openkylin
/
linux
mirror of https://gitee.com/openkylin/linux.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity

/spare/repo/libata-dev branch 'v2.6.13'

This commit is contained in:
Jeff Garzik 2005-08-29 16:12:36 -04:00
parent ff40c6d3d1 02b3e4e2d7
commit 2fca877b68
4486 changed files with 332650 additions and 143058 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
CREDITS
View File

@ -1624,10 +1624,10 @@ E: ajoshi@shell.unixbox.com
D: fbdev hacking
N: Jesper Juhl
E: juhl-lkml@dif.dk
D: Various small janitor fixes, cleanups etc.
E: jesper.juhl@gmail.com
D: Various fixes, cleanups and minor features.
S: Lemnosvej 1, 3.tv
S: 2300 Copenhagen S
S: 2300 Copenhagen S.
S: Denmark
N: Jozsef Kadlecsik
@ -2380,9 +2380,10 @@ E: tmolina@cablespeed.com
D: bug fixes, documentation, minor hackery
N: James Morris
E: jmorris@intercode.com.au
W: http://www.intercode.com.au/jmorris/
D: Netfilter, Linux Security Modules (LSM).
E: jmorris@namei.org
W: http://namei.org/
D: Netfilter, Linux Security Modules (LSM), SELinux, IPSec,
D: Crypto API, general networking, miscellaneous.
S: PO Box 707
S: Spit Junction NSW 2088
S: Australia
@ -2422,8 +2423,7 @@ S: Toronto, Ontario
S: Canada
N: Zwane Mwaikambo
E: zwane@linuxpower.ca
W: http://function.linuxpower.ca
E: zwane@arm.linux.org.uk
D: Various driver hacking
D: Lowlevel x86 kernel hacking
D: General debugging

2
Documentation/00-INDEX
View File

@ -138,6 +138,8 @@ java.txt
- info on the in-kernel binary support for Java(tm).
kbuild/
- directory with info about the kernel build process.
kdumpt.txt
- mini HowTo on getting the crash dump code to work.
kernel-doc-nano-HOWTO.txt
- mini HowTo on generation and location of kernel documentation files.
kernel-docs.txt

27
Documentation/Changes
View File

@ -44,9 +44,9 @@ running, the suggested command should tell you.
Again, keep in mind that this list assumes you are already
functionally running a Linux 2.4 kernel. Also, not all tools are
necessary on all systems; obviously, if you don't have any PCMCIA (PC
Card) hardware, for example, you probably needn't concern yourself
with pcmcia-cs.
necessary on all systems; obviously, if you don't have any ISDN
hardware, for example, you probably needn't concern yourself with
isdn4k-utils.
o Gnu C 2.95.3 # gcc --version
o Gnu make 3.79.1 # make --version
@ -57,13 +57,15 @@ o e2fsprogs 1.29 # tune2fs
o jfsutils 1.1.3 # fsck.jfs -V
o reiserfsprogs 3.6.3 # reiserfsck -V 2>&1|grep reiserfsprogs
o xfsprogs 2.6.0 # xfs_db -V
o pcmciautils 004
o pcmcia-cs 3.1.21 # cardmgr -V
o quota-tools 3.09 # quota -V
o PPP 2.4.0 # pppd --version
o isdn4k-utils 3.1pre1 # isdnctrl 2>&1|grep version
o nfs-utils 1.0.5 # showmount --version
o procps 3.2.0 # ps --version
o oprofile 0.5.3 # oprofiled --version
o oprofile 0.9 # oprofiled --version
o udev 058 # udevinfo -V
Kernel compilation
==================
@ -186,13 +188,20 @@ architecture independent and any version from 2.0.0 onward should
work correctly with this version of the XFS kernel code (2.6.0 or
later is recommended, due to some significant improvements).
PCMCIAutils
-----------
PCMCIAutils replaces pcmcia-cs (see below). It properly sets up
PCMCIA sockets at system startup and loads the appropriate modules
for 16-bit PCMCIA devices if the kernel is modularized and the hotplug
subsystem is used.
Pcmcia-cs
---------
PCMCIA (PC Card) support is now partially implemented in the main
kernel source. Pay attention when you recompile your kernel ;-).
Also, be sure to upgrade to the latest pcmcia-cs release.
kernel source. The "pcmciautils" package (see above) replaces pcmcia-cs
for newest kernels.
Quota-tools
-----------
@ -349,9 +358,13 @@ Xfsprogs
--------
o <ftp://oss.sgi.com/projects/xfs/download/>
Pcmciautils
-----------
o <ftp://ftp.kernel.org/pub/linux/utils/kernel/pcmcia/>
Pcmcia-cs
---------
o <ftp://pcmcia-cs.sourceforge.net/pub/pcmcia-cs/pcmcia-cs-3.1.21.tar.gz>
o <http://pcmcia-cs.sourceforge.net/>
Quota-tools
----------

2
Documentation/DocBook/Makefile
View File

@ -49,7 +49,7 @@ installmandocs: mandocs
KERNELDOC = scripts/kernel-doc
DOCPROC = scripts/basic/docproc
XMLTOFLAGS = -m Documentation/DocBook/stylesheet.xsl
XMLTOFLAGS = -m $(srctree)/Documentation/DocBook/stylesheet.xsl
#XMLTOFLAGS += --skip-validation
###

2
Documentation/DocBook/kernel-api.tmpl
View File

@ -266,7 +266,7 @@ X!Ekernel/module.c
<chapter id="hardware">
<title>Hardware Interfaces</title>
<sect1><title>Interrupt Handling</title>
!Iarch/i386/kernel/irq.c
!Ikernel/irq/manage.c
</sect1>
<sect1><title>Resources Management</title>

96
Documentation/DocBook/libata.tmpl
View File

@ -84,6 +84,14 @@ void (*port_disable) (struct ata_port *);
Called from ata_bus_probe() and ata_bus_reset() error paths,
as well as when unregistering from the SCSI module (rmmod, hot
unplug).
This function should do whatever needs to be done to take the
port out of use. In most cases, ata_port_disable() can be used
as this hook.
</para>
<para>
Called from ata_bus_probe() on a failed probe.
Called from ata_bus_reset() on a failed bus reset.
Called from ata_scsi_release().
</para>
</sect2>
@ -98,6 +106,13 @@ void (*dev_config) (struct ata_port *, struct ata_device *);
found. Typically used to apply device-specific fixups prior to
issue of SET FEATURES - XFER MODE, and prior to operation.
</para>
<para>
Called by ata_device_add() after ata_dev_identify() determines
a device is present.
</para>
<para>
This entry may be specified as NULL in ata_port_operations.
</para>
</sect2>
@ -135,6 +150,8 @@ void (*tf_read) (struct ata_port *ap, struct ata_taskfile *tf);
registers / DMA buffers. ->tf_read() is called to read the
hardware registers / DMA buffers, to obtain the current set of
taskfile register values.
Most drivers for taskfile-based hardware (PIO or MMIO) use
ata_tf_load() and ata_tf_read() for these hooks.
</para>
</sect2>
@ -147,6 +164,8 @@ void (*exec_command)(struct ata_port *ap, struct ata_taskfile *tf);
<para>
causes an ATA command, previously loaded with
->tf_load(), to be initiated in hardware.
Most drivers for taskfile-based hardware use ata_exec_command()
for this hook.
</para>
</sect2>
@ -161,6 +180,10 @@ Allow low-level driver to filter ATA PACKET commands, returning a status
indicating whether or not it is OK to use DMA for the supplied PACKET
command.
</para>
<para>
This hook may be specified as NULL, in which case libata will
assume that atapi dma can be supported.
</para>
</sect2>
@ -175,6 +198,14 @@ u8 (*check_err)(struct ata_port *ap);
Reads the Status/AltStatus/Error ATA shadow register from
hardware. On some hardware, reading the Status register has
the side effect of clearing the interrupt condition.
Most drivers for taskfile-based hardware use
ata_check_status() for this hook.
</para>
<para>
Note that because this is called from ata_device_add(), at
least a dummy function that clears device interrupts must be
provided for all drivers, even if the controller doesn't
actually have a taskfile status register.
</para>
</sect2>
@ -188,7 +219,13 @@ void (*dev_select)(struct ata_port *ap, unsigned int device);
Issues the low-level hardware command(s) that causes one of N
hardware devices to be considered 'selected' (active and
available for use) on the ATA bus. This generally has no
meaning on FIS-based devices.
meaning on FIS-based devices.
</para>
<para>
Most drivers for taskfile-based hardware use
ata_std_dev_select() for this hook. Controllers which do not
support second drives on a port (such as SATA contollers) will
use ata_noop_dev_select().
</para>
</sect2>
@ -204,6 +241,8 @@ void (*phy_reset) (struct ata_port *ap);
for device presence (PATA and SATA), typically a soft reset
(SRST) will be performed. Drivers typically use the helper
functions ata_bus_reset() or sata_phy_reset() for this hook.
Many SATA drivers use sata_phy_reset() or call it from within
their own phy_reset() functions.
</para>
</sect2>
@ -227,6 +266,25 @@ PCI IDE DMA Status register.
These hooks are typically either no-ops, or simply not implemented, in
FIS-based drivers.
</para>
<para>
Most legacy IDE drivers use ata_bmdma_setup() for the bmdma_setup()
hook. ata_bmdma_setup() will write the pointer to the PRD table to
the IDE PRD Table Address register, enable DMA in the DMA Command
register, and call exec_command() to begin the transfer.
</para>
<para>
Most legacy IDE drivers use ata_bmdma_start() for the bmdma_start()
hook. ata_bmdma_start() will write the ATA_DMA_START flag to the DMA
Command register.
</para>
<para>
Many legacy IDE drivers use ata_bmdma_stop() for the bmdma_stop()
hook. ata_bmdma_stop() clears the ATA_DMA_START flag in the DMA
command register.
</para>
<para>
Many legacy IDE drivers use ata_bmdma_status() as the bmdma_status() hook.
</para>
</sect2>
@ -250,6 +308,10 @@ int (*qc_issue) (struct ata_queued_cmd *qc);
helper function ata_qc_issue_prot() for taskfile protocol-based
dispatch. More advanced drivers implement their own ->qc_issue.
</para>
<para>
ata_qc_issue_prot() calls ->tf_load(), ->bmdma_setup(), and
->bmdma_start() as necessary to initiate a transfer.
</para>
</sect2>
@ -279,6 +341,21 @@ void (*irq_clear) (struct ata_port *);
before the interrupt handler is registered, to be sure hardware
is quiet.
</para>
<para>
The second argument, dev_instance, should be cast to a pointer
to struct ata_host_set.
</para>
<para>
Most legacy IDE drivers use ata_interrupt() for the
irq_handler hook, which scans all ports in the host_set,
determines which queued command was active (if any), and calls
ata_host_intr(ap,qc).
</para>
<para>
Most legacy IDE drivers use ata_bmdma_irq_clear() for the
irq_clear() hook, which simply clears the interrupt and error
flags in the DMA status register.
</para>
</sect2>
@ -292,6 +369,7 @@ void (*scr_write) (struct ata_port *ap, unsigned int sc_reg,
<para>
Read and write standard SATA phy registers. Currently only used
if ->phy_reset hook called the sata_phy_reset() helper function.
sc_reg is one of SCR_STATUS, SCR_CONTROL, SCR_ERROR, or SCR_ACTIVE.
</para>
</sect2>
@ -307,17 +385,29 @@ void (*host_stop) (struct ata_host_set *host_set);
->port_start() is called just after the data structures for each
port are initialized. Typically this is used to alloc per-port
DMA buffers / tables / rings, enable DMA engines, and similar
tasks.
tasks. Some drivers also use this entry point as a chance to
allocate driver-private memory for ap->private_data.
</para>
<para>
Many drivers use ata_port_start() as this hook or call
it from their own port_start() hooks. ata_port_start()
allocates space for a legacy IDE PRD table and returns.
</para>
<para>
->port_stop() is called after ->host_stop(). It's sole function
is to release DMA/memory resources, now that they are no longer
actively being used.
actively being used. Many drivers also free driver-private
data from port at this time.
</para>
<para>
Many drivers use ata_port_stop() as this hook, which frees the
PRD table.
</para>
<para>
->host_stop() is called after all ->port_stop() calls
have completed. The hook must finalize hardware shutdown, release DMA
and other resources, etc.
This hook may be specified as NULL, in which case it is not called.
</para>
</sect2>

1
Documentation/DocBook/stylesheet.xsl
View File

@ -2,4 +2,5 @@
<stylesheet xmlns="http://www.w3.org/1999/XSL/Transform" version="1.0">
<param name="chunk.quietly">1</param>
<param name="funcsynopsis.style">ansi</param>
<param name="funcsynopsis.tabular.threshold">80</param>
</stylesheet>

144
Documentation/IPMI.txt
View File

@ -25,9 +25,10 @@ subject and I can't cover it all here!
Configuration
-------------
The LinuxIPMI driver is modular, which means you have to pick several
The Linux IPMI driver is modular, which means you have to pick several
things to have it work right depending on your hardware. Most of
these are available in the 'Character Devices' menu.
these are available in the 'Character Devices' menu then the IPMI
menu.
No matter what, you must pick 'IPMI top-level message handler' to use
IPMI. What you do beyond that depends on your needs and hardware.
@ -35,33 +36,30 @@ IPMI. What you do beyond that depends on your needs and hardware.
The message handler does not provide any user-level interfaces.
Kernel code (like the watchdog) can still use it. If you need access
from userland, you need to select 'Device interface for IPMI' if you
want access through a device driver. Another interface is also
available, you may select 'IPMI sockets' in the 'Networking Support'
main menu. This provides a socket interface to IPMI. You may select
both of these at the same time, they will both work together.
want access through a device driver.
The driver interface depends on your hardware. If you have a board
with a standard interface (These will generally be either "KCS",
"SMIC", or "BT", consult your hardware manual), choose the 'IPMI SI
handler' option. A driver also exists for direct I2C access to the
IPMI management controller. Some boards support this, but it is
unknown if it will work on every board. For this, choose 'IPMI SMBus
handler', but be ready to try to do some figuring to see if it will
work.
There is also a KCS-only driver interface supplied, but it is
depracated in favor of the SI interface.
The driver interface depends on your hardware. If your system
properly provides the SMBIOS info for IPMI, the driver will detect it
and just work. If you have a board with a standard interface (These
will generally be either "KCS", "SMIC", or "BT", consult your hardware
manual), choose the 'IPMI SI handler' option. A driver also exists
for direct I2C access to the IPMI management controller. Some boards
support this, but it is unknown if it will work on every board. For
this, choose 'IPMI SMBus handler', but be ready to try to do some
figuring to see if it will work on your system if the SMBIOS/APCI
information is wrong or not present. It is fairly safe to have both
these enabled and let the drivers auto-detect what is present.
You should generally enable ACPI on your system, as systems with IPMI
should have ACPI tables describing them.
can have ACPI tables describing them.
If you have a standard interface and the board manufacturer has done
their job correctly, the IPMI controller should be automatically
detect (via ACPI or SMBIOS tables) and should just work. Sadly, many
boards do not have this information. The driver attempts standard
defaults, but they may not work. If you fall into this situation, you
need to read the section below named 'The SI Driver' on how to
hand-configure your system.
detected (via ACPI or SMBIOS tables) and should just work. Sadly,
many boards do not have this information. The driver attempts
standard defaults, but they may not work. If you fall into this
situation, you need to read the section below named 'The SI Driver' or
"The SMBus Driver" on how to hand-configure your system.
IPMI defines a standard watchdog timer. You can enable this with the
'IPMI Watchdog Timer' config option. If you compile the driver into
@ -73,6 +71,18 @@ closed (by default it is disabled on close). Go into the 'Watchdog
Cards' menu, enable 'Watchdog Timer Support', and enable the option
'Disable watchdog shutdown on close'.
IPMI systems can often be powered off using IPMI commands. Select
'IPMI Poweroff' to do this. The driver will auto-detect if the system
can be powered off by IPMI. It is safe to enable this even if your
system doesn't support this option. This works on ATCA systems, the
Radisys CPI1 card, and any IPMI system that supports standard chassis
management commands.
If you want the driver to put an event into the event log on a panic,
enable the 'Generate a panic event to all BMCs on a panic' option. If
you want the whole panic string put into the event log using OEM
events, enable the 'Generate OEM events containing the panic string'
option.
Basic Design
------------
@ -80,7 +90,7 @@ Basic Design
The Linux IPMI driver is designed to be very modular and flexible, you
only need to take the pieces you need and you can use it in many
different ways. Because of that, it's broken into many chunks of
code. These chunks are:
code. These chunks (by module name) are:
ipmi_msghandler - This is the central piece of software for the IPMI
system. It handles all messages, message timing, and responses. The
@ -93,18 +103,26 @@ ipmi_devintf - This provides a userland IOCTL interface for the IPMI
driver, each open file for this device ties in to the message handler
as an IPMI user.
ipmi_si - A driver for various system interfaces. This supports
KCS, SMIC, and may support BT in the future. Unless you have your own
custom interface, you probably need to use this.
ipmi_si - A driver for various system interfaces. This supports KCS,
SMIC, and BT interfaces. Unless you have an SMBus interface or your
own custom interface, you probably need to use this.
ipmi_smb - A driver for accessing BMCs on the SMBus. It uses the
I2C kernel driver's SMBus interfaces to send and receive IPMI messages
over the SMBus.
af_ipmi - A network socket interface to IPMI. This doesn't take up
a character device in your system.
ipmi_watchdog - IPMI requires systems to have a very capable watchdog
timer. This driver implements the standard Linux watchdog timer
interface on top of the IPMI message handler.
Note that the KCS-only interface ahs been removed.
ipmi_poweroff - Some systems support the ability to be turned off via
IPMI commands.
These are all individually selectable via configuration options.
Note that the KCS-only interface has been removed. The af_ipmi driver
is no longer supported and has been removed because it was impossible
to do 32 bit emulation on 64-bit kernels with it.
Much documentation for the interface is in the include files. The
IPMI include files are:
@ -424,7 +442,7 @@ at module load time (for a module) with:
modprobe ipmi_smb.o
addr=<adapter1>,<i2caddr1>[,<adapter2>,<i2caddr2>[,...]]
dbg=<flags1>,<flags2>...
[defaultprobe=0] [dbg_probe=1]
[defaultprobe=1] [dbg_probe=1]
The addresses are specified in pairs, the first is the adapter ID and the
second is the I2C address on that adapter.
@ -532,3 +550,67 @@ Once you open the watchdog timer, you must write a 'V' character to the
device to close it, or the timer will not stop. This is a new semantic
for the driver, but makes it consistent with the rest of the watchdog
drivers in Linux.
Panic Timeouts
--------------
The OpenIPMI driver supports the ability to put semi-custom and custom
events in the system event log if a panic occurs. if you enable the
'Generate a panic event to all BMCs on a panic' option, you will get
one event on a panic in a standard IPMI event format. If you enable
the 'Generate OEM events containing the panic string' option, you will
also get a bunch of OEM events holding the panic string.
The field settings of the events are:
* Generator ID: 0x21 (kernel)
* EvM Rev: 0x03 (this event is formatting in IPMI 1.0 format)
* Sensor Type: 0x20 (OS critical stop sensor)
* Sensor #: The first byte of the panic string (0 if no panic string)
* Event Dir | Event Type: 0x6f (Assertion, sensor-specific event info)
* Event Data 1: 0xa1 (Runtime stop in OEM bytes 2 and 3)
* Event data 2: second byte of panic string
* Event data 3: third byte of panic string
See the IPMI spec for the details of the event layout. This event is
always sent to the local management controller. It will handle routing
the message to the right place
Other OEM events have the following format:
Record ID (bytes 0-1): Set by the SEL.
Record type (byte 2): 0xf0 (OEM non-timestamped)
byte 3: The slave address of the card saving the panic
byte 4: A sequence number (starting at zero)
The rest of the bytes (11 bytes) are the panic string. If the panic string
is longer than 11 bytes, multiple messages will be sent with increasing
sequence numbers.
Because you cannot send OEM events using the standard interface, this
function will attempt to find an SEL and add the events there. It
will first query the capabilities of the local management controller.
If it has an SEL, then they will be stored in the SEL of the local
management controller. If not, and the local management controller is
an event generator, the event receiver from the local management
controller will be queried and the events sent to the SEL on that
device. Otherwise, the events go nowhere since there is nowhere to
send them.
Poweroff
--------
If the poweroff capability is selected, the IPMI driver will install
a shutdown function into the standard poweroff function pointer. This
is in the ipmi_poweroff module. When the system requests a powerdown,
it will send the proper IPMI commands to do this. This is supported on
several platforms.
There is a module parameter named "poweroff_control" that may either be zero
(do a power down) or 2 (do a power cycle, power the system off, then power
it on in a few seconds). Setting ipmi_poweroff.poweroff_control=x will do
the same thing on the kernel command line. The parameter is also available
via the proc filesystem in /proc/ipmi/poweroff_control. Note that if the
system does not support power cycling, it will always to the power off.
Note that if you have ACPI enabled, the system will prefer using ACPI to
power off.

14
Documentation/SubmittingDrivers
View File

@ -13,13 +13,14 @@ Allocating Device Numbers
-------------------------
Major and minor numbers for block and character devices are allocated
by the Linux assigned name and number authority (currently better
known as H Peter Anvin). The site is http://www.lanana.org/. This
by the Linux assigned name and number authority (currently this is
Torben Mathiasen). The site is http://www.lanana.org/. This
also deals with allocating numbers for devices that are not going to
be submitted to the mainstream kernel.
See Documentation/devices.txt for more information on this.
If you don't use assigned numbers then when you device is submitted it will
get given an assigned number even if that is different from values you may
If you don't use assigned numbers then when your device is submitted it will
be given an assigned number even if that is different from values you may
have shipped to customers before.
Who To Submit Drivers To
@ -32,7 +33,8 @@ Linux 2.2:
If the code area has a general maintainer then please submit it to
the maintainer listed in MAINTAINERS in the kernel file. If the
maintainer does not respond or you cannot find the appropriate
maintainer then please contact Alan Cox <alan@lxorguk.ukuu.org.uk>
maintainer then please contact the 2.2 kernel maintainer:
Marc-Christian Petersen <m.c.p@wolk-project.de>.
Linux 2.4:
The same rules apply as 2.2. The final contact point for Linux 2.4
@ -48,7 +50,7 @@ What Criteria Determine Acceptance
Licensing: The code must be released to us under the
GNU General Public License. We don't insist on any kind
of exclusively GPL licensing, and if you wish the driver
of exclusive GPL licensing, and if you wish the driver
to be useful to other communities such as BSD you may well
wish to release under multiple licenses.

66
Documentation/SubmittingPatches
View File

@ -35,7 +35,7 @@ not in any lower subdirectory.
To create a patch for a single file, it is often sufficient to do:
SRCTREE= linux-2.4
SRCTREE= linux-2.6
MYFILE= drivers/net/mydriver.c
cd $SRCTREE
@ -48,17 +48,18 @@ To create a patch for multiple files, you should unpack a "vanilla",
or unmodified kernel source tree, and generate a diff against your
own source tree. For example:
MYSRC= /devel/linux-2.4
MYSRC= /devel/linux-2.6
tar xvfz linux-2.4.0-test11.tar.gz
mv linux linux-vanilla
wget http://www.moses.uklinux.net/patches/dontdiff
diff -uprN -X dontdiff linux-vanilla $MYSRC > /tmp/patch
rm -f dontdiff
tar xvfz linux-2.6.12.tar.gz
mv linux-2.6.12 linux-2.6.12-vanilla
diff -uprN -X linux-2.6.12-vanilla/Documentation/dontdiff \
linux-2.6.12-vanilla $MYSRC > /tmp/patch
"dontdiff" is a list of files which are generated by the kernel during
the build process, and should be ignored in any diff(1)-generated
patch. dontdiff is maintained by Tigran Aivazian <tigran@veritas.com>
patch. The "dontdiff" file is included in the kernel tree in
2.6.12 and later. For earlier kernel versions, you can get it
from <http://www.xenotime.net/linux/doc/dontdiff>.
Make sure your patch does not include any extra files which do not
belong in a patch submission. Make sure to review your patch -after-
@ -66,18 +67,20 @@ generated it with diff(1), to ensure accuracy.
If your changes produce a lot of deltas, you may want to look into
splitting them into individual patches which modify things in
logical stages, this will facilitate easier reviewing by other
logical stages. This will facilitate easier reviewing by other
kernel developers, very important if you want your patch accepted.
There are a number of scripts which can aid in this;
There are a number of scripts which can aid in this:
Quilt:
http://savannah.nongnu.org/projects/quilt
Randy Dunlap's patch scripts:
http://developer.osdl.org/rddunlap/scripts/patching-scripts.tgz
http://www.xenotime.net/linux/scripts/patching-scripts-002.tar.gz
Andrew Morton's patch scripts:
http://www.zip.com.au/~akpm/linux/patches/patch-scripts-0.16
http://www.zip.com.au/~akpm/linux/patches/patch-scripts-0.20
2) Describe your changes.
@ -132,21 +135,6 @@ which require discussion or do not have a clear advantage should
usually be sent first to linux-kernel. Only after the patch is
discussed should the patch then be submitted to Linus.
For small patches you may want to CC the Trivial Patch Monkey
trivial@rustcorp.com.au set up by Rusty Russell; which collects "trivial"
patches. Trivial patches must qualify for one of the following rules:
Spelling fixes in documentation
Spelling fixes which could break grep(1).
Warning fixes (cluttering with useless warnings is bad)
Compilation fixes (only if they are actually correct)
Runtime fixes (only if they actually fix things)
Removing use of deprecated functions/macros (eg. check_region).
Contact detail and documentation fixes
Non-portable code replaced by portable code (even in arch-specific,
since people copy, as long as it's trivial)
Any fix by the author/maintainer of the file. (ie. patch monkey
in re-transmission mode)
5) Select your CC (e-mail carbon copy) list.
@ -161,6 +149,11 @@ USB, framebuffer devices, the VFS, the SCSI subsystem, etc. See the
MAINTAINERS file for a mailing list that relates specifically to
your change.
If changes affect userland-kernel interfaces, please send
the MAN-PAGES maintainer (as listed in the MAINTAINERS file)
a man-pages patch, or at least a notification of the change,
so that some information makes its way into the manual pages.
Even if the maintainer did not respond in step #4, make sure to ALWAYS
copy the maintainer when you change their code.
@ -178,6 +171,8 @@ patches. Trivial patches must qualify for one of the following rules:
since people copy, as long as it's trivial)
Any fix by the author/maintainer of the file. (ie. patch monkey
in re-transmission mode)
URL: <http://www.kernel.org/pub/linux/kernel/people/rusty/trivial/>
@ -299,13 +294,24 @@ can certify the below:
then you just add a line saying
Signed-off-by: Random J Developer <random@developer.org>
Signed-off-by: Random J Developer <random@developer.example.org>
Some people also put extra tags at the end. They'll just be ignored for
now, but you can do this to mark internal company procedures or just
point out some special detail about the sign-off.
12) More references for submitting patches
Andrew Morton, "The perfect patch" (tpp).
<http://www.zip.com.au/~akpm/linux/patches/stuff/tpp.txt>
Jeff Garzik, "Linux kernel patch submission format."
<http://linux.yyz.us/patch-format.html>
-----------------------------------
SECTION 2 - HINTS, TIPS, AND TRICKS
-----------------------------------
@ -374,7 +380,5 @@ and 'extern __inline__'.
4) Don't over-design.
Don't try to anticipate nebulous future cases which may or may not
be useful: "Make it as simple as you can, and no simpler"
be useful: "Make it as simple as you can, and no simpler."

38
Documentation/acpi-hotkey.txt Normal file
View File

@ -0,0 +1,38 @@
driver/acpi/hotkey.c implement:
1. /proc/acpi/hotkey/event_config
(event based hotkey or event config interface):
a. add a event based hotkey(event) :
echo "0:bus::action:method:num:num" > event_config
b. delete a event based hotkey(event):
echo "1:::::num:num" > event_config
c. modify a event based hotkey(event):
echo "2:bus::action:method:num:num" > event_config
2. /proc/acpi/hotkey/poll_config
(polling based hotkey or event config interface):
a.add a polling based hotkey(event) :
echo "0:bus:method:action:method:num" > poll_config
this adding command will create a proc file
/proc/acpi/hotkey/method, which is used to get
result of polling.
b.delete a polling based hotkey(event):
echo "1:::::num" > event_config
c.modify a polling based hotkey(event):
echo "2:bus:method:action:method:num" > poll_config
3./proc/acpi/hotkey/action
(interface to call aml method associated with a
specific hotkey(event))
echo "event_num:event_type:event_argument" >
/proc/acpi/hotkey/action.
The result of the execution of this aml method is
attached to /proc/acpi/hotkey/poll_method, which is dnyamically
created. Please use command "cat /proc/acpi/hotkey/polling_method"
to retrieve it.
Note: Use cmdline "acpi_generic_hotkey" to over-ride
loading any platform specific drivers.

93
Documentation/arm/Samsung-S3C24XX/USB-Host.txt Normal file
View File

@ -0,0 +1,93 @@
S3C24XX USB Host support
========================
Introduction
------------
This document details the S3C2410/S3C2440 in-built OHCI USB host support.
Configuration
-------------
Enable at least the following kernel options:
menuconfig:
Device Drivers --->
USB support --->
<*> Support for Host-side USB
<*> OHCI HCD support
.config:
CONFIG_USB
CONFIG_USB_OHCI_HCD
Once these options are configured, the standard set of USB device
drivers can be configured and used.
Board Support
-------------
The driver attaches to a platform device, which will need to be
added by the board specific support file in linux/arch/arm/mach-s3c2410,
such as mach-bast.c or mach-smdk2410.c
The platform device's platform_data field is only needed if the
board implements extra power control or over-current monitoring.
The OHCI driver does not ensure the state of the S3C2410's MISCCTRL
register, so if both ports are to be used for the host, then it is
the board support file's responsibility to ensure that the second
port is configured to be connected to the OHCI core.
Platform Data
-------------
See linux/include/asm-arm/arch-s3c2410/usb-control.h for the
descriptions of the platform device data. An implementation
can be found in linux/arch/arm/mach-s3c2410/usb-simtec.c .
The `struct s3c2410_hcd_info` contains a pair of functions
that get called to enable over-current detection, and to
control the port power status.
The ports are numbered 0 and 1.
power_control:
Called to enable or disable the power on the port.
enable_oc:
Called to enable or disable the over-current monitoring.
This should claim or release the resources being used to
check the power condition on the port, such as an IRQ.
report_oc:
The OHCI driver fills this field in for the over-current code
to call when there is a change to the over-current state on
an port. The ports argument is a bitmask of 1 bit per port,
with bit X being 1 for an over-current on port X.
The function s3c2410_usb_report_oc() has been provided to
ensure this is called correctly.
port[x]:
This is struct describes each port, 0 or 1. The platform driver
should set the flags field of each port to S3C_HCDFLG_USED if
the port is enabled.
Document Author
---------------
Ben Dooks, (c) 2005 Simtec Electronics

10
Documentation/basic_profiling.txt
View File

@ -27,9 +27,13 @@ dump output readprofile -m /boot/System.map > captured_profile
Oprofile
--------
Get the source (I use 0.8) from http://oprofile.sourceforge.net/
and add "idle=poll" to the kernel command line
Get the source (see Changes for required version) from
http://oprofile.sourceforge.net/ and add "idle=poll" to the kernel command
line.
Configure with CONFIG_PROFILING=y and CONFIG_OPROFILE=y & reboot on new kernel
./configure --with-kernel-support
make install
@ -46,7 +50,7 @@ start opcontrol --start
stop opcontrol --stop
dump output opreport > output_file
To only report on the kernel, run opreport /boot/vmlinux > output_file
To only report on the kernel, run opreport -l /boot/vmlinux > output_file
A reset is needed to clear old statistics, which survive a reboot.

176
Documentation/block/ioprio.txt Normal file
View File

@ -0,0 +1,176 @@
Block io priorities
===================
Intro
-----
With the introduction of cfq v3 (aka cfq-ts or time sliced cfq), basic io
priorities is supported for reads on files. This enables users to io nice
processes or process groups, similar to what has been possible to cpu
scheduling for ages. This document mainly details the current possibilites
with cfq, other io schedulers do not support io priorities so far.
Scheduling classes
------------------
CFQ implements three generic scheduling classes that determine how io is
served for a process.
IOPRIO_CLASS_RT: This is the realtime io class. This scheduling class is given
higher priority than any other in the system, processes from this class are
given first access to the disk every time. Thus it needs to be used with some
care, one io RT process can starve the entire system. Within the RT class,
there are 8 levels of class data that determine exactly how much time this
process needs the disk for on each service. In the future this might change
to be more directly mappable to performance, by passing in a wanted data
rate instead.
IOPRIO_CLASS_BE: This is the best-effort scheduling class, which is the default
for any process that hasn't set a specific io priority. The class data
determines how much io bandwidth the process will get, it's directly mappable
to the cpu nice levels just more coarsely implemented. 0 is the highest
BE prio level, 7 is the lowest. The mapping between cpu nice level and io
nice level is determined as: io_nice = (cpu_nice + 20) / 5.
IOPRIO_CLASS_IDLE: This is the idle scheduling class, processes running at this
level only get io time when no one else needs the disk. The idle class has no
class data, since it doesn't really apply here.
Tools
-----
See below for a sample ionice tool. Usage:
# ionice -c<class> -n<level> -p<pid>
If pid isn't given, the current process is assumed. IO priority settings
are inherited on fork, so you can use ionice to start the process at a given
level:
# ionice -c2 -n0 /bin/ls
will run ls at the best-effort scheduling class at the highest priority.
For a running process, you can give the pid instead:
# ionice -c1 -n2 -p100
will change pid 100 to run at the realtime scheduling class, at priority 2.
---> snip ionice.c tool <---
#include <stdio.h>
#include <stdlib.h>
#include <errno.h>
#include <getopt.h>
#include <unistd.h>
#include <sys/ptrace.h>
#include <asm/unistd.h>
extern int sys_ioprio_set(int, int, int);
extern int sys_ioprio_get(int, int);
#if defined(__i386__)
#define __NR_ioprio_set 289
#define __NR_ioprio_get 290
#elif defined(__ppc__)
#define __NR_ioprio_set 273
#define __NR_ioprio_get 274
#elif defined(__x86_64__)
#define __NR_ioprio_set 251
#define __NR_ioprio_get 252
#elif defined(__ia64__)
#define __NR_ioprio_set 1274
#define __NR_ioprio_get 1275
#else
#error "Unsupported arch"
#endif
_syscall3(int, ioprio_set, int, which, int, who, int, ioprio);
_syscall2(int, ioprio_get, int, which, int, who);
enum {
IOPRIO_CLASS_NONE,
IOPRIO_CLASS_RT,
IOPRIO_CLASS_BE,
IOPRIO_CLASS_IDLE,
};
enum {
IOPRIO_WHO_PROCESS = 1,
IOPRIO_WHO_PGRP,
IOPRIO_WHO_USER,
};
#define IOPRIO_CLASS_SHIFT 13
const char *to_prio[] = { "none", "realtime", "best-effort", "idle", };
int main(int argc, char *argv[])
{
int ioprio = 4, set = 0, ioprio_class = IOPRIO_CLASS_BE;
int c, pid = 0;
while ((c = getopt(argc, argv, "+n:c:p:")) != EOF) {
switch (c) {
case 'n':
ioprio = strtol(optarg, NULL, 10);
set = 1;
break;
case 'c':
ioprio_class = strtol(optarg, NULL, 10);
set = 1;
break;
case 'p':
pid = strtol(optarg, NULL, 10);
break;
}
}
switch (ioprio_class) {
case IOPRIO_CLASS_NONE:
ioprio_class = IOPRIO_CLASS_BE;
break;
case IOPRIO_CLASS_RT:
case IOPRIO_CLASS_BE:
break;
case IOPRIO_CLASS_IDLE:
ioprio = 7;
break;
default:
printf("bad prio class %d\n", ioprio_class);
return 1;
}
if (!set) {
if (!pid && argv[optind])
pid = strtol(argv[optind], NULL, 10);
ioprio = ioprio_get(IOPRIO_WHO_PROCESS, pid);
printf("pid=%d, %d\n", pid, ioprio);
if (ioprio == -1)
perror("ioprio_get");
else {
ioprio_class = ioprio >> IOPRIO_CLASS_SHIFT;
ioprio = ioprio & 0xff;
printf("%s: prio %d\n", to_prio[ioprio_class], ioprio);
}
} else {
if (ioprio_set(IOPRIO_WHO_PROCESS, pid, ioprio | ioprio_class << IOPRIO_CLASS_SHIFT) == -1) {
perror("ioprio_set");
return 1;
}
if (argv[optind])
execvp(argv[optind], &argv[optind]);
}
return 0;
}
---> snip ionice.c tool <---
March 11 2005, Jens Axboe <axboe@suse.de>

1
Documentation/cciss.txt
View File

@ -17,6 +17,7 @@ This driver is known to work with the following cards:
* SA P600
* SA P800
* SA E400
* SA E300
If nodes are not already created in the /dev/cciss directory, run as root:

16
Documentation/cdrom/sbpcd
View File

@ -419,6 +419,7 @@ into the file "track01":
*/
#include <stdio.h>
#include <sys/ioctl.h>
#include <sys/types.h>
#include <linux/cdrom.h>
static struct cdrom_tochdr hdr;
@ -429,7 +430,7 @@ static int datafile, drive;
static int i, j, limit, track, err;
static char filename[32];
main(int argc, char *argv[])
int main(int argc, char *argv[])
{
/*
* open /dev/cdrom
@ -516,6 +517,7 @@ entry[track+1].cdte_addr.lba=entry[track].cdte_addr.lba+300;
}
arg.addr.lba++;
}
return 0;
}
/*===================== end program ========================================*/
@ -564,15 +566,16 @@ Appendix -- the "cdtester" utility:
#include <stdio.h>
#include <malloc.h>
#include <sys/ioctl.h>
#include <sys/types.h>
#include <linux/cdrom.h>
#ifdef AZT_PRIVATE_IOCTLS
#include <linux/../../drivers/cdrom/aztcd.h>
#endif AZT_PRIVATE_IOCTLS
#endif /* AZT_PRIVATE_IOCTLS */
#ifdef SBP_PRIVATE_IOCTLS
#include <linux/../../drivers/cdrom/sbpcd.h>
#include <linux/fs.h>
#endif SBP_PRIVATE_IOCTLS
#endif /* SBP_PRIVATE_IOCTLS */
struct cdrom_tochdr hdr;
struct cdrom_tochdr tocHdr;
@ -590,7 +593,7 @@ union
struct cdrom_msf msf;
unsigned char buf[CD_FRAMESIZE_RAW];
} azt;
#endif AZT_PRIVATE_IOCTLS
#endif /* AZT_PRIVATE_IOCTLS */
int i, i1, i2, i3, j, k;
unsigned char sequence=0;
unsigned char command[80];
@ -738,7 +741,7 @@ void display(int size,unsigned char *buffer)
}
}
main(int argc, char *argv[])
int main(int argc, char *argv[])
{
printf("\nTesting tool for a CDROM driver's audio functions V0.1\n");
printf("(C) 1995 Eberhard Moenkeberg <emoenke@gwdg.de>\n");
@ -1046,12 +1049,13 @@ main(int argc, char *argv[])
rc=ioctl(drive,CDROMAUDIOBUFSIZ,j);
printf("%d frames granted.\n",rc);
break;
#endif SBP_PRIVATE_IOCTLS
#endif /* SBP_PRIVATE_IOCTLS */
default:
printf("unknown command: \"%s\".\n",command);
break;
}
}
return 0;
}
/*==========================================================================*/

14
Documentation/cpu-freq/governors.txt
View File

@ -9,6 +9,7 @@
Dominik Brodowski <linux@brodo.de>
some additions and corrections by Nico Golde <nico@ngolde.de>
@ -25,6 +26,7 @@ Contents:
2.1 Performance
2.2 Powersave
2.3 Userspace
2.4 Ondemand
3. The Governor Interface in the CPUfreq Core
@ -86,7 +88,7 @@ highest frequency within the borders of scaling_min_freq and
scaling_max_freq.
2.1 Powersave
2.2 Powersave
-------------
The CPUfreq governor "powersave" sets the CPU statically to the
@ -94,7 +96,7 @@ lowest frequency within the borders of scaling_min_freq and
scaling_max_freq.
2.2 Userspace
2.3 Userspace
-------------
The CPUfreq governor "userspace" allows the user, or any userspace
@ -103,6 +105,14 @@ by making a sysfs file "scaling_setspeed" available in the CPU-device
directory.
2.4 Ondemand
------------
The CPUfreq govenor "ondemand" sets the CPU depending on the
current usage. To do this the CPU must have the capability to
switch the frequency very fast.
3. The Governor Interface in the CPUfreq Core
=============================================

16
Documentation/cpusets.txt
View File

@ -51,6 +51,14 @@ mems_allowed vector.
If a cpuset is cpu or mem exclusive, no other cpuset, other than a direct
ancestor or descendent, may share any of the same CPUs or Memory Nodes.
A cpuset that is cpu exclusive has a sched domain associated with it.
The sched domain consists of all cpus in the current cpuset that are not
part of any exclusive child cpusets.
This ensures that the scheduler load balacing code only balances
against the cpus that are in the sched domain as defined above and not
all of the cpus in the system. This removes any overhead due to
load balancing code trying to pull tasks outside of the cpu exclusive
cpuset only to be prevented by the tasks' cpus_allowed mask.
User level code may create and destroy cpusets by name in the cpuset
virtual file system, manage the attributes and permissions of these
@ -84,6 +92,9 @@ This can be especially valuable on:
and a database), or
* NUMA systems running large HPC applications with demanding
performance characteristics.
* Also cpu_exclusive cpusets are useful for servers running orthogonal
workloads such as RT applications requiring low latency and HPC
applications that are throughput sensitive
These subsets, or "soft partitions" must be able to be dynamically
adjusted, as the job mix changes, without impacting other concurrently
@ -125,6 +136,8 @@ Cpusets extends these two mechanisms as follows:
- A cpuset may be marked exclusive, which ensures that no other
cpuset (except direct ancestors and descendents) may contain
any overlapping CPUs or Memory Nodes.
Also a cpu_exclusive cpuset would be associated with a sched
domain.
- You can list all the tasks (by pid) attached to any cpuset.
The implementation of cpusets requires a few, simple hooks
@ -136,6 +149,9 @@ into the rest of the kernel, none in performance critical paths:
allowed in that tasks cpuset.
- in sched.c migrate_all_tasks(), to keep migrating tasks within
the CPUs allowed by their cpuset, if possible.
- in sched.c, a new API partition_sched_domains for handling
sched domain changes associated with cpu_exclusive cpusets
and related changes in both sched.c and arch/ia64/kernel/domain.c
- in the mbind and set_mempolicy system calls, to mask the requested
Memory Nodes by what's allowed in that tasks cpuset.
- in page_alloc, to restrict memory to allowed nodes.

