Skip to content

Commit

Permalink
---
Browse files Browse the repository at this point in the history
yaml
---
r: 42075
b: refs/heads/master
c: 7e45b0e
h: refs/heads/master
i:
  42073: 1b79684
  42071: a213542
v: v3
  • Loading branch information
Alan Cox authored and Jeff Garzik committed Dec 2, 2006
1 parent 4b32ac8 commit 87986cb
Show file tree
Hide file tree
Showing 972 changed files with 36,586 additions and 60,538 deletions.
2 changes: 1 addition & 1 deletion [refs]
Original file line number Diff line number Diff line change
@@ -1,2 +1,2 @@
---
refs/heads/master: a80958f4849316a18c06f75b9e850ccecbf20df8
refs/heads/master: 7e45b0e5f6f68d04311ba7813f5f5457c9119640
3 changes: 0 additions & 3 deletions trunk/CREDITS
Original file line number Diff line number Diff line change
Expand Up @@ -2598,9 +2598,6 @@ S: Ucitelska 1576
S: Prague 8
S: 182 00 Czech Republic

N: Rick Payne
D: RFC2385 Support for TCP

N: Barak A. Pearlmutter
E: bap@cs.unm.edu
W: http://www.cs.unm.edu/~bap/
Expand Down
2 changes: 2 additions & 0 deletions trunk/Documentation/00-INDEX
Original file line number Diff line number Diff line change
Expand Up @@ -104,6 +104,8 @@ firmware_class/
- request_firmware() hotplug interface info.
floppy.txt
- notes and driver options for the floppy disk driver.
ftape.txt
- notes about the floppy tape device driver.
hayes-esp.txt
- info on using the Hayes ESP serial driver.
highuid.txt
Expand Down
38 changes: 32 additions & 6 deletions trunk/Documentation/feature-removal-schedule.txt
Original file line number Diff line number Diff line change
Expand Up @@ -70,6 +70,18 @@ Who: Dominik Brodowski <linux@brodo.de>

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

What: ip_queue and ip6_queue (old ipv4-only and ipv6-only netfilter queue)
When: December 2005
Why: This interface has been obsoleted by the new layer3-independent
"nfnetlink_queue". The Kernel interface is compatible, so the old
ip[6]tables "QUEUE" targets still work and will transparently handle
all packets into nfnetlink queue number 0. Userspace users will have
to link against API-compatible library on top of libnfnetlink_queue
instead of the current 'libipq'.
Who: Harald Welte <laforge@netfilter.org>

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

What: remove EXPORT_SYMBOL(kernel_thread)
When: August 2006
Files: arch/*/kernel/*_ksyms.c
Expand Down Expand Up @@ -215,6 +227,21 @@ Who: Patrick McHardy <kaber@trash.net>

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

What: frame diverter
When: November 2006
Why: The frame diverter is included in most distribution kernels, but is
broken. It does not correctly handle many things:
- IPV6
- non-linear skb's
- network device RCU on removal
- input frames not correctly checked for protocol errors
It also adds allocation overhead even if not enabled.
It is not clear if anyone is still using it.
Who: Stephen Hemminger <shemminger@osdl.org>

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


What: PHYSDEVPATH, PHYSDEVBUS, PHYSDEVDRIVER in the uevent environment
When: October 2008
Why: The stacking of class devices makes these values misleading and
Expand All @@ -234,11 +261,10 @@ Who: Jean Delvare <khali@linux-fr.org>

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

What: IPv4 only connection tracking/NAT/helpers
When: 2.6.22
Why: The new layer 3 independant connection tracking replaces the old
IPv4 only version. After some stabilization of the new code the
old one will be removed.
Who: Patrick McHardy <kaber@trash.net>
What: ftape
When: 2.6.20
Why: Orphaned for ages. SMP bugs long unfixed. Few users left
in the world.
Who: Jeff Garzik <jeff@garzik.org>

---------------------------
307 changes: 307 additions & 0 deletions trunk/Documentation/ftape.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,307 @@
Intro
=====

This file describes some issues involved when using the "ftape"
floppy tape device driver that comes with the Linux kernel.

ftape has a home page at

http://ftape.dot-heine.de/

which contains further information about ftape. Please cross check
this WWW address against the address given (if any) in the MAINTAINERS
file located in the top level directory of the Linux kernel source
tree.

NOTE: This is an unmaintained set of drivers, and it is not guaranteed to work.
If you are interested in taking over maintenance, contact Claus-Justus Heine
<ch@dot-heine.de>, the former maintainer.

