or 89355/16/63 or 16383/16/63

So is it o.k. that on the harddisk is printed CHS=16383/16/63 and linux
always uses 89355/16/63???

One information on the harddisk-label which could be important on that:
In LBA-mode there are
90 069 840 sectors

I hope someone could help me.

Greetings

Knuth Posern

 
 
 

How do I set the physical geometry?

Post by Dances With Cro » Thu, 24 Aug 2000 09:00:16



Quote:>I've got a IBM 45GB Deskstar DTLA 307045 with CHS=16383/16/63 (printed
>on the label of the harddisk).
>I suppose these CHS-values are the values of the physical geometry?!
[snip]
>One information on the harddisk-label which could be important on that:
>In LBA-mode there are 90 069 840 sectors

Modern hard disks do not have anything like a physical geometry in the
old-style sense of the term.  That "16383/16/63" printed on the drive is
shorthand for "This disk is too big for the ancient disk-size-reporting
method that BIOS 'normal mode' used back in the day."

What exactly is the problem that requires you to pass geometry arguments
to the kernel at boot time?  Is your machine's BIOS so old that it can't
be upgraded to something that can handle large disks?  Machines built
after 1997 or so should have no problems with most disks.  If the BIOS
is too old, you can probably find out what the Linux kernel thinks the
geometry of the disk is by booting from a rescue system like Tom's
RootBoot (http://www.toms.net/rb/ ) and running "fdisk".  That will say
something like:
  Disk /dev/hda 255 heads, 63 secotrs, 5606 cylinders
Also, Linux kernels prior to 2.2.14 had problems with accessing IDE
disks > 32G.  See if you can get a newer kernel--RH 6.2, SuSE 6.4,
Mandrake 7.1, et al shipped with kernels capable of handling IDE disks
up to 137G.

--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin /  Those who do not understand Unix are
http://www.brainbench.com     /   condemned to reinvent it, poorly.
-----------------------------/           --Henry Spencer

 
 
 

How do I set the physical geometry?

Post by Norman Levi » Sun, 31 Dec 1899 09:00:00




> >I've got a IBM 45GB Deskstar DTLA 307045 with CHS=16383/16/63 (printed
> >on the label of the harddisk).
> >I suppose these CHS-values are the values of the physical geometry?!
> [snip]
> >One information on the harddisk-label which could be important on that:
> >In LBA-mode there are 90 069 840 sectors

> Modern hard disks do not have anything like a physical geometry in the
> old-style sense of the term.  

** until disk go solid state or bubble memory, they certainly have
a physical geometry.

That "16383/16/63" printed on the drive is

Quote:> shorthand for "This disk is too big for the ancient disk-size-reporting
> method that BIOS 'normal mode' used back in the day."

** true, that size is probably larger than most bios can handle.

BUT, if you pretended that you had 32 heads instead of 16, you could
address all of the disk with 1/2 the number of cylinders.  So
16383/16/63  can look like
 8162/32/63  which can look like
 4081/64/63  which can look like
 2042/128/63 which can look like
 1021/256/63

Quote:

> What exactly is the problem that requires you to pass geometry arguments
> to the kernel at boot time?  Is your machine's BIOS so old that it can't
> be upgraded to something that can handle large disks?  Machines built
> after 1997 or so should have no problems with most disks.  If the BIOS
> is too old, you can probably find out what the Linux kernel thinks the
> geometry of the disk is by booting from a rescue system like Tom's
> RootBoot (http://www.toms.net/rb/ ) and running "fdisk".  That will say
> something like:
>   Disk /dev/hda 255 heads, 63 secotrs, 5606 cylinders
> Also, Linux kernels prior to 2.2.14 had problems with accessing IDE
> disks > 32G.  See if you can get a newer kernel--RH 6.2, SuSE 6.4,
> Mandrake 7.1, et al shipped with kernels capable of handling IDE disks
> up to 137G.

> --
> Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
> Brainbench MVP for Linux Admin /  Those who do not understand Unix are
> http://www.brainbench.com     /   condemned to reinvent it, poorly.
> -----------------------------/           --Henry Spencer

--
Norman Levin
 
 
 

How do I set the physical geometry?

Post by emc.. » Sun, 31 Dec 1899 09:00:00



> > Modern hard disks do not have anything like a physical geometry in the
> > old-style sense of the term.  
> ** until disk go solid state or bubble memory, they certainly have
> a physical geometry.

He never said they didn't have a physical geometry; just that it's
nothing so simple as "cylinders, heads, and sectors per track."

By the way, devices with no moving parts also have a physical
geometry.  RAM, for example.

--

"misfortune, n. The kind of fortune that never misses."
        - Ambrose Bierce, The Devil's Dictionary

 
 
 

How do I set the physical geometry?

Post by Villy Kru » Sun, 31 Dec 1899 09:00:00




>> > Modern hard disks do not have anything like a physical geometry in the
>> > old-style sense of the term.  

>> ** until disk go solid state or bubble memory, they certainly have
>> a physical geometry.

>He never said they didn't have a physical geometry; just that it's
>nothing so simple as "cylinders, heads, and sectors per track."

Just consider that disks nowadays have more sectors per track on the
outer tracks than on the inner tracks.  The outer tracks are longer
and therefore there are room for more bits.  That is also how you
can try to multiply the CHS values and won't get a value different
from the total disk size.

The appearent geometry of ide disks is largely dictated by the number
of bits allocated to each of cyls, heads, and sectors.  Read the
HOWTO/mini/Large-Disk for the*details.

--
Villy

 
 
 

1. Benefits from computing physical IDE disk geometry?

I'm excited about the new I/O scheduler (proposed?) in the 2.5.x kernel, but
I have to admit to a considerable amount of ignorance of its actual
behavior.  Thus, if it already does what I'm talking about, please feel free
to ignore this post.  :)

Any good SCSI drive knows the physical geometry of the disk and can
therefore optimally schedule reads and writes.  Although necessary features,
like read queueing, are also available in the current SATA spec, I'm not
sure most drives will implement it, at least not very well.

So, what if one were to write a program which would perform a bunch of
seek-time tests to estimate an IDE disk's physical geometry?  It could then
make that information available to the kernel to use to reorder accesses
more optimally.  Additionally, discrepancies from expected seek times could
be logged in the kernel and used to further improve efficiency over time.
If it were good enough, many of the advantages of using SCSI disks would
become less significant.

Ideas?

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

2. How to detect a program is being traced

3. HELP: Linux/DOS IDE geometry problems

4. how many time the scheduler give to a process?

5. Set Geometry from command line

6. backspace (delete) in netscape

7. where to set harddisk geometry

8. zip 100 problem

9. Ghostview default term geometry --where to set?

10. lilo.conf - How do you set drive geometry?

11. why doesn't grub's "geometry" command retain setting?

12. Setting defaults for emacs font/geometry

13. Setting geometry important for SCSI disk?