1
Documentation/devices.txt
View File

@ -94,6 +94,7 @@ Your cooperation is appreciated.
9 = /dev/urandom Faster, less secure random number gen.
10 = /dev/aio Asyncronous I/O notification interface
11 = /dev/kmsg Writes to this come out as printk's
12 = /dev/oldmem Access to crash dump from kexec kernel
1 block RAM disk
0 = /dev/ram0 First RAM disk
1 = /dev/ram1 Second RAM disk

4
Documentation/dontdiff
View File

@ -41,6 +41,7 @@ COPYING
CREDITS
CVS
ChangeSet
Image
Kerntypes
MODS.txt
Module.symvers
@ -103,6 +104,8 @@ logo_*.c
logo_*_clut224.c
logo_*_mono.c
lxdialog
mach-types
mach-types.h
make_times_h
map
maui_boot.h
@ -111,6 +114,7 @@ mkdep
mktables
modpost
modversions.h*
offset.h
offsets.h
oui.c*
parse.c*

285
Documentation/dvb/README.dibusb
View File

@ -1,285 +0,0 @@
Documentation for dib3000* frontend drivers and dibusb device driver
====================================================================
Copyright (C) 2004-5 Patrick Boettcher (patrick.boettcher@desy.de),
dibusb and dib3000mb/mc drivers based on GPL code, which has
Copyright (C) 2004 Amaury Demol for DiBcom (ademol@dibcom.fr)
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License as
published by the Free Software Foundation, version 2.
Supported devices USB1.1
========================
Produced and reselled by Twinhan:
---------------------------------
- TwinhanDTV USB-Ter DVB-T Device (VP7041)
http://www.twinhan.com/product_terrestrial_3.asp
- TwinhanDTV Magic Box (VP7041e)
http://www.twinhan.com/product_terrestrial_4.asp
- HAMA DVB-T USB device
http://www.hama.de/portal/articleId*110620/action*2598
- CTS Portable (Chinese Television System) (2)
http://www.2cts.tv/ctsportable/
- Unknown USB DVB-T device with vendor ID Hyper-Paltek
Produced and reselled by KWorld:
--------------------------------
- KWorld V-Stream XPERT DTV DVB-T USB
http://www.kworld.com.tw/en/product/DVBT-USB/DVBT-USB.html
- JetWay DTV DVB-T USB
http://www.jetway.com.tw/evisn/product/lcd-tv/DVT-USB/dtv-usb.htm
- ADSTech Instant TV DVB-T USB
http://www.adstech.com/products/PTV-333/intro/PTV-333_intro.asp?pid=PTV-333
Others:
-------
- Ultima Electronic/Artec T1 USB TVBOX (AN2135, AN2235, AN2235 with Panasonic Tuner)
http://82.161.246.249/products-tvbox.html
- Compro Videomate DVB-U2000 - DVB-T USB (2)
http://www.comprousa.com/products/vmu2000.htm
- Grandtec USB DVB-T
http://www.grand.com.tw/
- Avermedia AverTV DVBT USB (2)
http://www.avermedia.com/
- DiBcom USB DVB-T reference device (non-public)
Supported devices USB2.0
========================
- Twinhan MagicBox II (2)
http://www.twinhan.com/product_terrestrial_7.asp
- Hanftek UMT-010 (1)
http://www.globalsources.com/si/6008819757082/ProductDetail/Digital-TV/product_id-100046529
- Typhoon/Yakumo/HAMA DVB-T mobile USB2.0 (1)
http://www.yakumo.de/produkte/index.php?pid=1&ag=DVB-T
- Artec T1 USB TVBOX (FX2) (2)
- Hauppauge WinTV NOVA-T USB2
http://www.hauppauge.com/
- KWorld/ADSTech Instant DVB-T USB2.0 (DiB3000M-B)
- DiBcom USB2.0 DVB-T reference device (non-public)
1) It is working almost.
2) No test reports received yet.
0. NEWS:
2005-02-11 - added support for the KWorld/ADSTech Instant DVB-T USB2.0. Thanks a lot to Joachim von Caron
2005-02-02 - added support for the Hauppauge Win-TV Nova-T USB2
2005-01-31 - distorted streaming is finally gone for USB1.1 devices
2005-01-13 - moved the mirrored pid_filter_table back to dvb-dibusb
- first almost working version for HanfTek UMT-010
- found out, that Yakumo/HAMA/Typhoon are predessors of the HanfTek UMT-010
2005-01-10 - refactoring completed, now everything is very delightful
- tuner quirks for some weird devices (Artec T1 AN2235 device has sometimes a
Panasonic Tuner assembled). Tunerprobing implemented. Thanks a lot to Gunnar Wittich.
2004-12-29 - after several days of struggling around bug of no returning URBs fixed.
2004-12-26 - refactored the dibusb-driver, splitted into separate files
- i2c-probing enabled
2004-12-06 - possibility for demod i2c-address probing
- new usb IDs (Compro,Artec)
2004-11-23 - merged changes from DiB3000MC_ver2.1
- revised the debugging
- possibility to deliver the complete TS for USB2.0
2004-11-21 - first working version of the dib3000mc/p frontend driver.
2004-11-12 - added additional remote control keys. Thanks to Uwe Hanke.
2004-11-07 - added remote control support. Thanks to David Matthews.
2004-11-05 - added support for a new devices (Grandtec/Avermedia/Artec)
- merged my changes (for dib3000mb/dibusb) to the FE_REFACTORING, because it became HEAD
- moved transfer control (pid filter, fifo control) from usb driver to frontend, it seems
better settled there (added xfer_ops-struct)
- created a common files for frontends (mc/p/mb)
2004-09-28 - added support for a new device (Unkown, vendor ID is Hyper-Paltek)
2004-09-20 - added support for a new device (Compro DVB-U2000), thanks
to Amaury Demol for reporting
- changed usb TS transfer method (several urbs, stopping transfer
before setting a new pid)
2004-09-13 - added support for a new device (Artec T1 USB TVBOX), thanks
to Christian Motschke for reporting
2004-09-05 - released the dibusb device and dib3000mb-frontend driver
(old news for vp7041.c)
2004-07-15 - found out, by accident, that the device has a TUA6010XS for
PLL
2004-07-12 - figured out, that the driver should also work with the
CTS Portable (Chinese Television System)
2004-07-08 - firmware-extraction-2.422-problem solved, driver is now working
properly with firmware extracted from 2.422
- #if for 2.6.4 (dvb), compile issue
- changed firmware handling, see vp7041.txt sec 1.1
2004-07-02 - some tuner modifications, v0.1, cleanups, first public
2004-06-28 - now using the dvb_dmx_swfilter_packets, everything
runs fine now
2004-06-27 - able to watch and switching channels (pre-alpha)
- no section filtering yet
2004-06-06 - first TS received, but kernel oops :/
2004-05-14 - firmware loader is working
2004-05-11 - start writing the driver
1. How to use?
NOTE: This driver was developed using Linux 2.6.6.,
it is working with 2.6.7 and above.
Linux 2.4.x support is not planned, but patches are very welcome.
NOTE: I'm using Debian testing, so the following explaination (especially
the hotplug-path) needn't match your system, but probably it will :).
The driver is included in the kernel since Linux 2.6.10.
1.1. Firmware
The USB driver needs to download a firmware to start working.
You can either use "get_dvb_firmware dibusb" to download the firmware or you
can get it directly via
for USB1.1 (AN2135)
http://www.linuxtv.org/downloads/firmware/dvb-dibusb-5.0.0.11.fw
for USB1.1 (AN2235) (a few Artec T1 devices)
http://www.linuxtv.org/downloads/firmware/dvb-dibusb-an2235-1.fw
for USB2.0 (FX2) Hauppauge, DiBcom
http://www.linuxtv.org/downloads/firmware/dvb-dibusb-6.0.0.5.fw
for USB2.0 ADSTech/Kworld USB2.0
http://www.linuxtv.org/downloads/firmware/dvb-dibusb-adstech-usb2-1.fw
for USB2.0 HanfTek
http://www.linuxtv.org/downloads/firmware/dvb-dibusb-an2235-1.fw
1.2. Compiling
Since the driver is in the linux kernel, activating the driver in
your favorite config-environment should sufficient. I recommend
to compile the driver as module. Hotplug does the rest.
1.3. Loading the drivers
Hotplug is able to load the driver, when it is needed (because you plugged
in the device).
If you want to enable debug output, you have to load the driver manually and
from withing the dvb-kernel cvs repository.
first have a look, which debug level are available:
modinfo dib3000mb
modinfo dib3000-common
modinfo dib3000mc
modinfo dvb-dibusb
modprobe dib3000-common debug=<level>
modprobe dib3000mb debug=<level>
modprobe dib3000mc debug=<level>
modprobe dvb-dibusb debug=<level>
should do the trick.
When the driver is loaded successfully, the firmware file was in
the right place and the device is connected, the "Power"-LED should be
turned on.
At this point you should be able to start a dvb-capable application. For myself
I used mplayer, dvbscan, tzap and kaxtv, they are working. Using the device
in vdr is working now also.
2. Known problems and bugs
- Don't remove the USB device while running an DVB application, your system will die.
2.1. Adding support for devices
It is not possible to determine the range of devices based on the DiBcom
reference designs. This is because the reference design of DiBcom can be sold
to thirds, without telling DiBcom (so done with the Twinhan VP7041 and
the HAMA device).
When you think you have a device like this and the driver does not recognizes it,
please send the ****load*.inf and the ****cap*.inf of the Windows driver to me.
Sometimes the Vendor or Product ID is identical to the ones of Twinhan, even
though it is not a Twinhan device (e.g. HAMA), then please send me the name
of the device. I will add it to this list in order to make this clear to
others.
If you are familar with C you can also add the VID and PID of the device to
the dvb-dibusb-core.c-file and create a patch and send it over to me or to
the linux-dvb mailing list, _after_ you have tried compiling and modprobing
it.
2.2. USB1.1 Bandwidth limitation
Most of the currently supported devices are USB1.1 and thus they have a
maximum bandwidth of about 5-6 MBit/s when connected to a USB2.0 hub.
This is not enough for receiving the complete transport stream of a
DVB-T channel (which can be about 16 MBit/s). Normally this is not a
problem, if you only want to watch TV (this does not apply for HDTV),
but watching a channel while recording another channel on the same
frequency simply does not work very well. This applies to all USB1.1
DVB-T devices, not just dibusb)
Update: For the USB1.1 and VDR some work has been done (patches and comments
are still very welcome). Maybe the problem is solved in the meantime because I
now use the dmx_sw_filter function instead of dmx_sw_filter_packet. I hope the
linux-dvb software filter is able to get the best of the garbled TS.
The bug, where the TS is distorted by a heavy usage of the device is gone
definitely. All dibusb-devices I was using (Twinhan, Kworld, DiBcom) are
working like charm now with VDR. Sometimes I even was able to record a channel
and watch another one.
2.3. Comments
Patches, comments and suggestions are very very welcome.
3. Acknowledgements
Amaury Demol (ademol@dibcom.fr) and Francois Kanounnikoff from DiBcom for
providing specs, code and help, on which the dvb-dibusb, dib3000mb and
dib3000mc are based.
David Matthews for identifying a new device type (Artec T1 with AN2235)
and for extending dibusb with remote control event handling. Thank you.
Alex Woods for frequently answering question about usb and dvb
stuff, a big thank you.
Bernd Wagner for helping with huge bug reports and discussions.
Gunnar Wittich and Joachim von Caron for their trust for giving me
root-shells on their machines to implement support for new devices.
Some guys on the linux-dvb mailing list for encouraging me
Peter Schildmann >peter.schildmann-nospam-at-web.de< for his
user-level firmware loader, which saves a lot of time
(when writing the vp7041 driver)
Ulf Hermenau for helping me out with traditional chinese.
André Smoktun and Christian Frömmel for supporting me with
hardware and listening to my problems very patient

232
Documentation/dvb/README.dvb-usb Normal file
View File

@ -0,0 +1,232 @@
Documentation for dvb-usb-framework module and its devices
Idea behind the dvb-usb-framework
=================================
In March 2005 I got the new Twinhan USB2.0 DVB-T device. They provided specs and a firmware.
Quite keen I wanted to put the driver (with some quirks of course) into dibusb.
After reading some specs and doing some USB snooping, it realized, that the
dibusb-driver would be a complete mess afterwards. So I decided to do it in a
different way: With the help of a dvb-usb-framework.
The framework provides generic functions (mostly kernel API calls), such as:
- Transport Stream URB handling in conjunction with dvb-demux-feed-control
(bulk and isoc are supported)
- registering the device for the DVB-API
- registering an I2C-adapter if applicable
- remote-control/input-device handling
- firmware requesting and loading (currently just for the Cypress USB
controllers)
- other functions/methods which can be shared by several drivers (such as
functions for bulk-control-commands)
- TODO: a I2C-chunker. It creates device-specific chunks of register-accesses
depending on length of a register and the number of values that can be
multi-written and multi-read.
The source code of the particular DVB USB devices does just the communication
with the device via the bus. The connection between the DVB-API-functionality
is done via callbacks, assigned in a static device-description (struct
dvb_usb_device) each device-driver has to have.
For an example have a look in drivers/media/dvb/dvb-usb/vp7045*.
Objective is to migrate all the usb-devices (dibusb, cinergyT2, maybe the
ttusb; flexcop-usb already benefits from the generic flexcop-device) to use
the dvb-usb-lib.
TODO: dynamic enabling and disabling of the pid-filter in regard to number of
feeds requested.
Supported devices
========================
See the LinuxTV DVB Wiki at www.linuxtv.org for a complete list of
cards/drivers/firmwares:
http://www.linuxtv.org/wiki/index.php/DVB_USB
0. History & News:
2005-06-30 - added support for WideView WT-220U (Thanks to Steve Chang)
2005-05-30 - added basic isochronous support to the dvb-usb-framework
added support for Conexant Hybrid reference design and Nebula DigiTV USB
2005-04-17 - all dibusb devices ported to make use of the dvb-usb-framework
2005-04-02 - re-enabled and improved remote control code.
2005-03-31 - ported the Yakumo/Hama/Typhoon DVB-T USB2.0 device to dvb-usb.
2005-03-30 - first commit of the dvb-usb-module based on the dibusb-source. First device is a new driver for the
TwinhanDTV Alpha / MagicBox II USB2.0-only DVB-T device.
(change from dvb-dibusb to dvb-usb)
2005-03-28 - added support for the AVerMedia AverTV DVB-T USB2.0 device (Thanks to Glen Harris and Jiun-Kuei Jung, AVerMedia)
2005-03-14 - added support for the Typhoon/Yakumo/HAMA DVB-T mobile USB2.0
2005-02-11 - added support for the KWorld/ADSTech Instant DVB-T USB2.0. Thanks a lot to Joachim von Caron
2005-02-02 - added support for the Hauppauge Win-TV Nova-T USB2
2005-01-31 - distorted streaming is gone for USB1.1 devices
2005-01-13 - moved the mirrored pid_filter_table back to dvb-dibusb
- first almost working version for HanfTek UMT-010
- found out, that Yakumo/HAMA/Typhoon are predecessors of the HanfTek UMT-010
2005-01-10 - refactoring completed, now everything is very delightful
- tuner quirks for some weird devices (Artec T1 AN2235 device has sometimes a
Panasonic Tuner assembled). Tunerprobing implemented. Thanks a lot to Gunnar Wittich.
2004-12-29 - after several days of struggling around bug of no returning URBs fixed.
2004-12-26 - refactored the dibusb-driver, splitted into separate files
- i2c-probing enabled
2004-12-06 - possibility for demod i2c-address probing
- new usb IDs (Compro, Artec)
2004-11-23 - merged changes from DiB3000MC_ver2.1
- revised the debugging
- possibility to deliver the complete TS for USB2.0
2004-11-21 - first working version of the dib3000mc/p frontend driver.
2004-11-12 - added additional remote control keys. Thanks to Uwe Hanke.
2004-11-07 - added remote control support. Thanks to David Matthews.
2004-11-05 - added support for a new devices (Grandtec/Avermedia/Artec)
- merged my changes (for dib3000mb/dibusb) to the FE_REFACTORING, because it became HEAD
- moved transfer control (pid filter, fifo control) from usb driver to frontend, it seems
better settled there (added xfer_ops-struct)
- created a common files for frontends (mc/p/mb)
2004-09-28 - added support for a new device (Unkown, vendor ID is Hyper-Paltek)
2004-09-20 - added support for a new device (Compro DVB-U2000), thanks
to Amaury Demol for reporting
- changed usb TS transfer method (several urbs, stopping transfer
before setting a new pid)
2004-09-13 - added support for a new device (Artec T1 USB TVBOX), thanks
to Christian Motschke for reporting
2004-09-05 - released the dibusb device and dib3000mb-frontend driver
(old news for vp7041.c)
2004-07-15 - found out, by accident, that the device has a TUA6010XS for
PLL
2004-07-12 - figured out, that the driver should also work with the
CTS Portable (Chinese Television System)
2004-07-08 - firmware-extraction-2.422-problem solved, driver is now working
properly with firmware extracted from 2.422
- #if for 2.6.4 (dvb), compile issue
- changed firmware handling, see vp7041.txt sec 1.1
2004-07-02 - some tuner modifications, v0.1, cleanups, first public
2004-06-28 - now using the dvb_dmx_swfilter_packets, everything
runs fine now
2004-06-27 - able to watch and switching channels (pre-alpha)
- no section filtering yet
2004-06-06 - first TS received, but kernel oops :/
2004-05-14 - firmware loader is working
2004-05-11 - start writing the driver
1. How to use?
1.1. Firmware
Most of the USB drivers need to download a firmware to the device before start
working.
Have a look at the Wikipage for the DVB-USB-drivers to find out, which firmware
you need for your device:
http://www.linuxtv.org/wiki/index.php/DVB_USB
1.2. Compiling
Since the driver is in the linux kernel, activating the driver in
your favorite config-environment should sufficient. I recommend
to compile the driver as module. Hotplug does the rest.
If you use dvb-kernel enter the build-2.6 directory run 'make' and 'insmod.sh
load' afterwards.
1.3. Loading the drivers
Hotplug is able to load the driver, when it is needed (because you plugged
in the device).
If you want to enable debug output, you have to load the driver manually and
from withing the dvb-kernel cvs repository.
first have a look, which debug level are available:
modinfo dvb-usb
modinfo dvb-usb-vp7045
etc.
modprobe dvb-usb debug=<level>
modprobe dvb-usb-vp7045 debug=<level>
etc.
should do the trick.
When the driver is loaded successfully, the firmware file was in
the right place and the device is connected, the "Power"-LED should be
turned on.
At this point you should be able to start a dvb-capable application. I'm use
(t|s)zap, mplayer and dvbscan to test the basics. VDR-xine provides the
long-term test scenario.
2. Known problems and bugs
- Don't remove the USB device while running an DVB application, your system
will go crazy or die most likely.
2.1. Adding support for devices
TODO
2.2. USB1.1 Bandwidth limitation
A lot of the currently supported devices are USB1.1 and thus they have a
maximum bandwidth of about 5-6 MBit/s when connected to a USB2.0 hub.
This is not enough for receiving the complete transport stream of a
DVB-T channel (which is about 16 MBit/s). Normally this is not a
problem, if you only want to watch TV (this does not apply for HDTV),
but watching a channel while recording another channel on the same
frequency simply does not work very well. This applies to all USB1.1
DVB-T devices, not just the dvb-usb-devices)
The bug, where the TS is distorted by a heavy usage of the device is gone
definitely. All dvb-usb-devices I was using (Twinhan, Kworld, DiBcom) are
working like charm now with VDR. Sometimes I even was able to record a channel
and watch another one.
2.3. Comments
Patches, comments and suggestions are very very welcome.
3. Acknowledgements
Amaury Demol (ademol@dibcom.fr) and Francois Kanounnikoff from DiBcom for
providing specs, code and help, on which the dvb-dibusb, dib3000mb and
dib3000mc are based.
David Matthews for identifying a new device type (Artec T1 with AN2235)
and for extending dibusb with remote control event handling. Thank you.
Alex Woods for frequently answering question about usb and dvb
stuff, a big thank you.
Bernd Wagner for helping with huge bug reports and discussions.
Gunnar Wittich and Joachim von Caron for their trust for providing
root-shells on their machines to implement support for new devices.
Allan Third and Michael Hutchinson for their help to write the Nebula
digitv-driver.
Glen Harris for bringing up, that there is a new dibusb-device and Jiun-Kuei
Jung from AVerMedia who kindly provided a special firmware to get the device
up and running in Linux.
Jennifer Chen, Jeff and Jack from Twinhan for kindly supporting by
writing the vp7045-driver.
Steve Chang from WideView for providing information for new devices and
firmware files.
Michael Paxton for submitting remote control keymaps.
Some guys on the linux-dvb mailing list for encouraging me.
Peter Schildmann >peter.schildmann-nospam-at-web.de< for his
user-level firmware loader, which saves a lot of time
(when writing the vp7041 driver)
Ulf Hermenau for helping me out with traditional chinese.
André Smoktun and Christian Frömmel for supporting me with
hardware and listening to my problems very patiently.

90
Documentation/dvb/bt8xx.txt
View File

@ -1,69 +1,55 @@
How to get the Nebula, PCTV and Twinhan DST cards working
=========================================================
How to get the Nebula Electronics DigiTV, Pinnacle PCTV Sat, Twinhan DST + clones working
=========================================================================================
This class of cards has a bt878a as the PCI interface, and
require the bttv driver.
1) General information
======================
Please pay close attention to the warning about the bttv module
options below for the DST card.
This class of cards has a bt878a chip as the PCI interface.
The different card drivers require the bttv driver to provide the means
to access the i2c bus and the gpio pins of the bt8xx chipset.
1) General informations
=======================
2) Compilation rules for Kernel >= 2.6.12
=========================================
These drivers require the bttv driver to provide the means to access
the i2c bus and the gpio pins of the bt8xx chipset.
Enable the following options:
Because of this, you need to enable
"Device drivers" => "Multimedia devices"
=> "Video For Linux" => "BT848 Video For Linux"
Furthermore you need to enable
=> "Video For Linux" => "BT848 Video For Linux"
"Device drivers" => "Multimedia devices" => "Digital Video Broadcasting Devices"
=> "DVB for Linux" "DVB Core Support" "Nebula/Pinnacle PCTV/TwinHan PCI Cards"
=> "DVB for Linux" "DVB Core Support" "Nebula/Pinnacle PCTV/TwinHan PCI Cards"
2) Loading Modules
==================
3) Loading Modules, described by two approaches
===============================================
In general you need to load the bttv driver, which will handle the gpio and
i2c communication for us, plus the common dvb-bt8xx device driver.
The frontends for Nebula (nxt6000), Pinnacle PCTV (cx24110) and
TwinHan (dst) are loaded automatically by the dvb-bt8xx device driver.
i2c communication for us, plus the common dvb-bt8xx device driver,
which is called the backend.
The frontends for Nebula DigiTV (nxt6000), Pinnacle PCTV Sat (cx24110),
TwinHan DST + clones (dst and dst-ca) are loaded automatically by the backend.
For further details about TwinHan DST + clones see /Documentation/dvb/ci.txt.
3a) Nebula / Pinnacle PCTV
3a) The manual approach
-----------------------
Loading modules:
modprobe bttv
modprobe dvb-bt8xx
Unloading modules:
modprobe -r dvb-bt8xx
modprobe -r bttv
3b) The automatic approach
--------------------------
$ modprobe bttv (normally bttv is being loaded automatically by kmod)
$ modprobe dvb-bt8xx (or just place dvb-bt8xx in /etc/modules for automatic loading)
If not already done by installation, place a line either in
/etc/modules.conf or in /etc/modprobe.conf containing this text:
alias char-major-81 bttv
Then place a line in /etc/modules containing this text:
dvb-bt8xx
3b) TwinHan and Clones
--------------------------
$ modprobe bttv i2c_hw=1 card=0x71
$ modprobe dvb-bt8xx
$ modprobe dst
The value 0x71 will override the PCI type detection for dvb-bt8xx,
which is necessary for TwinHan cards.
If you're having an older card (blue color circuit) and card=0x71 locks
your machine, try using 0x68, too. If that does not work, ask on the
mailing list.
The DST module takes a couple of useful parameters.
verbose takes values 0 to 5. These values control the verbosity level.
debug takes values 0 and 1. You can either disable or enable debugging.
dst_addons takes values 0 and 0x20. A value of 0 means it is a FTA card.
0x20 means it has a Conditional Access slot.
The autodected values are determined bythe cards 'response
string' which you can see in your logs e.g.
dst_get_device_id: Recognise [DSTMCI]
Reboot your system and have fun!
--
Authors: Richard Walker, Jamie Honan, Michael Hunold, Manu Abraham
Authors: Richard Walker, Jamie Honan, Michael Hunold, Manu Abraham, Uwe Bugla

16
Documentation/fb/vesafb.txt
View File

@ -144,7 +144,21 @@ vgapal Use the standard vga registers for palette changes.
This is the default.
pmipal Use the protected mode interface for palette changes.
mtrr setup memory type range registers for the vesafb framebuffer.
mtrr:n setup memory type range registers for the vesafb framebuffer
where n:
0 - disabled (equivalent to nomtrr)
1 - uncachable
2 - write-back
3 - write-combining (default)
4 - write-through
If you see the following in dmesg, choose the type that matches the
old one. In this example, use "mtrr:2".
...
mtrr: type mismatch for e0000000,8000000 old: write-back new: write-combining
...
nomtrr disable mtrr
vremap:n
remap 'n' MiB of video RAM. If 0 or not specified, remap memory

52
Documentation/feature-removal-schedule.txt
View File

@ -43,6 +43,14 @@ Who: Randy Dunlap <rddunlap@osdl.org>
---------------------------
What: RAW driver (CONFIG_RAW_DRIVER)
When: December 2005
Why: declared obsolete since kernel 2.6.3
O_DIRECT can be used instead
Who: Adrian Bunk <bunk@stusta.de>
---------------------------
What: register_ioctl32_conversion() / unregister_ioctl32_conversion()
When: April 2005
Why: Replaced by ->compat_ioctl in file_operations and other method
@ -66,6 +74,14 @@ Who: Paul E. McKenney <paulmck@us.ibm.com>
---------------------------
What: remove verify_area()
When: July 2006
Files: Various uaccess.h headers.
Why: Deprecated and redundant. access_ok() should be used instead.
Who: Jesper Juhl <juhl-lkml@dif.dk>
---------------------------
What: IEEE1394 Audio and Music Data Transmission Protocol driver,
Connection Management Procedures driver
When: November 2005
@ -83,3 +99,39 @@ Why: Deprecated in favour of the new ioctl-based rawiso interface, which is
more efficient. You should really be using libraw1394 for raw1394
access anyway.
Who: Jody McIntyre <scjody@steamballoon.com>
---------------------------
What: register_serial/unregister_serial
When: September 2005
Why: This interface does not allow serial ports to be registered against
a struct device, and as such does not allow correct power management
of such ports. 8250-based ports should use serial8250_register_port
and serial8250_unregister_port, or platform devices instead.
Who: Russell King <rmk@arm.linux.org.uk>
---------------------------
What: i2c sysfs name change: in1_ref, vid deprecated in favour of cpu0_vid
When: November 2005
Files: drivers/i2c/chips/adm1025.c, drivers/i2c/chips/adm1026.c
Why: Match the other drivers' name for the same function, duplicate names
will be available until removal of old names.
Who: Grant Coady <gcoady@gmail.com>
---------------------------
What: PCMCIA control ioctl (needed for pcmcia-cs [cardmgr, cardctl])
When: November 2005
Files: drivers/pcmcia/: pcmcia_ioctl.c
Why: With the 16-bit PCMCIA subsystem now behaving (almost) like a
normal hotpluggable bus, and with it using the default kernel
infrastructure (hotplug, driver core, sysfs) keeping the PCMCIA
control ioctl needed by cardmgr and cardctl from pcmcia-cs is
unnecessary, and makes further cleanups and integration of the
PCMCIA subsystem into the Linux kernel device driver model more
difficult. The features provided by cardmgr and cardctl are either
handled by the kernel itself now or are available in the new
pcmciautils package available at
http://kernel.org/pub/linux/utils/kernel/pcmcia/
Who: Dominik Brodowski <linux@brodo.de>

2
Documentation/filesystems/ext2.txt
View File

@ -58,6 +58,8 @@ noacl Don't support POSIX ACLs.
nobh Do not attach buffer_heads to file pagecache.
xip Use execute in place (no caching) if possible
grpquota,noquota,quota,usrquota Quota options are silently ignored by ext2.

151
Documentation/filesystems/inotify.txt Normal file
View File

@ -0,0 +1,151 @@
inotify
a powerful yet simple file change notification system
Document started 15 Mar 2005 by Robert Love <rml@novell.com>
(i) User Interface
Inotify is controlled by a set of three system calls and normal file I/O on a
returned file descriptor.
First step in using inotify is to initialise an inotify instance:
int fd = inotify_init ();
Each instance is associated with a unique, ordered queue.
Change events are managed by "watches". A watch is an (object,mask) pair where
the object is a file or directory and the mask is a bit mask of one or more
inotify events that the application wishes to receive. See <linux/inotify.h>
for valid events. A watch is referenced by a watch descriptor, or wd.
Watches are added via a path to the file.
Watches on a directory will return events on any files inside of the directory.
Adding a watch is simple:
int wd = inotify_add_watch (fd, path, mask);
Where "fd" is the return value from inotify_init(), path is the path to the
object to watch, and mask is the watch mask (see <linux/inotify.h>).
You can update an existing watch in the same manner, by passing in a new mask.
An existing watch is removed via
int ret = inotify_rm_watch (fd, wd);
Events are provided in the form of an inotify_event structure that is read(2)
from a given inotify instance. The filename is of dynamic length and follows
the struct. It is of size len. The filename is padded with null bytes to
ensure proper alignment. This padding is reflected in len.
You can slurp multiple events by passing a large buffer, for example
size_t len = read (fd, buf, BUF_LEN);
Where "buf" is a pointer to an array of "inotify_event" structures at least
BUF_LEN bytes in size. The above example will return as many events as are
available and fit in BUF_LEN.
Each inotify instance fd is also select()- and poll()-able.
You can find the size of the current event queue via the standard FIONREAD
ioctl on the fd returned by inotify_init().
All watches are destroyed and cleaned up on close.
(ii)
Prototypes:
int inotify_init (void);
int inotify_add_watch (int fd, const char *path, __u32 mask);
int inotify_rm_watch (int fd, __u32 mask);
(iii) Internal Kernel Implementation
Each inotify instance is associated with an inotify_device structure.
Each watch is associated with an inotify_watch structure. Watches are chained
off of each associated device and each associated inode.
See fs/inotify.c for the locking and lifetime rules.
(iv) Rationale
Q: What is the design decision behind not tying the watch to the open fd of
the watched object?
A: Watches are associated with an open inotify device, not an open file.
This solves the primary problem with dnotify: keeping the file open pins
the file and thus, worse, pins the mount. Dnotify is therefore infeasible
for use on a desktop system with removable media as the media cannot be
unmounted. Watching a file should not require that it be open.
Q: What is the design decision behind using an-fd-per-instance as opposed to
an fd-per-watch?
A: An fd-per-watch quickly consumes more file descriptors than are allowed,
more fd's than are feasible to manage, and more fd's than are optimally
select()-able. Yes, root can bump the per-process fd limit and yes, users
can use epoll, but requiring both is a silly and extraneous requirement.
A watch consumes less memory than an open file, separating the number
spaces is thus sensible. The current design is what user-space developers
want: Users initialize inotify, once, and add n watches, requiring but one
fd and no twiddling with fd limits. Initializing an inotify instance two
thousand times is silly. If we can implement user-space's preferences
cleanly--and we can, the idr layer makes stuff like this trivial--then we
should.
There are other good arguments. With a single fd, there is a single
item to block on, which is mapped to a single queue of events. The single
fd returns all watch events and also any potential out-of-band data. If
every fd was a separate watch,
- There would be no way to get event ordering. Events on file foo and
file bar would pop poll() on both fd's, but there would be no way to tell
which happened first. A single queue trivially gives you ordering. Such
ordering is crucial to existing applications such as Beagle. Imagine
"mv a b ; mv b a" events without ordering.
- We'd have to maintain n fd's and n internal queues with state,
versus just one. It is a lot messier in the kernel. A single, linear
queue is the data structure that makes sense.
- User-space developers prefer the current API. The Beagle guys, for
example, love it. Trust me, I asked. It is not a surprise: Who'd want
to manage and block on 1000 fd's via select?
- No way to get out of band data.
- 1024 is still too low. ;-)
When you talk about designing a file change notification system that
scales to 1000s of directories, juggling 1000s of fd's just does not seem
the right interface. It is too heavy.
Additionally, it _is_ possible to more than one instance and
juggle more than one queue and thus more than one associated fd. There
need not be a one-fd-per-process mapping; it is one-fd-per-queue and a
process can easily want more than one queue.
Q: Why the system call approach?
A: The poor user-space interface is the second biggest problem with dnotify.
Signals are a terrible, terrible interface for file notification. Or for
anything, for that matter. The ideal solution, from all perspectives, is a
file descriptor-based one that allows basic file I/O and poll/select.
Obtaining the fd and managing the watches could have been done either via a
device file or a family of new system calls. We decided to implement a
family of system calls because that is the preffered approach for new kernel
interfaces. The only real difference was whether we wanted to use open(2)
and ioctl(2) or a couple of new system calls. System calls beat ioctls.

29
Documentation/filesystems/ntfs.txt
View File

@ -21,7 +21,7 @@ Overview
========
Linux-NTFS comes with a number of user-space programs known as ntfsprogs.
These include mkntfs, a full-featured ntfs file system format utility,
These include mkntfs, a full-featured ntfs filesystem format utility,
ntfsundelete used for recovering files that were unintentionally deleted
from an NTFS volume and ntfsresize which is used to resize an NTFS partition.
See the web site for more information.
@ -149,7 +149,14 @@ case_sensitive=<BOOL> If case_sensitive is specified, treat all file names as
name, if it exists. If case_sensitive, you will need
to provide the correct case of the short file name.
errors=opt What to do when critical file system errors are found.
disable_sparse=<BOOL> If disable_sparse is specified, creation of sparse
regions, i.e. holes, inside files is disabled for the
volume (for the duration of this mount only). By
default, creation of sparse regions is enabled, which
is consistent with the behaviour of traditional Unix
filesystems.
errors=opt What to do when critical filesystem errors are found.
Following values can be used for "opt":
continue: DEFAULT, try to clean-up as much as
possible, e.g. marking a corrupt inode as
@ -432,6 +439,24 @@ ChangeLog
Note, a technical ChangeLog aimed at kernel hackers is in fs/ntfs/ChangeLog.
2.1.23:
- Stamp the user space journal, aka transaction log, aka $UsnJrnl, if
it is present and active thus telling Windows and applications using
the transaction log that changes can have happened on the volume
which are not recorded in $UsnJrnl.
- Detect the case when Windows has been hibernated (suspended to disk)
and if this is the case do not allow (re)mounting read-write to
prevent data corruption when you boot back into the suspended
Windows session.
- Implement extension of resident files using the normal file write
code paths, i.e. most very small files can be extended to be a little
bit bigger but not by much.
- Add new mount option "disable_sparse". (See list of mount options
above for details.)
- Improve handling of ntfs volumes with errors and strange boot sectors
in particular.
- Fix various bugs including a nasty deadlock that appeared in recent
kernels (around 2.6.11-2.6.12 timeframe).
2.1.22:
- Improve handling of ntfs volumes with errors.
- Fix various bugs and race conditions.