Contents
========

A minus 1: Ftape documentation

A. Changes
1. Goal
2. I/O Block Size
3. Write Access when not at EOD (End Of Data) or BOT (Begin Of Tape)
4. Formatting
5. Interchanging cartridges with other operating systems

B. Debugging Output
1. Introduction
2. Tuning the debugging output

C. Boot and load time configuration
1. Setting boot time parameters
2. Module load time parameters
3. Ftape boot- and load time options
4. Example kernel parameter setting
5. Example module parameter setting

D. Support and contacts

*******************************************************************************

A minus 1. Ftape documentation
==============================

Unluckily, the ftape-HOWTO is out of date. This really needs to be
changed. Up to date documentation as well as recent development
versions of ftape and useful links to related topics can be found at
the ftape home page at

http://ftape.dot-heine.de/

*******************************************************************************

A. Changes
==========

1. Goal
~~~~
The goal of all that incompatibilities was to give ftape an interface
that resembles the interface provided by SCSI tape drives as close
as possible. Thus any Unix backup program that is known to work
with SCSI tape drives should also work.

The concept of a fixed block size for read/write transfers is
rather unrelated to this SCSI tape compatibility at the file system
interface level. It developed out of a feature of zftape, a
block wise user transparent on-the-fly compression. That compression
support will not be dropped in future releases for compatibility
reasons with previous releases of zftape.

2. I/O Block Size
~~~~~~~~~~~~~~
The block size defaults to 10k which is the default block size of
GNU tar.

The block size can be tuned either during kernel configuration or
at runtime with the MTIOCTOP ioctl using the MTSETBLK operation
(i.e. do "mt -f /dev/qft0" setblk #BLKSZ). A block size of 0
switches to variable block size mode i.e. "mt setblk 0" switches
off the block size restriction. However, this disables zftape's
built in on-the-fly compression which doesn't work with variable
block size mode.

The BLKSZ parameter must be given as a byte count and must be a
multiple of 32k or 0, i.e. use "mt setblk 32768" to switch to a
block size of 32k.

The typical symptom of a block size mismatch is an "invalid
argument" error message.

3. Write Access when not at EOD (End Of Data) or BOT (Begin Of Tape)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
zftape (the file system interface of ftape-3.x) denies write access
to the tape cartridge when it isn't positioned either at BOT or
EOD.

4. Formatting
~~~~~~~~~~
ftape DOES support formatting of floppy tape cartridges. You need the
`ftformat' program that is shipped with the modules version of ftape.
Please get the latest version of ftape from

ftp://sunsite.unc.edu/pub/Linux/kernel/tapes

or from the ftape home page at

http://ftape.dot-heine.de/

`ftformat' is contained in the `./contrib/' subdirectory of that
separate ftape package.

5. Interchanging cartridges with other operating systems
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The internal emulation of Unix tape device file marks has changed
completely. ftape now uses the volume table segment as specified
by the QIC-40/80/3010/3020/113 standards to emulate file marks. As
a consequence there is limited support to interchange cartridges
with other operating systems.

To be more precise: ftape will detect volumes written by other OS's
programs and other OS's programs will detect volumes written by
ftape.

However, it isn't possible to extract the data dumped to the tape
by some MSDOS program with ftape. This exceeds the scope of a
kernel device driver. If you need such functionality, then go ahead
and write a user space utility that is able to do that. ftape already
provides all kernel level support necessary to do that.

*******************************************************************************

B. Debugging Output
================

1. Introduction
~~~~~~~~~~~~
The ftape driver can be very noisy in that is can print lots of
debugging messages to the kernel log files and the system console.
While this is useful for debugging it might be annoying during
normal use and enlarges the size of the driver by several kilobytes.

To reduce the size of the driver you can trim the maximal amount of
debugging information available during kernel configuration. Please
refer to the kernel configuration script and its on-line help
functionality.

The amount of debugging output maps to the "tracing" boot time
option and the "ft_tracing" modules option as follows:

0 bugs
1 + errors (with call-stack dump)
2 + warnings
3 + information
4 + more information
5 + program flow
6 + fdc/dma info
7 + data flow
8 + everything else

