Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.debian.user > #224268

Re: Installing Buster (improvements)

From "Thomas Schmitt" <scdbackup@gmx.net>
Newsgroups linux.debian.user
Subject Re: Installing Buster (improvements)
Date 2020-07-01 17:20 +0200
Message-ID <AnMGJ-1Ju-15@gated-at.bofh.it> (permalink)
References <AnK2d-8pg-1@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


Hi,

Klaus Jantzen wrote:
> Do you know how much space you are talking
> about when you specify the space e.g. from cylinder  30312/sector 301/head
> 45 to cylinder 30800/sector 20/head 4 on an SSD?

This does not depend on the storage device, at least not in this century.

In your case it rather depends on the partition editor's guessing of how
many Heads there are per Cylinder (H/C) and how many Sectors per Head (S/H).
Then an address of Cylinder/Head/Sector is converted to Logical Block Adress
(LBA) by

  LBA = ( ( Cylinder * H/C ) + Head ) * S/H + Sector - 1

The size is then the difference between start and end LBA plus 1.

---------------------------------------------------------------------------

The C/H/S addressing scheme still exists only because MBR partition tables
have fields for start and end C/H/S. But that scheme ends at 8,422,686,720
bytes, due to the width of those fields.
So everything above the ~8 GB limit has to be computed as Logical Block
Adress from the C/H/S input before it can be written into the partition
table on the storage device.

The big question is the pair of conversion numbers H/C and S/H.
By tradition it is guessed and handed out by the BIOS to those who care.
But that guess is not guaranteed to be stable from machine to machine.

If the table already has entries which associate LBA with valid looking
C/H/S, then it is possible to deduce matching candidate pairs. Often only
one of the 16065 possible pairs matches all existing entries with valid
C/H/S numbers.
But some partition editors still guess differently, even if their guess
does not yield the existing LBAs.


Have a nice day :)

Thomas

Back to linux.debian.user | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread


Thread

Installing Buster (improvements) Klaus Jantzen <k.d.jantzen@mailbox.org> - 2020-06-30 22:40 +0200
  Re: Installing Buster (improvements) Peter Ehlert <peter@sdi-baja.com> - 2020-06-30 23:50 +0200
    Re: Installing Buster (improvements) Klaus Jantzen <k.d.jantzen@mailbox.org> - 2020-07-01 14:30 +0200
      Re: Installing Buster (improvements) "Thomas Schmitt" <scdbackup@gmx.net> - 2020-07-01 17:20 +0200
      Re: Installing Buster (improvements) David Wright <deblis@lionunicorn.co.uk> - 2020-07-02 03:10 +0200
  Re: Installing Buster (improvements) "Thomas Schmitt" <scdbackup@gmx.net> - 2020-07-01 08:30 +0200

csiph-web