67
Documentation/filesystems/xip.txt Normal file
View File

@ -0,0 +1,67 @@
Execute-in-place for file mappings
----------------------------------
Motivation
----------
File mappings are performed by mapping page cache pages to userspace. In
addition, read&write type file operations also transfer data from/to the page
cache.
For memory backed storage devices that use the block device interface, the page
cache pages are in fact copies of the original storage. Various approaches
exist to work around the need for an extra copy. The ramdisk driver for example
does read the data into the page cache, keeps a reference, and discards the
original data behind later on.
Execute-in-place solves this issue the other way around: instead of keeping
data in the page cache, the need to have a page cache copy is eliminated
completely. With execute-in-place, read&write type operations are performed
directly from/to the memory backed storage device. For file mappings, the
storage device itself is mapped directly into userspace.
This implementation was initialy written for shared memory segments between
different virtual machines on s390 hardware to allow multiple machines to
share the same binaries and libraries.
Implementation
--------------
Execute-in-place is implemented in three steps: block device operation,
address space operation, and file operations.
A block device operation named direct_access is used to retrieve a
reference (pointer) to a block on-disk. The reference is supposed to be
cpu-addressable, physical address and remain valid until the release operation
is performed. A struct block_device reference is used to address the device,
and a sector_t argument is used to identify the individual block. As an
alternative, memory technology devices can be used for this.
The block device operation is optional, these block devices support it as of
today:
- dcssblk: s390 dcss block device driver
An address space operation named get_xip_page is used to retrieve reference
to a struct page. To address the target page, a reference to an address_space,
and a sector number is provided. A 3rd argument indicates whether the
function should allocate blocks if needed.
This address space operation is mutually exclusive with readpage&writepage that
do page cache read/write operations.
The following filesystems support it as of today:
- ext2: the second extended filesystem, see Documentation/filesystems/ext2.txt
A set of file operations that do utilize get_xip_page can be found in
mm/filemap_xip.c . The following file operation implementations are provided:
- aio_read/aio_write
- readv/writev
- sendfile
The generic file operations do_sync_read/do_sync_write can be used to implement
classic synchronous IO calls.
Shortcomings
------------
This implementation is limited to storage devices that are cpu addressable at
all times (no highmem or such). It works well on rom/ram, but enhancements are
needed to make it work with flash in read+write mode.
Putting the Linux kernel and/or its modules on a xip filesystem does not mean
they are not copied.

111
Documentation/hwmon/adm1021 Normal file
View File

@ -0,0 +1,111 @@
Kernel driver adm1021
=====================
Supported chips:
* Analog Devices ADM1021
Prefix: 'adm1021'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Analog Devices website
* Analog Devices ADM1021A/ADM1023
Prefix: 'adm1023'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Analog Devices website
* Genesys Logic GL523SM
Prefix: 'gl523sm'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet:
* Intel Xeon Processor
Prefix: - any other - may require 'force_adm1021' parameter
Addresses scanned: none
Datasheet: Publicly available at Intel website
* Maxim MAX1617
Prefix: 'max1617'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Maxim website
* Maxim MAX1617A
Prefix: 'max1617a'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Maxim website
* National Semiconductor LM84
Prefix: 'lm84'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the National Semiconductor website
* Philips NE1617
Prefix: 'max1617' (probably detected as a max1617)
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Philips website
* Philips NE1617A
Prefix: 'max1617' (probably detected as a max1617)
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Philips website
* TI THMC10
Prefix: 'thmc10'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the TI website
* Onsemi MC1066
Prefix: 'mc1066'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the Onsemi website
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>
Module Parameters
-----------------
* read_only: int
Don't set any values, read only mode
Description
-----------
The chips supported by this driver are very similar. The Maxim MAX1617 is
the oldest; it has the problem that it is not very well detectable. The
MAX1617A solves that. The ADM1021 is a straight clone of the MAX1617A.
Ditto for the THMC10. From here on, we will refer to all these chips as
ADM1021-clones.
The ADM1021 and MAX1617A reports a die code, which is a sort of revision
code. This can help us pinpoint problems; it is not very useful
otherwise.
ADM1021-clones implement two temperature sensors. One of them is internal,
and measures the temperature of the chip itself; the other is external and
is realised in the form of a transistor-like device. A special alarm
indicates whether the remote sensor is connected.
Each sensor has its own low and high limits. When they are crossed, the
corresponding alarm is set and remains on as long as the temperature stays
out of range. Temperatures are measured in degrees Celsius. Measurements
are possible between -65 and +127 degrees, with a resolution of one degree.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may already
have disappeared!
This driver only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values. It is possible to make
ADM1021-clones do faster measurements, but there is really no good reason
for that.
Xeon support
------------
Some Xeon processors have real max1617, adm1021, or compatible chips
within them, with two temperature sensors.
Other Xeons have chips with only one sensor.
If you have a Xeon, and the adm1021 module loads, and both temperatures
appear valid, then things are good.
If the adm1021 module doesn't load, you should try this:
modprobe adm1021 force_adm1021=BUS,ADDRESS
ADDRESS can only be 0x18, 0x1a, 0x29, 0x2b, 0x4c, or 0x4e.
If you have dual Xeons you may have appear to have two separate
adm1021-compatible chips, or two single-temperature sensors, at distinct
addresses.

51
Documentation/hwmon/adm1025 Normal file
View File

@ -0,0 +1,51 @@
Kernel driver adm1025
=====================
Supported chips:
* Analog Devices ADM1025, ADM1025A
Prefix: 'adm1025'
Addresses scanned: I2C 0x2c - 0x2e
Datasheet: Publicly available at the Analog Devices website
* Philips NE1619
Prefix: 'ne1619'
Addresses scanned: I2C 0x2c - 0x2d
Datasheet: Publicly available at the Philips website
The NE1619 presents some differences with the original ADM1025:
* Only two possible addresses (0x2c - 0x2d).
* No temperature offset register, but we don't use it anyway.
* No INT mode for pin 16. We don't play with it anyway.
Authors:
Chen-Yuan Wu <gwu@esoft.com>,
Jean Delvare <khali@linux-fr.org>
Description
-----------
(This is from Analog Devices.) The ADM1025 is a complete system hardware
monitor for microprocessor-based systems, providing measurement and limit
comparison of various system parameters. Five voltage measurement inputs
are provided, for monitoring +2.5V, +3.3V, +5V and +12V power supplies and
the processor core voltage. The ADM1025 can monitor a sixth power-supply
voltage by measuring its own VCC. One input (two pins) is dedicated to a
remote temperature-sensing diode and an on-chip temperature sensor allows
ambient temperature to be monitored.
One specificity of this chip is that the pin 11 can be hardwired in two
different manners. It can act as the +12V power-supply voltage analog
input, or as the a fifth digital entry for the VID reading (bit 4). It's
kind of strange since both are useful, and the reason for designing the
chip that way is obscure at least to me. The bit 5 of the configuration
register can be used to define how the chip is hardwired. Please note that
it is not a choice you have to make as the user. The choice was already
made by your motherboard's maker. If the configuration bit isn't set
properly, you'll have a wrong +12V reading or a wrong VID reading. The way
the driver handles that is to preserve this bit through the initialization
process, assuming that the BIOS set it up properly beforehand. If it turns
out not to be true in some cases, we'll provide a module parameter to force
modes.
This driver also supports the ADM1025A, which differs from the ADM1025
only in that it has "open-drain VID inputs while the ADM1025 has on-chip
100k pull-ups on the VID inputs". It doesn't make any difference for us.

93
Documentation/hwmon/adm1026 Normal file
View File

@ -0,0 +1,93 @@
Kernel driver adm1026
=====================
Supported chips:
* Analog Devices ADM1026
Prefix: 'adm1026'
Addresses scanned: I2C 0x2c, 0x2d, 0x2e
Datasheet: Publicly available at the Analog Devices website
http://www.analog.com/en/prod/0,,766_825_ADM1026,00.html
Authors:
Philip Pokorny <ppokorny@penguincomputing.com> for Penguin Computing
Justin Thiessen <jthiessen@penguincomputing.com>
Module Parameters
-----------------
* gpio_input: int array (min = 1, max = 17)
List of GPIO pins (0-16) to program as inputs
* gpio_output: int array (min = 1, max = 17)
List of GPIO pins (0-16) to program as outputs
* gpio_inverted: int array (min = 1, max = 17)
List of GPIO pins (0-16) to program as inverted
* gpio_normal: int array (min = 1, max = 17)
List of GPIO pins (0-16) to program as normal/non-inverted
* gpio_fan: int array (min = 1, max = 8)
List of GPIO pins (0-7) to program as fan tachs
Description
-----------
This driver implements support for the Analog Devices ADM1026. Analog
Devices calls it a "complete thermal system management controller."
The ADM1026 implements three (3) temperature sensors, 17 voltage sensors,
16 general purpose digital I/O lines, eight (8) fan speed sensors (8-bit),
an analog output and a PWM output along with limit, alarm and mask bits for
all of the above. There is even 8k bytes of EEPROM memory on chip.
Temperatures are measured in degrees Celsius. There are two external
sensor inputs and one internal sensor. Each sensor has a high and low
limit. If the limit is exceeded, an interrupt (#SMBALERT) can be
generated. The interrupts can be masked. In addition, there are over-temp
limits for each sensor. If this limit is exceeded, the #THERM output will
be asserted. The current temperature and limits have a resolution of 1
degree.
Fan rotation speeds are reported in RPM (rotations per minute) but measured
in counts of a 22.5kHz internal clock. Each fan has a high limit which
corresponds to a minimum fan speed. If the limit is exceeded, an interrupt
can be generated. Each fan can be programmed to divide the reference clock
by 1, 2, 4 or 8. Not all RPM values can accurately be represented, so some
rounding is done. With a divider of 8, the slowest measurable speed of a
two pulse per revolution fan is 661 RPM.
There are 17 voltage sensors. An alarm is triggered if the voltage has
crossed a programmable minimum or maximum limit. Note that minimum in this
case always means 'closest to zero'; this is important for negative voltage
measurements. Several inputs have integrated attenuators so they can measure
higher voltages directly. 3.3V, 5V, 12V, -12V and battery voltage all have
dedicated inputs. There are several inputs scaled to 0-3V full-scale range
for SCSI terminator power. The remaining inputs are not scaled and have
a 0-2.5V full-scale range. A 2.5V or 1.82V reference voltage is provided
for negative voltage measurements.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may already
have disappeared! Note that in the current implementation, all hardware
registers are read whenever any data is read (unless it is less than 2.0
seconds since the last update). This means that you can easily miss
once-only alarms.
The ADM1026 measures continuously. Analog inputs are measured about 4
times a second. Fan speed measurement time depends on fan speed and
divisor. It can take as long as 1.5 seconds to measure all fan speeds.
The ADM1026 has the ability to automatically control fan speed based on the
temperature sensor inputs. Both the PWM output and the DAC output can be
used to control fan speed. Usually only one of these two outputs will be
used. Write the minimum PWM or DAC value to the appropriate control
register. Then set the low temperature limit in the tmin values for each
temperature sensor. The range of control is fixed at 20 °C, and the
largest difference between current and tmin of the temperature sensors sets
the control output. See the datasheet for several example circuits for
controlling fan speed with the PWM and DAC outputs. The fan speed sensors
do not have PWM compensation, so it is probably best to control the fan
voltage from the power lead rather than on the ground lead.
The datasheet shows an example application with VID signals attached to
GPIO lines. Unfortunately, the chip may not be connected to the VID lines
in this way. The driver assumes that the chips *is* connected this way to
get a VID voltage.

35
Documentation/hwmon/adm1031 Normal file
View File

@ -0,0 +1,35 @@
Kernel driver adm1031
=====================
Supported chips:
* Analog Devices ADM1030
Prefix: 'adm1030'
Addresses scanned: I2C 0x2c to 0x2e
Datasheet: Publicly available at the Analog Devices website
http://products.analog.com/products/info.asp?product=ADM1030
* Analog Devices ADM1031
Prefix: 'adm1031'
Addresses scanned: I2C 0x2c to 0x2e
Datasheet: Publicly available at the Analog Devices website
http://products.analog.com/products/info.asp?product=ADM1031
Authors:
Alexandre d'Alton <alex@alexdalton.org>
Jean Delvare <khali@linux-fr.org>
Description
-----------
The ADM1030 and ADM1031 are digital temperature sensors and fan controllers.
They sense their own temperature as well as the temperature of up to one
(ADM1030) or two (ADM1031) external diodes.
All temperature values are given in degrees Celsius. Resolution is 0.5
degree for the local temperature, 0.125 degree for the remote temperatures.
Each temperature channel has its own high and low limits, plus a critical
limit.
The ADM1030 monitors a single fan speed, while the ADM1031 monitors up to
two. Each fan channel has its own low speed limit.

177
Documentation/hwmon/adm9240 Normal file
View File

@ -0,0 +1,177 @@
Kernel driver adm9240
=====================
Supported chips:
* Analog Devices ADM9240
Prefix: 'adm9240'
Addresses scanned: I2C 0x2c - 0x2f
Datasheet: Publicly available at the Analog Devices website
http://www.analog.com/UploadedFiles/Data_Sheets/79857778ADM9240_0.pdf
* Dallas Semiconductor DS1780
Prefix: 'ds1780'
Addresses scanned: I2C 0x2c - 0x2f
Datasheet: Publicly available at the Dallas Semiconductor (Maxim) website
http://pdfserv.maxim-ic.com/en/ds/DS1780.pdf
* National Semiconductor LM81
Prefix: 'lm81'
Addresses scanned: I2C 0x2c - 0x2f
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/ds.cgi/LM/LM81.pdf
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>,
Michiel Rook <michiel@grendelproject.nl>,
Grant Coady <gcoady@gmail.com> with guidance
from Jean Delvare <khali@linux-fr.org>
Interface
---------
The I2C addresses listed above assume BIOS has not changed the
chip MSB 5-bit address. Each chip reports a unique manufacturer
identification code as well as the chip revision/stepping level.
Description
-----------
[From ADM9240] The ADM9240 is a complete system hardware monitor for
microprocessor-based systems, providing measurement and limit comparison
of up to four power supplies and two processor core voltages, plus
temperature, two fan speeds and chassis intrusion. Measured values can
be read out via an I2C-compatible serial System Management Bus, and values
for limit comparisons can be programmed in over the same serial bus. The
high speed successive approximation ADC allows frequent sampling of all
analog channels to ensure a fast interrupt response to any out-of-limit
measurement.
The ADM9240, DS1780 and LM81 are register compatible, the following
details are common to the three chips. Chip differences are described
after this section.
Measurements
------------
The measurement cycle
The adm9240 driver will take a measurement reading no faster than once
each two seconds. User-space may read sysfs interface faster than the
measurement update rate and will receive cached data from the most
recent measurement.
ADM9240 has a very fast 320us temperature and voltage measurement cycle
with independent fan speed measurement cycles counting alternating rising
edges of the fan tacho inputs.
DS1780 measurement cycle is about once per second including fan speed.
LM81 measurement cycle is about once per 400ms including fan speed.
The LM81 12-bit extended temperature measurement mode is not supported.
Temperature
-----------
On chip temperature is reported as degrees Celsius as 9-bit signed data
with resolution of 0.5 degrees Celsius. High and low temperature limits
are 8-bit signed data with resolution of one degree Celsius.
Temperature alarm is asserted once the temperature exceeds the high limit,
and is cleared when the temperature falls below the temp1_max_hyst value.
Fan Speed
---------
Two fan tacho inputs are provided, the ADM9240 gates an internal 22.5kHz
clock via a divider to an 8-bit counter. Fan speed (rpm) is calculated by:
rpm = (22500 * 60) / (count * divider)
Automatic fan clock divider
* User sets 0 to fan_min limit
- low speed alarm is disabled
- fan clock divider not changed
- auto fan clock adjuster enabled for valid fan speed reading
* User sets fan_min limit too low
- low speed alarm is enabled
- fan clock divider set to max
- fan_min set to register value 254 which corresponds
to 664 rpm on adm9240
- low speed alarm will be asserted if fan speed is
less than minimum measurable speed
- auto fan clock adjuster disabled
* User sets reasonable fan speed
- low speed alarm is enabled
- fan clock divider set to suit fan_min
- auto fan clock adjuster enabled: adjusts fan_min
* User sets unreasonably high low fan speed limit
- resolution of the low speed limit may be reduced
- alarm will be asserted
- auto fan clock adjuster enabled: adjusts fan_min
* fan speed may be displayed as zero until the auto fan clock divider
adjuster brings fan speed clock divider back into chip measurement
range, this will occur within a few measurement cycles.
Analog Output
-------------
An analog output provides a 0 to 1.25 volt signal intended for an external
fan speed amplifier circuit. The analog output is set to maximum value on
power up or reset. This doesn't do much on the test Intel SE440BX-2.
Voltage Monitor
Voltage (IN) measurement is internally scaled:
nr label nominal maximum resolution
mV mV mV
0 +2.5V 2500 3320 13.0
1 Vccp1 2700 3600 14.1
2 +3.3V 3300 4380 17.2
3 +5V 5000 6640 26.0
4 +12V 12000 15940 62.5
5 Vccp2 2700 3600 14.1
The reading is an unsigned 8-bit value, nominal voltage measurement is
represented by a reading of 192, being 3/4 of the measurement range.
An alarm is asserted for any voltage going below or above the set limits.
The driver reports and accepts voltage limits scaled to the above table.
VID Monitor
-----------
The chip has five inputs to read the 5-bit VID and reports the mV value
based on detected CPU type.
Chassis Intrusion
-----------------
An alarm is asserted when the CI pin goes active high. The ADM9240
Datasheet has an example of an external temperature sensor driving
this pin. On an Intel SE440BX-2 the Chassis Intrusion header is
connected to a normally open switch.
The ADM9240 provides an internal open drain on this line, and may output
a 20 ms active low pulse to reset an external Chassis Intrusion latch.
Clear the CI latch by writing value 1 to the sysfs chassis_clear file.
Alarm flags reported as 16-bit word
bit label comment
--- ------------- --------------------------
0 +2.5 V_Error high or low limit exceeded
1 VCCP_Error high or low limit exceeded
2 +3.3 V_Error high or low limit exceeded
3 +5 V_Error high or low limit exceeded
4 Temp_Error temperature error
6 FAN1_Error fan low limit exceeded
7 FAN2_Error fan low limit exceeded
8 +12 V_Error high or low limit exceeded
9 VCCP2_Error high or low limit exceeded
12 Chassis_Error CI pin went high
Remaining bits are reserved and thus undefined. It is important to note
that alarm bits may be cleared on read, user-space may latch alarms and
provide the end-user with a method to clear alarm memory.

72
Documentation/hwmon/asb100 Normal file
View File

@ -0,0 +1,72 @@
Kernel driver asb100
====================
Supported Chips:
* Asus ASB100 and ASB100-A "Bach"
Prefix: 'asb100'
Addresses scanned: I2C 0x2d
Datasheet: none released
Author: Mark M. Hoffman <mhoffman@lightlink.com>
Description
-----------
This driver implements support for the Asus ASB100 and ASB100-A "Bach".
These are custom ASICs available only on Asus mainboards. Asus refuses to
supply a datasheet for these chips. Thanks go to many people who helped
investigate their hardware, including:
Vitaly V. Bursov
Alexander van Kaam (author of MBM for Windows)
Bertrik Sikken
The ASB100 implements seven voltage sensors, three fan rotation speed
sensors, four temperature sensors, VID lines and alarms. In addition to
these, the ASB100-A also implements a single PWM controller for fans 2 and
3 (i.e. one setting controls both.) If you have a plain ASB100, the PWM
controller will simply not work (or maybe it will for you... it doesn't for
me).
Temperatures are measured and reported in degrees Celsius.
Fan speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit.
Voltage sensors (also known as IN sensors) report values in volts.
The VID lines encode the core voltage value: the voltage level your
processor should work with. This is hardcoded by the mainboard and/or
processor itself. It is a value in volts.
Alarms: (TODO question marks indicate may or may not work)
0x0001 => in0 (?)
0x0002 => in1 (?)
0x0004 => in2
0x0008 => in3
0x0010 => temp1 (1)
0x0020 => temp2
0x0040 => fan1
0x0080 => fan2
0x0100 => in4
0x0200 => in5 (?) (2)
0x0400 => in6 (?) (2)
0x0800 => fan3
0x1000 => chassis switch
0x2000 => temp3
Alarm Notes:
(1) This alarm will only trigger if the hysteresis value is 127C.
I.e. it behaves the same as w83781d.
(2) The min and max registers for these values appear to
be read-only or otherwise stuck at 0x00.
TODO:
* Experiment with fan divisors > 8.
* Experiment with temp. sensor types.
* Are there really 13 voltage inputs? Probably not...
* Cleanups, no doubt...

108
Documentation/hwmon/ds1621 Normal file
View File

@ -0,0 +1,108 @@
Kernel driver ds1621
====================
Supported chips:
* Dallas Semiconductor DS1621
Prefix: 'ds1621'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the Dallas Semiconductor website
http://www.dalsemi.com/
* Dallas Semiconductor DS1625
Prefix: 'ds1621'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the Dallas Semiconductor website
http://www.dalsemi.com/
Authors:
Christian W. Zuckschwerdt <zany@triq.net>
valuable contributions by Jan M. Sendler <sendler@sendler.de>
ported to 2.6 by Aurelien Jarno <aurelien@aurel32.net>
with the help of Jean Delvare <khali@linux-fr.org>
Module Parameters
------------------
* polarity int
Output's polarity: 0 = active high, 1 = active low
Description
-----------
The DS1621 is a (one instance) digital thermometer and thermostat. It has
both high and low temperature limits which can be user defined (i.e.
programmed into non-volatile on-chip registers). Temperature range is -55
degree Celsius to +125 in 0.5 increments. You may convert this into a
Fahrenheit range of -67 to +257 degrees with 0.9 steps. If polarity
parameter is not provided, original value is used.
As for the thermostat, behavior can also be programmed using the polarity
toggle. On the one hand ("heater"), the thermostat output of the chip,
Tout, will trigger when the low limit temperature is met or underrun and
stays high until the high limit is met or exceeded. On the other hand
("cooler"), vice versa. That way "heater" equals "active low", whereas
"conditioner" equals "active high". Please note that the DS1621 data sheet
is somewhat misleading in this point since setting the polarity bit does
not simply invert Tout.
A second thing is that, during extensive testing, Tout showed a tolerance
of up to +/- 0.5 degrees even when compared against precise temperature
readings. Be sure to have a high vs. low temperature limit gap of al least
1.0 degree Celsius to avoid Tout "bouncing", though!
As for alarms, you can read the alarm status of the DS1621 via the 'alarms'
/sys file interface. The result consists mainly of bit 6 and 5 of the
configuration register of the chip; bit 6 (0x40 or 64) is the high alarm
bit and bit 5 (0x20 or 32) the low one. These bits are set when the high or
low limits are met or exceeded and are reset by the module as soon as the
respective temperature ranges are left.
The alarm registers are in no way suitable to find out about the actual
status of Tout. They will only tell you about its history, whether or not
any of the limits have ever been met or exceeded since last power-up or
reset. Be aware: When testing, it showed that the status of Tout can change
with neither of the alarms set.
Temperature conversion of the DS1621 takes up to 1000ms; internal access to
non-volatile registers may last for 10ms or below.
High Accuracy Temperature Reading
---------------------------------
As said before, the temperature issued via the 9-bit i2c-bus data is
somewhat arbitrary. Internally, the temperature conversion is of a
different kind that is explained (not so...) well in the DS1621 data sheet.
To cut the long story short: Inside the DS1621 there are two oscillators,
both of them biassed by a temperature coefficient.
Higher resolution of the temperature reading can be achieved using the
internal projection, which means taking account of REG_COUNT and REG_SLOPE
(the driver manages them):
Taken from Dallas Semiconductors App Note 068: 'Increasing Temperature
Resolution on the DS1620' and App Note 105: 'High Resolution Temperature
Measurement with Dallas Direct-to-Digital Temperature Sensors'
- Read the 9-bit temperature and strip the LSB (Truncate the .5 degs)
- The resulting value is TEMP_READ.
- Then, read REG_COUNT.
- And then, REG_SLOPE.
TEMP = TEMP_READ - 0.25 + ((REG_SLOPE - REG_COUNT) / REG_SLOPE)
Note that this is what the DONE bit in the DS1621 configuration register is
good for: Internally, one temperature conversion takes up to 1000ms. Before
that conversion is complete you will not be able to read valid things out
of REG_COUNT and REG_SLOPE. The DONE bit, as you may have guessed by now,
tells you whether the conversion is complete ("done", in plain English) and
thus, whether the values you read are good or not.
The DS1621 has two modes of operation: "Continuous" conversion, which can
be understood as the default stand-alone mode where the chip gets the
temperature and controls external devices via its Tout pin or tells other
i2c's about it if they care. The other mode is called "1SHOT", that means
that it only figures out about the temperature when it is explicitly told
to do so; this can be seen as power saving mode.
Now if you want to read REG_COUNT and REG_SLOPE, you have to either stop
the continuous conversions until the contents of these registers are valid,
or, in 1SHOT mode, you have to have one conversion made.

169
Documentation/hwmon/fscher Normal file
View File

@ -0,0 +1,169 @@
Kernel driver fscher
====================
Supported chips:
* Fujitsu-Siemens Hermes chip
Prefix: 'fscher'
Addresses scanned: I2C 0x73
Authors:
Reinhard Nissl <rnissl@gmx.de> based on work
from Hermann Jung <hej@odn.de>,
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>
Description
-----------
This driver implements support for the Fujitsu-Siemens Hermes chip. It is
described in the 'Register Set Specification BMC Hermes based Systemboard'
from Fujitsu-Siemens.
The Hermes chip implements a hardware-based system management, e.g. for
controlling fan speed and core voltage. There is also a watchdog counter on
the chip which can trigger an alarm and even shut the system down.
The chip provides three temperature values (CPU, motherboard and
auxiliary), three voltage values (+12V, +5V and battery) and three fans
(power supply, CPU and auxiliary).
Temperatures are measured in degrees Celsius. The resolution is 1 degree.
Fan rotation speeds are reported in RPM (rotations per minute). The value
can be divided by a programmable divider (1, 2 or 4) which is stored on
the chip.
Voltage sensors (also known as "in" sensors) report their values in volts.
All values are reported as final values from the driver. There is no need
for further calculations.
Detailed description
--------------------
Below you'll find a single line description of all the bit values. With
this information, you're able to decode e. g. alarms, wdog, etc. To make
use of the watchdog, you'll need to set the watchdog time and enable the
watchdog. After that it is necessary to restart the watchdog time within
the specified period of time, or a system reset will occur.
* revision
READING & 0xff = 0x??: HERMES revision identification
* alarms
READING & 0x80 = 0x80: CPU throttling active
READING & 0x80 = 0x00: CPU running at full speed
READING & 0x10 = 0x10: software event (see control:1)
READING & 0x10 = 0x00: no software event
READING & 0x08 = 0x08: watchdog event (see wdog:2)
READING & 0x08 = 0x00: no watchdog event
READING & 0x02 = 0x02: thermal event (see temp*:1)
READING & 0x02 = 0x00: no thermal event
READING & 0x01 = 0x01: fan event (see fan*:1)
READING & 0x01 = 0x00: no fan event
READING & 0x13 ! 0x00: ALERT LED is flashing
* control
READING & 0x01 = 0x01: software event
READING & 0x01 = 0x00: no software event
WRITING & 0x01 = 0x01: set software event
WRITING & 0x01 = 0x00: clear software event
* watchdog_control
READING & 0x80 = 0x80: power off on watchdog event while thermal event
READING & 0x80 = 0x00: watchdog power off disabled (just system reset enabled)
READING & 0x40 = 0x40: watchdog timebase 60 seconds (see also wdog:1)
READING & 0x40 = 0x00: watchdog timebase 2 seconds
READING & 0x10 = 0x10: watchdog enabled
READING & 0x10 = 0x00: watchdog disabled
WRITING & 0x80 = 0x80: enable "power off on watchdog event while thermal event"
WRITING & 0x80 = 0x00: disable "power off on watchdog event while thermal event"
WRITING & 0x40 = 0x40: set watchdog timebase to 60 seconds
WRITING & 0x40 = 0x00: set watchdog timebase to 2 seconds
WRITING & 0x20 = 0x20: disable watchdog
WRITING & 0x10 = 0x10: enable watchdog / restart watchdog time
* watchdog_state
READING & 0x02 = 0x02: watchdog system reset occurred
READING & 0x02 = 0x00: no watchdog system reset occurred
WRITING & 0x02 = 0x02: clear watchdog event
* watchdog_preset
READING & 0xff = 0x??: configured watch dog time in units (see wdog:3 0x40)
WRITING & 0xff = 0x??: configure watch dog time in units
* in* (0: +5V, 1: +12V, 2: onboard 3V battery)
READING: actual voltage value
* temp*_status (1: CPU sensor, 2: onboard sensor, 3: auxiliary sensor)
READING & 0x02 = 0x02: thermal event (overtemperature)
READING & 0x02 = 0x00: no thermal event
READING & 0x01 = 0x01: sensor is working
READING & 0x01 = 0x00: sensor is faulty
WRITING & 0x02 = 0x02: clear thermal event
* temp*_input (1: CPU sensor, 2: onboard sensor, 3: auxiliary sensor)
READING: actual temperature value
* fan*_status (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
READING & 0x04 = 0x04: fan event (fan fault)
READING & 0x04 = 0x00: no fan event
WRITING & 0x04 = 0x04: clear fan event
* fan*_div (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
Divisors 2,4 and 8 are supported, both for reading and writing
* fan*_pwm (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
READING & 0xff = 0x00: fan may be switched off
READING & 0xff = 0x01: fan must run at least at minimum speed (supply: 6V)
READING & 0xff = 0xff: fan must run at maximum speed (supply: 12V)
READING & 0xff = 0x??: fan must run at least at given speed (supply: 6V..12V)
WRITING & 0xff = 0x00: fan may be switched off
WRITING & 0xff = 0x01: fan must run at least at minimum speed (supply: 6V)
WRITING & 0xff = 0xff: fan must run at maximum speed (supply: 12V)
WRITING & 0xff = 0x??: fan must run at least at given speed (supply: 6V..12V)
* fan*_input (1: power supply fan, 2: CPU fan, 3: auxiliary fan)
READING: actual RPM value
Limitations
-----------
* Measuring fan speed
It seems that the chip counts "ripples" (typical fans produce 2 ripples per
rotation while VERAX fans produce 18) in a 9-bit register. This register is
read out every second, then the ripple prescaler (2, 4 or 8) is applied and
the result is stored in the 8 bit output register. Due to the limitation of
the counting register to 9 bits, it is impossible to measure a VERAX fan
properly (even with a prescaler of 8). At its maximum speed of 3500 RPM the
fan produces 1080 ripples per second which causes the counting register to
overflow twice, leading to only 186 RPM.
* Measuring input voltages
in2 ("battery") reports the voltage of the onboard lithium battery and not
+3.3V from the power supply.
* Undocumented features
Fujitsu-Siemens Computers has not documented all features of the chip so
far. Their software, System Guard, shows that there are a still some
features which cannot be controlled by this implementation.

74
Documentation/hwmon/gl518sm Normal file
View File

@ -0,0 +1,74 @@
Kernel driver gl518sm
=====================
Supported chips:
* Genesys Logic GL518SM release 0x00
Prefix: 'gl518sm'
Addresses scanned: I2C 0x2c and 0x2d
Datasheet: http://www.genesyslogic.com/pdf
* Genesys Logic GL518SM release 0x80
Prefix: 'gl518sm'
Addresses scanned: I2C 0x2c and 0x2d
Datasheet: http://www.genesyslogic.com/pdf
Authors:
Frodo Looijaard <frodol@dds.nl>,
Kyösti Mälkki <kmalkki@cc.hut.fi>
Hong-Gunn Chew <hglinux@gunnet.org>
Jean Delvare <khali@linux-fr.org>
Description
-----------
IMPORTANT:
For the revision 0x00 chip, the in0, in1, and in2 values (+5V, +3V,
and +12V) CANNOT be read. This is a limitation of the chip, not the driver.
This driver supports the Genesys Logic GL518SM chip. There are at least
two revision of this chip, which we call revision 0x00 and 0x80. Revision
0x80 chips support the reading of all voltages and revision 0x00 only
for VIN3.
The GL518SM implements one temperature sensor, two fan rotation speed
sensors, and four voltage sensors. It can report alarms through the
computer speakers.
Temperatures are measured in degrees Celsius. An alarm goes off while the
temperature is above the over temperature limit, and has not yet dropped
below the hysteresis limit. The alarm always reflects the current
situation. Measurements are guaranteed between -10 degrees and +110
degrees, with a accuracy of +/-3 degrees.
Rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. In
case when you have selected to turn fan1 off, no fan1 alarm is triggered.
Fan readings can be divided by a programmable divider (1, 2, 4 or 8) to
give the readings more range or accuracy. Not all RPM values can
accurately be represented, so some rounding is done. With a divider
of 2, the lowest representable value is around 1900 RPM.
Voltage sensors (also known as VIN sensors) report their values in volts.
An alarm is triggered if the voltage has crossed a programmable minimum or
maximum limit. Note that minimum in this case always means 'closest to
zero'; this is important for negative voltage measurements. The VDD input
measures voltages between 0.000 and 5.865 volt, with a resolution of 0.023
volt. The other inputs measure voltages between 0.000 and 4.845 volt, with
a resolution of 0.019 volt. Note that revision 0x00 chips do not support
reading the current voltage of any input except for VIN3; limit setting and
alarms work fine, though.
When an alarm is triggered, you can be warned by a beeping signal through your
computer speaker. It is possible to enable all beeping globally, or only the
beeping for some alarms.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once (except for temperature alarms). This means that the
cause for the alarm may already have disappeared! Note that in the current
implementation, all hardware registers are read whenever any data is read
(unless it is less than 1.5 seconds since the last update). This means that
you can easily miss once-only alarms.
The GL518SM only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.

96
Documentation/hwmon/it87 Normal file
View File

@ -0,0 +1,96 @@
Kernel driver it87
==================
Supported chips:
* IT8705F
Prefix: 'it87'
Addresses scanned: from Super I/O config space, or default ISA 0x290 (8 I/O ports)
Datasheet: Publicly available at the ITE website
http://www.ite.com.tw/
* IT8712F
Prefix: 'it8712'
Addresses scanned: I2C 0x28 - 0x2f
from Super I/O config space, or default ISA 0x290 (8 I/O ports)
Datasheet: Publicly available at the ITE website
http://www.ite.com.tw/
* SiS950 [clone of IT8705F]
Prefix: 'sis950'
Addresses scanned: from Super I/O config space, or default ISA 0x290 (8 I/O ports)
Datasheet: No longer be available
Author: Christophe Gauthron <chrisg@0-in.com>
Module Parameters
-----------------
* update_vbat: int
0 if vbat should report power on value, 1 if vbat should be updated after
each read. Default is 0. On some boards the battery voltage is provided
by either the battery or the onboard power supply. Only the first reading
at power on will be the actual battery voltage (which the chip does
automatically). On other boards the battery voltage is always fed to
the chip so can be read at any time. Excessive reading may decrease
battery life but no information is given in the datasheet.
* fix_pwm_polarity int
Force PWM polarity to active high (DANGEROUS). Some chips are
misconfigured by BIOS - PWM values would be inverted. This option tries
to fix this. Please contact your BIOS manufacturer and ask him for fix.
Description
-----------
This driver implements support for the IT8705F, IT8712F and SiS950 chips.
This driver also supports IT8712F, which adds SMBus access, and a VID
input, used to report the Vcore voltage of the Pentium processor.
The IT8712F additionally features VID inputs.
These chips are 'Super I/O chips', supporting floppy disks, infrared ports,
joysticks and other miscellaneous stuff. For hardware monitoring, they
include an 'environment controller' with 3 temperature sensors, 3 fan
rotation speed sensors, 8 voltage sensors, and associated alarms.
Temperatures are measured in degrees Celsius. An alarm is triggered once
when the Overtemperature Shutdown limit is crossed.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give the
readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in volts. An
alarm is triggered if the voltage has crossed a programmable minimum or
maximum limit. Note that minimum in this case always means 'closest to
zero'; this is important for negative voltage measurements. All voltage
inputs can measure voltages between 0 and 4.08 volts, with a resolution of
0.016 volt. The battery voltage in8 does not have limit registers.
The VID lines (IT8712F only) encode the core voltage value: the voltage
level your processor should work with. This is hardcoded by the mainboard
and/or processor itself. It is a value in volts.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may already
have disappeared! Note that in the current implementation, all hardware
registers are read whenever any data is read (unless it is less than 1.5
seconds since the last update). This means that you can easily miss
once-only alarms.
The IT87xx only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.
To change sensor N to a thermistor, 'echo 2 > tempN_type' where N is 1, 2,
or 3. To change sensor N to a thermal diode, 'echo 3 > tempN_type'.
Give 0 for unused sensor. Any other value is invalid. To configure this at
startup, consult lm_sensors's /etc/sensors.conf. (2 = thermistor;
3 = thermal diode)
The fan speed control features are limited to manual PWM mode. Automatic
"Smart Guardian" mode control handling is not implemented. However
if you want to go for "manual mode" just write 1 to pwmN_enable.

57
Documentation/hwmon/lm63 Normal file
View File

@ -0,0 +1,57 @@
Kernel driver lm63
==================
Supported chips:
* National Semiconductor LM63
Prefix: 'lm63'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/pf/LM/LM63.html
Author: Jean Delvare <khali@linux-fr.org>
Thanks go to Tyan and especially Alex Buckingham for setting up a remote
access to their S4882 test platform for this driver.
http://www.tyan.com/
Description
-----------
The LM63 is a digital temperature sensor with integrated fan monitoring
and control.
The LM63 is basically an LM86 with fan speed monitoring and control
capabilities added. It misses some of the LM86 features though:
- No low limit for local temperature.
- No critical limit for local temperature.
- Critical limit for remote temperature can be changed only once. We
will consider that the critical limit is read-only.
The datasheet isn't very clear about what the tachometer reading is.
An explanation from National Semiconductor: The two lower bits of the read
value have to be masked out. The value is still 16 bit in width.
All temperature values are given in degrees Celsius. Resolution is 1.0
degree for the local temperature, 0.125 degree for the remote temperature.
The fan speed is measured using a tachometer. Contrary to most chips which
store the value in an 8-bit register and have a selectable clock divider
to make sure that the result will fit in the register, the LM63 uses 16-bit
value for measuring the speed of the fan. It can measure fan speeds down to
83 RPM, at least in theory.
Note that the pin used for fan monitoring is shared with an alert out
function. Depending on how the board designer wanted to use the chip, fan
speed monitoring will or will not be possible. The proper chip configuration
is left to the BIOS, and the driver will blindly trust it.
A PWM output can be used to control the speed of the fan. The LM63 has two
PWM modes: manual and automatic. Automatic mode is not fully implemented yet
(you cannot define your custom PWM/temperature curve), and mode change isn't
supported either.
The lm63 driver will not update its values more frequently than every
second; reading them more often will do no harm, but will return 'old'
values.

65
Documentation/hwmon/lm75 Normal file
View File

@ -0,0 +1,65 @@
Kernel driver lm75
==================
Supported chips:
* National Semiconductor LM75
Prefix: 'lm75'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/
* Dallas Semiconductor DS75
Prefix: 'lm75'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the Dallas Semiconductor website
http://www.maxim-ic.com/
* Dallas Semiconductor DS1775
Prefix: 'lm75'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the Dallas Semiconductor website
http://www.maxim-ic.com/
* Maxim MAX6625, MAX6626
Prefix: 'lm75'
Addresses scanned: I2C 0x48 - 0x4b
Datasheet: Publicly available at the Maxim website
http://www.maxim-ic.com/
* Microchip (TelCom) TCN75
Prefix: 'lm75'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the Microchip website
http://www.microchip.com/
Author: Frodo Looijaard <frodol@dds.nl>
Description
-----------
The LM75 implements one temperature sensor. Limits can be set through the
Overtemperature Shutdown register and Hysteresis register. Each value can be
set and read to half-degree accuracy.
An alarm is issued (usually to a connected LM78) when the temperature
gets higher then the Overtemperature Shutdown value; it stays on until
the temperature falls below the Hysteresis value.
All temperatures are in degrees Celsius, and are guaranteed within a
range of -55 to +125 degrees.
The LM75 only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.
The LM75 is usually used in combination with LM78-like chips, to measure
the temperature of the processor(s).
The DS75, DS1775, MAX6625, and MAX6626 are supported as well.
They are not distinguished from an LM75. While most of these chips
have three additional bits of accuracy (12 vs. 9 for the LM75),
the additional bits are not supported. Not only that, but these chips will
not be detected if not in 9-bit precision mode (use the force parameter if
needed).
The TCN75 is supported as well, and is not distinguished from an LM75.
The LM75 is essentially an industry standard; there may be other
LM75 clones not listed here, with or without various enhancements,
that are supported.
The LM77 is not supported, contrary to what we pretended for a long time.
Both chips are simply not compatible, value encoding differs.