2. Tuning the debugging output
~~~~~~~~~~~~~~~~~~~~~~~~~~~
To reduce the amount of debugging output printed to the system
console you can

i) trim the debugging output at run-time with

mt -f /dev/nqft0 setdensity #DBGLVL

where "#DBGLVL" is a number between 0 and 9

ii) trim the debugging output at module load time with

modprobe ftape ft_tracing=#DBGLVL

Of course, this applies only if you have configured ftape to be
compiled as a module.

iii) trim the debugging output during system boot time. Add the
following to the kernel command line:

ftape=#DBGLVL,tracing

Please refer also to the next section if you don't know how to
set boot time parameters.

*******************************************************************************

C. Boot and load time configuration
================================

1. Setting boot time parameters
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Assuming that you use lilo, the LI)nux LO)ader, boot time kernel
parameters can be set by adding a line

append some_kernel_boot_time_parameter

to `/etc/lilo.conf' or at real boot time by typing in the options
at the prompt provided by LILO. I can't give you advice on how to
specify those parameters with other loaders as I don't use them.

For ftape, each "some_kernel_boot_time_parameter" looks like
"ftape=value,option". As an example, the debugging output can be
increased with

ftape=4,tracing

NOTE: the value precedes the option name.

2. Module load time parameters
~~~~~~~~~~~~~~~~~~~~~~~~~~~
Module parameters can be specified either directly when invoking
the program 'modprobe' at the shell prompt:

modprobe ftape ft_tracing=4

or by editing the file `/etc/modprobe.conf' in which case they take
effect each time when the module is loaded with `modprobe' (please
refer to the respective manual pages). Thus, you should add a line

options ftape ft_tracing=4

to `/etc/modprobe.conf` if you intend to increase the debugging
output of the driver.


3. Ftape boot- and load time options
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

i. Controlling the amount of debugging output
DBGLVL has to be replaced by a number between 0 and 8.

module | kernel command line
-----------------------|----------------------
ft_tracing=DBGLVL | ftape=DBGLVL,tracing

ii. Hardware setup
BASE is the base address of your floppy disk controller,
IRQ and DMA give its interrupt and DMA channel, respectively.
BOOL is an integer, "0" means "no"; any other value means
"yes". You don't need to specify anything if connecting your tape
drive to the standard floppy disk controller. All of these
values have reasonable defaults. The defaults can be modified
during kernel configuration, i.e. while running "make config",
"make menuconfig" or "make xconfig" in the top level directory
of the Linux kernel source tree. Please refer also to the on
line documentation provided during that kernel configuration
process.

ft_probe_fc10 is set to a non-zero value if you wish for ftape to
probe for a Colorado FC-10 or FC-20 controller.

ft_mach2 is set to a non-zero value if you wish for ftape to probe
for a Mountain MACH-2 controller.

module | kernel command line
-----------------------|----------------------
ft_fdc_base=BASE | ftape=BASE,ioport
ft_fdc_irq=IRQ | ftape=IRQ,irq
ft_fdc_dma=DMA | ftape=DMA,dma
ft_probe_fc10=BOOL | ftape=BOOL,fc10
ft_mach2=BOOL | ftape=BOOL,mach2
ft_fdc_threshold=THR | ftape=THR,threshold
ft_fdc_rate_limit=RATE | ftape=RATE,datarate

4. Example kernel parameter setting
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To configure ftape to probe for a Colorado FC-10/FC-20 controller
and to increase the amount of debugging output a little bit, add
the following line to `/etc/lilo.conf':

append ftape=1,fc10 ftape=4,tracing

5. Example module parameter setting
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To do the same, but with ftape compiled as a loadable kernel
module, add the following line to `/etc/modprobe.conf':

options ftape ft_probe_fc10=1 ft_tracing=4

*******************************************************************************

D. Support and contacts
====================

Ftape is distributed under the GNU General Public License. There is
absolutely no warranty for this software. However, you can reach
the current maintainer of the ftape package under the email address
given in the MAINTAINERS file which is located in the top level
directory of the Linux kernel source tree. There you'll find also
the relevant mailing list to use as a discussion forum and the web
page to query for the most recent documentation, related work and
development versions of ftape.

Changelog:
==========

~1996: Original Document

10-24-2004: General cleanup and updating, noting additional module options.
James Nelson <james4765@gmail.com>
3 changes: 3 additions & 0 deletions trunk/Documentation/kernel-parameters.txt
Original file line number Diff line number Diff line change
Expand Up @@ -557,6 +557,9 @@ and is between 256 and 4096 characters. It is defined in the file
floppy= [HW]
See Documentation/floppy.txt.

ftape= [HW] Floppy Tape subsystem debugging options.
See Documentation/ftape.txt.

gamecon.map[2|3]=
[HW,JOY] Multisystem joystick and NES/SNES/PSX pad
support via parallel port (up to 5 devices per port)
Expand Down
Loading

0 comments on commit 87986cb

Please sign in to comment.