22
Documentation/hwmon/lm77 Normal file
View File

@ -0,0 +1,22 @@
Kernel driver lm77
==================
Supported chips:
* National Semiconductor LM77
Prefix: 'lm77'
Addresses scanned: I2C 0x48 - 0x4b
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/
Author: Andras BALI <drewie@freemail.hu>
Description
-----------
The LM77 implements one temperature sensor. The temperature
sensor incorporates a band-gap type temperature sensor,
10-bit ADC, and a digital comparator with user-programmable upper
and lower limit values.
Limits can be set through the Overtemperature Shutdown register and
Hysteresis register.

82
Documentation/hwmon/lm78 Normal file
View File

@ -0,0 +1,82 @@
Kernel driver lm78
==================
Supported chips:
* National Semiconductor LM78
Prefix: 'lm78'
Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/
* National Semiconductor LM78-J
Prefix: 'lm78-j'
Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/
* National Semiconductor LM79
Prefix: 'lm79'
Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/
Author: Frodo Looijaard <frodol@dds.nl>
Description
-----------
This driver implements support for the National Semiconductor LM78, LM78-J
and LM79. They are described as 'Microprocessor System Hardware Monitors'.
There is almost no difference between the three supported chips. Functionally,
the LM78 and LM78-J are exactly identical. The LM79 has one more VID line,
which is used to report the lower voltages newer Pentium processors use.
From here on, LM7* means either of these three types.
The LM7* implements one temperature sensor, three fan rotation speed sensors,
seven voltage sensors, VID lines, alarms, and some miscellaneous stuff.
Temperatures are measured in degrees Celsius. An alarm is triggered once
when the Overtemperature Shutdown limit is crossed; it is triggered again
as soon as it drops below the Hysteresis value. A more useful behavior
can be found by setting the Hysteresis value to +127 degrees Celsius; in
this case, alarms are issued during all the time when the actual temperature
is above the Overtemperature Shutdown value. Measurements are guaranteed
between -55 and +125 degrees, with a resolution of 1 degree.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give
the readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in volts.
An alarm is triggered if the voltage has crossed a programmable minimum
or maximum limit. Note that minimum in this case always means 'closest to
zero'; this is important for negative voltage measurements. All voltage
inputs can measure voltages between 0 and 4.08 volts, with a resolution
of 0.016 volt.
The VID lines encode the core voltage value: the voltage level your processor
should work with. This is hardcoded by the mainboard and/or processor itself.
It is a value in volts. When it is unconnected, you will often find the
value 3.50 V here.
In addition to the alarms described above, there are a couple of additional
ones. There is a BTI alarm, which gets triggered when an external chip has
crossed its limits. Usually, this is connected to all LM75 chips; if at
least one crosses its limits, this bit gets set. The CHAS alarm triggers
if your computer case is open. The FIFO alarms should never trigger; it
indicates an internal error. The SMI_IN alarm indicates some other chip
has triggered an SMI interrupt. As we do not use SMI interrupts at all,
this condition usually indicates there is a problem with some other
device.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may
already have disappeared! Note that in the current implementation, all
hardware registers are read whenever any data is read (unless it is less
than 1.5 seconds since the last update). This means that you can easily
miss once-only alarms.
The LM7* only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.

56
Documentation/hwmon/lm80 Normal file
View File

@ -0,0 +1,56 @@
Kernel driver lm80
==================
Supported chips:
* National Semiconductor LM80
Prefix: 'lm80'
Addresses scanned: I2C 0x28 - 0x2f
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>
Description
-----------
This driver implements support for the National Semiconductor LM80.
It is described as a 'Serial Interface ACPI-Compatible Microprocessor
System Hardware Monitor'.
The LM80 implements one temperature sensor, two fan rotation speed sensors,
seven voltage sensors, alarms, and some miscellaneous stuff.
Temperatures are measured in degrees Celsius. There are two sets of limits
which operate independently. When the HOT Temperature Limit is crossed,
this will cause an alarm that will be reasserted until the temperature
drops below the HOT Hysteresis. The Overtemperature Shutdown (OS) limits
should work in the same way (but this must be checked; the datasheet
is unclear about this). Measurements are guaranteed between -55 and
+125 degrees. The current temperature measurement has a resolution of
0.0625 degrees; the limits have a resolution of 1 degree.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give
the readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in volts.
An alarm is triggered if the voltage has crossed a programmable minimum
or maximum limit. Note that minimum in this case always means 'closest to
zero'; this is important for negative voltage measurements. All voltage
inputs can measure voltages between 0 and 2.55 volts, with a resolution
of 0.01 volt.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may
already have disappeared! Note that in the current implementation, all
hardware registers are read whenever any data is read (unless it is less
than 2.0 seconds since the last update). This means that you can easily
miss once-only alarms.
The LM80 only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.

76
Documentation/hwmon/lm83 Normal file
View File

@ -0,0 +1,76 @@
Kernel driver lm83
==================
Supported chips:
* National Semiconductor LM83
Prefix: 'lm83'
Addresses scanned: I2C 0x18 - 0x1a, 0x29 - 0x2b, 0x4c - 0x4e
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/pf/LM/LM83.html
Author: Jean Delvare <khali@linux-fr.org>
Description
-----------
The LM83 is a digital temperature sensor. It senses its own temperature as
well as the temperature of up to three external diodes. It is compatible
with many other devices such as the LM84 and all other ADM1021 clones.
The main difference between the LM83 and the LM84 in that the later can
only sense the temperature of one external diode.
Using the adm1021 driver for a LM83 should work, but only two temperatures
will be reported instead of four.
The LM83 is only found on a handful of motherboards. Both a confirmed
list and an unconfirmed list follow. If you can confirm or infirm the
fact that any of these motherboards do actually have an LM83, please
contact us. Note that the LM90 can easily be misdetected as a LM83.
Confirmed motherboards:
SBS P014
Unconfirmed motherboards:
Gigabyte GA-8IK1100
Iwill MPX2
Soltek SL-75DRV5
The driver has been successfully tested by Magnus Forsström, who I'd
like to thank here. More testers will be of course welcome.
The fact that the LM83 is only scarcely used can be easily explained.
Most motherboards come with more than just temperature sensors for
health monitoring. They also have voltage and fan rotation speed
sensors. This means that temperature-only chips are usually used as
secondary chips coupled with another chip such as an IT8705F or similar
chip, which provides more features. Since systems usually need three
temperature sensors (motherboard, processor, power supply) and primary
chips provide some temperature sensors, the secondary chip, if needed,
won't have to handle more than two temperatures. Thus, ADM1021 clones
are sufficient, and there is no need for a four temperatures sensor
chip such as the LM83. The only case where using an LM83 would make
sense is on SMP systems, such as the above-mentioned Iwill MPX2,
because you want an additional temperature sensor for each additional
CPU.
On the SBS P014, this is different, since the LM83 is the only hardware
monitoring chipset. One temperature sensor is used for the motherboard
(actually measuring the LM83's own temperature), one is used for the
CPU. The two other sensors must be used to measure the temperature of
two other points of the motherboard. We suspect these points to be the
north and south bridges, but this couldn't be confirmed.
All temperature values are given in degrees Celsius. Local temperature
is given within a range of 0 to +85 degrees. Remote temperatures are
given within a range of 0 to +125 degrees. Resolution is 1.0 degree,
accuracy is guaranteed to 3.0 degrees (see the datasheet for more
details).
Each sensor has its own high limit, but the critical limit is common to
all four sensors. There is no hysteresis mechanism as found on most
recent temperature sensors.
The lm83 driver will not update its values more frequently than every
other second; reading them more often will do no harm, but will return
'old' values.

221
Documentation/hwmon/lm85 Normal file
View File

@ -0,0 +1,221 @@
Kernel driver lm85
==================
Supported chips:
* National Semiconductor LM85 (B and C versions)
Prefix: 'lm85'
Addresses scanned: I2C 0x2c, 0x2d, 0x2e
Datasheet: http://www.national.com/pf/LM/LM85.html
* Analog Devices ADM1027
Prefix: 'adm1027'
Addresses scanned: I2C 0x2c, 0x2d, 0x2e
Datasheet: http://www.analog.com/en/prod/0,,766_825_ADM1027,00.html
* Analog Devices ADT7463
Prefix: 'adt7463'
Addresses scanned: I2C 0x2c, 0x2d, 0x2e
Datasheet: http://www.analog.com/en/prod/0,,766_825_ADT7463,00.html
* SMSC EMC6D100, SMSC EMC6D101
Prefix: 'emc6d100'
Addresses scanned: I2C 0x2c, 0x2d, 0x2e
Datasheet: http://www.smsc.com/main/tools/discontinued/6d100.pdf
* SMSC EMC6D102
Prefix: 'emc6d102'
Addresses scanned: I2C 0x2c, 0x2d, 0x2e
Datasheet: http://www.smsc.com/main/catalog/emc6d102.html
Authors:
Philip Pokorny <ppokorny@penguincomputing.com>,
Frodo Looijaard <frodol@dds.nl>,
Richard Barrington <rich_b_nz@clear.net.nz>,
Margit Schubert-While <margitsw@t-online.de>,
Justin Thiessen <jthiessen@penguincomputing.com>
Description
-----------
This driver implements support for the National Semiconductor LM85 and
compatible chips including the Analog Devices ADM1027, ADT7463 and
SMSC EMC6D10x chips family.
The LM85 uses the 2-wire interface compatible with the SMBUS 2.0
specification. Using an analog to digital converter it measures three (3)
temperatures and five (5) voltages. It has four (4) 16-bit counters for
measuring fan speed. Five (5) digital inputs are provided for sampling the
VID signals from the processor to the VRM. Lastly, there are three (3) PWM
outputs that can be used to control fan speed.
The voltage inputs have internal scaling resistors so that the following
voltage can be measured without external resistors:
2.5V, 3.3V, 5V, 12V, and CPU core voltage (2.25V)
The temperatures measured are one internal diode, and two remote diodes.
Remote 1 is generally the CPU temperature. These inputs are designed to
measure a thermal diode like the one in a Pentium 4 processor in a socket
423 or socket 478 package. They can also measure temperature using a
transistor like the 2N3904.
A sophisticated control system for the PWM outputs is designed into the
LM85 that allows fan speed to be adjusted automatically based on any of the
three temperature sensors. Each PWM output is individually adjustable and
programmable. Once configured, the LM85 will adjust the PWM outputs in
response to the measured temperatures without further host intervention.
This feature can also be disabled for manual control of the PWM's.
Each of the measured inputs (voltage, temperature, fan speed) has
corresponding high/low limit values. The LM85 will signal an ALARM if any
measured value exceeds either limit.
The LM85 samples all inputs continuously. The lm85 driver will not read
the registers more often than once a second. Further, configuration data is
only read once each 5 minutes. There is twice as much config data as
measurements, so this would seem to be a worthwhile optimization.
Special Features
----------------
The LM85 has four fan speed monitoring modes. The ADM1027 has only two.
Both have special circuitry to compensate for PWM interactions with the
TACH signal from the fans. The ADM1027 can be configured to measure the
speed of a two wire fan, but the input conditioning circuitry is different
for 3-wire and 2-wire mode. For this reason, the 2-wire fan modes are not
exposed to user control. The BIOS should initialize them to the correct
mode. If you've designed your own ADM1027, you'll have to modify the
init_client function and add an insmod parameter to set this up.
To smooth the response of fans to changes in temperature, the LM85 has an
optional filter for smoothing temperatures. The ADM1027 has the same
config option but uses it to rate limit the changes to fan speed instead.
The ADM1027 and ADT7463 have a 10-bit ADC and can therefore measure
temperatures with 0.25 degC resolution. They also provide an offset to the
temperature readings that is automatically applied during measurement.
This offset can be used to zero out any errors due to traces and placement.
The documentation says that the offset is in 0.25 degC steps, but in
initial testing of the ADM1027 it was 1.00 degC steps. Analog Devices has
confirmed this "bug". The ADT7463 is reported to work as described in the
documentation. The current lm85 driver does not show the offset register.
The ADT7463 has a THERM asserted counter. This counter has a 22.76ms
resolution and a range of 5.8 seconds. The driver implements a 32-bit
accumulator of the counter value to extend the range to over a year. The
counter will stay at it's max value until read.
See the vendor datasheets for more information. There is application note
from National (AN-1260) with some additional information about the LM85.
The Analog Devices datasheet is very detailed and describes a procedure for
determining an optimal configuration for the automatic PWM control.
The SMSC EMC6D100 & EMC6D101 monitor external voltages, temperatures, and
fan speeds. They use this monitoring capability to alert the system to out
of limit conditions and can automatically control the speeds of multiple
fans in a PC or embedded system. The EMC6D101, available in a 24-pin SSOP
package, and the EMC6D100, available in a 28-pin SSOP package, are designed
to be register compatible. The EMC6D100 offers all the features of the
EMC6D101 plus additional voltage monitoring and system control features.
Unfortunately it is not possible to distinguish between the package
versions on register level so these additional voltage inputs may read
zero. The EMC6D102 features addtional ADC bits thus extending precision
of voltage and temperature channels.
Hardware Configurations
-----------------------
The LM85 can be jumpered for 3 different SMBus addresses. There are
no other hardware configuration options for the LM85.
The lm85 driver detects both LM85B and LM85C revisions of the chip. See the
datasheet for a complete description of the differences. Other than
identifying the chip, the driver behaves no differently with regard to
these two chips. The LM85B is recommended for new designs.
The ADM1027 and ADT7463 chips have an optional SMBALERT output that can be
used to signal the chipset in case a limit is exceeded or the temperature
sensors fail. Individual sensor interrupts can be masked so they won't
trigger SMBALERT. The SMBALERT output if configured replaces one of the other
functions (PWM2 or IN0). This functionality is not implemented in current
driver.
The ADT7463 also has an optional THERM output/input which can be connected
to the processor PROC_HOT output. If available, the autofan control
dynamic Tmin feature can be enabled to keep the system temperature within
spec (just?!) with the least possible fan noise.
Configuration Notes
-------------------
Besides standard interfaces driver adds following:
* Temperatures and Zones
Each temperature sensor is associated with a Zone. There are three
sensors and therefore three zones (# 1, 2 and 3). Each zone has the following
temperature configuration points:
* temp#_auto_temp_off - temperature below which fans should be off or spinning very low.
* temp#_auto_temp_min - temperature over which fans start to spin.
* temp#_auto_temp_max - temperature when fans spin at full speed.
* temp#_auto_temp_crit - temperature when all fans will run full speed.
* PWM Control
There are three PWM outputs. The LM85 datasheet suggests that the
pwm3 output control both fan3 and fan4. Each PWM can be individually
configured and assigned to a zone for it's control value. Each PWM can be
configured individually according to the following options.
* pwm#_auto_pwm_min - this specifies the PWM value for temp#_auto_temp_off
temperature. (PWM value from 0 to 255)
* pwm#_auto_pwm_freq - select base frequency of PWM output. You can select
in range of 10.0 to 94.0 Hz in .1 Hz units.
(Values 100 to 940).
The pwm#_auto_pwm_freq can be set to one of the following 8 values. Setting the
frequency to a value not on this list, will result in the next higher frequency
being selected. The actual device frequency may vary slightly from this
specification as designed by the manufacturer. Consult the datasheet for more
details. (PWM Frequency values: 100, 150, 230, 300, 380, 470, 620, 940)
* pwm#_auto_pwm_minctl - this flags selects for temp#_auto_temp_off temperature
the bahaviour of fans. Write 1 to let fans spinning at
pwm#_auto_pwm_min or write 0 to let them off.
NOTE: It has been reported that there is a bug in the LM85 that causes the flag
to be associated with the zones not the PWMs. This contradicts all the
published documentation. Setting pwm#_min_ctl in this case actually affects all
PWMs controlled by zone '#'.
* PWM Controlling Zone selection
* pwm#_auto_channels - controls zone that is associated with PWM
Configuration choices:
Value Meaning
------ ------------------------------------------------
1 Controlled by Zone 1
2 Controlled by Zone 2
3 Controlled by Zone 3
23 Controlled by higher temp of Zone 2 or 3
123 Controlled by highest temp of Zone 1, 2 or 3
0 PWM always 0% (off)
-1 PWM always 100% (full on)
-2 Manual control (write to 'pwm#' to set)
The National LM85's have two vendor specific configuration
features. Tach. mode and Spinup Control. For more details on these,
see the LM85 datasheet or Application Note AN-1260.
The Analog Devices ADM1027 has several vendor specific enhancements.
The number of pulses-per-rev of the fans can be set, Tach monitoring
can be optimized for PWM operation, and an offset can be applied to
the temperatures to compensate for systemic errors in the
measurements.
In addition to the ADM1027 features, the ADT7463 also has Tmin control
and THERM asserted counts. Automatic Tmin control acts to adjust the
Tmin value to maintain the measured temperature sensor at a specified
temperature. There isn't much documentation on this feature in the
ADT7463 data sheet. This is not supported by current driver.

73
Documentation/hwmon/lm87 Normal file
View File

@ -0,0 +1,73 @@
Kernel driver lm87
==================
Supported chips:
* National Semiconductor LM87
Prefix: 'lm87'
Addresses scanned: I2C 0x2c - 0x2f
Datasheet: http://www.national.com/pf/LM/LM87.html
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>,
Mark Studebaker <mdsxyz123@yahoo.com>,
Stephen Rousset <stephen.rousset@rocketlogix.com>,
Dan Eaton <dan.eaton@rocketlogix.com>,
Jean Delvare <khali@linux-fr.org>,
Original 2.6 port Jeff Oliver
Description
-----------
This driver implements support for the National Semiconductor LM87.
The LM87 implements up to three temperature sensors, up to two fan
rotation speed sensors, up to seven voltage sensors, alarms, and some
miscellaneous stuff.
Temperatures are measured in degrees Celsius. Each input has a high
and low alarm settings. A high limit produces an alarm when the value
goes above it, and an alarm is also produced when the value goes below
the low limit.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give
the readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in
volts. An alarm is triggered if the voltage has crossed a programmable
minimum or maximum limit. Note that minimum in this case always means
'closest to zero'; this is important for negative voltage measurements.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may
already have disappeared! Note that in the current implementation, all
hardware registers are read whenever any data is read (unless it is less
than 1.0 seconds since the last update). This means that you can easily
miss once-only alarms.
The lm87 driver only updates its values each 1.0 seconds; reading it more
often will do no harm, but will return 'old' values.
Hardware Configurations
-----------------------
The LM87 has four pins which can serve one of two possible functions,
depending on the hardware configuration.
Some functions share pins, so not all functions are available at the same
time. Which are depends on the hardware setup. This driver assumes that
the BIOS configured the chip correctly. In that respect, it differs from
the original driver (from lm_sensors for Linux 2.4), which would force the
LM87 to an arbitrary, compile-time chosen mode, regardless of the actual
chipset wiring.
For reference, here is the list of exclusive functions:
- in0+in5 (default) or temp3
- fan1 (default) or in6
- fan2 (default) or in7
- VID lines (default) or IRQ lines (not handled by this driver)

121
Documentation/hwmon/lm90 Normal file
View File

@ -0,0 +1,121 @@
Kernel driver lm90
==================
Supported chips:
* National Semiconductor LM90
Prefix: 'lm90'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/pf/LM/LM90.html
* National Semiconductor LM89
Prefix: 'lm99'
Addresses scanned: I2C 0x4c and 0x4d
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/pf/LM/LM89.html
* National Semiconductor LM99
Prefix: 'lm99'
Addresses scanned: I2C 0x4c and 0x4d
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/pf/LM/LM99.html
* National Semiconductor LM86
Prefix: 'lm86'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the National Semiconductor website
http://www.national.com/pf/LM/LM86.html
* Analog Devices ADM1032
Prefix: 'adm1032'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the Analog Devices website
http://products.analog.com/products/info.asp?product=ADM1032
* Analog Devices ADT7461
Prefix: 'adt7461'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the Analog Devices website
http://products.analog.com/products/info.asp?product=ADT7461
Note: Only if in ADM1032 compatibility mode
* Maxim MAX6657
Prefix: 'max6657'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the Maxim website
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2578
* Maxim MAX6658
Prefix: 'max6657'
Addresses scanned: I2C 0x4c
Datasheet: Publicly available at the Maxim website
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2578
* Maxim MAX6659
Prefix: 'max6657'
Addresses scanned: I2C 0x4c, 0x4d (unsupported 0x4e)
Datasheet: Publicly available at the Maxim website
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2578
Author: Jean Delvare <khali@linux-fr.org>
Description
-----------
The LM90 is a digital temperature sensor. It senses its own temperature as
well as the temperature of up to one external diode. It is compatible
with many other devices such as the LM86, the LM89, the LM99, the ADM1032,
the MAX6657, MAX6658 and the MAX6659 all of which are supported by this driver.
Note that there is no easy way to differentiate between the last three
variants. The extra address and features of the MAX6659 are not supported by
this driver. Additionally, the ADT7461 is supported if found in ADM1032
compatibility mode.
The specificity of this family of chipsets over the ADM1021/LM84
family is that it features critical limits with hysteresis, and an
increased resolution of the remote temperature measurement.
The different chipsets of the family are not strictly identical, although
very similar. This driver doesn't handle any specific feature for now,
but could if there ever was a need for it. For reference, here comes a
non-exhaustive list of specific features:
LM90:
* Filter and alert configuration register at 0xBF.
* ALERT is triggered by temperatures over critical limits.
LM86 and LM89:
* Same as LM90
* Better external channel accuracy
LM99:
* Same as LM89
* External temperature shifted by 16 degrees down
ADM1032:
* Consecutive alert register at 0x22.
* Conversion averaging.
* Up to 64 conversions/s.
* ALERT is triggered by open remote sensor.
ADT7461
* Extended temperature range (breaks compatibility)
* Lower resolution for remote temperature
MAX6657 and MAX6658:
* Remote sensor type selection
MAX6659
* Selectable address
* Second critical temperature limit
* Remote sensor type selection
All temperature values are given in degrees Celsius. Resolution
is 1.0 degree for the local temperature, 0.125 degree for the remote
temperature.
Each sensor has its own high and low limits, plus a critical limit.
Additionally, there is a relative hysteresis value common to both critical
values. To make life easier to user-space applications, two absolute values
are exported, one for each channel, but these values are of course linked.
Only the local hysteresis can be set from user-space, and the same delta
applies to the remote hysteresis.
The lm90 driver will not update its values more frequently than every
other second; reading them more often will do no harm, but will return
'old' values.

37
Documentation/hwmon/lm92 Normal file
View File

@ -0,0 +1,37 @@
Kernel driver lm92
==================
Supported chips:
* National Semiconductor LM92
Prefix: 'lm92'
Addresses scanned: I2C 0x48 - 0x4b
Datasheet: http://www.national.com/pf/LM/LM92.html
* National Semiconductor LM76
Prefix: 'lm92'
Addresses scanned: none, force parameter needed
Datasheet: http://www.national.com/pf/LM/LM76.html
* Maxim MAX6633/MAX6634/MAX6635
Prefix: 'lm92'
Addresses scanned: I2C 0x48 - 0x4b
MAX6633 with address in 0x40 - 0x47, 0x4c - 0x4f needs force parameter
and MAX6634 with address in 0x4c - 0x4f needs force parameter
Datasheet: http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3074
Authors:
Abraham van der Merwe <abraham@2d3d.co.za>
Jean Delvare <khali@linux-fr.org>
Description
-----------
This driver implements support for the National Semiconductor LM92
temperature sensor.
Each LM92 temperature sensor supports a single temperature sensor. There are
alarms for high, low, and critical thresholds. There's also an hysteresis to
control the thresholds for resetting alarms.
Support was added later for the LM76 and Maxim MAX6633/MAX6634/MAX6635,
which are mostly compatible. They have not all been tested, so you
may need to use the force parameter.

29
Documentation/hwmon/max1619 Normal file
View File

@ -0,0 +1,29 @@
Kernel driver max1619
=====================
Supported chips:
* Maxim MAX1619
Prefix: 'max1619'
Addresses scanned: I2C 0x18-0x1a, 0x29-0x2b, 0x4c-0x4e
Datasheet: Publicly available at the Maxim website
http://pdfserv.maxim-ic.com/en/ds/MAX1619.pdf
Authors:
Alexey Fisher <fishor@mail.ru>,
Jean Delvare <khali@linux-fr.org>
Description
-----------
The MAX1619 is a digital temperature sensor. It senses its own temperature as
well as the temperature of up to one external diode.
All temperature values are given in degrees Celsius. Resolution
is 1.0 degree for the local temperature and for the remote temperature.
Only the external sensor has high and low limits.
The max1619 driver will not update its values more frequently than every
other second; reading them more often will do no harm, but will return
'old' values.

189
Documentation/hwmon/pc87360 Normal file
View File

@ -0,0 +1,189 @@
Kernel driver pc87360
=====================
Supported chips:
* National Semiconductor PC87360, PC87363, PC87364, PC87365 and PC87366
Prefixes: 'pc87360', 'pc87363', 'pc87364', 'pc87365', 'pc87366'
Addresses scanned: none, address read from Super I/O config space
Datasheets:
http://www.national.com/pf/PC/PC87360.html
http://www.national.com/pf/PC/PC87363.html
http://www.national.com/pf/PC/PC87364.html
http://www.national.com/pf/PC/PC87365.html
http://www.national.com/pf/PC/PC87366.html
Authors: Jean Delvare <khali@linux-fr.org>
Thanks to Sandeep Mehta, Tonko de Rooy and Daniel Ceregatti for testing.
Thanks to Rudolf Marek for helping me investigate conversion issues.
Module Parameters
-----------------
* init int
Chip initialization level:
0: None
*1: Forcibly enable internal voltage and temperature channels, except in9
2: Forcibly enable all voltage and temperature channels, except in9
3: Forcibly enable all voltage and temperature channels, including in9
Note that this parameter has no effect for the PC87360, PC87363 and PC87364
chips.
Also note that for the PC87366, initialization levels 2 and 3 don't enable
all temperature channels, because some of them share pins with each other,
so they can't be used at the same time.
Description
-----------
The National Semiconductor PC87360 Super I/O chip contains monitoring and
PWM control circuitry for two fans. The PC87363 chip is similar, and the
PC87364 chip has monitoring and PWM control for a third fan.
The National Semiconductor PC87365 and PC87366 Super I/O chips are complete
hardware monitoring chipsets, not only controlling and monitoring three fans,
but also monitoring eleven voltage inputs and two (PC87365) or up to four
(PC87366) temperatures.
Chip #vin #fan #pwm #temp devid
PC87360 - 2 2 - 0xE1
PC87363 - 2 2 - 0xE8
PC87364 - 3 3 - 0xE4
PC87365 11 3 3 2 0xE5
PC87366 11 3 3 3-4 0xE9
The driver assumes that no more than one chip is present, and one of the
standard Super I/O addresses is used (0x2E/0x2F or 0x4E/0x4F)
Fan Monitoring
--------------
Fan rotation speeds are reported in RPM (revolutions per minute). An alarm
is triggered if the rotation speed has dropped below a programmable limit.
A different alarm is triggered if the fan speed is too low to be measured.
Fan readings are affected by a programmable clock divider, giving the
readings more range or accuracy. Usually, users have to learn how it works,
but this driver implements dynamic clock divider selection, so you don't
have to care no more.
For reference, here are a few values about clock dividers:
slowest accuracy highest
measurable around 3000 accurate
divider speed (RPM) RPM (RPM) speed (RPM)
1 1882 18 6928
2 941 37 4898
4 470 74 3464
8 235 150 2449
For the curious, here is how the values above were computed:
* slowest measurable speed: clock/(255*divider)
* accuracy around 3000 RPM: 3000^2/clock
* highest accurate speed: sqrt(clock*100)
The clock speed for the PC87360 family is 480 kHz. I arbitrarily chose 100
RPM as the lowest acceptable accuracy.
As mentioned above, you don't have to care about this no more.
Note that not all RPM values can be represented, even when the best clock
divider is selected. This is not only true for the measured speeds, but
also for the programmable low limits, so don't be surprised if you try to
set, say, fan1_min to 2900 and it finally reads 2909.
Fan Control
-----------
PWM (pulse width modulation) values range from 0 to 255, with 0 meaning
that the fan is stopped, and 255 meaning that the fan goes at full speed.
Be extremely careful when changing PWM values. Low PWM values, even
non-zero, can stop the fan, which may cause irreversible damage to your
hardware if temperature increases too much. When changing PWM values, go
step by step and keep an eye on temperatures.
One user reported problems with PWM. Changing PWM values would break fan
speed readings. No explanation nor fix could be found.
Temperature Monitoring
----------------------
Temperatures are reported in degrees Celsius. Each temperature measured has
associated low, high and overtemperature limits, each of which triggers an
alarm when crossed.
The first two temperature channels are external. The third one (PC87366
only) is internal.
The PC87366 has three additional temperature channels, based on
thermistors (as opposed to thermal diodes for the first three temperature
channels). For technical reasons, these channels are held by the VLM
(voltage level monitor) logical device, not the TMS (temperature
measurement) one. As a consequence, these temperatures are exported as
voltages, and converted into temperatures in user-space.
Note that these three additional channels share their pins with the
external thermal diode channels, so you (physically) can't use them all at
the same time. Although it should be possible to mix the two sensor types,
the documents from National Semiconductor suggest that motherboard
manufacturers should choose one type and stick to it. So you will more
likely have either channels 1 to 3 (thermal diodes) or 3 to 6 (internal
thermal diode, and thermistors).
Voltage Monitoring
------------------
Voltages are reported relatively to a reference voltage, either internal or
external. Some of them (in7:Vsb, in8:Vdd and in10:AVdd) are divided by two
internally, you will have to compensate in sensors.conf. Others (in0 to in6)
are likely to be divided externally. The meaning of each of these inputs as
well as the values of the resistors used for division is left to the
motherboard manufacturers, so you will have to document yourself and edit
sensors.conf accordingly. National Semiconductor has a document with
recommended resistor values for some voltages, but this still leaves much
room for per motherboard specificities, unfortunately. Even worse,
motherboard manufacturers don't seem to care about National Semiconductor's
recommendations.
Each voltage measured has associated low and high limits, each of which
triggers an alarm when crossed.
When available, VID inputs are used to provide the nominal CPU Core voltage.
The driver will default to VRM 9.0, but this can be changed from user-space.
The chipsets can handle two sets of VID inputs (on dual-CPU systems), but
the driver will only export one for now. This may change later if there is
a need.
General Remarks
---------------
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may already
have disappeared! Note that all hardware registers are read whenever any
data is read (unless it is less than 2 seconds since the last update, in
which case cached values are returned instead). As a consequence, when
a once-only alarm triggers, it may take 2 seconds for it to show, and 2
more seconds for it to disappear.
Monitoring of in9 isn't enabled at lower init levels (<3) because that
channel measures the battery voltage (Vbat). It is a known fact that
repeatedly sampling the battery voltage reduces its lifetime. National
Semiconductor smartly designed their chipset so that in9 is sampled only
once every 1024 sampling cycles (that is every 34 minutes at the default
sampling rate), so the effect is attenuated, but still present.
Limitations
-----------
The datasheets suggests that some values (fan mins, fan dividers)
shouldn't be changed once the monitoring has started, but we ignore that
recommendation. We'll reconsider if it actually causes trouble.

106
Documentation/hwmon/sis5595 Normal file
View File

@ -0,0 +1,106 @@
Kernel driver sis5595
=====================
Supported chips:
* Silicon Integrated Systems Corp. SiS5595 Southbridge Hardware Monitor
Prefix: 'sis5595'
Addresses scanned: ISA in PCI-space encoded address
Datasheet: Publicly available at the Silicon Integrated Systems Corp. site.
Authors:
Kyösti Mälkki <kmalkki@cc.hut.fi>,
Mark D. Studebaker <mdsxyz123@yahoo.com>,
Aurelien Jarno <aurelien@aurel32.net> 2.6 port
SiS southbridge has a LM78-like chip integrated on the same IC.
This driver is a customized copy of lm78.c
Supports following revisions:
Version PCI ID PCI Revision
1 1039/0008 AF or less
2 1039/0008 B0 or greater
Note: these chips contain a 0008 device which is incompatible with the
5595. We recognize these by the presence of the listed
"blacklist" PCI ID and refuse to load.
NOT SUPPORTED PCI ID BLACKLIST PCI ID
540 0008 0540
550 0008 0550
5513 0008 5511
5581 0008 5597
5582 0008 5597
5597 0008 5597
630 0008 0630
645 0008 0645
730 0008 0730
735 0008 0735
Module Parameters
-----------------
force_addr=0xaddr Set the I/O base address. Useful for boards
that don't set the address in the BIOS. Does not do a
PCI force; the device must still be present in lspci.
Don't use this unless the driver complains that the
base address is not set.
Example: 'modprobe sis5595 force_addr=0x290'
Description
-----------
The SiS5595 southbridge has integrated hardware monitor functions. It also
has an I2C bus, but this driver only supports the hardware monitor. For the
I2C bus driver see i2c-sis5595.
The SiS5595 implements zero or one temperature sensor, two fan speed
sensors, four or five voltage sensors, and alarms.
On the first version of the chip, there are four voltage sensors and one
temperature sensor.
On the second version of the chip, the temperature sensor (temp) and the
fifth voltage sensor (in4) share a pin which is configurable, but not
through the driver. Sorry. The driver senses the configuration of the pin,
which was hopefully set by the BIOS.
Temperatures are measured in degrees Celsius. An alarm is triggered once
when the max is crossed; it is also triggered when it drops below the min
value. Measurements are guaranteed between -55 and +125 degrees, with a
resolution of 1 degree.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give
the readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in volts. An
alarm is triggered if the voltage has crossed a programmable minimum or
maximum limit. Note that minimum in this case always means 'closest to
zero'; this is important for negative voltage measurements. All voltage
inputs can measure voltages between 0 and 4.08 volts, with a resolution of
0.016 volt.
In addition to the alarms described above, there is a BTI alarm, which gets
triggered when an external chip has crossed its limits. Usually, this is
connected to some LM75-like chip; if at least one crosses its limits, this
bit gets set.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may already
have disappeared! Note that in the current implementation, all hardware
registers are read whenever any data is read (unless it is less than 1.5
seconds since the last update). This means that you can easily miss
once-only alarms.
The SiS5595 only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.
Problems
--------
Some chips refuse to be enabled. We don't know why.
The driver will recognize this and print a message in dmesg.

46
Documentation/i2c/chips/smsc47b397.txt → Documentation/hwmon/smsc47b397
View File

@ -1,7 +1,19 @@
Kernel driver smsc47b397
========================
Supported chips:
* SMSC LPC47B397-NC
Prefix: 'smsc47b397'
Addresses scanned: none, address read from Super I/O config space
Datasheet: In this file
Authors: Mark M. Hoffman <mhoffman@lightlink.com>
Utilitek Systems, Inc.
November 23, 2004
The following specification describes the SMSC LPC47B397-NC sensor chip
(for which there is no public datasheet available). This document was
(for which there is no public datasheet available). This document was
provided by Craig Kelly (In-Store Broadcast Network) and edited/corrected
by Mark M. Hoffman <mhoffman@lightlink.com>.
@ -10,10 +22,10 @@ by Mark M. Hoffman <mhoffman@lightlink.com>.
Methods for detecting the HP SIO and reading the thermal data on a dc7100.
The thermal information on the dc7100 is contained in the SIO Hardware Monitor
(HWM). The information is accessed through an index/data pair. The index/data
pair is located at the HWM Base Address + 0 and the HWM Base Address + 1. The
(HWM). The information is accessed through an index/data pair. The index/data
pair is located at the HWM Base Address + 0 and the HWM Base Address + 1. The
HWM Base address can be obtained from Logical Device 8, registers 0x60 (MSB)
and 0x61 (LSB). Currently we are using 0x480 for the HWM Base Address and
and 0x61 (LSB). Currently we are using 0x480 for the HWM Base Address and
0x480 and 0x481 for the index/data pair.
Reading temperature information.
@ -50,7 +62,7 @@ Reading the tach LSB locks the tach MSB.
The LSB Must be read first.
How to convert the tach reading to RPM.
The tach reading (TCount) is given by: (Tach MSB * 256) + (Tach LSB)
The tach reading (TCount) is given by: (Tach MSB * 256) + (Tach LSB)
The SIO counts the number of 90kHz (11.111us) pulses per revolution.
RPM = 60/(TCount * 11.111us)
@ -72,20 +84,20 @@ To program the configuration registers, the following sequence must be followed:
Enter Configuration Mode
To place the chip into the Configuration State The config key (0x55) is written
to the CONFIG PORT (0x2E).
to the CONFIG PORT (0x2E).
Configuration Mode
In configuration mode, the INDEX PORT is located at the CONFIG PORT address and
the DATA PORT is at INDEX PORT address + 1.
The desired configuration registers are accessed in two steps:
The desired configuration registers are accessed in two steps:
a. Write the index of the Logical Device Number Configuration Register
(i.e., 0x07) to the INDEX PORT and then write the number of the
desired logical device to the DATA PORT.
b. Write the address of the desired configuration register within the
logical device to the INDEX PORT and then write or read the config-
uration register through the DATA PORT.
uration register through the DATA PORT.
Note: If accessing the Global Configuration Registers, step (a) is not required.
@ -96,18 +108,18 @@ The chip returns to the RUN State. (This is important).
Programming Example
The following is an example of how to read the SIO Device ID located at 0x20
; ENTER CONFIGURATION MODE
; ENTER CONFIGURATION MODE
MOV DX,02EH
MOV AX,055H
OUT DX,AL
; GLOBAL CONFIGURATION REGISTER
; GLOBAL CONFIGURATION REGISTER
MOV DX,02EH
MOV AL,20H
OUT DX,AL
OUT DX,AL
; READ THE DATA
MOV DX,02FH
IN AL,DX
; EXIT CONFIGURATION MODE
; EXIT CONFIGURATION MODE
MOV DX,02EH
MOV AX,0AAH
OUT DX,AL
@ -122,12 +134,12 @@ Obtaining the HWM Base Address.
The following is an example of how to read the HWM Base Address located in
Logical Device 8.
; ENTER CONFIGURATION MODE
; ENTER CONFIGURATION MODE
MOV DX,02EH
MOV AX,055H
OUT DX,AL
; CONFIGURE REGISTER CRE0,
; LOGICAL DEVICE 8
; CONFIGURE REGISTER CRE0,
; LOGICAL DEVICE 8
MOV DX,02EH
MOV AL,07H
OUT DX,AL ;Point to LD# Config Reg
@ -135,12 +147,12 @@ MOV DX,02FH
MOV AL, 08H
OUT DX,AL;Point to Logical Device 8
;
MOV DX,02EH
MOV DX,02EH
MOV AL,60H
OUT DX,AL ; Point to HWM Base Addr MSB
MOV DX,02FH
IN AL,DX ; Get MSB of HWM Base Addr
; EXIT CONFIGURATION MODE
; EXIT CONFIGURATION MODE
MOV DX,02EH
MOV AX,0AAH
OUT DX,AL

52
Documentation/hwmon/smsc47m1 Normal file
View File

@ -0,0 +1,52 @@
Kernel driver smsc47m1
======================
Supported chips:
* SMSC LPC47B27x, LPC47M10x, LPC47M13x, LPC47M14x, LPC47M15x and LPC47M192
Addresses scanned: none, address read from Super I/O config space
Prefix: 'smsc47m1'
Datasheets:
http://www.smsc.com/main/datasheets/47b27x.pdf
http://www.smsc.com/main/datasheets/47m10x.pdf
http://www.smsc.com/main/tools/discontinued/47m13x.pdf
http://www.smsc.com/main/datasheets/47m14x.pdf
http://www.smsc.com/main/tools/discontinued/47m15x.pdf
http://www.smsc.com/main/datasheets/47m192.pdf
Authors:
Mark D. Studebaker <mdsxyz123@yahoo.com>,
With assistance from Bruce Allen <ballen@uwm.edu>, and his
fan.c program: http://www.lsc-group.phys.uwm.edu/%7Eballen/driver/
Gabriele Gorla <gorlik@yahoo.com>,
Jean Delvare <khali@linux-fr.org>
Description
-----------
The Standard Microsystems Corporation (SMSC) 47M1xx Super I/O chips
contain monitoring and PWM control circuitry for two fans.
The 47M15x and 47M192 chips contain a full 'hardware monitoring block'
in addition to the fan monitoring and control. The hardware monitoring
block is not supported by the driver.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give
the readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
PWM values are from 0 to 255.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may
already have disappeared! Note that in the current implementation, all
hardware registers are read whenever any data is read (unless it is less
than 1.5 seconds since the last update). This means that you can easily
miss once-only alarms.
**********************
The lm_sensors project gratefully acknowledges the support of
Intel in the development of this driver.

0
Documentation/i2c/sysfs-interface → Documentation/hwmon/sysfs-interface
View File

39
Documentation/hwmon/userspace-tools Normal file
View File

@ -0,0 +1,39 @@
Introduction
------------
Most mainboards have sensor chips to monitor system health (like temperatures,
voltages, fans speed). They are often connected through an I2C bus, but some
are also connected directly through the ISA bus.
The kernel drivers make the data from the sensor chips available in the /sys
virtual filesystem. Userspace tools are then used to display or set or the
data in a more friendly manner.
Lm-sensors
----------
Core set of utilites that will allow you to obtain health information,
setup monitoring limits etc. You can get them on their homepage
http://www.lm-sensors.nu/ or as a package from your Linux distribution.
If from website:
Get lmsensors from project web site. Please note, you need only userspace
part, so compile with "make user_install" target.
General hints to get things working:
0) get lm-sensors userspace utils
1) compile all drivers in I2C section as modules in your kernel
2) run sensors-detect script, it will tell you what modules you need to load.
3) load them and run "sensors" command, you should see some results.
4) fix sensors.conf, labels, limits, fan divisors
5) if any more problems consult FAQ, or documentation
Other utilites
--------------
If you want some graphical indicators of system health look for applications
like: gkrellm, ksensors, xsensors, wmtemp, wmsensors, wmgtemp, ksysguardd,
hardware-monitor
If you are server administrator you can try snmpd or mrtgutils.

65
Documentation/hwmon/via686a Normal file
View File

@ -0,0 +1,65 @@
Kernel driver via686a
=====================
Supported chips:
* Via VT82C686A, VT82C686B Southbridge Integrated Hardware Monitor
Prefix: 'via686a'
Addresses scanned: ISA in PCI-space encoded address
Datasheet: On request through web form (http://www.via.com.tw/en/support/datasheets/)
Authors:
Kyösti Mälkki <kmalkki@cc.hut.fi>,
Mark D. Studebaker <mdsxyz123@yahoo.com>
Bob Dougherty <bobd@stanford.edu>
(Some conversion-factor data were contributed by
Jonathan Teh Soon Yew <j.teh@iname.com>
and Alex van Kaam <darkside@chello.nl>.)
Module Parameters
-----------------
force_addr=0xaddr Set the I/O base address. Useful for Asus A7V boards
that don't set the address in the BIOS. Does not do a
PCI force; the via686a must still be present in lspci.
Don't use this unless the driver complains that the
base address is not set.
Example: 'modprobe via686a force_addr=0x6000'
Description
-----------
The driver does not distinguish between the chips and reports
all as a 686A.
The Via 686a southbridge has integrated hardware monitor functionality.
It also has an I2C bus, but this driver only supports the hardware monitor.
For the I2C bus driver, see <file:Documentation/i2c/busses/i2c-viapro>
The Via 686a implements three temperature sensors, two fan rotation speed
sensors, five voltage sensors and alarms.
Temperatures are measured in degrees Celsius. An alarm is triggered once
when the Overtemperature Shutdown limit is crossed; it is triggered again
as soon as it drops below the hysteresis value.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8) to give
the readings more range or accuracy. Not all RPM values can accurately be
represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in volts.
An alarm is triggered if the voltage has crossed a programmable minimum
or maximum limit. Voltages are internally scalled, so each voltage channel
has a different resolution and range.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may
already have disappeared! Note that in the current implementation, all
hardware registers are read whenever any data is read (unless it is less
than 1.5 seconds since the last update). This means that you can easily
miss once-only alarms.
The driver only updates its values each 1.5 seconds; reading it more often
will do no harm, but will return 'old' values.

66
Documentation/hwmon/w83627hf Normal file
View File

@ -0,0 +1,66 @@
Kernel driver w83627hf
======================
Supported chips:
* Winbond W83627HF (ISA accesses ONLY)
Prefix: 'w83627hf'
Addresses scanned: ISA address retrieved from Super I/O registers
Datasheet: http://www.winbond.com/PDF/sheet/w83627hf.pdf
* Winbond W83627THF
Prefix: 'w83627thf'
Addresses scanned: ISA address retrieved from Super I/O registers
Datasheet: http://www.winbond.com/PDF/sheet/w83627thf.pdf
* Winbond W83697HF
Prefix: 'w83697hf'
Addresses scanned: ISA address retrieved from Super I/O registers
Datasheet: http://www.winbond.com/PDF/sheet/697hf.pdf
* Winbond W83637HF
Prefix: 'w83637hf'
Addresses scanned: ISA address retrieved from Super I/O registers
Datasheet: http://www.winbond.com/PDF/sheet/w83637hf.pdf
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>,
Mark Studebaker <mdsxyz123@yahoo.com>,
Bernhard C. Schrenk <clemy@clemy.org>
Module Parameters
-----------------
* force_addr: int
Initialize the ISA address of the sensors
* force_i2c: int
Initialize the I2C address of the sensors
* init: int
(default is 1)
Use 'init=0' to bypass initializing the chip.
Try this if your computer crashes when you load the module.
Description
-----------
This driver implements support for ISA accesses *only* for
the Winbond W83627HF, W83627THF, W83697HF and W83637HF Super I/O chips.
We will refer to them collectively as Winbond chips.
This driver supports ISA accesses, which should be more reliable
than i2c accesses. Also, for Tyan boards which contain both a
Super I/O chip and a second i2c-only Winbond chip (often a W83782D),
using this driver will avoid i2c address conflicts and complex
initialization that were required in the w83781d driver.
If you really want i2c accesses for these Super I/O chips,
use the w83781d driver. However this is not the preferred method
now that this ISA driver has been developed.
Technically, the w83627thf does not support a VID reading. However, it's
possible or even likely that your mainboard maker has routed these signals
to a specific set of general purpose IO pins (the Asus P4C800-E is one such
board). The w83627thf driver now interprets these as VID. If the VID on
your board doesn't work, first see doc/vid in the lm_sensors package. If
that still doesn't help, email us at lm-sensors@lm-sensors.org.
For further information on this driver see the w83781d driver
documentation.

402
Documentation/hwmon/w83781d Normal file
View File

@ -0,0 +1,402 @@
Kernel driver w83781d
=====================
Supported chips:
* Winbond W83781D
Prefix: 'w83781d'
Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
Datasheet: http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/w83781d.pdf
* Winbond W83782D
Prefix: 'w83782d'
Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
Datasheet: http://www.winbond.com/PDF/sheet/w83782d.pdf
* Winbond W83783S
Prefix: 'w83783s'
Addresses scanned: I2C 0x2d
Datasheet: http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/w83783s.pdf
* Winbond W83627HF
Prefix: 'w83627hf'
Addresses scanned: I2C 0x20 - 0x2f, ISA 0x290 (8 I/O ports)
Datasheet: http://www.winbond.com/PDF/sheet/w83627hf.pdf
* Asus AS99127F
Prefix: 'as99127f'
Addresses scanned: I2C 0x28 - 0x2f
Datasheet: Unavailable from Asus
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>,
Mark Studebaker <mdsxyz123@yahoo.com>
Module parameters
-----------------
* init int
(default 1)
Use 'init=0' to bypass initializing the chip.
Try this if your computer crashes when you load the module.
force_subclients=bus,caddr,saddr,saddr
This is used to force the i2c addresses for subclients of
a certain chip. Typical usage is `force_subclients=0,0x2d,0x4a,0x4b'
to force the subclients of chip 0x2d on bus 0 to i2c addresses
0x4a and 0x4b. This parameter is useful for certain Tyan boards.
Description
-----------
This driver implements support for the Winbond W83781D, W83782D, W83783S,
W83627HF chips, and the Asus AS99127F chips. We will refer to them
collectively as W8378* chips.
There is quite some difference between these chips, but they are similar
enough that it was sensible to put them together in one driver.
The W83627HF chip is assumed to be identical to the ISA W83782D.
The Asus chips are similar to an I2C-only W83782D.
Chip #vin #fanin #pwm #temp wchipid vendid i2c ISA
as99127f 7 3 0 3 0x31 0x12c3 yes no
as99127f rev.2 (type_name = as99127f) 0x31 0x5ca3 yes no
w83781d 7 3 0 3 0x10-1 0x5ca3 yes yes
w83627hf 9 3 2 3 0x21 0x5ca3 yes yes(LPC)
w83782d 9 3 2-4 3 0x30 0x5ca3 yes yes
w83783s 5-6 3 2 1-2 0x40 0x5ca3 yes no
Detection of these chips can sometimes be foiled because they can be in
an internal state that allows no clean access. If you know the address
of the chip, use a 'force' parameter; this will put them into a more
well-behaved state first.
The W8378* implements temperature sensors (three on the W83781D and W83782D,
two on the W83783S), three fan rotation speed sensors, voltage sensors
(seven on the W83781D, nine on the W83782D and six on the W83783S), VID
lines, alarms with beep warnings, and some miscellaneous stuff.
Temperatures are measured in degrees Celsius. There is always one main
temperature sensor, and one (W83783S) or two (W83781D and W83782D) other
sensors. An alarm is triggered for the main sensor once when the
Overtemperature Shutdown limit is crossed; it is triggered again as soon as
it drops below the Hysteresis value. A more useful behavior
can be found by setting the Hysteresis value to +127 degrees Celsius; in
this case, alarms are issued during all the time when the actual temperature
is above the Overtemperature Shutdown value. The driver sets the
hysteresis value for temp1 to 127 at initialization.
For the other temperature sensor(s), an alarm is triggered when the
temperature gets higher then the Overtemperature Shutdown value; it stays
on until the temperature falls below the Hysteresis value. But on the
W83781D, there is only one alarm that functions for both other sensors!
Temperatures are guaranteed within a range of -55 to +125 degrees. The
main temperature sensors has a resolution of 1 degree; the other sensor(s)
of 0.5 degree.
Fan rotation speeds are reported in RPM (rotations per minute). An alarm is
triggered if the rotation speed has dropped below a programmable limit. Fan
readings can be divided by a programmable divider (1, 2, 4 or 8 for the
W83781D; 1, 2, 4, 8, 16, 32, 64 or 128 for the others) to give
the readings more range or accuracy. Not all RPM values can accurately
be represented, so some rounding is done. With a divider of 2, the lowest
representable value is around 2600 RPM.
Voltage sensors (also known as IN sensors) report their values in volts.
An alarm is triggered if the voltage has crossed a programmable minimum
or maximum limit. Note that minimum in this case always means 'closest to
zero'; this is important for negative voltage measurements. All voltage
inputs can measure voltages between 0 and 4.08 volts, with a resolution
of 0.016 volt.
The VID lines encode the core voltage value: the voltage level your processor
should work with. This is hardcoded by the mainboard and/or processor itself.
It is a value in volts. When it is unconnected, you will often find the
value 3.50 V here.
The W83782D and W83783S temperature conversion machine understands about
several kinds of temperature probes. You can program the so-called
beta value in the sensor files. '1' is the PII/Celeron diode, '2' is the
TN3904 transistor, and 3435 the default thermistor value. Other values
are (not yet) supported.
In addition to the alarms described above, there is a CHAS alarm on the
chips which triggers if your computer case is open.
When an alarm goes off, you can be warned by a beeping signal through
your computer speaker. It is possible to enable all beeping globally,
or only the beeping for some alarms.
If an alarm triggers, it will remain triggered until the hardware register
is read at least once. This means that the cause for the alarm may
already have disappeared! Note that in the current implementation, all
hardware registers are read whenever any data is read (unless it is less
than 1.5 seconds since the last update). This means that you can easily
miss once-only alarms.
The chips only update values each 1.5 seconds; reading them more often
will do no harm, but will return 'old' values.
AS99127F PROBLEMS
-----------------
The as99127f support was developed without the benefit of a datasheet.
In most cases it is treated as a w83781d (although revision 2 of the
AS99127F looks more like a w83782d).
This support will be BETA until a datasheet is released.
One user has reported problems with fans stopping
occasionally.
Note that the individual beep bits are inverted from the other chips.
The driver now takes care of this so that user-space applications
don't have to know about it.
Known problems:
- Problems with diode/thermistor settings (supported?)
- One user reports fans stopping under high server load.
- Revision 2 seems to have 2 PWM registers but we don't know
how to handle them. More details below.
These will not be fixed unless we get a datasheet.
If you have problems, please lobby Asus to release a datasheet.
Unfortunately several others have without success.
Please do not send mail to us asking for better as99127f support.
We have done the best we can without a datasheet.
Please do not send mail to the author or the sensors group asking for
a datasheet or ideas on how to convince Asus. We can't help.
NOTES:
-----
783s has no in1 so that in[2-6] are compatible with the 781d/782d.
783s pin is programmable for -5V or temp1; defaults to -5V,
no control in driver so temp1 doesn't work.
782d and 783s datasheets differ on which is pwm1 and which is pwm2.
We chose to follow 782d.
782d and 783s pin is programmable for fan3 input or pwm2 output;
defaults to fan3 input.
If pwm2 is enabled (with echo 255 1 > pwm2), then
fan3 will report 0.
782d has pwm1-2 for ISA, pwm1-4 for i2c. (pwm3-4 share pins with
the ISA pins)
Data sheet updates:
------------------
- PWM clock registers:
000: master / 512
001: master / 1024
010: master / 2048
011: master / 4096
100: master / 8192
Answers from Winbond tech support
---------------------------------
>
> 1) In the W83781D data sheet section 7.2 last paragraph, it talks about
> reprogramming the R-T table if the Beta of the thermistor is not
> 3435K. The R-T table is described briefly in section 8.20.
> What formulas do I use to program a new R-T table for a given Beta?
>
We are sorry that the calculation for R-T table value is
confidential. If you have another Beta value of thermistor, we can help
to calculate the R-T table for you. But you should give us real R-T
Table which can be gotten by thermistor vendor. Therefore we will calculate
them and obtain 32-byte data, and you can fill the 32-byte data to the
register in Bank0.CR51 of W83781D.
> 2) In the W83782D data sheet, it mentions that pins 38, 39, and 40 are
> programmable to be either thermistor or Pentium II diode inputs.
> How do I program them for diode inputs? I can't find any register
> to program these to be diode inputs.
--> You may program Bank0 CR[5Dh] and CR[59h] registers.
CR[5Dh] bit 1(VTIN1) bit 2(VTIN2) bit 3(VTIN3)
thermistor 0 0 0
diode 1 1 1
(error) CR[59h] bit 4(VTIN1) bit 2(VTIN2) bit 3(VTIN3)
(right) CR[59h] bit 4(VTIN1) bit 5(VTIN2) bit 6(VTIN3)
PII thermal diode 1 1 1
2N3904 diode 0 0 0
Asus Clones
-----------
We have no datasheets for the Asus clones (AS99127F and ASB100 Bach).
Here are some very useful information that were given to us by Alex Van
Kaam about how to detect these chips, and how to read their values. He
also gives advice for another Asus chipset, the Mozart-2 (which we
don't support yet). Thanks Alex!
I reworded some parts and added personal comments.
# Detection:
AS99127F rev.1, AS99127F rev.2 and ASB100:
- I2C address range: 0x29 - 0x2F
- If register 0x58 holds 0x31 then we have an Asus (either ASB100 or
AS99127F)
- Which one depends on register 0x4F (manufacturer ID):
0x06 or 0x94: ASB100
0x12 or 0xC3: AS99127F rev.1
0x5C or 0xA3: AS99127F rev.2
Note that 0x5CA3 is Winbond's ID (WEC), which let us think Asus get their
AS99127F rev.2 direct from Winbond. The other codes mean ATT and DVC,
respectively. ATT could stand for Asustek something (although it would be
very badly chosen IMHO), I don't know what DVC could stand for. Maybe
these codes simply aren't meant to be decoded that way.
Mozart-2:
- I2C address: 0x77
- If register 0x58 holds 0x56 or 0x10 then we have a Mozart-2
- Of the Mozart there are 3 types:
0x58=0x56, 0x4E=0x94, 0x4F=0x36: Asus ASM58 Mozart-2
0x58=0x56, 0x4E=0x94, 0x4F=0x06: Asus AS2K129R Mozart-2
0x58=0x10, 0x4E=0x5C, 0x4F=0xA3: Asus ??? Mozart-2
You can handle all 3 the exact same way :)
# Temperature sensors:
ASB100:
- sensor 1: register 0x27
- sensor 2 & 3 are the 2 LM75's on the SMBus
- sensor 4: register 0x17
Remark: I noticed that on Intel boards sensor 2 is used for the CPU
and 4 is ignored/stuck, on AMD boards sensor 4 is the CPU and sensor 2 is
either ignored or a socket temperature.
AS99127F (rev.1 and 2 alike):
- sensor 1: register 0x27
- sensor 2 & 3 are the 2 LM75's on the SMBus
Remark: Register 0x5b is suspected to be temperature type selector. Bit 1
would control temp1, bit 3 temp2 and bit 5 temp3.
Mozart-2:
- sensor 1: register 0x27
- sensor 2: register 0x13
# Fan sensors:
ASB100, AS99127F (rev.1 and 2 alike):
- 3 fans, identical to the W83781D
Mozart-2:
- 2 fans only, 1350000/RPM/div
- fan 1: register 0x28, divisor on register 0xA1 (bits 4-5)
- fan 2: register 0x29, divisor on register 0xA1 (bits 6-7)
# Voltages:
This is where there is a difference between AS99127F rev.1 and 2.
Remark: The difference is similar to the difference between
W83781D and W83782D.
ASB100:
in0=r(0x20)*0.016
in1=r(0x21)*0.016
in2=r(0x22)*0.016
in3=r(0x23)*0.016*1.68
in4=r(0x24)*0.016*3.8
in5=r(0x25)*(-0.016)*3.97
in6=r(0x26)*(-0.016)*1.666
AS99127F rev.1:
in0=r(0x20)*0.016
in1=r(0x21)*0.016
in2=r(0x22)*0.016
in3=r(0x23)*0.016*1.68
in4=r(0x24)*0.016*3.8
in5=r(0x25)*(-0.016)*3.97
in6=r(0x26)*(-0.016)*1.503
AS99127F rev.2:
in0=r(0x20)*0.016
in1=r(0x21)*0.016
in2=r(0x22)*0.016
in3=r(0x23)*0.016*1.68
in4=r(0x24)*0.016*3.8
in5=(r(0x25)*0.016-3.6)*5.14+3.6
in6=(r(0x26)*0.016-3.6)*3.14+3.6
Mozart-2:
in0=r(0x20)*0.016
in1=255
in2=r(0x22)*0.016
in3=r(0x23)*0.016*1.68
in4=r(0x24)*0.016*4
in5=255
in6=255
# PWM
Additional info about PWM on the AS99127F (may apply to other Asus
chips as well) by Jean Delvare as of 2004-04-09:
AS99127F revision 2 seems to have two PWM registers at 0x59 and 0x5A,
and a temperature sensor type selector at 0x5B (which basically means
that they swapped registers 0x59 and 0x5B when you compare with Winbond
chips).
Revision 1 of the chip also has the temperature sensor type selector at
0x5B, but PWM registers have no effect.
We don't know exactly how the temperature sensor type selection works.
Looks like bits 1-0 are for temp1, bits 3-2 for temp2 and bits 5-4 for
temp3, although it is possible that only the most significant bit matters
each time. So far, values other than 0 always broke the readings.
PWM registers seem to be split in two parts: bit 7 is a mode selector,
while the other bits seem to define a value or threshold.
When bit 7 is clear, bits 6-0 seem to hold a threshold value. If the value
is below a given limit, the fan runs at low speed. If the value is above
the limit, the fan runs at full speed. We have no clue as to what the limit
represents. Note that there seem to be some inertia in this mode, speed
changes may need some time to trigger. Also, an hysteresis mechanism is
suspected since walking through all the values increasingly and then
decreasingly led to slightly different limits.
When bit 7 is set, bits 3-0 seem to hold a threshold value, while bits 6-4
would not be significant. If the value is below a given limit, the fan runs
at full speed, while if it is above the limit it runs at low speed (so this
is the contrary of the other mode, in a way). Here again, we don't know
what the limit is supposed to represent.
One remarkable thing is that the fans would only have two or three
different speeds (transitional states left apart), not a whole range as
you usually get with PWM.
As a conclusion, you can write 0x00 or 0x8F to the PWM registers to make
fans run at low speed, and 0x7F or 0x80 to make them run at full speed.
Please contact us if you can figure out how it is supposed to work. As
long as we don't know more, the w83781d driver doesn't handle PWM on
AS99127F chips at all.
Additional info about PWM on the AS99127F rev.1 by Hector Martin:
I've been fiddling around with the (in)famous 0x59 register and
found out the following values do work as a form of coarse pwm:
0x80 - seems to turn fans off after some time(1-2 minutes)... might be
some form of auto-fan-control based on temp? hmm (Qfan? this mobo is an
old ASUS, it isn't marketed as Qfan. Maybe some beta pre-attemp at Qfan
that was dropped at the BIOS)
0x81 - off
0x82 - slightly "on-ner" than off, but my fans do not get to move. I can
hear the high-pitched PWM sound that motors give off at too-low-pwm.
0x83 - now they do move. Estimate about 70% speed or so.
0x84-0x8f - full on
Changing the high nibble doesn't seem to do much except the high bit
(0x80) must be set for PWM to work, else the current pwm doesn't seem to
change.
My mobo is an ASUS A7V266-E. This behavior is similar to what I got
with speedfan under Windows, where 0-15% would be off, 15-2x% (can't
remember the exact value) would be 70% and higher would be full on.

39
Documentation/hwmon/w83l785ts Normal file
View File

@ -0,0 +1,39 @@
Kernel driver w83l785ts
=======================
Supported chips:
* Winbond W83L785TS-S
Prefix: 'w83l785ts'
Addresses scanned: I2C 0x2e
Datasheet: Publicly available at the Winbond USA website
http://www.winbond-usa.com/products/winbond_products/pdfs/PCIC/W83L785TS-S.pdf
Authors:
Jean Delvare <khali@linux-fr.org>
Description
-----------
The W83L785TS-S is a digital temperature sensor. It senses the
temperature of a single external diode. The high limit is
theoretically defined as 85 or 100 degrees C through a combination
of external resistors, so the user cannot change it. Values seen so
far suggest that the two possible limits are actually 95 and 110
degrees C. The datasheet is rather poor and obviously inaccurate
on several points including this one.
All temperature values are given in degrees Celsius. Resolution
is 1.0 degree. See the datasheet for details.
The w83l785ts driver will not update its values more frequently than
every other second; reading them more often will do no harm, but will
return 'old' values.
Known Issues
------------
On some systems (Asus), the BIOS is known to interfere with the driver
and cause read errors. The driver will retry a given number of times
(5 by default) and then give up, returning the old value (or 0 if
there is no old value). It seems to work well enough so that you should
not notice anything. Thanks to James Bolt for helping test this feature.

2
Documentation/i2c/busses/i2c-sis69x
View File

@ -42,7 +42,7 @@ I suspect that this driver could be made to work for the following SiS
chipsets as well: 635, and 635T. If anyone owns a board with those chips
AND is willing to risk crashing & burning an otherwise well-behaved kernel
in the name of progress... please contact me at <mhoffman@lightlink.com> or
via the project's mailing list: <sensors@stimpy.netroedge.com>. Please
via the project's mailing list: <lm-sensors@lm-sensors.org>. Please
send bug reports and/or success stories as well.

96
Documentation/i2c/chips/eeprom Normal file
View File

@ -0,0 +1,96 @@
Kernel driver eeprom
====================
Supported chips:
* Any EEPROM chip in the designated address range
Prefix: 'eeprom'
Addresses scanned: I2C 0x50 - 0x57
Datasheets: Publicly available from:
Atmel (www.atmel.com),
Catalyst (www.catsemi.com),
Fairchild (www.fairchildsemi.com),
Microchip (www.microchip.com),
Philips (www.semiconductor.philips.com),
Rohm (www.rohm.com),
ST (www.st.com),
Xicor (www.xicor.com),
and others.
Chip Size (bits) Address
24C01 1K 0x50 (shadows at 0x51 - 0x57)
24C01A 1K 0x50 - 0x57 (Typical device on DIMMs)
24C02 2K 0x50 - 0x57
24C04 4K 0x50, 0x52, 0x54, 0x56
(additional data at 0x51, 0x53, 0x55, 0x57)
24C08 8K 0x50, 0x54 (additional data at 0x51, 0x52,
0x53, 0x55, 0x56, 0x57)
24C16 16K 0x50 (additional data at 0x51 - 0x57)
Sony 2K 0x57
Atmel 34C02B 2K 0x50 - 0x57, SW write protect at 0x30-37
Catalyst 34FC02 2K 0x50 - 0x57, SW write protect at 0x30-37
Catalyst 34RC02 2K 0x50 - 0x57, SW write protect at 0x30-37
Fairchild 34W02 2K 0x50 - 0x57, SW write protect at 0x30-37
Microchip 24AA52 2K 0x50 - 0x57, SW write protect at 0x30-37
ST M34C02 2K 0x50 - 0x57, SW write protect at 0x30-37
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>,
Jean Delvare <khali@linux-fr.org>,
Greg Kroah-Hartman <greg@kroah.com>,
IBM Corp.
Description
-----------
This is a simple EEPROM module meant to enable reading the first 256 bytes
of an EEPROM (on a SDRAM DIMM for example). However, it will access serial
EEPROMs on any I2C adapter. The supported devices are generically called
24Cxx, and are listed above; however the numbering for these
industry-standard devices may vary by manufacturer.
This module was a programming exercise to get used to the new project
organization laid out by Frodo, but it should be at least completely
effective for decoding the contents of EEPROMs on DIMMs.
DIMMS will typically contain a 24C01A or 24C02, or the 34C02 variants.
The other devices will not be found on a DIMM because they respond to more
than one address.
DDC Monitors may contain any device. Often a 24C01, which responds to all 8
addresses, is found.
Recent Sony Vaio laptops have an EEPROM at 0x57. We couldn't get the
specification, so it is guess work and far from being complete.
The Microchip 24AA52/24LCS52, ST M34C02, and others support an additional
software write protect register at 0x30 - 0x37 (0x20 less than the memory
location). The chip responds to "write quick" detection at this address but
does not respond to byte reads. If this register is present, the lower 128
bytes of the memory array are not write protected. Any byte data write to
this address will write protect the memory array permanently, and the
device will no longer respond at the 0x30-37 address. The eeprom driver
does not support this register.
Lacking functionality:
* Full support for larger devices (24C04, 24C08, 24C16). These are not
typically found on a PC. These devices will appear as separate devices at
multiple addresses.
* Support for really large devices (24C32, 24C64, 24C128, 24C256, 24C512).
These devices require two-byte address fields and are not supported.
* Enable Writing. Again, no technical reason why not, but making it easy
to change the contents of the EEPROMs (on DIMMs anyway) also makes it easy
to disable the DIMMs (potentially preventing the computer from booting)
until the values are restored somehow.
Use:
After inserting the module (and any other required SMBus/i2c modules), you
should have some EEPROM directories in /sys/bus/i2c/devices/* of names such
as "0-0050". Inside each of these is a series of files, the eeprom file
contains the binary data from EEPROM.

66
Documentation/i2c/chips/max6875 Normal file
View File

@ -0,0 +1,66 @@
Kernel driver max6875
=====================
Supported chips:
* Maxim MAX6874, MAX6875
Prefix: 'max6875'
Addresses scanned: 0x50, 0x52
Datasheet:
http://pdfserv.maxim-ic.com/en/ds/MAX6874-MAX6875.pdf
Author: Ben Gardner <bgardner@wabtec.com>
Module Parameters
-----------------
* allow_write int
Set to non-zero to enable write permission:
*0: Read only
1: Read and write
Description
-----------
The Maxim MAX6875 is an EEPROM-programmable power-supply sequencer/supervisor.
It provides timed outputs that can be used as a watchdog, if properly wired.
It also provides 512 bytes of user EEPROM.
At reset, the MAX6875 reads the configuration EEPROM into its configuration
registers. The chip then begins to operate according to the values in the
registers.
The Maxim MAX6874 is a similar, mostly compatible device, with more intputs
and outputs:
vin gpi vout
MAX6874 6 4 8
MAX6875 4 3 5
MAX6874 chips can have four different addresses (as opposed to only two for
the MAX6875). The additional addresses (0x54 and 0x56) are not probed by
this driver by default, but the probe module parameter can be used if
needed.
See the datasheet for details on how to program the EEPROM.
Sysfs entries
-------------
eeprom_user - 512 bytes of user-defined EEPROM space. Only writable if
allow_write was set and register 0x43 is 0.
eeprom_config - 70 bytes of config EEPROM. Note that changes will not get
loaded into register space until a power cycle or device reset.
reg_config - 70 bytes of register space. Any changes take affect immediately.
General Remarks
---------------
A typical application will require that the EEPROMs be programmed once and
never altered afterwards.

47
Documentation/i2c/chips/pca9539 Normal file
View File

@ -0,0 +1,47 @@
Kernel driver pca9539
=====================
Supported chips:
* Philips PCA9539
Prefix: 'pca9539'
Addresses scanned: 0x74 - 0x77
Datasheet:
http://www.semiconductors.philips.com/acrobat/datasheets/PCA9539_2.pdf
Author: Ben Gardner <bgardner@wabtec.com>
Description
-----------
The Philips PCA9539 is a 16 bit low power I/O device.
All 16 lines can be individually configured as an input or output.
The input sense can also be inverted.
The 16 lines are split between two bytes.
Sysfs entries
-------------
Each is a byte that maps to the 8 I/O bits.
A '0' suffix is for bits 0-7, while '1' is for bits 8-15.
input[01] - read the current value
output[01] - sets the output value
direction[01] - direction of each bit: 1=input, 0=output
invert[01] - toggle the input bit sense
input reads the actual state of the line and is always available.
The direction defaults to input for all channels.
General Remarks
---------------
Note that each output, direction, and invert entry controls 8 lines.
You should use the read, modify, write sequence.
For example. to set output bit 0 of 1.
val=$(cat output0)
val=$(( $val | 1 ))
echo $val > output0

69
Documentation/i2c/chips/pcf8574 Normal file
View File

@ -0,0 +1,69 @@
Kernel driver pcf8574
=====================
Supported chips:
* Philips PCF8574
Prefix: 'pcf8574'
Addresses scanned: I2C 0x20 - 0x27
Datasheet: Publicly available at the Philips Semiconductors website
http://www.semiconductors.philips.com/pip/PCF8574P.html
* Philips PCF8574A
Prefix: 'pcf8574a'
Addresses scanned: I2C 0x38 - 0x3f
Datasheet: Publicly available at the Philips Semiconductors website
http://www.semiconductors.philips.com/pip/PCF8574P.html
Authors:
Frodo Looijaard <frodol@dds.nl>,
Philip Edelbrock <phil@netroedge.com>,
Dan Eaton <dan.eaton@rocketlogix.com>,
Aurelien Jarno <aurelien@aurel32.net>,
Jean Delvare <khali@linux-fr.org>,
Description
-----------
The PCF8574(A) is an 8-bit I/O expander for the I2C bus produced by Philips
Semiconductors. It is designed to provide a byte I2C interface to up to 16
separate devices (8 x PCF8574 and 8 x PCF8574A).
This device consists of a quasi-bidirectional port. Each of the eight I/Os
can be independently used as an input or output. To setup an I/O as an
input, you have to write a 1 to the corresponding output.
For more informations see the datasheet.
Accessing PCF8574(A) via /sys interface
-------------------------------------
! Be careful !
The PCF8574(A) is plainly impossible to detect ! Stupid chip.
So every chip with address in the interval [20..27] and [38..3f] are
detected as PCF8574(A). If you have other chips in this address
range, the workaround is to load this module after the one
for your others chips.
On detection (i.e. insmod, modprobe et al.), directories are being
created for each detected PCF8574(A):
/sys/bus/i2c/devices/<0>-<1>/
where <0> is the bus the chip was detected on (e. g. i2c-0)
and <1> the chip address ([20..27] or [38..3f]):
(example: /sys/bus/i2c/devices/1-0020/)
Inside these directories, there are two files each:
read and write (and one file with chip name).
The read file is read-only. Reading gives you the current I/O input
if the corresponding output is set as 1, otherwise the current output
value, that is to say 0.
The write file is read/write. Writing a value outputs it on the I/O
port. Reading returns the last written value.
On module initialization the chip is configured as eight inputs (all
outputs to 1), so you can connect any circuit to the PCF8574(A) without
being afraid of short-circuit.

90
Documentation/i2c/chips/pcf8591 Normal file
View File

@ -0,0 +1,90 @@
Kernel driver pcf8591
=====================
Supported chips:
* Philips PCF8591
Prefix: 'pcf8591'
Addresses scanned: I2C 0x48 - 0x4f
Datasheet: Publicly available at the Philips Semiconductor website
http://www.semiconductors.philips.com/pip/PCF8591P.html
Authors:
Aurelien Jarno <aurelien@aurel32.net>
valuable contributions by Jan M. Sendler <sendler@sendler.de>,
Jean Delvare <khali@linux-fr.org>
Description
-----------
The PCF8591 is an 8-bit A/D and D/A converter (4 analog inputs and one
analog output) for the I2C bus produced by Philips Semiconductors. It
is designed to provide a byte I2C interface to up to 4 separate devices.
The PCF8591 has 4 analog inputs programmable as single-ended or
differential inputs :
- mode 0 : four single ended inputs
Pins AIN0 to AIN3 are single ended inputs for channels 0 to 3
- mode 1 : three differential inputs
Pins AIN3 is the common negative differential input
Pins AIN0 to AIN2 are positive differential inputs for channels 0 to 2
- mode 2 : single ended and differential mixed
Pins AIN0 and AIN1 are single ended inputs for channels 0 and 1
Pins AIN2 is the positive differential input for channel 3
Pins AIN3 is the negative differential input for channel 3
- mode 3 : two differential inputs
Pins AIN0 is the positive differential input for channel 0
Pins AIN1 is the negative differential input for channel 0
Pins AIN2 is the positive differential input for channel 1
Pins AIN3 is the negative differential input for channel 1
See the datasheet for details.
Module parameters
-----------------
* input_mode int
Analog input mode:
0 = four single ended inputs
1 = three differential inputs
2 = single ended and differential mixed
3 = two differential inputs
Accessing PCF8591 via /sys interface
-------------------------------------
! Be careful !
The PCF8591 is plainly impossible to detect ! Stupid chip.
So every chip with address in the interval [48..4f] is
detected as PCF8591. If you have other chips in this address
range, the workaround is to load this module after the one
for your others chips.
On detection (i.e. insmod, modprobe et al.), directories are being
created for each detected PCF8591:
/sys/bus/devices/<0>-<1>/
where <0> is the bus the chip was detected on (e. g. i2c-0)
and <1> the chip address ([48..4f])
Inside these directories, there are such files:
in0, in1, in2, in3, out0_enable, out0_output, name
Name contains chip name.
The in0, in1, in2 and in3 files are RO. Reading gives the value of the
corresponding channel. Depending on the current analog inputs configuration,
files in2 and/or in3 do not exist. Values range are from 0 to 255 for single
ended inputs and -128 to +127 for differential inputs (8-bit ADC).
The out0_enable file is RW. Reading gives "1" for analog output enabled and
"0" for analog output disabled. Writing accepts "0" and "1" accordingly.
The out0_output file is RW. Writing a number between 0 and 255 (8-bit DAC), send
the value to the digital-to-analog converter. Note that a voltage will
only appears on AOUT pin if aout0_enable equals 1. Reading returns the last
value written.

15
Documentation/i2c/dev-interface
View File

@ -14,9 +14,12 @@ C example
=========
So let's say you want to access an i2c adapter from a C program. The
first thing to do is `#include <linux/i2c.h>" and "#include <linux/i2c-dev.h>.
Yes, I know, you should never include kernel header files, but until glibc
knows about i2c, there is not much choice.
first thing to do is "#include <linux/i2c-dev.h>". Please note that
there are two files named "i2c-dev.h" out there, one is distributed
with the Linux kernel and is meant to be included from kernel
driver code, the other one is distributed with lm_sensors and is
meant to be included from user-space programs. You obviously want
the second one here.
Now, you have to decide which adapter you want to access. You should
inspect /sys/class/i2c-dev/ to decide this. Adapter numbers are assigned
@ -78,7 +81,7 @@ Full interface description
==========================
The following IOCTLs are defined and fully supported
(see also i2c-dev.h and i2c.h):
(see also i2c-dev.h):
ioctl(file,I2C_SLAVE,long addr)
Change slave address. The address is passed in the 7 lower bits of the
@ -97,10 +100,10 @@ ioctl(file,I2C_PEC,long select)
ioctl(file,I2C_FUNCS,unsigned long *funcs)
Gets the adapter functionality and puts it in *funcs.
ioctl(file,I2C_RDWR,struct i2c_ioctl_rdwr_data *msgset)
ioctl(file,I2C_RDWR,struct i2c_rdwr_ioctl_data *msgset)
Do combined read/write transaction without stop in between.
The argument is a pointer to a struct i2c_ioctl_rdwr_data {
The argument is a pointer to a struct i2c_rdwr_ioctl_data {
struct i2c_msg *msgs; /* ptr to array of simple messages */
int nmsgs; /* number of messages to exchange */

2
Documentation/i2c/porting-clients
View File

@ -57,7 +57,7 @@ Technical changes:
Documentation/i2c/sysfs-interface for the individual files. Also
convert the units these files read and write to the specified ones.
If you need to add a new type of file, please discuss it on the
sensors mailing list <sensors@stimpy.netroedge.com> by providing a
sensors mailing list <lm-sensors@lm-sensors.org> by providing a
patch to the Documentation/i2c/sysfs-interface file.
* [Attach] For I2C drivers, the attach function should make sure

69
Documentation/i2c/writing-clients
View File

@ -27,7 +27,6 @@ address.
static struct i2c_driver foo_driver = {
.owner = THIS_MODULE,
.name = "Foo version 2.3 driver",
.id = I2C_DRIVERID_FOO, /* from i2c-id.h, optional */
.flags = I2C_DF_NOTIFY,
.attach_adapter = &foo_attach_adapter,
.detach_client = &foo_detach_client,
@ -37,12 +36,6 @@ static struct i2c_driver foo_driver = {
The name can be chosen freely, and may be upto 40 characters long. Please
use something descriptive here.
If used, the id should be a unique ID. The range 0xf000 to 0xffff is
reserved for local use, and you can use one of those until you start
distributing the driver, at which time you should contact the i2c authors
to get your own ID(s). Note that most of the time you don't need an ID
at all so you can just omit it.
Don't worry about the flags field; just put I2C_DF_NOTIFY into it. This
means that your driver will be notified when new adapters are found.
This is almost always what you want.
@ -171,45 +164,31 @@ The following lists are used internally:
normal_i2c: filled in by the module writer.
A list of I2C addresses which should normally be examined.
normal_i2c_range: filled in by the module writer.
A list of pairs of I2C addresses, each pair being an inclusive range of
addresses which should normally be examined.
probe: insmod parameter.
A list of pairs. The first value is a bus number (-1 for any I2C bus),
the second is the address. These addresses are also probed, as if they
were in the 'normal' list.
probe_range: insmod parameter.
A list of triples. The first value is a bus number (-1 for any I2C bus),
the second and third are addresses. These form an inclusive range of
addresses that are also probed, as if they were in the 'normal' list.
ignore: insmod parameter.
A list of pairs. The first value is a bus number (-1 for any I2C bus),
the second is the I2C address. These addresses are never probed.
This parameter overrules 'normal' and 'probe', but not the 'force' lists.
ignore_range: insmod parameter.
A list of triples. The first value is a bus number (-1 for any I2C bus),
the second and third are addresses. These form an inclusive range of
I2C addresses that are never probed.
This parameter overrules 'normal' and 'probe', but not the 'force' lists.
force: insmod parameter.
A list of pairs. The first value is a bus number (-1 for any I2C bus),
the second is the I2C address. A device is blindly assumed to be on
the given address, no probing is done.
Fortunately, as a module writer, you just have to define the `normal'
and/or `normal_range' parameters. The complete declaration could look
like this:
Fortunately, as a module writer, you just have to define the `normal_i2c'
parameter. The complete declaration could look like this:
/* Scan 0x20 to 0x2f, 0x37, and 0x40 to 0x4f */
static unsigned short normal_i2c[] = { 0x37,I2C_CLIENT_END };
static unsigned short normal_i2c_range[] = { 0x20, 0x2f, 0x40, 0x4f,
I2C_CLIENT_END };
/* Scan 0x37, and 0x48 to 0x4f */
static unsigned short normal_i2c[] = { 0x37, 0x48, 0x49, 0x4a, 0x4b, 0x4c,
0x4d, 0x4e, 0x4f, I2C_CLIENT_END };
/* Magic definition of all other variables and things */
I2C_CLIENT_INSMOD;
Note that you *have* to call the two defined variables `normal_i2c' and
`normal_i2c_range', without any prefix!
Note that you *have* to call the defined variable `normal_i2c',
without any prefix!
Probing classes (sensors)
@ -223,39 +202,17 @@ The following lists are used internally. They are all lists of integers.
normal_i2c: filled in by the module writer. Terminated by SENSORS_I2C_END.
A list of I2C addresses which should normally be examined.
normal_i2c_range: filled in by the module writer. Terminated by
SENSORS_I2C_END
A list of pairs of I2C addresses, each pair being an inclusive range of
addresses which should normally be examined.
normal_isa: filled in by the module writer. Terminated by SENSORS_ISA_END.
A list of ISA addresses which should normally be examined.
normal_isa_range: filled in by the module writer. Terminated by
SENSORS_ISA_END
A list of triples. The first two elements are ISA addresses, being an
range of addresses which should normally be examined. The third is the
modulo parameter: only addresses which are 0 module this value relative
to the first address of the range are actually considered.
probe: insmod parameter. Initialize this list with SENSORS_I2C_END values.
A list of pairs. The first value is a bus number (SENSORS_ISA_BUS for
the ISA bus, -1 for any I2C bus), the second is the address. These
addresses are also probed, as if they were in the 'normal' list.
probe_range: insmod parameter. Initialize this list with SENSORS_I2C_END
values.
A list of triples. The first value is a bus number (SENSORS_ISA_BUS for
the ISA bus, -1 for any I2C bus), the second and third are addresses.
These form an inclusive range of addresses that are also probed, as
if they were in the 'normal' list.
ignore: insmod parameter. Initialize this list with SENSORS_I2C_END values.
A list of pairs. The first value is a bus number (SENSORS_ISA_BUS for
the ISA bus, -1 for any I2C bus), the second is the I2C address. These
addresses are never probed. This parameter overrules 'normal' and
'probe', but not the 'force' lists.
ignore_range: insmod parameter. Initialize this list with SENSORS_I2C_END
values.
A list of triples. The first value is a bus number (SENSORS_ISA_BUS for
the ISA bus, -1 for any I2C bus), the second and third are addresses.
These form an inclusive range of I2C addresses that are never probed.
This parameter overrules 'normal' and 'probe', but not the 'force' lists.
Also used is a list of pointers to sensors_force_data structures:
force_data: insmod parameters. A list, ending with an element of which
@ -269,16 +226,14 @@ Also used is a list of pointers to sensors_force_data structures:
So we have a generic insmod variabled `force', and chip-specific variables
`force_CHIPNAME'.
Fortunately, as a module writer, you just have to define the `normal'
and/or `normal_range' parameters, and define what chip names are used.
Fortunately, as a module writer, you just have to define the `normal_i2c'
and `normal_isa' parameters, and define what chip names are used.
The complete declaration could look like this:
/* Scan i2c addresses 0x20 to 0x2f, 0x37, and 0x40 to 0x4f
static unsigned short normal_i2c[] = {0x37,SENSORS_I2C_END};
static unsigned short normal_i2c_range[] = {0x20,0x2f,0x40,0x4f,
SENSORS_I2C_END};
/* Scan i2c addresses 0x37, and 0x48 to 0x4f */
static unsigned short normal_i2c[] = { 0x37, 0x48, 0x49, 0x4a, 0x4b, 0x4c,
0x4d, 0x4e, 0x4f, I2C_CLIENT_END };
/* Scan ISA address 0x290 */
static unsigned int normal_isa[] = {0x0290,SENSORS_ISA_END};
static unsigned int normal_isa_range[] = {SENSORS_ISA_END};
/* Define chips foo and bar, as well as all module parameters and things */
SENSORS_INSMOD_2(foo,bar);

114
Documentation/infiniband/core_locking.txt Normal file
View File

@ -0,0 +1,114 @@
INFINIBAND MIDLAYER LOCKING
This guide is an attempt to make explicit the locking assumptions
made by the InfiniBand midlayer. It describes the requirements on
both low-level drivers that sit below the midlayer and upper level
protocols that use the midlayer.
Sleeping and interrupt context
With the following exceptions, a low-level driver implementation of
all of the methods in struct ib_device may sleep. The exceptions
are any methods from the list:
create_ah
modify_ah
query_ah
destroy_ah
bind_mw
post_send
post_recv
poll_cq
req_notify_cq
map_phys_fmr
which may not sleep and must be callable from any context.
The corresponding functions exported to upper level protocol
consumers:
ib_create_ah
ib_modify_ah
ib_query_ah
ib_destroy_ah
ib_bind_mw
ib_post_send
ib_post_recv
ib_req_notify_cq
ib_map_phys_fmr
are therefore safe to call from any context.
In addition, the function
ib_dispatch_event
used by low-level drivers to dispatch asynchronous events through
the midlayer is also safe to call from any context.
Reentrancy
All of the methods in struct ib_device exported by a low-level
driver must be fully reentrant. The low-level driver is required to
perform all synchronization necessary to maintain consistency, even
if multiple function calls using the same object are run
simultaneously.
The IB midlayer does not perform any serialization of function calls.
Because low-level drivers are reentrant, upper level protocol
consumers are not required to perform any serialization. However,
some serialization may be required to get sensible results. For
example, a consumer may safely call ib_poll_cq() on multiple CPUs
simultaneously. However, the ordering of the work completion
information between different calls of ib_poll_cq() is not defined.
Callbacks
A low-level driver must not perform a callback directly from the
same callchain as an ib_device method call. For example, it is not
allowed for a low-level driver to call a consumer's completion event
handler directly from its post_send method. Instead, the low-level
driver should defer this callback by, for example, scheduling a
tasklet to perform the callback.
The low-level driver is responsible for ensuring that multiple
completion event handlers for the same CQ are not called
simultaneously. The driver must guarantee that only one CQ event
handler for a given CQ is running at a time. In other words, the
following situation is not allowed:
CPU1 CPU2
low-level driver ->
consumer CQ event callback:
/* ... */
ib_req_notify_cq(cq, ...);
low-level driver ->
/* ... */ consumer CQ event callback:
/* ... */
return from CQ event handler
The context in which completion event and asynchronous event
callbacks run is not defined. Depending on the low-level driver, it
may be process context, softirq context, or interrupt context.
Upper level protocol consumers may not sleep in a callback.
Hot-plug
A low-level driver announces that a device is ready for use by
consumers when it calls ib_register_device(), all initialization
must be complete before this call. The device must remain usable
until the driver's call to ib_unregister_device() has returned.
A low-level driver must call ib_register_device() and
ib_unregister_device() from process context. It must not hold any
semaphores that could cause deadlock if a consumer calls back into
the driver across these calls.
An upper level protocol consumer may begin using an IB device as
soon as the add method of its struct ib_client is called for that
device. A consumer must finish all cleanup and free all resources
relating to a device before returning from the remove method.
A consumer is permitted to sleep in its add and remove methods.

51
Documentation/infiniband/user_mad.txt
View File

@ -28,13 +28,37 @@ Creating MAD agents
Receiving MADs
MADs are received using read(). The buffer passed to read() must be
large enough to hold at least one struct ib_user_mad. For example:
MADs are received using read(). The receive side now supports
RMPP. The buffer passed to read() must be at least one
struct ib_user_mad + 256 bytes. For example:
struct ib_user_mad mad;
ret = read(fd, &mad, sizeof mad);
if (ret != sizeof mad)
If the buffer passed is not large enough to hold the received
MAD (RMPP), the errno is set to ENOSPC and the length of the
buffer needed is set in mad.length.
Example for normal MAD (non RMPP) reads:
struct ib_user_mad *mad;
mad = malloc(sizeof *mad + 256);
ret = read(fd, mad, sizeof *mad + 256);
if (ret != sizeof mad + 256) {
perror("read");
free(mad);
}
Example for RMPP reads:
struct ib_user_mad *mad;
mad = malloc(sizeof *mad + 256);
ret = read(fd, mad, sizeof *mad + 256);
if (ret == -ENOSPC)) {
length = mad.length;
free(mad);
mad = malloc(sizeof *mad + length);
ret = read(fd, mad, sizeof *mad + length);
}
if (ret < 0) {
perror("read");
free(mad);
}
In addition to the actual MAD contents, the other struct ib_user_mad
fields will be filled in with information on the received MAD. For
@ -50,18 +74,21 @@ Sending MADs
MADs are sent using write(). The agent ID for sending should be
filled into the id field of the MAD, the destination LID should be
filled into the lid field, and so on. For example:
filled into the lid field, and so on. The send side does support
RMPP so arbitrary length MAD can be sent. For example:
struct ib_user_mad mad;
struct ib_user_mad *mad;
/* fill in mad.data */
mad = malloc(sizeof *mad + mad_length);
mad.id = my_agent; /* req.id from agent registration */
mad.lid = my_dest; /* in network byte order... */
/* fill in mad->data */
mad->hdr.id = my_agent; /* req.id from agent registration */
mad->hdr.lid = my_dest; /* in network byte order... */
/* etc. */
ret = write(fd, &mad, sizeof mad);
if (ret != sizeof mad)
ret = write(fd, &mad, sizeof *mad + mad_length);
if (ret != sizeof *mad + mad_length)
perror("write");
Setting IsSM Capability Bit

69
Documentation/infiniband/user_verbs.txt Normal file
View File

@ -0,0 +1,69 @@
USERSPACE VERBS ACCESS
The ib_uverbs module, built by enabling CONFIG_INFINIBAND_USER_VERBS,
enables direct userspace access to IB hardware via "verbs," as
described in chapter 11 of the InfiniBand Architecture Specification.
To use the verbs, the libibverbs library, available from
<http://openib.org/>, is required. libibverbs contains a
device-independent API for using the ib_uverbs interface.
libibverbs also requires appropriate device-dependent kernel and
userspace driver for your InfiniBand hardware. For example, to use
a Mellanox HCA, you will need the ib_mthca kernel module and the
libmthca userspace driver be installed.
User-kernel communication
Userspace communicates with the kernel for slow path, resource
management operations via the /dev/infiniband/uverbsN character
devices. Fast path operations are typically performed by writing
directly to hardware registers mmap()ed into userspace, with no
system call or context switch into the kernel.
Commands are sent to the kernel via write()s on these device files.
The ABI is defined in drivers/infiniband/include/ib_user_verbs.h.
The structs for commands that require a response from the kernel
contain a 64-bit field used to pass a pointer to an output buffer.
Status is returned to userspace as the return value of the write()
system call.
Resource management
Since creation and destruction of all IB resources is done by
commands passed through a file descriptor, the kernel can keep track
of which resources are attached to a given userspace context. The
ib_uverbs module maintains idr tables that are used to translate
between kernel pointers and opaque userspace handles, so that kernel
pointers are never exposed to userspace and userspace cannot trick
the kernel into following a bogus pointer.
This also allows the kernel to clean up when a process exits and
prevent one process from touching another process's resources.
Memory pinning
Direct userspace I/O requires that memory regions that are potential
I/O targets be kept resident at the same physical address. The
ib_uverbs module manages pinning and unpinning memory regions via
get_user_pages() and put_page() calls. It also accounts for the
amount of memory pinned in the process's locked_vm, and checks that
unprivileged processes do not exceed their RLIMIT_MEMLOCK limit.
Pages that are pinned multiple times are counted each time they are
pinned, so the value of locked_vm may be an overestimate of the
number of pages pinned by a process.
/dev files
To create the appropriate character device files automatically with
udev, a rule like
KERNEL="uverbs*", NAME="infiniband/%k"
can be used. This will create device nodes named
/dev/infiniband/uverbs0
and so on. Since the InfiniBand userspace verbs should be safe for
use by non-privileged processes, it may be useful to add an
appropriate MODE or GROUP to the udev rule.

179
Documentation/kdump/gdbmacros.txt Normal file
View File

@ -0,0 +1,179 @@
#
# This file contains a few gdb macros (user defined commands) to extract
# useful information from kernel crashdump (kdump) like stack traces of
# all the processes or a particular process and trapinfo.
#
# These macros can be used by copying this file in .gdbinit (put in home
# directory or current directory) or by invoking gdb command with
# --command=<command-file-name> option
#
# Credits:
# Alexander Nyberg <alexn@telia.com>
# V Srivatsa <vatsa@in.ibm.com>
# Maneesh Soni <maneesh@in.ibm.com>
#
define bttnobp
set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
set $pid_off=((size_t)&((struct task_struct *)0)->pids[1].pid_list.next)
set $init_t=&init_task
set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
while ($next_t != $init_t)
set $next_t=(struct task_struct *)$next_t
printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm
printf "===================\n"
set var $stackp = $next_t.thread.esp
set var $stack_top = ($stackp & ~4095) + 4096
while ($stackp < $stack_top)
if (*($stackp) > _stext && *($stackp) < _sinittext)
info symbol *($stackp)
end
set $stackp += 4
end
set $next_th=(((char *)$next_t->pids[1].pid_list.next) - $pid_off)
while ($next_th != $next_t)
set $next_th=(struct task_struct *)$next_th
printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm
printf "===================\n"
set var $stackp = $next_t.thread.esp
set var $stack_top = ($stackp & ~4095) + 4096
while ($stackp < $stack_top)
if (*($stackp) > _stext && *($stackp) < _sinittext)
info symbol *($stackp)
end
set $stackp += 4
end
set $next_th=(((char *)$next_th->pids[1].pid_list.next) - $pid_off)
end
set $next_t=(char *)($next_t->tasks.next) - $tasks_off
end
end
document bttnobp
dump all thread stack traces on a kernel compiled with !CONFIG_FRAME_POINTER
end
define btt
set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
set $pid_off=((size_t)&((struct task_struct *)0)->pids[1].pid_list.next)
set $init_t=&init_task
set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
while ($next_t != $init_t)
set $next_t=(struct task_struct *)$next_t
printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm
printf "===================\n"
set var $stackp = $next_t.thread.esp
set var $stack_top = ($stackp & ~4095) + 4096
set var $stack_bot = ($stackp & ~4095)
set $stackp = *($stackp)
while (($stackp < $stack_top) && ($stackp > $stack_bot))
set var $addr = *($stackp + 4)
info symbol $addr
set $stackp = *($stackp)
end
set $next_th=(((char *)$next_t->pids[1].pid_list.next) - $pid_off)
while ($next_th != $next_t)
set $next_th=(struct task_struct *)$next_th
printf "\npid %d; comm %s:\n", $next_t.pid, $next_t.comm
printf "===================\n"
set var $stackp = $next_t.thread.esp
set var $stack_top = ($stackp & ~4095) + 4096
set var $stack_bot = ($stackp & ~4095)
set $stackp = *($stackp)
while (($stackp < $stack_top) && ($stackp > $stack_bot))
set var $addr = *($stackp + 4)
info symbol $addr
set $stackp = *($stackp)
end
set $next_th=(((char *)$next_th->pids[1].pid_list.next) - $pid_off)
end
set $next_t=(char *)($next_t->tasks.next) - $tasks_off
end
end
document btt
dump all thread stack traces on a kernel compiled with CONFIG_FRAME_POINTER
end
define btpid
set var $pid = $arg0
set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
set $pid_off=((size_t)&((struct task_struct *)0)->pids[1].pid_list.next)
set $init_t=&init_task
set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
set var $pid_task = 0
while ($next_t != $init_t)
set $next_t=(struct task_struct *)$next_t
if ($next_t.pid == $pid)
set $pid_task = $next_t
end
set $next_th=(((char *)$next_t->pids[1].pid_list.next) - $pid_off)
while ($next_th != $next_t)
set $next_th=(struct task_struct *)$next_th
if ($next_th.pid == $pid)
set $pid_task = $next_th
end
set $next_th=(((char *)$next_th->pids[1].pid_list.next) - $pid_off)
end
set $next_t=(char *)($next_t->tasks.next) - $tasks_off
end
printf "\npid %d; comm %s:\n", $pid_task.pid, $pid_task.comm
printf "===================\n"
set var $stackp = $pid_task.thread.esp
set var $stack_top = ($stackp & ~4095) + 4096
set var $stack_bot = ($stackp & ~4095)
set $stackp = *($stackp)
while (($stackp < $stack_top) && ($stackp > $stack_bot))
set var $addr = *($stackp + 4)
info symbol $addr
set $stackp = *($stackp)
end
end
document btpid
backtrace of pid
end
define trapinfo
set var $pid = $arg0
set $tasks_off=((size_t)&((struct task_struct *)0)->tasks)
set $pid_off=((size_t)&((struct task_struct *)0)->pids[1].pid_list.next)
set $init_t=&init_task
set $next_t=(((char *)($init_t->tasks).next) - $tasks_off)
set var $pid_task = 0
while ($next_t != $init_t)
set $next_t=(struct task_struct *)$next_t
if ($next_t.pid == $pid)
set $pid_task = $next_t
end
set $next_th=(((char *)$next_t->pids[1].pid_list.next) - $pid_off)
while ($next_th != $next_t)
set $next_th=(struct task_struct *)$next_th
if ($next_th.pid == $pid)
set $pid_task = $next_th
end
set $next_th=(((char *)$next_th->pids[1].pid_list.next) - $pid_off)
end
set $next_t=(char *)($next_t->tasks.next) - $tasks_off
end
printf "Trapno %ld, cr2 0x%lx, error_code %ld\n", $pid_task.thread.trap_no, \
$pid_task.thread.cr2, $pid_task.thread.error_code
end
document trapinfo
Run info threads and lookup pid of thread #1
'trapinfo <pid>' will tell you by which trap & possibly
addresthe kernel paniced.
end

141
Documentation/kdump/kdump.txt Normal file
View File

@ -0,0 +1,141 @@
Documentation for kdump - the kexec-based crash dumping solution
================================================================
DESIGN
======
Kdump uses kexec to reboot to a second kernel whenever a dump needs to be taken.
This second kernel is booted with very little memory. The first kernel reserves
the section of memory that the second kernel uses. This ensures that on-going
DMA from the first kernel does not corrupt the second kernel.
All the necessary information about Core image is encoded in ELF format and
stored in reserved area of memory before crash. Physical address of start of
ELF header is passed to new kernel through command line parameter elfcorehdr=.
On i386, the first 640 KB of physical memory is needed to boot, irrespective
of where the kernel loads. Hence, this region is backed up by kexec just before
rebooting into the new kernel.
In the second kernel, "old memory" can be accessed in two ways.
- The first one is through a /dev/oldmem device interface. A capture utility
can read the device file and write out the memory in raw format. This is raw
dump of memory and analysis/capture tool should be intelligent enough to
determine where to look for the right information. ELF headers (elfcorehdr=)
can become handy here.
- The second interface is through /proc/vmcore. This exports the dump as an ELF
format file which can be written out using any file copy command
(cp, scp, etc). Further, gdb can be used to perform limited debugging on
the dump file. This method ensures methods ensure that there is correct
ordering of the dump pages (corresponding to the first 640 KB that has been
relocated).
SETUP
=====
1) Download http://www.xmission.com/~ebiederm/files/kexec/kexec-tools-1.101.tar.gz
and apply http://lse.sourceforge.net/kdump/patches/kexec-tools-1.101-kdump.patch
and after that build the source.
2) Download and build the appropriate (latest) kexec/kdump (-mm) kernel
patchset and apply it to the vanilla kernel tree.
Two kernels need to be built in order to get this feature working.
A) First kernel:
a) Enable "kexec system call" feature (in Processor type and features).
CONFIG_KEXEC=y
b) This kernel's physical load address should be the default value of
0x100000 (0x100000, 1 MB) (in Processor type and features).
CONFIG_PHYSICAL_START=0x100000
c) Enable "sysfs file system support" (in Pseudo filesystems).
CONFIG_SYSFS=y
d) Boot into first kernel with the command line parameter "crashkernel=Y@X".
Use appropriate values for X and Y. Y denotes how much memory to reserve
for the second kernel, and X denotes at what physical address the reserved
memory section starts. For example: "crashkernel=64M@16M".
B) Second kernel:
a) Enable "kernel crash dumps" feature (in Processor type and features).
CONFIG_CRASH_DUMP=y
b) Specify a suitable value for "Physical address where the kernel is
loaded" (in Processor type and features). Typically this value
should be same as X (See option d) above, e.g., 16 MB or 0x1000000.
CONFIG_PHYSICAL_START=0x1000000
c) Enable "/proc/vmcore support" (Optional, in Pseudo filesystems).
CONFIG_PROC_VMCORE=y
d) Disable SMP support and build a UP kernel (Until it is fixed).
CONFIG_SMP=n
e) Enable "Local APIC support on uniprocessors".
CONFIG_X86_UP_APIC=y
f) Enable "IO-APIC support on uniprocessors"
CONFIG_X86_UP_IOAPIC=y
Note: i) Options a) and b) depend upon "Configure standard kernel features
(for small systems)" (under General setup).
ii) Option a) also depends on CONFIG_HIGHMEM (under Processor
type and features).
iii) Both option a) and b) are under "Processor type and features".
3) Boot into the first kernel. You are now ready to try out kexec-based crash
dumps.
4) Load the second kernel to be booted using:
kexec -p <second-kernel> --crash-dump --args-linux --append="root=<root-dev>
init 1 irqpoll"
Note: i) <second-kernel> has to be a vmlinux image. bzImage will not work,
as of now.
ii) By default ELF headers are stored in ELF32 format (for i386). This
is sufficient to represent the physical memory up to 4GB. To store
headers in ELF64 format, specifiy "--elf64-core-headers" on the
kexec command line additionally.
iii) Specify "irqpoll" as command line parameter. This reduces driver
initialization failures in second kernel due to shared interrupts.
5) System reboots into the second kernel when a panic occurs. A module can be
written to force the panic or "ALT-SysRq-c" can be used initiate a crash
dump for testing purposes.
6) Write out the dump file using
cp /proc/vmcore <dump-file>
Dump memory can also be accessed as a /dev/oldmem device for a linear/raw
view. To create the device, type:
mknod /dev/oldmem c 1 12
Use "dd" with suitable options for count, bs and skip to access specific
portions of the dump.
Entire memory: dd if=/dev/oldmem of=oldmem.001
ANALYSIS
========
Limited analysis can be done using gdb on the dump file copied out of
/proc/vmcore. Use vmlinux built with -g and run
gdb vmlinux <dump-file>
Stack trace for the task on processor 0, register display, memory display
work fine.
Note: gdb cannot analyse core files generated in ELF64 format for i386.
TODO
====
1) Provide a kernel pages filtering mechanism so that core file size is not
insane on systems having huge memory banks.
2) Modify "crash" tool to make it recognize this dump.
CONTACT
=======
Vivek Goyal (vgoyal@in.ibm.com)
Maneesh Soni (maneesh@in.ibm.com)

38
Documentation/kernel-parameters.txt
View File

@ -37,7 +37,7 @@ restrictions referred to are that the relevant option is valid if:
IA-32 IA-32 aka i386 architecture is enabled.
IA-64 IA-64 architecture is enabled.
IOSCHED More than one I/O scheduler is enabled.
IP_PNP IP DCHP, BOOTP, or RARP is enabled.
IP_PNP IP DHCP, BOOTP, or RARP is enabled.
ISAPNP ISA PnP code is enabled.
ISDN Appropriate ISDN support is enabled.
JOY Appropriate joystick support is enabled.
@ -159,6 +159,11 @@ running once the system is up.
acpi_fake_ecdt [HW,ACPI] Workaround failure due to BIOS lacking ECDT
acpi_generic_hotkey [HW,ACPI]
Allow consolidated generic hotkey driver to
over-ride platform specific driver.
See also Documentation/acpi-hotkey.txt.
ad1816= [HW,OSS]
Format: <io>,<irq>,<dma>,<dma2>
See also Documentation/sound/oss/AD1816.
@ -358,6 +363,10 @@ running once the system is up.
cpia_pp= [HW,PPT]
Format: { parport<nr> | auto | none }
crashkernel=nn[KMG]@ss[KMG]
[KNL] Reserve a chunk of physical memory to
hold a kernel to switch to with kexec on panic.
cs4232= [HW,OSS]
Format: <io>,<irq>,<dma>,<dma2>,<mpuio>,<mpuirq>
@ -447,6 +456,10 @@ running once the system is up.
Format: {"as"|"cfq"|"deadline"|"noop"}
See Documentation/block/as-iosched.txt
and Documentation/block/deadline-iosched.txt for details.
elfcorehdr= [IA-32]
Specifies physical address of start of kernel core image
elf header.
See Documentation/kdump.txt for details.
enforcing [SELINUX] Set initial enforcing status.
Format: {"0" | "1"}
@ -548,6 +561,9 @@ running once the system is up.
i810= [HW,DRM]
i8k.ignore_dmi [HW] Continue probing hardware even if DMI data
indicates that the driver is running on unsupported
hardware.
i8k.force [HW] Activate i8k driver even if SMM BIOS signature
does not match list of supported models.
i8k.power_status
@ -611,6 +627,17 @@ running once the system is up.
ips= [HW,SCSI] Adaptec / IBM ServeRAID controller
See header of drivers/scsi/ips.c.
irqfixup [HW]
When an interrupt is not handled search all handlers
for it. Intended to get systems with badly broken
firmware running.
irqpoll [HW]
When an interrupt is not handled search all handlers
for it. Also check all handlers each timer
interrupt. Intended to get systems with badly broken
firmware running.
isapnp= [ISAPNP]
Format: <RDP>, <reset>, <pci_scan>, <verbosity>
@ -736,6 +763,9 @@ running once the system is up.
maxcpus= [SMP] Maximum number of processors that an SMP kernel
should make use of
max_addr=[KMG] [KNL,BOOT,ia64] All physical memory greater than or
equal to this physical address is ignored.
max_luns= [SCSI] Maximum number of LUNs to probe
Should be between 1 and 2^32-1.
@ -1019,6 +1049,10 @@ running once the system is up.
irqmask=0xMMMM [IA-32] Set a bit mask of IRQs allowed to be assigned
automatically to PCI devices. You can make the kernel
exclude IRQs of your ISA cards this way.
pirqaddr=0xAAAAA [IA-32] Specify the physical address
of the PIRQ table (normally generated
by the BIOS) if it is outside the
F0000h-100000h range.
lastbus=N [IA-32] Scan all buses till bus #N. Can be useful
if the kernel is unable to find your secondary buses
and you want to tell it explicitly which ones they are.
@ -1104,7 +1138,7 @@ running once the system is up.
See Documentation/ramdisk.txt.
psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to
probe for (bare|imps|exps).
probe for (bare|imps|exps|lifebook|any).
psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports
per second.
psmouse.resetafter=

317
Documentation/keys.txt
View File

@ -22,6 +22,7 @@ This document has the following sections:
- New procfs files
- Userspace system call interface
- Kernel services
- Notes on accessing payload contents
- Defining a key type
- Request-key callback service
- Key access filesystem
@ -45,27 +46,26 @@ Each key has a number of attributes:
- State.
(*) Each key is issued a serial number of type key_serial_t that is unique
for the lifetime of that key. All serial numbers are positive non-zero
32-bit integers.
(*) Each key is issued a serial number of type key_serial_t that is unique for
the lifetime of that key. All serial numbers are positive non-zero 32-bit
integers.
Userspace programs can use a key's serial numbers as a way to gain access
to it, subject to permission checking.
(*) Each key is of a defined "type". Types must be registered inside the
kernel by a kernel service (such as a filesystem) before keys of that
type can be added or used. Userspace programs cannot define new types
directly.
kernel by a kernel service (such as a filesystem) before keys of that type
can be added or used. Userspace programs cannot define new types directly.
Key types are represented in the kernel by struct key_type. This defines
a number of operations that can be performed on a key of that type.
Key types are represented in the kernel by struct key_type. This defines a
number of operations that can be performed on a key of that type.
Should a type be removed from the system, all the keys of that type will
be invalidated.
(*) Each key has a description. This should be a printable string. The key
type provides an operation to perform a match between the description on
a key and a criterion string.
type provides an operation to perform a match between the description on a
key and a criterion string.
(*) Each key has an owner user ID, a group ID and a permissions mask. These
are used to control what a process may do to a key from userspace, and
@ -74,10 +74,10 @@ Each key has a number of attributes:
(*) Each key can be set to expire at a specific time by the key type's
instantiation function. Keys can also be immortal.
(*) Each key can have a payload. This is a quantity of data that represent
the actual "key". In the case of a keyring, this is a list of keys to
which the keyring links; in the case of a user-defined key, it's an
arbitrary blob of data.
(*) Each key can have a payload. This is a quantity of data that represent the
actual "key". In the case of a keyring, this is a list of keys to which
the keyring links; in the case of a user-defined key, it's an arbitrary
blob of data.
Having a payload is not required; and the payload can, in fact, just be a
value stored in the struct key itself.
@ -92,8 +92,8 @@ Each key has a number of attributes:
(*) Each key can be in one of a number of basic states:
(*) Uninstantiated. The key exists, but does not have any data
attached. Keys being requested from userspace will be in this state.
(*) Uninstantiated. The key exists, but does not have any data attached.
Keys being requested from userspace will be in this state.
(*) Instantiated. This is the normal state. The key is fully formed, and
has data attached.
@ -140,10 +140,10 @@ The key service provides a number of features besides keys:
clone, fork, vfork or execve occurs. A new keyring is created only when
required.
The process-specific keyring is replaced with an empty one in the child
on clone, fork, vfork unless CLONE_THREAD is supplied, in which case it
is shared. execve also discards the process's process keyring and creates
a new one.
The process-specific keyring is replaced with an empty one in the child on
clone, fork, vfork unless CLONE_THREAD is supplied, in which case it is
shared. execve also discards the process's process keyring and creates a
new one.
The session-specific keyring is persistent across clone, fork, vfork and
execve, even when the latter executes a set-UID or set-GID binary. A
@ -177,11 +177,11 @@ The key service provides a number of features besides keys:
If a system call that modifies a key or keyring in some way would put the
user over quota, the operation is refused and error EDQUOT is returned.
(*) There's a system call interface by which userspace programs can create
and manipulate keys and keyrings.
(*) There's a system call interface by which userspace programs can create and
manipulate keys and keyrings.
(*) There's a kernel interface by which services can register types and
search for keys.
(*) There's a kernel interface by which services can register types and search
for keys.
(*) There's a way for the a search done from the kernel to call back to
userspace to request a key that can't be found in a process's keyrings.
@ -194,9 +194,9 @@ The key service provides a number of features besides keys:
KEY ACCESS PERMISSIONS
======================
Keys have an owner user ID, a group access ID, and a permissions mask. The
mask has up to eight bits each for user, group and other access. Only five of
each set of eight bits are defined. These permissions granted are:
Keys have an owner user ID, a group access ID, and a permissions mask. The mask
has up to eight bits each for user, group and other access. Only five of each
set of eight bits are defined. These permissions granted are:
(*) View
@ -210,8 +210,8 @@ each set of eight bits are defined. These permissions granted are:
(*) Write
This permits a key's payload to be instantiated or updated, or it allows
a link to be added to or removed from a keyring.
This permits a key's payload to be instantiated or updated, or it allows a
link to be added to or removed from a keyring.
(*) Search
@ -238,8 +238,8 @@ about the status of the key service:
(*) /proc/keys
This lists all the keys on the system, giving information about their
type, description and permissions. The payload of the key is not
available this way:
type, description and permissions. The payload of the key is not available
this way:
SERIAL FLAGS USAGE EXPY PERM UID GID TYPE DESCRIPTION: SUMMARY
00000001 I----- 39 perm 1f0000 0 0 keyring _uid_ses.0: 1/4
@ -318,21 +318,21 @@ The main syscalls are:
If a key of the same type and description as that proposed already exists
in the keyring, this will try to update it with the given payload, or it
will return error EEXIST if that function is not supported by the key
type. The process must also have permission to write to the key to be
able to update it. The new key will have all user permissions granted and
no group or third party permissions.
type. The process must also have permission to write to the key to be able
to update it. The new key will have all user permissions granted and no
group or third party permissions.
Otherwise, this will attempt to create a new key of the specified type
and description, and to instantiate it with the supplied payload and
attach it to the keyring. In this case, an error will be generated if the
process does not have permission to write to the keyring.
Otherwise, this will attempt to create a new key of the specified type and
description, and to instantiate it with the supplied payload and attach it
to the keyring. In this case, an error will be generated if the process
does not have permission to write to the keyring.
The payload is optional, and the pointer can be NULL if not required by
the type. The payload is plen in size, and plen can be zero for an empty
payload.
A new keyring can be generated by setting type "keyring", the keyring
name as the description (or NULL) and setting the payload to NULL.
A new keyring can be generated by setting type "keyring", the keyring name
as the description (or NULL) and setting the payload to NULL.
User defined keys can be created by specifying type "user". It is
recommended that a user defined key's description by prefixed with a type
@ -369,9 +369,9 @@ The keyctl syscall functions are:
key_serial_t keyctl(KEYCTL_GET_KEYRING_ID, key_serial_t id,
int create);
The special key specified by "id" is looked up (with the key being
created if necessary) and the ID of the key or keyring thus found is
returned if it exists.
The special key specified by "id" is looked up (with the key being created
if necessary) and the ID of the key or keyring thus found is returned if
it exists.
If the key does not yet exist, the key will be created if "create" is
non-zero; and the error ENOKEY will be returned if "create" is zero.
@ -402,8 +402,8 @@ The keyctl syscall functions are:
This will try to update the specified key with the given payload, or it
will return error EOPNOTSUPP if that function is not supported by the key
type. The process must also have permission to write to the key to be
able to update it.
type. The process must also have permission to write to the key to be able
to update it.
The payload is of length plen, and may be absent or empty as for
add_key().
@ -422,8 +422,8 @@ The keyctl syscall functions are:
long keyctl(KEYCTL_CHOWN, key_serial_t key, uid_t uid, gid_t gid);
This function permits a key's owner and group ID to be changed. Either
one of uid or gid can be set to -1 to suppress that change.
This function permits a key's owner and group ID to be changed. Either one
of uid or gid can be set to -1 to suppress that change.
Only the superuser can change a key's owner to something other than the
key's current owner. Similarly, only the superuser can change a key's
@ -484,12 +484,12 @@ The keyctl syscall functions are:
long keyctl(KEYCTL_LINK, key_serial_t keyring, key_serial_t key);
This function creates a link from the keyring to the key. The process
must have write permission on the keyring and must have link permission
on the key.
This function creates a link from the keyring to the key. The process must
have write permission on the keyring and must have link permission on the
key.
Should the keyring not be a keyring, error ENOTDIR will result; and if
the keyring is full, error ENFILE will result.
Should the keyring not be a keyring, error ENOTDIR will result; and if the
keyring is full, error ENFILE will result.
The link procedure checks the nesting of the keyrings, returning ELOOP if
it appears to deep or EDEADLK if the link would introduce a cycle.
@ -503,8 +503,8 @@ The keyctl syscall functions are:
specified key, and removes it if found. Subsequent links to that key are
ignored. The process must have write permission on the keyring.
If the keyring is not a keyring, error ENOTDIR will result; and if the
key is not present, error ENOENT will be the result.
If the keyring is not a keyring, error ENOTDIR will result; and if the key
is not present, error ENOENT will be the result.
(*) Search a keyring tree for a key:
@ -513,9 +513,9 @@ The keyctl syscall functions are:
const char *type, const char *description,
key_serial_t dest_keyring);
This searches the keyring tree headed by the specified keyring until a
key is found that matches the type and description criteria. Each keyring
is checked for keys before recursion into its children occurs.
This searches the keyring tree headed by the specified keyring until a key
is found that matches the type and description criteria. Each keyring is
checked for keys before recursion into its children occurs.
The process must have search permission on the top level keyring, or else
error EACCES will result. Only keyrings that the process has search
@ -549,8 +549,8 @@ The keyctl syscall functions are:
As much of the data as can be fitted into the buffer will be copied to
userspace if the buffer pointer is not NULL.
On a successful return, the function will always return the amount of
data available rather than the amount copied.
On a successful return, the function will always return the amount of data
available rather than the amount copied.
(*) Instantiate a partially constructed key.
@ -568,8 +568,8 @@ The keyctl syscall functions are:
it, and the key must be uninstantiated.
If a keyring is specified (non-zero), the key will also be linked into
that keyring, however all the constraints applying in KEYCTL_LINK apply
in this case too.
that keyring, however all the constraints applying in KEYCTL_LINK apply in
this case too.
The payload and plen arguments describe the payload data as for add_key().
@ -587,8 +587,39 @@ The keyctl syscall functions are:
it, and the key must be uninstantiated.
If a keyring is specified (non-zero), the key will also be linked into
that keyring, however all the constraints applying in KEYCTL_LINK apply
in this case too.
that keyring, however all the constraints applying in KEYCTL_LINK apply in
this case too.
(*) Set the default request-key destination keyring.
long keyctl(KEYCTL_SET_REQKEY_KEYRING, int reqkey_defl);
This sets the default keyring to which implicitly requested keys will be
attached for this thread. reqkey_defl should be one of these constants:
CONSTANT VALUE NEW DEFAULT KEYRING
====================================== ====== =======================
KEY_REQKEY_DEFL_NO_CHANGE -1 No change
KEY_REQKEY_DEFL_DEFAULT 0 Default[1]
KEY_REQKEY_DEFL_THREAD_KEYRING 1 Thread keyring
KEY_REQKEY_DEFL_PROCESS_KEYRING 2 Process keyring
KEY_REQKEY_DEFL_SESSION_KEYRING 3 Session keyring
KEY_REQKEY_DEFL_USER_KEYRING 4 User keyring
KEY_REQKEY_DEFL_USER_SESSION_KEYRING 5 User session keyring
KEY_REQKEY_DEFL_GROUP_KEYRING 6 Group keyring
The old default will be returned if successful and error EINVAL will be
returned if reqkey_defl is not one of the above values.
The default keyring can be overridden by the keyring indicated to the
request_key() system call.
Note that this setting is inherited across fork/exec.
[1] The default default is: the thread keyring if there is one, otherwise
the process keyring if there is one, otherwise the session keyring if
there is one, otherwise the user default session keyring.
===============
@ -601,17 +632,14 @@ be broken down into two areas: keys and key types.
Dealing with keys is fairly straightforward. Firstly, the kernel service
registers its type, then it searches for a key of that type. It should retain
the key as long as it has need of it, and then it should release it. For a
filesystem or device file, a search would probably be performed during the
open call, and the key released upon close. How to deal with conflicting keys
due to two different users opening the same file is left to the filesystem
author to solve.
filesystem or device file, a search would probably be performed during the open
call, and the key released upon close. How to deal with conflicting keys due to
two different users opening the same file is left to the filesystem author to
solve.
When accessing a key's payload data, key->lock should be at least read locked,
or else the data may be changed by an update being performed from userspace
whilst the driver or filesystem is trying to access it. If no update method is
supplied, then the key's payload may be accessed without holding a lock as
there is no way to change it, provided it can be guaranteed that the key's
type definition won't go away.
When accessing a key's payload contents, certain precautions must be taken to
prevent access vs modification races. See the section "Notes on accessing
payload contents" for more information.
(*) To search for a key, call:
@ -629,6 +657,9 @@ type definition won't go away.
Should the function fail error ENOKEY, EKEYEXPIRED or EKEYREVOKED will be
returned.
If successful, the key will have been attached to the default keyring for
implicitly obtained request-key keys, as set by KEYCTL_SET_REQKEY_KEYRING.
(*) When it is no longer required, the key should be released using:
@ -690,6 +721,54 @@ type definition won't go away.
void unregister_key_type(struct key_type *type);
===================================
NOTES ON ACCESSING PAYLOAD CONTENTS
===================================
The simplest payload is just a number in key->payload.value. In this case,
there's no need to indulge in RCU or locking when accessing the payload.
More complex payload contents must be allocated and a pointer to them set in
key->payload.data. One of the following ways must be selected to access the
data:
(1) Unmodifyable key type.
If the key type does not have a modify method, then the key's payload can
be accessed without any form of locking, provided that it's known to be
instantiated (uninstantiated keys cannot be "found").
(2) The key's semaphore.
The semaphore could be used to govern access to the payload and to control
the payload pointer. It must be write-locked for modifications and would
have to be read-locked for general access. The disadvantage of doing this
is that the accessor may be required to sleep.
(3) RCU.
RCU must be used when the semaphore isn't already held; if the semaphore
is held then the contents can't change under you unexpectedly as the
semaphore must still be used to serialise modifications to the key. The
key management code takes care of this for the key type.
However, this means using:
rcu_read_lock() ... rcu_dereference() ... rcu_read_unlock()
to read the pointer, and:
rcu_dereference() ... rcu_assign_pointer() ... call_rcu()
to set the pointer and dispose of the old contents after a grace period.
Note that only the key type should ever modify a key's payload.
Furthermore, an RCU controlled payload must hold a struct rcu_head for the
use of call_rcu() and, if the payload is of variable size, the length of
the payload. key->datalen cannot be relied upon to be consistent with the
payload just dereferenced if the key's semaphore is not held.
===================
DEFINING A KEY TYPE
===================
@ -717,15 +796,15 @@ The structure has a number of fields, some of which are mandatory:
int key_payload_reserve(struct key *key, size_t datalen);
With the revised data length. Error EDQUOT will be returned if this is
not viable.
With the revised data length. Error EDQUOT will be returned if this is not
viable.
(*) int (*instantiate)(struct key *key, const void *data, size_t datalen);
This method is called to attach a payload to a key during construction.
The payload attached need not bear any relation to the data passed to
this function.
The payload attached need not bear any relation to the data passed to this
function.
If the amount of data attached to the key differs from the size in
keytype->def_datalen, then key_payload_reserve() should be called.
@ -734,38 +813,47 @@ The structure has a number of fields, some of which are mandatory:
The fact that KEY_FLAG_INSTANTIATED is not set in key->flags prevents
anything else from gaining access to the key.
This method may sleep if it wishes.
It is safe to sleep in this method.
(*) int (*duplicate)(struct key *key, const struct key *source);
If this type of key can be duplicated, then this method should be
provided. It is called to copy the payload attached to the source into
the new key. The data length on the new key will have been updated and
the quota adjusted already.
provided. It is called to copy the payload attached to the source into the
new key. The data length on the new key will have been updated and the
quota adjusted already.
This method will be called with the source key's semaphore read-locked to
prevent its payload from being changed. It is safe to sleep here.
prevent its payload from being changed, thus RCU constraints need not be
applied to the source key.
This method does not have to lock the destination key in order to attach a
payload. The fact that KEY_FLAG_INSTANTIATED is not set in key->flags
prevents anything else from gaining access to the key.
It is safe to sleep in this method.
(*) int (*update)(struct key *key, const void *data, size_t datalen);
If this type of key can be updated, then this method should be
provided. It is called to update a key's payload from the blob of data
provided.
If this type of key can be updated, then this method should be provided.
It is called to update a key's payload from the blob of data provided.
key_payload_reserve() should be called if the data length might change
before any changes are actually made. Note that if this succeeds, the
type is committed to changing the key because it's already been altered,
so all memory allocation must be done first.
before any changes are actually made. Note that if this succeeds, the type
is committed to changing the key because it's already been altered, so all
memory allocation must be done first.
key_payload_reserve() should be called with the key->lock write locked,
and the changes to the key's attached payload should be made before the
key is locked.
The key will have its semaphore write-locked before this method is called,
but this only deters other writers; any changes to the key's payload must
be made under RCU conditions, and call_rcu() must be used to dispose of
the old payload.
The key will have its semaphore write-locked before this method is
called. Any changes to the key should be made with the key's rwlock
write-locked also. It is safe to sleep here.
key_payload_reserve() should be called before the changes are made, but
after all allocations and other potentially failing function calls are
made.
It is safe to sleep in this method.
(*) int (*match)(const struct key *key, const void *desc);
@ -782,12 +870,12 @@ The structure has a number of fields, some of which are mandatory:
(*) void (*destroy)(struct key *key);
This method is optional. It is called to discard the payload data on a
key when it is being destroyed.
This method is optional. It is called to discard the payload data on a key
when it is being destroyed.
This method does not need to lock the key; it can consider the key as
being inaccessible. Note that the key's type may have changed before this
function is called.
This method does not need to lock the key to access the payload; it can
consider the key as being inaccessible at this time. Note that the key's
type may have been changed before this function is called.
It is not safe to sleep in this method; the caller may hold spinlocks.
@ -797,26 +885,31 @@ The structure has a number of fields, some of which are mandatory:
This method is optional. It is called during /proc/keys reading to
summarise a key's description and payload in text form.
This method will be called with the key's rwlock read-locked. This will
prevent the key's payload and state changing; also the description should
not change. This also means it is not safe to sleep in this method.
This method will be called with the RCU read lock held. rcu_dereference()
should be used to read the payload pointer if the payload is to be
accessed. key->datalen cannot be trusted to stay consistent with the
contents of the payload.
The description will not change, though the key's state may.
It is not safe to sleep in this method; the RCU read lock is held by the
caller.
(*) long (*read)(const struct key *key, char __user *buffer, size_t buflen);
This method is optional. It is called by KEYCTL_READ to translate the
key's payload into something a blob of data for userspace to deal
with. Ideally, the blob should be in the same format as that passed in to
the instantiate and update methods.
key's payload into something a blob of data for userspace to deal with.
Ideally, the blob should be in the same format as that passed in to the
instantiate and update methods.
If successful, the blob size that could be produced should be returned
rather than the size copied.
This method will be called with the key's semaphore read-locked. This
will prevent the key's payload changing. It is not necessary to also
read-lock key->lock when accessing the key's payload. It is safe to sleep
in this method, such as might happen when the userspace buffer is
accessed.
This method will be called with the key's semaphore read-locked. This will
prevent the key's payload changing. It is not necessary to use RCU locking
when accessing the key's payload. It is safe to sleep in this method, such
as might happen when the userspace buffer is accessed.
============================
@ -853,8 +946,8 @@ If it returns with the key remaining in the unconstructed state, the key will
be marked as being negative, it will be added to the session keyring, and an
error will be returned to the key requestor.
Supplementary information may be provided from whoever or whatever invoked
this service. This will be passed as the <callout_info> parameter. If no such
Supplementary information may be provided from whoever or whatever invoked this
service. This will be passed as the <callout_info> parameter. If no such
information was made available, then "-" will be passed as this parameter
instead.

588
Documentation/kprobes.txt Normal file
View File

@ -0,0 +1,588 @@
Title : Kernel Probes (Kprobes)
Authors : Jim Keniston <jkenisto@us.ibm.com>
: Prasanna S Panchamukhi <prasanna@in.ibm.com>
CONTENTS
1. Concepts: Kprobes, Jprobes, Return Probes
2. Architectures Supported
3. Configuring Kprobes
4. API Reference
5. Kprobes Features and Limitations
6. Probe Overhead
7. TODO
8. Kprobes Example
9. Jprobes Example
10. Kretprobes Example
1. Concepts: Kprobes, Jprobes, Return Probes
Kprobes enables you to dynamically break into any kernel routine and
collect debugging and performance information non-disruptively. You
can trap at almost any kernel code address, specifying a handler
routine to be invoked when the breakpoint is hit.
There are currently three types of probes: kprobes, jprobes, and
kretprobes (also called return probes). A kprobe can be inserted
on virtually any instruction in the kernel. A jprobe is inserted at
the entry to a kernel function, and provides convenient access to the
function's arguments. A return probe fires when a specified function
returns.
In the typical case, Kprobes-based instrumentation is packaged as
a kernel module. The module's init function installs ("registers")
one or more probes, and the exit function unregisters them. A
registration function such as register_kprobe() specifies where
the probe is to be inserted and what handler is to be called when
the probe is hit.
The next three subsections explain how the different types of
probes work. They explain certain things that you'll need to
know in order to make the best use of Kprobes -- e.g., the
difference between a pre_handler and a post_handler, and how
to use the maxactive and nmissed fields of a kretprobe. But
if you're in a hurry to start using Kprobes, you can skip ahead
to section 2.
1.1 How Does a Kprobe Work?
When a kprobe is registered, Kprobes makes a copy of the probed
instruction and replaces the first byte(s) of the probed instruction
with a breakpoint instruction (e.g., int3 on i386 and x86_64).
When a CPU hits the breakpoint instruction, a trap occurs, the CPU's
registers are saved, and control passes to Kprobes via the
notifier_call_chain mechanism. Kprobes executes the "pre_handler"
associated with the kprobe, passing the handler the addresses of the
kprobe struct and the saved registers.
Next, Kprobes single-steps its copy of the probed instruction.
(It would be simpler to single-step the actual instruction in place,
but then Kprobes would have to temporarily remove the breakpoint
instruction. This would open a small time window when another CPU
could sail right past the probepoint.)
After the instruction is single-stepped, Kprobes executes the
"post_handler," if any, that is associated with the kprobe.
Execution then continues with the instruction following the probepoint.
1.2 How Does a Jprobe Work?
A jprobe is implemented using a kprobe that is placed on a function's
entry point. It employs a simple mirroring principle to allow
seamless access to the probed function's arguments. The jprobe
handler routine should have the same signature (arg list and return
type) as the function being probed, and must always end by calling
the Kprobes function jprobe_return().
Here's how it works. When the probe is hit, Kprobes makes a copy of
the saved registers and a generous portion of the stack (see below).
Kprobes then points the saved instruction pointer at the jprobe's
handler routine, and returns from the trap. As a result, control
passes to the handler, which is presented with the same register and
stack contents as the probed function. When it is done, the handler
calls jprobe_return(), which traps again to restore the original stack
contents and processor state and switch to the probed function.
By convention, the callee owns its arguments, so gcc may produce code
that unexpectedly modifies that portion of the stack. This is why
Kprobes saves a copy of the stack and restores it after the jprobe
handler has run. Up to MAX_STACK_SIZE bytes are copied -- e.g.,
64 bytes on i386.
Note that the probed function's args may be passed on the stack
or in registers (e.g., for x86_64 or for an i386 fastcall function).
The jprobe will work in either case, so long as the handler's
prototype matches that of the probed function.
1.3 How Does a Return Probe Work?
When you call register_kretprobe(), Kprobes establishes a kprobe at
the entry to the function. When the probed function is called and this
probe is hit, Kprobes saves a copy of the return address, and replaces
the return address with the address of a "trampoline." The trampoline
is an arbitrary piece of code -- typically just a nop instruction.
At boot time, Kprobes registers a kprobe at the trampoline.
When the probed function executes its return instruction, control
passes to the trampoline and that probe is hit. Kprobes' trampoline
handler calls the user-specified handler associated with the kretprobe,
then sets the saved instruction pointer to the saved return address,
and that's where execution resumes upon return from the trap.
While the probed function is executing, its return address is
stored in an object of type kretprobe_instance. Before calling
register_kretprobe(), the user sets the maxactive field of the
kretprobe struct to specify how many instances of the specified
function can be probed simultaneously. register_kretprobe()
pre-allocates the indicated number of kretprobe_instance objects.
For example, if the function is non-recursive and is called with a
spinlock held, maxactive = 1 should be enough. If the function is
non-recursive and can never relinquish the CPU (e.g., via a semaphore
or preemption), NR_CPUS should be enough. If maxactive <= 0, it is
set to a default value. If CONFIG_PREEMPT is enabled, the default
is max(10, 2*NR_CPUS). Otherwise, the default is NR_CPUS.
It's not a disaster if you set maxactive too low; you'll just miss
some probes. In the kretprobe struct, the nmissed field is set to
zero when the return probe is registered, and is incremented every
time the probed function is entered but there is no kretprobe_instance
object available for establishing the return probe.
2. Architectures Supported
Kprobes, jprobes, and return probes are implemented on the following
architectures:
- i386
- x86_64 (AMD-64, E64MT)
- ppc64
- ia64 (Support for probes on certain instruction types is still in progress.)
- sparc64 (Return probes not yet implemented.)
3. Configuring Kprobes
When configuring the kernel using make menuconfig/xconfig/oldconfig,
ensure that CONFIG_KPROBES is set to "y". Under "Kernel hacking",
look for "Kprobes". You may have to enable "Kernel debugging"
(CONFIG_DEBUG_KERNEL) before you can enable Kprobes.
You may also want to ensure that CONFIG_KALLSYMS and perhaps even
CONFIG_KALLSYMS_ALL are set to "y", since kallsyms_lookup_name()
is a handy, version-independent way to find a function's address.
If you need to insert a probe in the middle of a function, you may find
it useful to "Compile the kernel with debug info" (CONFIG_DEBUG_INFO),
so you can use "objdump -d -l vmlinux" to see the source-to-object
code mapping.
4. API Reference
The Kprobes API includes a "register" function and an "unregister"
function for each type of probe. Here are terse, mini-man-page
specifications for these functions and the associated probe handlers
that you'll write. See the latter half of this document for examples.
4.1 register_kprobe
#include <linux/kprobes.h>
int register_kprobe(struct kprobe *kp);
Sets a breakpoint at the address kp->addr. When the breakpoint is
hit, Kprobes calls kp->pre_handler. After the probed instruction
is single-stepped, Kprobe calls kp->post_handler. If a fault
occurs during execution of kp->pre_handler or kp->post_handler,
or during single-stepping of the probed instruction, Kprobes calls
kp->fault_handler. Any or all handlers can be NULL.
register_kprobe() returns 0 on success, or a negative errno otherwise.
User's pre-handler (kp->pre_handler):
#include <linux/kprobes.h>
#include <linux/ptrace.h>
int pre_handler(struct kprobe *p, struct pt_regs *regs);
Called with p pointing to the kprobe associated with the breakpoint,
and regs pointing to the struct containing the registers saved when
the breakpoint was hit. Return 0 here unless you're a Kprobes geek.
User's post-handler (kp->post_handler):
#include <linux/kprobes.h>
#include <linux/ptrace.h>
void post_handler(struct kprobe *p, struct pt_regs *regs,
unsigned long flags);
p and regs are as described for the pre_handler. flags always seems
to be zero.
User's fault-handler (kp->fault_handler):
#include <linux/kprobes.h>
#include <linux/ptrace.h>
int fault_handler(struct kprobe *p, struct pt_regs *regs, int trapnr);
p and regs are as described for the pre_handler. trapnr is the
architecture-specific trap number associated with the fault (e.g.,
on i386, 13 for a general protection fault or 14 for a page fault).
Returns 1 if it successfully handled the exception.
4.2 register_jprobe
#include <linux/kprobes.h>
int register_jprobe(struct jprobe *jp)
Sets a breakpoint at the address jp->kp.addr, which must be the address
of the first instruction of a function. When the breakpoint is hit,
Kprobes runs the handler whose address is jp->entry.
The handler should have the same arg list and return type as the probed
function; and just before it returns, it must call jprobe_return().
(The handler never actually returns, since jprobe_return() returns
control to Kprobes.) If the probed function is declared asmlinkage,
fastcall, or anything else that affects how args are passed, the
handler's declaration must match.
register_jprobe() returns 0 on success, or a negative errno otherwise.
4.3 register_kretprobe
#include <linux/kprobes.h>
int register_kretprobe(struct kretprobe *rp);
Establishes a return probe for the function whose address is
rp->kp.addr. When that function returns, Kprobes calls rp->handler.
You must set rp->maxactive appropriately before you call
register_kretprobe(); see "How Does a Return Probe Work?" for details.
register_kretprobe() returns 0 on success, or a negative errno
otherwise.
User's return-probe handler (rp->handler):
#include <linux/kprobes.h>
#include <linux/ptrace.h>
int kretprobe_handler(struct kretprobe_instance *ri, struct pt_regs *regs);
regs is as described for kprobe.pre_handler. ri points to the
kretprobe_instance object, of which the following fields may be
of interest:
- ret_addr: the return address
- rp: points to the corresponding kretprobe object
- task: points to the corresponding task struct
The handler's return value is currently ignored.
4.4 unregister_*probe
#include <linux/kprobes.h>
void unregister_kprobe(struct kprobe *kp);
void unregister_jprobe(struct jprobe *jp);
void unregister_kretprobe(struct kretprobe *rp);
Removes the specified probe. The unregister function can be called
at any time after the probe has been registered.
5. Kprobes Features and Limitations
As of Linux v2.6.12, Kprobes allows multiple probes at the same
address. Currently, however, there cannot be multiple jprobes on
the same function at the same time.
In general, you can install a probe anywhere in the kernel.
In particular, you can probe interrupt handlers. Known exceptions
are discussed in this section.
For obvious reasons, it's a bad idea to install a probe in
the code that implements Kprobes (mostly kernel/kprobes.c and
arch/*/kernel/kprobes.c). A patch in the v2.6.13 timeframe instructs
Kprobes to reject such requests.
If you install a probe in an inline-able function, Kprobes makes
no attempt to chase down all inline instances of the function and
install probes there. gcc may inline a function without being asked,
so keep this in mind if you're not seeing the probe hits you expect.
A probe handler can modify the environment of the probed function
-- e.g., by modifying kernel data structures, or by modifying the
contents of the pt_regs struct (which are restored to the registers
upon return from the breakpoint). So Kprobes can be used, for example,
to install a bug fix or to inject faults for testing. Kprobes, of
course, has no way to distinguish the deliberately injected faults
from the accidental ones. Don't drink and probe.
Kprobes makes no attempt to prevent probe handlers from stepping on
each other -- e.g., probing printk() and then calling printk() from a
probe handler. As of Linux v2.6.12, if a probe handler hits a probe,
that second probe's handlers won't be run in that instance.
In Linux v2.6.12 and previous versions, Kprobes' data structures are
protected by a single lock that is held during probe registration and
unregistration and while handlers are run. Thus, no two handlers
can run simultaneously. To improve scalability on SMP systems,
this restriction will probably be removed soon, in which case
multiple handlers (or multiple instances of the same handler) may
run concurrently on different CPUs. Code your handlers accordingly.
Kprobes does not use semaphores or allocate memory except during
registration and unregistration.
Probe handlers are run with preemption disabled. Depending on the
architecture, handlers may also run with interrupts disabled. In any
case, your handler should not yield the CPU (e.g., by attempting to
acquire a semaphore).
Since a return probe is implemented by replacing the return
address with the trampoline's address, stack backtraces and calls
to __builtin_return_address() will typically yield the trampoline's
address instead of the real return address for kretprobed functions.
(As far as we can tell, __builtin_return_address() is used only
for instrumentation and error reporting.)
If the number of times a function is called does not match the
number of times it returns, registering a return probe on that
function may produce undesirable results. We have the do_exit()
and do_execve() cases covered. do_fork() is not an issue. We're
unaware of other specific cases where this could be a problem.
6. Probe Overhead
On a typical CPU in use in 2005, a kprobe hit takes 0.5 to 1.0
microseconds to process. Specifically, a benchmark that hits the same
probepoint repeatedly, firing a simple handler each time, reports 1-2
million hits per second, depending on the architecture. A jprobe or
return-probe hit typically takes 50-75% longer than a kprobe hit.
When you have a return probe set on a function, adding a kprobe at
the entry to that function adds essentially no overhead.
Here are sample overhead figures (in usec) for different architectures.
k = kprobe; j = jprobe; r = return probe; kr = kprobe + return probe
on same function; jr = jprobe + return probe on same function
i386: Intel Pentium M, 1495 MHz, 2957.31 bogomips
k = 0.57 usec; j = 1.00; r = 0.92; kr = 0.99; jr = 1.40
x86_64: AMD Opteron 246, 1994 MHz, 3971.48 bogomips
k = 0.49 usec; j = 0.76; r = 0.80; kr = 0.82; jr = 1.07
ppc64: POWER5 (gr), 1656 MHz (SMT disabled, 1 virtual CPU per physical CPU)
k = 0.77 usec; j = 1.31; r = 1.26; kr = 1.45; jr = 1.99
7. TODO
a. SystemTap (http://sourceware.org/systemtap): Work in progress
to provide a simplified programming interface for probe-based
instrumentation.
b. Improved SMP scalability: Currently, work is in progress to handle
multiple kprobes in parallel.
c. Kernel return probes for sparc64.
d. Support for other architectures.
e. User-space probes.
8. Kprobes Example
Here's a sample kernel module showing the use of kprobes to dump a
stack trace and selected i386 registers when do_fork() is called.
----- cut here -----
/*kprobe_example.c*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/kprobes.h>
#include <linux/kallsyms.h>
#include <linux/sched.h>
/*For each probe you need to allocate a kprobe structure*/
static struct kprobe kp;
/*kprobe pre_handler: called just before the probed instruction is executed*/
int handler_pre(struct kprobe *p, struct pt_regs *regs)
{
printk("pre_handler: p->addr=0x%p, eip=%lx, eflags=0x%lx\n",
p->addr, regs->eip, regs->eflags);
dump_stack();
return 0;
}
/*kprobe post_handler: called after the probed instruction is executed*/
void handler_post(struct kprobe *p, struct pt_regs *regs, unsigned long flags)
{
printk("post_handler: p->addr=0x%p, eflags=0x%lx\n",
p->addr, regs->eflags);
}
/* fault_handler: this is called if an exception is generated for any
* instruction within the pre- or post-handler, or when Kprobes
* single-steps the probed instruction.
*/
int handler_fault(struct kprobe *p, struct pt_regs *regs, int trapnr)
{
printk("fault_handler: p->addr=0x%p, trap #%dn",
p->addr, trapnr);
/* Return 0 because we don't handle the fault. */
return 0;
}
int init_module(void)
{
int ret;
kp.pre_handler = handler_pre;
kp.post_handler = handler_post;
kp.fault_handler = handler_fault;
kp.addr = (kprobe_opcode_t*) kallsyms_lookup_name("do_fork");
/* register the kprobe now */
if (!kp.addr) {
printk("Couldn't find %s to plant kprobe\n", "do_fork");
return -1;
}
if ((ret = register_kprobe(&kp) < 0)) {
printk("register_kprobe failed, returned %d\n", ret);
return -1;
}
printk("kprobe registered\n");
return 0;
}
void cleanup_module(void)
{
unregister_kprobe(&kp);
printk("kprobe unregistered\n");
}
MODULE_LICENSE("GPL");
----- cut here -----
You can build the kernel module, kprobe-example.ko, using the following
Makefile:
----- cut here -----
obj-m := kprobe-example.o
KDIR := /lib/modules/$(shell uname -r)/build
PWD := $(shell pwd)
default:
$(MAKE) -C $(KDIR) SUBDIRS=$(PWD) modules
clean:
rm -f *.mod.c *.ko *.o
----- cut here -----
$ make
$ su -
...
# insmod kprobe-example.ko
You will see the trace data in /var/log/messages and on the console
whenever do_fork() is invoked to create a new process.
9. Jprobes Example
Here's a sample kernel module showing the use of jprobes to dump
the arguments of do_fork().
----- cut here -----
/*jprobe-example.c */
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/fs.h>
#include <linux/uio.h>
#include <linux/kprobes.h>
#include <linux/kallsyms.h>
/*
* Jumper probe for do_fork.
* Mirror principle enables access to arguments of the probed routine
* from the probe handler.
*/
/* Proxy routine having the same arguments as actual do_fork() routine */
long jdo_fork(unsigned long clone_flags, unsigned long stack_start,
struct pt_regs *regs, unsigned long stack_size,
int __user * parent_tidptr, int __user * child_tidptr)
{
printk("jprobe: clone_flags=0x%lx, stack_size=0x%lx, regs=0x%p\n",
clone_flags, stack_size, regs);
/* Always end with a call to jprobe_return(). */
jprobe_return();
/*NOTREACHED*/
return 0;
}
static struct jprobe my_jprobe = {
.entry = (kprobe_opcode_t *) jdo_fork
};
int init_module(void)
{
int ret;
my_jprobe.kp.addr = (kprobe_opcode_t *) kallsyms_lookup_name("do_fork");
if (!my_jprobe.kp.addr) {
printk("Couldn't find %s to plant jprobe\n", "do_fork");
return -1;
}
if ((ret = register_jprobe(&my_jprobe)) <0) {
printk("register_jprobe failed, returned %d\n", ret);
return -1;
}
printk("Planted jprobe at %p, handler addr %p\n",
my_jprobe.kp.addr, my_jprobe.entry);
return 0;
}
void cleanup_module(void)
{
unregister_jprobe(&my_jprobe);
printk("jprobe unregistered\n");
}
MODULE_LICENSE("GPL");
----- cut here -----
Build and insert the kernel module as shown in the above kprobe
example. You will see the trace data in /var/log/messages and on
the console whenever do_fork() is invoked to create a new process.
(Some messages may be suppressed if syslogd is configured to
eliminate duplicate messages.)
10. Kretprobes Example
Here's a sample kernel module showing the use of return probes to
report failed calls to sys_open().
----- cut here -----
/*kretprobe-example.c*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/kprobes.h>
#include <linux/kallsyms.h>
static const char *probed_func = "sys_open";
/* Return-probe handler: If the probed function fails, log the return value. */
static int ret_handler(struct kretprobe_instance *ri, struct pt_regs *regs)
{
// Substitute the appropriate register name for your architecture --
// e.g., regs->rax for x86_64, regs->gpr[3] for ppc64.
int retval = (int) regs->eax;
if (retval < 0) {
printk("%s returns %d\n", probed_func, retval);
}
return 0;
}
static struct kretprobe my_kretprobe = {
.handler = ret_handler,
/* Probe up to 20 instances concurrently. */
.maxactive = 20
};
int init_module(void)
{
int ret;
my_kretprobe.kp.addr =
(kprobe_opcode_t *) kallsyms_lookup_name(probed_func);
if (!my_kretprobe.kp.addr) {
printk("Couldn't find %s to plant return probe\n", probed_func);
return -1;
}
if ((ret = register_kretprobe(&my_kretprobe)) < 0) {
printk("register_kretprobe failed, returned %d\n", ret);
return -1;
}
printk("Planted return probe at %p\n", my_kretprobe.kp.addr);
return 0;
}
void cleanup_module(void)
{
unregister_kretprobe(&my_kretprobe);
printk("kretprobe unregistered\n");
/* nmissed > 0 suggests that maxactive was set too low. */
printk("Missed probing %d instances of %s\n",
my_kretprobe.nmissed, probed_func);
}
MODULE_LICENSE("GPL");
----- cut here -----
Build and insert the kernel module as shown in the above kprobe
example. You will see the trace data in /var/log/messages and on the
console whenever sys_open() returns a negative value. (Some messages
may be suppressed if syslogd is configured to eliminate duplicate
messages.)
For additional information on Kprobes, refer to the following URLs:
http://www-106.ibm.com/developerworks/library/l-kprobes.html?ca=dgr-lnxw42Kprobe
http://www.redhat.com/magazine/005mar05/features/kprobes/

4
Documentation/networking/00-INDEX
View File

@ -114,9 +114,7 @@ tuntap.txt
vortex.txt
- info on using 3Com Vortex (3c590, 3c592, 3c595, 3c597) Ethernet cards.
wan-router.txt
- Wan router documentation
wanpipe.txt
- WANPIPE(tm) Multiprotocol WAN Driver for Linux WAN Router
- WAN router documentation
wavelan.txt
- AT&T GIS (nee NCR) WaveLAN card: An Ethernet-like radio transceiver
x25.txt

998
Documentation/networking/bonding.txt
View File

File diff suppressed because it is too large Load Diff

94
Documentation/networking/dmfe.txt
View File

@ -1,59 +1,65 @@
dmfe.c: Version 1.28 01/18/2000
Davicom DM9102(A)/DM9132/DM9801 fast ethernet driver for Linux.
A Davicom DM9102(A)/DM9132/DM9801 fast ethernet driver for Linux.
Copyright (C) 1997 Sten Wang
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
A. Compiler command:
This driver provides kernel support for Davicom DM9102(A)/DM9132/DM9801 ethernet cards ( CNET
10/100 ethernet cards uses Davicom chipset too, so this driver supports CNET cards too ).If you
didn't compile this driver as a module, it will automatically load itself on boot and print a
line similar to :
A-1: For normal single or multiple processor kernel
"gcc -DMODULE -D__KERNEL__ -I/usr/src/linux/net/inet -Wall
-Wstrict-prototypes -O6 -c dmfe.c"
dmfe: Davicom DM9xxx net driver, version 1.36.4 (2002-01-17)
A-2: For single or multiple processor with kernel module version function
"gcc -DMODULE -DMODVERSIONS -D__KERNEL__ -I/usr/src/linux/net/inet
-Wall -Wstrict-prototypes -O6 -c dmfe.c"
If you compiled this driver as a module, you have to load it on boot.You can load it with command :
insmod dmfe
This way it will autodetect the device mode.This is the suggested way to load the module.Or you can pass
a mode= setting to module while loading, like :
insmod dmfe mode=0 # Force 10M Half Duplex
insmod dmfe mode=1 # Force 100M Half Duplex
insmod dmfe mode=4 # Force 10M Full Duplex
insmod dmfe mode=5 # Force 100M Full Duplex
Next you should configure your network interface with a command similar to :
ifconfig eth0 172.22.3.18
^^^^^^^^^^^
Your IP Adress
Then you may have to modify the default routing table with command :
route add default eth0
B. The following steps teach you how to activate a DM9102 board:
1. Used the upper compiler command to compile dmfe.c
2. Insert dmfe module into kernel
"insmod dmfe" ;;Auto Detection Mode (Suggest)
"insmod dmfe mode=0" ;;Force 10M Half Duplex
"insmod dmfe mode=1" ;;Force 100M Half Duplex
"insmod dmfe mode=4" ;;Force 10M Full Duplex
"insmod dmfe mode=5" ;;Force 100M Full Duplex
3. Config a dm9102 network interface
"ifconfig eth0 172.22.3.18"
^^^^^^^^^^^ Your IP address
4. Activate the IP routing table. For some distributions, it is not
necessary. You can type "route" to check.
"route add default eth0"
Now your ethernet card should be up and running.
5. Well done. Your DM9102 adapter is now activated.
TODO:
Implement pci_driver::suspend() and pci_driver::resume() power management methods.
Check on 64 bit boxes.
Check and fix on big endian boxes.
Test and make sure PCI latency is now correct for all cases.
C. Object files description:
1. dmfe_rh61.o: For Redhat 6.1
Authors:
If you can make sure your kernel version, you can rename
to dmfe.o and directly use it without re-compiling.
Sten Wang <sten_wang@davicom.com.tw > : Original Author
Tobias Ringstrom <tori@unhappy.mine.nu> : Current Maintainer
Contributors:
Author: Sten Wang, 886-3-5798797-8517, E-mail: sten_wang@davicom.com.tw
Marcelo Tosatti <marcelo@conectiva.com.br>
Alan Cox <alan@redhat.com>
Jeff Garzik <jgarzik@pobox.com>
Vojtech Pavlik <vojtech@suse.cz>

145
Documentation/networking/fib_trie.txt Normal file
View File

@ -0,0 +1,145 @@
LC-trie implementation notes.
Node types
----------
leaf
An end node with data. This has a copy of the relevant key, along
with 'hlist' with routing table entries sorted by prefix length.
See struct leaf and struct leaf_info.
trie node or tnode
An internal node, holding an array of child (leaf or tnode) pointers,
indexed through a subset of the key. See Level Compression.
A few concepts explained
------------------------
Bits (tnode)
The number of bits in the key segment used for indexing into the
child array - the "child index". See Level Compression.
Pos (tnode)
The position (in the key) of the key segment used for indexing into
the child array. See Path Compression.
Path Compression / skipped bits
Any given tnode is linked to from the child array of its parent, using
a segment of the key specified by the parent's "pos" and "bits"
In certain cases, this tnode's own "pos" will not be immediately
adjacent to the parent (pos+bits), but there will be some bits
in the key skipped over because they represent a single path with no
deviations. These "skipped bits" constitute Path Compression.
Note that the search algorithm will simply skip over these bits when
searching, making it necessary to save the keys in the leaves to
verify that they actually do match the key we are searching for.
Level Compression / child arrays
the trie is kept level balanced moving, under certain conditions, the
children of a full child (see "full_children") up one level, so that
instead of a pure binary tree, each internal node ("tnode") may
contain an arbitrarily large array of links to several children.
Conversely, a tnode with a mostly empty child array (see empty_children)
may be "halved", having some of its children moved downwards one level,
in order to avoid ever-increasing child arrays.
empty_children
the number of positions in the child array of a given tnode that are
NULL.
full_children
the number of children of a given tnode that aren't path compressed.
(in other words, they aren't NULL or leaves and their "pos" is equal
to this tnode's "pos"+"bits").
(The word "full" here is used more in the sense of "complete" than
as the opposite of "empty", which might be a tad confusing.)
Comments
---------
We have tried to keep the structure of the code as close to fib_hash as
possible to allow verification and help up reviewing.
fib_find_node()
A good start for understanding this code. This function implements a
straightforward trie lookup.
fib_insert_node()
Inserts a new leaf node in the trie. This is bit more complicated than
fib_find_node(). Inserting a new node means we might have to run the
level compression algorithm on part of the trie.
trie_leaf_remove()
Looks up a key, deletes it and runs the level compression algorithm.
trie_rebalance()
The key function for the dynamic trie after any change in the trie
it is run to optimize and reorganize. Tt will walk the trie upwards
towards the root from a given tnode, doing a resize() at each step
to implement level compression.
resize()
Analyzes a tnode and optimizes the child array size by either inflating
or shrinking it repeatedly until it fullfills the criteria for optimal
level compression. This part follows the original paper pretty closely
and there may be some room for experimentation here.
inflate()
Doubles the size of the child array within a tnode. Used by resize().
halve()
Halves the size of the child array within a tnode - the inverse of
inflate(). Used by resize();
fn_trie_insert(), fn_trie_delete(), fn_trie_select_default()
The route manipulation functions. Should conform pretty closely to the
corresponding functions in fib_hash.
fn_trie_flush()
This walks the full trie (using nextleaf()) and searches for empty
leaves which have to be removed.
fn_trie_dump()
Dumps the routing table ordered by prefix length. This is somewhat
slower than the corresponding fib_hash function, as we have to walk the
entire trie for each prefix length. In comparison, fib_hash is organized
as one "zone"/hash per prefix length.
Locking
-------
fib_lock is used for an RW-lock in the same way that this is done in fib_hash.
However, the functions are somewhat separated for other possible locking
scenarios. It might conceivably be possible to run trie_rebalance via RCU
to avoid read_lock in the fn_trie_lookup() function.
Main lookup mechanism
---------------------
fn_trie_lookup() is the main lookup function.
The lookup is in its simplest form just like fib_find_node(). We descend the
trie, key segment by key segment, until we find a leaf. check_leaf() does
the fib_semantic_match in the leaf's sorted prefix hlist.
If we find a match, we are done.
If we don't find a match, we enter prefix matching mode. The prefix length,
starting out at the same as the key length, is reduced one step at a time,
and we backtrack upwards through the trie trying to find a longest matching
prefix. The goal is always to reach a leaf and get a positive result from the
fib_semantic_match mechanism.
Inside each tnode, the search for longest matching prefix consists of searching
through the child array, chopping off (zeroing) the least significant "1" of
the child index until we find a match or the child index consists of nothing but
zeros.
At this point we backtrack (t->stats.backtrack++) up the trie, continuing to
chop off part of the key in order to find the longest matching prefix.
At this point we will repeatedly descend subtries to look for a match, and there
are some optimizations available that can provide us with "shortcuts" to avoid
descending into dead ends. Look for "HL_OPTIMIZE" sections in the code.
To alleviate any doubts about the correctness of the route selection process,
a new netlink operation has been added. Look for NETLINK_FIB_LOOKUP, which
gives userland access to fib_lookup().

56
Documentation/networking/ip-sysctl.txt
View File

@ -304,57 +304,6 @@ tcp_low_latency - BOOLEAN
changed would be a Beowulf compute cluster.
Default: 0
tcp_westwood - BOOLEAN
Enable TCP Westwood+ congestion control algorithm.
TCP Westwood+ is a sender-side only modification of the TCP Reno
protocol stack that optimizes the performance of TCP congestion
control. It is based on end-to-end bandwidth estimation to set
congestion window and slow start threshold after a congestion
episode. Using this estimation, TCP Westwood+ adaptively sets a
slow start threshold and a congestion window which takes into
account the bandwidth used at the time congestion is experienced.
TCP Westwood+ significantly increases fairness wrt TCP Reno in
wired networks and throughput over wireless links.
Default: 0
tcp_vegas_cong_avoid - BOOLEAN
Enable TCP Vegas congestion avoidance algorithm.
TCP Vegas is a sender-side only change to TCP that anticipates
the onset of congestion by estimating the bandwidth. TCP Vegas
adjusts the sending rate by modifying the congestion
window. TCP Vegas should provide less packet loss, but it is
not as aggressive as TCP Reno.
Default:0
tcp_bic - BOOLEAN
Enable BIC TCP congestion control algorithm.
BIC-TCP is a sender-side only change that ensures a linear RTT
fairness under large windows while offering both scalability and
bounded TCP-friendliness. The protocol combines two schemes
called additive increase and binary search increase. When the
congestion window is large, additive increase with a large
increment ensures linear RTT fairness as well as good
scalability. Under small congestion windows, binary search
increase provides TCP friendliness.
Default: 0
tcp_bic_low_window - INTEGER
Sets the threshold window (in packets) where BIC TCP starts to
adjust the congestion window. Below this threshold BIC TCP behaves
the same as the default TCP Reno.
Default: 14
tcp_bic_fast_convergence - BOOLEAN
Forces BIC TCP to more quickly respond to changes in congestion
window. Allows two flows sharing the same connection to converge
more rapidly.
Default: 1
tcp_default_win_scale - INTEGER
Sets the minimum window scale TCP will negotiate for on all
conections.
Default: 7
tcp_tso_win_divisor - INTEGER
This allows control over what percentage of the congestion window
can be consumed by a single TSO frame.
@ -368,6 +317,11 @@ tcp_frto - BOOLEAN
where packet loss is typically due to random radio interference
rather than intermediate router congestion.
tcp_congestion_control - STRING
Set the congestion control algorithm to be used for new
connections. The algorithm "reno" is always available, but
additional choices may be available based on kernel configuration.
somaxconn - INTEGER
Limit of socket listen() backlog, known in userspace as SOMAXCONN.
Defaults to 128. See also tcp_max_syn_backlog for additional tuning

69
Documentation/networking/tcp.txt
View File

@ -1,5 +1,72 @@
How the new TCP output machine [nyi] works.
TCP protocol
============
Last updated: 21 June 2005
Contents
========
- Congestion control
- How the new TCP output machine [nyi] works
Congestion control
==================
The following variables are used in the tcp_sock for congestion control:
snd_cwnd The size of the congestion window
snd_ssthresh Slow start threshold. We are in slow start if
snd_cwnd is less than this.
snd_cwnd_cnt A counter used to slow down the rate of increase
once we exceed slow start threshold.
snd_cwnd_clamp This is the maximum size that snd_cwnd can grow to.
snd_cwnd_stamp Timestamp for when congestion window last validated.
snd_cwnd_used Used as a highwater mark for how much of the
congestion window is in use. It is used to adjust
snd_cwnd down when the link is limited by the
application rather than the network.
As of 2.6.13, Linux supports pluggable congestion control algorithms.
A congestion control mechanism can be registered through functions in
tcp_cong.c. The functions used by the congestion control mechanism are
registered via passing a tcp_congestion_ops struct to
tcp_register_congestion_control. As a minimum name, ssthresh,
cong_avoid, min_cwnd must be valid.
Private data for a congestion control mechanism is stored in tp->ca_priv.
tcp_ca(tp) returns a pointer to this space. This is preallocated space - it
is important to check the size of your private data will fit this space, or
alternatively space could be allocated elsewhere and a pointer to it could
be stored here.
There are three kinds of congestion control algorithms currently: The
simplest ones are derived from TCP reno (highspeed, scalable) and just
provide an alternative the congestion window calculation. More complex
ones like BIC try to look at other events to provide better
heuristics. There are also round trip time based algorithms like
Vegas and Westwood+.
Good TCP congestion control is a complex problem because the algorithm
needs to maintain fairness and performance. Please review current
research and RFC's before developing new modules.
The method that is used to determine which congestion control mechanism is
determined by the setting of the sysctl net.ipv4.tcp_congestion_control.
The default congestion control will be the last one registered (LIFO);
so if you built everything as modules. the default will be reno. If you
build with the default's from Kconfig, then BIC will be builtin (not a module)
and it will end up the default.
If you really want a particular default value then you will need
to set it with the sysctl. If you use a sysctl, the module will be autoloaded
if needed and you will get the expected protocol. If you ask for an
unknown congestion method, then the sysctl attempt will fail.
If you remove a tcp congestion control module, then you will get the next
available one. Since reno can not be built as a module, and can not be
deleted, it will always be available.
How the new TCP output machine [nyi] works.
===========================================
Data is kept on a single queue. The skb->users flag tells us if the frame is
one that has been queued already. To add a frame we throw it on the end. Ack

622
Documentation/networking/wanpipe.txt
View File

@ -1,622 +0,0 @@
------------------------------------------------------------------------------
Linux WAN Router Utilities Package
------------------------------------------------------------------------------
Version 2.2.1
Mar 28, 2001
Author: Nenad Corbic <ncorbic@sangoma.com>
Copyright (c) 1995-2001 Sangoma Technologies Inc.
------------------------------------------------------------------------------
INTRODUCTION
Wide Area Networks (WANs) are used to interconnect Local Area Networks (LANs)
and/or stand-alone hosts over vast distances with data transfer rates
significantly higher than those achievable with commonly used dial-up
connections.
Usually an external device called `WAN router' sitting on your local network
or connected to your machine's serial port provides physical connection to
WAN. Although router's job may be as simple as taking your local network
traffic, converting it to WAN format and piping it through the WAN link, these
devices are notoriously expensive, with prices as much as 2 - 5 times higher
then the price of a typical PC box.
Alternatively, considering robustness and multitasking capabilities of Linux,
an internal router can be built (most routers use some sort of stripped down
Unix-like operating system anyway). With a number of relatively inexpensive WAN
interface cards available on the market, a perfectly usable router can be
built for less than half a price of an external router. Yet a Linux box
acting as a router can still be used for other purposes, such as fire-walling,
running FTP, WWW or DNS server, etc.
This kernel module introduces the notion of a WAN Link Driver (WLD) to Linux
operating system and provides generic hardware-independent services for such
drivers. Why can existing Linux network device interface not be used for
this purpose? Well, it can. However, there are a few key differences between
a typical network interface (e.g. Ethernet) and a WAN link.
Many WAN protocols, such as X.25 and frame relay, allow for multiple logical
connections (known as `virtual circuits' in X.25 terminology) over a single
physical link. Each such virtual circuit may (and almost always does) lead
to a different geographical location and, therefore, different network. As a
result, it is the virtual circuit, not the physical link, that represents a
route and, therefore, a network interface in Linux terms.
To further complicate things, virtual circuits are usually volatile in nature
(excluding so called `permanent' virtual circuits or PVCs). With almost no
time required to set up and tear down a virtual circuit, it is highly desirable
to implement on-demand connections in order to minimize network charges. So
unlike a typical network driver, the WAN driver must be able to handle multiple
network interfaces and cope as multiple virtual circuits come into existence
and go away dynamically.
Last, but not least, WAN configuration is much more complex than that of say
Ethernet and may well amount to several dozens of parameters. Some of them
are "link-wide" while others are virtual circuit-specific. The same holds
true for WAN statistics which is by far more extensive and extremely useful
when troubleshooting WAN connections. Extending the ifconfig utility to suit
these needs may be possible, but does not seem quite reasonable. Therefore, a
WAN configuration utility and corresponding application programmer's interface
is needed for this purpose.
Most of these problems are taken care of by this module. Its goal is to
provide a user with more-or-less standard look and feel for all WAN devices and
assist a WAN device driver writer by providing common services, such as:
o User-level interface via /proc file system
o Centralized configuration
o Device management (setup, shutdown, etc.)
o Network interface management (dynamic creation/destruction)
o Protocol encapsulation/decapsulation
To ba able to use the Linux WAN Router you will also need a WAN Tools package
available from
ftp.sangoma.com/pub/linux/current_wanpipe/wanpipe-X.Y.Z.tgz
where vX.Y.Z represent the wanpipe version number.
For technical questions and/or comments please e-mail to ncorbic@sangoma.com.
For general inquiries please contact Sangoma Technologies Inc. by
Hotline: 1-800-388-2475 (USA and Canada, toll free)
Phone: (905) 474-1990 ext: 106
Fax: (905) 474-9223
E-mail: dm@sangoma.com (David Mandelstam)
WWW: http://www.sangoma.com
INSTALLATION
Please read the WanpipeForLinux.pdf manual on how to
install the WANPIPE tools and drivers properly.
After installing wanpipe package: /usr/local/wanrouter/doc.
On the ftp.sangoma.com : /linux/current_wanpipe/doc
COPYRIGHT AND LICENSING INFORMATION
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; either version 2, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc., 675 Mass
Ave, Cambridge, MA 02139, USA.
ACKNOWLEDGEMENTS
This product is based on the WANPIPE(tm) Multiprotocol WAN Router developed
by Sangoma Technologies Inc. for Linux 2.0.x and 2.2.x. Success of the WANPIPE
together with the next major release of Linux kernel in summer 1996 commanded
adequate changes to the WANPIPE code to take full advantage of new Linux
features.
Instead of continuing developing proprietary interface tied to Sangoma WAN
cards, we decided to separate all hardware-independent code into a separate
module and defined two levels of interfaces - one for user-level applications
and another for kernel-level WAN drivers. WANPIPE is now implemented as a
WAN driver compliant with the WAN Link Driver interface. Also a general
purpose WAN configuration utility and a set of shell scripts was developed to
support WAN router at the user level.
Many useful ideas concerning hardware-independent interface implementation
were given by Mike McLagan <mike.mclagan@linux.org> and his implementation
of the Frame Relay router and drivers for Sangoma cards (dlci/sdla).
With the new implementation of the APIs being incorporated into the WANPIPE,
a special thank goes to Alan Cox in providing insight into BSD sockets.
Special thanks to all the WANPIPE users who performed field-testing, reported
bugs and made valuable comments and suggestions that help us to improve this
product.
NEW IN THIS RELEASE
o Updated the WANCFG utility
Calls the pppconfig to configure the PPPD
for async connections.
o Added the PPPCONFIG utility
Used to configure the PPPD dameon for the
WANPIPE Async PPP and standard serial port.
The wancfg calls the pppconfig to configure
the pppd.
o Fixed the PCI autodetect feature.
The SLOT 0 was used as an autodetect option
however, some high end PC's slot numbers start
from 0.
o This release has been tested with the new backupd
daemon release.
PRODUCT COMPONENTS AND RELATED FILES
/etc: (or user defined)
wanpipe1.conf default router configuration file
/lib/modules/X.Y.Z/misc:
wanrouter.o router kernel loadable module
af_wanpipe.o wanpipe api socket module
/lib/modules/X.Y.Z/net:
sdladrv.o Sangoma SDLA support module
wanpipe.o Sangoma WANPIPE(tm) driver module
/proc/net/wanrouter
Config reads current router configuration
Status reads current router status
{name} reads WAN driver statistics
/usr/sbin:
wanrouter wanrouter start-up script
wanconfig wanrouter configuration utility
sdladump WANPIPE adapter memory dump utility
fpipemon Monitor for Frame Relay
cpipemon Monitor for Cisco HDLC
ppipemon Monitor for PPP
xpipemon Monitor for X25
wpkbdmon WANPIPE keyboard led monitor/debugger
/usr/local/wanrouter:
README this file
COPYING GNU General Public License
Setup installation script
Filelist distribution definition file
wanrouter.rc meta-configuration file
(used by the Setup and wanrouter script)
/usr/local/wanrouter/doc:
wanpipeForLinux.pdf WAN Router User's Manual
/usr/local/wanrouter/patches:
wanrouter-v2213.gz patch for Linux kernels 2.2.11 up to 2.2.13.
wanrouter-v2214.gz patch for Linux kernel 2.2.14.
wanrouter-v2215.gz patch for Linux kernels 2.2.15 to 2.2.17.
wanrouter-v2218.gz patch for Linux kernels 2.2.18 and up.
wanrouter-v240.gz patch for Linux kernel 2.4.0.
wanrouter-v242.gz patch for Linux kernel 2.4.2 and up.
wanrouter-v2034.gz patch for Linux kernel 2.0.34
wanrouter-v2036.gz patch for Linux kernel 2.0.36 and up.
/usr/local/wanrouter/patches/kdrivers:
Sources of the latest WANPIPE device drivers.
These are used to UPGRADE the linux kernel to the newest
version if the kernel source has already been pathced with
WANPIPE drivers.
/usr/local/wanrouter/samples:
interface sample interface configuration file
wanpipe1.cpri CHDLC primary port
wanpipe2.csec CHDLC secondary port
wanpipe1.fr Frame Relay protocol
wanpipe1.ppp PPP protocol )
wanpipe1.asy CHDLC ASYNC protocol
wanpipe1.x25 X25 protocol
wanpipe1.stty Sync TTY driver (Used by Kernel PPPD daemon)
wanpipe1.atty Async TTY driver (Used by Kernel PPPD daemon)
wanrouter.rc sample meta-configuration file
/usr/local/wanrouter/util:
* wan-tools utilities source code
/usr/local/wanrouter/api/x25:
* x25 api sample programs.
/usr/local/wanrouter/api/chdlc:
* chdlc api sample programs.
/usr/local/wanrouter/api/fr:
* fr api sample programs.
/usr/local/wanrouter/config/wancfg:
wancfg WANPIPE GUI configuration program.
Creates wanpipe#.conf files.
/usr/local/wanrouter/config/cfgft1:
cfgft1 GUI CSU/DSU configuration program.
/usr/include/linux:
wanrouter.h router API definitions
wanpipe.h WANPIPE API definitions
sdladrv.h SDLA support module API definitions
sdlasfm.h SDLA firmware module definitions
if_wanpipe.h WANPIPE Socket definitions
if_wanpipe_common.h WANPIPE Socket/Driver common definitions.
sdlapci.h WANPIPE PCI definitions
/usr/src/linux/net/wanrouter:
* wanrouter source code
/var/log:
wanrouter wanrouter start-up log (created by the Setup script)
/var/lock: (or /var/lock/subsys for RedHat)
wanrouter wanrouter lock file (created by the Setup script)
/usr/local/wanrouter/firmware:
fr514.sfm Frame relay firmware for Sangoma S508/S514 card
cdual514.sfm Dual Port Cisco HDLC firmware for Sangoma S508/S514 card
ppp514.sfm PPP Firmware for Sangoma S508 and S514 cards
x25_508.sfm X25 Firmware for Sangoma S508 card.
REVISION HISTORY
1.0.0 December 31, 1996 Initial version
1.0.1 January 30, 1997 Status and statistics can be read via /proc
filesystem entries.
1.0.2 April 30, 1997 Added UDP management via monitors.
1.0.3 June 3, 1997 UDP management for multiple boards using Frame
Relay and PPP
Enabled continuous transmission of Configure
Request Packet for PPP (for 508 only)
Connection Timeout for PPP changed from 900 to 0
Flow Control Problem fixed for Frame Relay
1.0.4 July 10, 1997 S508/FT1 monitoring capability in fpipemon and
ppipemon utilities.
Configurable TTL for UDP packets.
Multicast and Broadcast IP source addresses are
silently discarded.
1.0.5 July 28, 1997 Configurable T391,T392,N391,N392,N393 for Frame
Relay in router.conf.
Configurable Memory Address through router.conf
for Frame Relay, PPP and X.25. (commenting this
out enables auto-detection).
Fixed freeing up received buffers using kfree()
for Frame Relay and X.25.
Protect sdla_peek() by calling save_flags(),
cli() and restore_flags().
Changed number of Trace elements from 32 to 20
Added DLCI specific data monitoring in FPIPEMON.
2.0.0 Nov 07, 1997 Implemented protection of RACE conditions by
critical flags for FRAME RELAY and PPP.
DLCI List interrupt mode implemented.
IPX support in FRAME RELAY and PPP.
IPX Server Support (MARS)
More driver specific stats included in FPIPEMON
and PIPEMON.
2.0.1 Nov 28, 1997 Bug Fixes for version 2.0.0.
Protection of "enable_irq()" while
"disable_irq()" has been enabled from any other
routine (for Frame Relay, PPP and X25).
Added additional Stats for Fpipemon and Ppipemon
Improved Load Sharing for multiple boards
2.0.2 Dec 09, 1997 Support for PAP and CHAP for ppp has been
implemented.
2.0.3 Aug 15, 1998 New release supporting Cisco HDLC, CIR for Frame
relay, Dynamic IP assignment for PPP and Inverse
Arp support for Frame-relay. Man Pages are
included for better support and a new utility
for configuring FT1 cards.
2.0.4 Dec 09, 1998 Dual Port support for Cisco HDLC.
Support for HDLC (LAPB) API.
Supports BiSync Streaming code for S502E
and S503 cards.
Support for Streaming HDLC API.
Provides a BSD socket interface for
creating applications using BiSync
streaming.
2.0.5 Aug 04, 1999 CHDLC initializatin bug fix.
PPP interrupt driven driver:
Fix to the PPP line hangup problem.
New PPP firmware
Added comments to the startup SYSTEM ERROR messages
Xpipemon debugging application for the X25 protocol
New USER_MANUAL.txt
Fixed the odd boundary 4byte writes to the board.
BiSync Streaming code has been taken out.
Available as a patch.
Streaming HDLC API has been taken out.
Available as a patch.
2.0.6 Aug 17, 1999 Increased debugging in statup scripts
Fixed insallation bugs from 2.0.5
Kernel patch works for both 2.2.10 and 2.2.11 kernels.
There is no functional difference between the two packages
2.0.7 Aug 26, 1999 o Merged X25API code into WANPIPE.
o Fixed a memeory leak for X25API
o Updated the X25API code for 2.2.X kernels.
o Improved NEM handling.
2.1.0 Oct 25, 1999 o New code for S514 PCI Card
o New CHDLC and Frame Relay drivers
o PPP and X25 are not supported in this release
2.1.1 Nov 30, 1999 o PPP support for S514 PCI Cards
2.1.3 Apr 06, 2000 o Socket based x25api
o Socket based chdlc api
o Socket based fr api
o Dual Port Receive only CHDLC support.
o Asynchronous CHDLC support (Secondary Port)
o cfgft1 GUI csu/dsu configurator
o wancfg GUI configuration file
configurator.
o Architectual directory changes.
beta-2.1.4 Jul 2000 o Dynamic interface configuration:
Network interfaces reflect the state
of protocol layer. If the protocol becomes
disconnected, driver will bring down
the interface. Once the protocol reconnects
the interface will be brought up.
Note: This option is turned off by default.
o Dynamic wanrouter setup using 'wanconfig':
wanconfig utility can be used to
shutdown,restart,start or reconfigure
a virtual circuit dynamically.
Frame Relay: Each DLCI can be:
created,stopped,restarted and reconfigured
dynamically using wanconfig.
ex: wanconfig card wanpipe1 dev wp1_fr16 up
o Wanrouter startup via command line arguments:
wanconfig also supports wanrouter startup via command line
arguments. Thus, there is no need to create a wanpipe#.conf
configuration file.
o Socket based x25api update/bug fixes.
Added support for LCN numbers greater than 255.
Option to pass up modem messages.
Provided a PCI IRQ check, so a single S514
card is guaranteed to have a non-sharing interrupt.
o Fixes to the wancfg utility.
o New FT1 debugging support via *pipemon utilities.
o Frame Relay ARP support Enabled.
beta3-2.1.4 Jul 2000 o X25 M_BIT Problem fix.
o Added the Multi-Port PPP
Updated utilites for the Multi-Port PPP.
2.1.4 Aut 2000
o In X25API:
Maximum packet an application can send
to the driver has been extended to 4096 bytes.
Fixed the x25 startup bug. Enable
communications only after all interfaces
come up. HIGH SVC/PVC is used to calculate
the number of channels.
Enable protocol only after all interfaces
are enabled.
o Added an extra state to the FT1 config, kernel module.
o Updated the pipemon debuggers.
o Blocked the Multi-Port PPP from running on kernels
2.2.16 or greater, due to syncppp kernel module
change.
beta1-2.1.5 Nov 15 2000
o Fixed the MulitPort PPP Support for kernels 2.2.16 and above.
2.2.X kernels only
o Secured the driver UDP debugging calls
- All illegal netowrk debugging calls are reported to
the log.
- Defined a set of allowed commands, all other denied.
o Cpipemon
- Added set FT1 commands to the cpipemon. Thus CSU/DSU
configuraiton can be performed using cpipemon.
All systems that cannot run cfgft1 GUI utility should
use cpipemon to configure the on board CSU/DSU.
o Keyboard Led Monitor/Debugger
- A new utilty /usr/sbin/wpkbdmon uses keyboard leds
to convey operatinal statistic information of the
Sangoma WANPIPE cards.
NUM_LOCK = Line State (On=connected, Off=disconnected)
CAPS_LOCK = Tx data (On=transmitting, Off=no tx data)
SCROLL_LOCK = Rx data (On=receiving, Off=no rx data
o Hardware probe on module load and dynamic device allocation
- During WANPIPE module load, all Sangoma cards are probed
and found information is printed in the /var/log/messages.
- If no cards are found, the module load fails.
- Appropriate number of devices are dynamically loaded
based on the number of Sangoma cards found.
Note: The kernel configuraiton option
CONFIG_WANPIPE_CARDS has been taken out.
o Fixed the Frame Relay and Chdlc network interfaces so they are
compatible with libpcap libraries. Meaning, tcpdump, snort,
ethereal, and all other packet sniffers and debuggers work on
all WANPIPE netowrk interfaces.
- Set the network interface encoding type to ARPHRD_PPP.
This tell the sniffers that data obtained from the
network interface is in pure IP format.
Fix for 2.2.X kernels only.
o True interface encoding option for Frame Relay and CHDLC
- The above fix sets the network interface encoding
type to ARPHRD_PPP, however some customers use
the encoding interface type to determine the
protocol running. Therefore, the TURE ENCODING
option will set the interface type back to the
original value.
NOTE: If this option is used with Frame Relay and CHDLC
libpcap library support will be broken.
i.e. tcpdump will not work.
Fix for 2.2.x Kernels only.
o Ethernet Bridgind over Frame Relay
- The Frame Relay bridging has been developed by
Kristian Hoffmann and Mark Wells.
- The Linux kernel bridge is used to send ethernet
data over the frame relay links.
For 2.2.X Kernels only.
o Added extensive 2.0.X support. Most new features of
2.1.5 for protocols Frame Relay, PPP and CHDLC are
supported under 2.0.X kernels.
beta1-2.2.0 Dec 30 2000
o Updated drivers for 2.4.X kernels.
o Updated drivers for SMP support.
o X25API is now able to share PCI interrupts.
o Took out a general polling routine that was used
only by X25API.
o Added appropriate locks to the dynamic reconfiguration
code.
o Fixed a bug in the keyboard debug monitor.
beta2-2.2.0 Jan 8 2001
o Patches for 2.4.0 kernel
o Patches for 2.2.18 kernel
o Minor updates to PPP and CHLDC drivers.
Note: No functinal difference.
beta3-2.2.9 Jan 10 2001
o I missed the 2.2.18 kernel patches in beta2-2.2.0
release. They are included in this release.
Stable Release
2.2.0 Feb 01 2001
o Bug fix in wancfg GUI configurator.
The edit function didn't work properly.
bata1-2.2.1 Feb 09 2001
o WANPIPE TTY Driver emulation.
Two modes of operation Sync and Async.
Sync: Using the PPPD daemon, kernel SyncPPP layer
and the Wanpipe sync TTY driver: a PPP protocol
connection can be established via Sangoma adapter, over
a T1 leased line.
The 2.4.0 kernel PPP layer supports MULTILINK
protocol, that can be used to bundle any number of Sangoma
adapters (T1 lines) into one, under a single IP address.
Thus, efficiently obtaining multiple T1 throughput.
NOTE: The remote side must also implement MULTILINK PPP
protocol.
Async:Using the PPPD daemon, kernel AsyncPPP layer
and the WANPIPE async TTY driver: a PPP protocol
connection can be established via Sangoma adapter and
a modem, over a telephone line.
Thus, the WANPIPE async TTY driver simulates a serial
TTY driver that would normally be used to interface the
MODEM to the linux kernel.
o WANPIPE PPP Backup Utility
This utility will monitor the state of the PPP T1 line.
In case of failure, a dial up connection will be established
via pppd daemon, ether via a serial tty driver (serial port),
or a WANPIPE async TTY driver (in case serial port is unavailable).
Furthermore, while in dial up mode, the primary PPP T1 link
will be monitored for signs of life.
If the PPP T1 link comes back to life, the dial up connection
will be shutdown and T1 line re-established.
o New Setup installation script.
Option to UPGRADE device drivers if the kernel source has
already been patched with WANPIPE.
Option to COMPILE WANPIPE modules against the currently
running kernel, thus no need for manual kernel and module
re-compilatin.
o Updates and Bug Fixes to wancfg utility.
bata2-2.2.1 Feb 20 2001
o Bug fixes to the CHDLC device drivers.
The driver had compilation problems under kernels
2.2.14 or lower.
o Bug fixes to the Setup installation script.
The device drivers compilation options didn't work
properly.
o Update to the wpbackupd daemon.
Optimized the cross-over times, between the primary
link and the backup dialup.
beta3-2.2.1 Mar 02 2001
o Patches for 2.4.2 kernel.
o Bug fixes to util/ make files.
o Bug fixes to the Setup installation script.
o Took out the backupd support and made it into
as separate package.
beta4-2.2.1 Mar 12 2001
o Fix to the Frame Relay Device driver.
IPSAC sends a packet of zero length
header to the frame relay driver. The
driver tries to push its own 2 byte header
into the packet, which causes the driver to
crash.
o Fix the WANPIPE re-configuration code.
Bug was found by trying to run the cfgft1 while the
interface was already running.
o Updates to cfgft1.
Writes a wanpipe#.cfgft1 configuration file
once the CSU/DSU is configured. This file can
holds the current CSU/DSU configuration.
>>>>>> END OF README <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

14
Documentation/pci.txt
View File

@ -266,20 +266,6 @@ port an old driver to the new PCI interface. They are no longer present
in the kernel as they aren't compatible with hotplug or PCI domains or
having sane locking.
pcibios_present() and Since ages, you don't need to test presence
pci_present() of PCI subsystem when trying to talk to it.
If it's not there, the list of PCI devices
is empty and all functions for searching for
devices just return NULL.
pcibios_(read|write)_* Superseded by their pci_(read|write)_*
counterparts.
pcibios_find_* Superseded by their pci_get_* counterparts.
pci_for_each_dev() Superseded by pci_get_device()
pci_for_each_dev_reverse() Superseded by pci_find_device_reverse()
pci_for_each_bus() Superseded by pci_find_next_bus()
pci_find_device() Superseded by pci_get_device()
pci_find_subsys() Superseded by pci_get_subsys()
pci_find_slot() Superseded by pci_get_slot()
pcibios_find_class() Superseded by pci_get_class()
pci_find_class() Superseded by pci_get_class()
pci_(read|write)_*_nodev() Superseded by pci_bus_(read|write)_*()

63
Documentation/pcmcia/devicetable.txt Normal file
View File

@ -0,0 +1,63 @@
Matching of PCMCIA devices to drivers is done using one or more of the
following criteria:
- manufactor ID
- card ID
- product ID strings _and_ hashes of these strings
- function ID
- device function (actual and pseudo)
You should use the helpers in include/pcmcia/device_id.h for generating the
struct pcmcia_device_id[] entries which match devices to drivers.
If you want to match product ID strings, you also need to pass the crc32
hashes of the string to the macro, e.g. if you want to match the product ID
string 1, you need to use
PCMCIA_DEVICE_PROD_ID1("some_string", 0x(hash_of_some_string)),
If the hash is incorrect, the kernel will inform you about this in "dmesg"
upon module initialization, and tell you of the correct hash.
You can determine the hash of the product ID strings by catting the file
"modalias" in the sysfs directory of the PCMCIA device. It generates a string
in the following form:
pcmcia:m0149cC1ABf06pfn00fn00pa725B842DpbF1EFEE84pc0877B627pd00000000
The hex value after "pa" is the hash of product ID string 1, after "pb" for
string 2 and so on.
Alternatively, you can use this small tool to determine the crc32 hash.
simply pass the string you want to evaluate as argument to this program,
e.g.
$ ./crc32hash "Dual Speed"
-------------------------------------------------------------------------
/* crc32hash.c - derived from linux/lib/crc32.c, GNU GPL v2 */
#include <string.h>
#include <stdio.h>
#include <ctype.h>
#include <stdlib.h>
unsigned int crc32(unsigned char const *p, unsigned int len)
{
int i;
unsigned int crc = 0;
while (len--) {
crc ^= *p++;
for (i = 0; i < 8; i++)
crc = (crc >> 1) ^ ((crc & 1) ? 0xedb88320 : 0);
}
return crc;
}
int main(int argc, char **argv) {
unsigned int result;
if (argc != 2) {
printf("no string passed as argument\n");
return -1;
}
result = crc32(argv[1], strlen(argv[1]));
printf("0x%x\n", result);
return 0;
}

67
Documentation/pcmcia/driver-changes.txt Normal file
View File

@ -0,0 +1,67 @@
This file details changes in 2.6 which affect PCMCIA card driver authors:
* event handler initialization in struct pcmcia_driver (as of 2.6.13)
The event handler is notified of all events, and must be initialized
as the event() callback in the driver's struct pcmcia_driver.
* pcmcia/version.h should not be used (as of 2.6.13)
This file will be removed eventually.
* in-kernel device<->driver matching (as of 2.6.13)
PCMCIA devices and their correct drivers can now be matched in
kernelspace. See 'devicetable.txt' for details.
* Device model integration (as of 2.6.11)
A struct pcmcia_device is registered with the device model core,
and can be used (e.g. for SET_NETDEV_DEV) by using
handle_to_dev(client_handle_t * handle).
* Convert internal I/O port addresses to unsigned long (as of 2.6.11)
ioaddr_t should be replaced by kio_addr_t in PCMCIA card drivers.
* irq_mask and irq_list parameters (as of 2.6.11)
The irq_mask and irq_list parameters should no longer be used in
PCMCIA card drivers. Instead, it is the job of the PCMCIA core to
determine which IRQ should be used. Therefore, link->irq.IRQInfo2
is ignored.
* client->PendingEvents is gone (as of 2.6.11)
client->PendingEvents is no longer available.
* client->Attributes are gone (as of 2.6.11)
client->Attributes is unused, therefore it is removed from all
PCMCIA card drivers
* core functions no longer available (as of 2.6.11)
The following functions have been removed from the kernel source
because they are unused by all in-kernel drivers, and no external
driver was reported to rely on them:
pcmcia_get_first_region()
pcmcia_get_next_region()
pcmcia_modify_window()
pcmcia_set_event_mask()
pcmcia_get_first_window()
pcmcia_get_next_window()
* device list iteration upon module removal (as of 2.6.10)
It is no longer necessary to iterate on the driver's internal
client list and call the ->detach() function upon module removal.
* Resource management. (as of 2.6.8)
Although the PCMCIA subsystem will allocate resources for cards,
it no longer marks these resources busy. This means that driver
authors are now responsible for claiming your resources as per
other drivers in Linux. You should use request_region() to mark
your IO regions in-use, and request_mem_region() to mark your
memory regions in-use. The name argument should be a pointer to
your driver name. Eg, for pcnet_cs, name should point to the
string "pcnet_cs".
* CardServices is gone
CardServices() in 2.4 is just a big switch statement to call various
services. In 2.6, all of those entry points are exported and called
directly (except for pcmcia_report_error(), just use cs_error() instead).
* struct pcmcia_driver
You need to use struct pcmcia_driver and pcmcia_{un,}register_driver
instead of {un,}register_pccard_driver

3
Documentation/power/kernel_threads.txt
View File

@ -12,8 +12,7 @@ refrigerator. Code to do this looks like this:
do {
hub_events();
wait_event_interruptible(khubd_wait, !list_empty(&hub_event_list));
if (current->flags & PF_FREEZE)
refrigerator(PF_FREEZE);
try_to_freeze();
} while (!signal_pending(current));
from drivers/usb/core/hub.c::hub_thread()

38
Documentation/power/pci.txt
View File

@ -291,6 +291,44 @@ a request to enable wake events from D3, two calls should be made to
pci_enable_wake (one for both D3hot and D3cold).
A reference implementation
-------------------------
.suspend()
{
/* driver specific operations */
/* Disable IRQ */
free_irq();
/* If using MSI */
pci_disable_msi();
pci_save_state();
pci_enable_wake();
/* Disable IO/bus master/irq router */
pci_disable_device();
pci_set_power_state(pci_choose_state());
}
.resume()
{
pci_set_power_state(PCI_D0);
pci_restore_state();
/* device's irq possibly is changed, driver should take care */
pci_enable_device();
pci_set_master();
/* if using MSI, device's vector possibly is changed */
pci_enable_msi();
request_irq();
/* driver specific operations; */
}
This is a typical implementation. Drivers can slightly change the order
of the operations in the implementation, ignore some operations or add
more deriver specific operations in it, but drivers should do something like
this on the whole.
5. Resources
~~~~~~~~~~~~

84
Documentation/power/swsusp.txt
View File

@ -164,11 +164,11 @@ place where the thread is safe to be frozen (no kernel semaphores
should be held at that point and it must be safe to sleep there), and
add:
if (current->flags & PF_FREEZE)
refrigerator(PF_FREEZE);
try_to_freeze();
If the thread is needed for writing the image to storage, you should
instead set the PF_NOFREEZE process flag when creating the thread.
instead set the PF_NOFREEZE process flag when creating the thread (and
be very carefull).
Q: What is the difference between between "platform", "shutdown" and
@ -233,3 +233,81 @@ A: Try running
cat `cat /proc/[0-9]*/maps | grep / | sed 's:.* /:/:' | sort -u` > /dev/null
after resume. swapoff -a; swapon -a may also be usefull.
Q: What happens to devices during swsusp? They seem to be resumed
during system suspend?
A: That's correct. We need to resume them if we want to write image to
disk. Whole sequence goes like
Suspend part
~~~~~~~~~~~~
running system, user asks for suspend-to-disk
user processes are stopped
suspend(PMSG_FREEZE): devices are frozen so that they don't interfere
with state snapshot
state snapshot: copy of whole used memory is taken with interrupts disabled
resume(): devices are woken up so that we can write image to swap
write image to swap
suspend(PMSG_SUSPEND): suspend devices so that we can power off
turn the power off
Resume part
~~~~~~~~~~~
(is actually pretty similar)
running system, user asks for suspend-to-disk
user processes are stopped (in common case there are none, but with resume-from-initrd, noone knows)
read image from disk
suspend(PMSG_FREEZE): devices are frozen so that they don't interfere
with image restoration
image restoration: rewrite memory with image
resume(): devices are woken up so that system can continue
thaw all user processes
Q: What is this 'Encrypt suspend image' for?
A: First of all: it is not a replacement for dm-crypt encrypted swap.
It cannot protect your computer while it is suspended. Instead it does
protect from leaking sensitive data after resume from suspend.
Think of the following: you suspend while an application is running
that keeps sensitive data in memory. The application itself prevents
the data from being swapped out. Suspend, however, must write these
data to swap to be able to resume later on. Without suspend encryption
your sensitive data are then stored in plaintext on disk. This means
that after resume your sensitive data are accessible to all
applications having direct access to the swap device which was used
for suspend. If you don't need swap after resume these data can remain
on disk virtually forever. Thus it can happen that your system gets
broken in weeks later and sensitive data which you thought were
encrypted and protected are retrieved and stolen from the swap device.
To prevent this situation you should use 'Encrypt suspend image'.
During suspend a temporary key is created and this key is used to
encrypt the data written to disk. When, during resume, the data was
read back into memory the temporary key is destroyed which simply
means that all data written to disk during suspend are then
inaccessible so they can't be stolen later on. The only thing that
you must then take care of is that you call 'mkswap' for the swap
partition used for suspend as early as possible during regular
boot. This asserts that any temporary key from an oopsed suspend or
from a failed or aborted resume is erased from the swap device.
As a rule of thumb use encrypted swap to protect your data while your
system is shut down or suspended. Additionally use the encrypted
suspend image to prevent sensitive data from being stolen after
resume.

4
Documentation/power/video.txt
View File

@ -83,8 +83,10 @@ Compaq Armada E500 - P3-700 none (1) (S1 also works OK)
Compaq Evo N620c vga=normal, s3_bios (2)
Dell 600m, ATI R250 Lf none (1), but needs xorg-x11-6.8.1.902-1
Dell D600, ATI RV250 vga=normal and X, or try vbestate (6)
Dell D610 vga=normal and X (possibly vbestate (6) too, but not tested)
Dell Inspiron 4000 ??? (*)
Dell Inspiron 500m ??? (*)
Dell Inspiron 510m ???
Dell Inspiron 600m ??? (*)
Dell Inspiron 8200 ??? (*)
Dell Inspiron 8500 ??? (*)
@ -115,6 +117,7 @@ IBM Thinkpad X40 Type 2371-7JG s3_bios,s3_mode (4)
Medion MD4220 ??? (*)
Samsung P35 vbetool needed (6)
Sharp PC-AR10 (ATI rage) none (1)
Sony Vaio PCG-C1VRX/K s3_bios (2)
Sony Vaio PCG-F403 ??? (*)
Sony Vaio PCG-N505SN ??? (*)
Sony Vaio vgn-s260 X or boot-radeon can init it (5)
@ -123,6 +126,7 @@ Toshiba Satellite 4030CDT s3_mode (3)
Toshiba Satellite 4080XCDT s3_mode (3)
Toshiba Satellite 4090XCDT ??? (*)
Toshiba Satellite P10-554 s3_bios,s3_mode (4)(****)
Toshiba M30 (2) xor X with nvidia driver using internal AGP
Uniwill 244IIO ??? (*)

19
Documentation/power/video_extension.txt
View File

@ -1,13 +1,16 @@
This driver implement the ACPI Extensions For Display Adapters
for integrated graphics devices on motherboard, as specified in
ACPI 2.0 Specification, Appendix B, allowing to perform some basic
control like defining the video POST device, retrieving EDID information
or to setup a video output, etc. Note that this is an ref. implementation only.
It may or may not work for your integrated video device.
ACPI video extensions
~~~~~~~~~~~~~~~~~~~~~
This driver implement the ACPI Extensions For Display Adapters for
integrated graphics devices on motherboard, as specified in ACPI 2.0
Specification, Appendix B, allowing to perform some basic control like
defining the video POST device, retrieving EDID information or to
setup a video output, etc. Note that this is an ref. implementation
only. It may or may not work for your integrated video device.
Interfaces exposed to userland through /proc/acpi/video:
VGA/info : display the supported video bus device capability like ,Video ROM, CRT/LCD/TV.
VGA/info : display the supported video bus device capability like Video ROM, CRT/LCD/TV.
VGA/ROM : Used to get a copy of the display devices' ROM data (up to 4k).
VGA/POST_info : Used to determine what options are implemented.
VGA/POST : Used to get/set POST device.
@ -15,7 +18,7 @@ VGA/DOS : Used to get/set ownership of output switching:
Please refer ACPI spec B.4.1 _DOS
VGA/CRT : CRT output
VGA/LCD : LCD output
VGA/TV : TV output
VGA/TVO : TV output
VGA/*/brightness : Used to get/set brightness of output device
Notify event through /proc/acpi/event:

79
Documentation/s390/s390dbf.txt
View File

@ -12,8 +12,8 @@ where log records can be stored efficiently in memory, where each component
One purpose of this is to inspect the debug logs after a production system crash
in order to analyze the reason for the crash.
If the system still runs but only a subcomponent which uses dbf failes,
it is possible to look at the debug logs on a live system via the Linux proc
filesystem.
it is possible to look at the debug logs on a live system via the Linux
debugfs filesystem.
The debug feature may also very useful for kernel and driver development.
Design:
@ -52,16 +52,18 @@ Each debug entry contains the following data:
- Flag, if entry is an exception or not
The debug logs can be inspected in a live system through entries in
the proc-filesystem. Under the path /proc/s390dbf there is
the debugfs-filesystem. Under the toplevel directory "s390dbf" there is
a directory for each registered component, which is named like the
corresponding component.
corresponding component. The debugfs normally should be mounted to
/sys/kernel/debug therefore the debug feature can be accessed unter
/sys/kernel/debug/s390dbf.
The content of the directories are files which represent different views
to the debug log. Each component can decide which views should be
used through registering them with the function debug_register_view().
Predefined views for hex/ascii, sprintf and raw binary data are provided.
It is also possible to define other views. The content of
a view can be inspected simply by reading the corresponding proc file.
a view can be inspected simply by reading the corresponding debugfs file.
All debug logs have an an actual debug level (range from 0 to 6).
The default level is 3. Event and Exception functions have a 'level'
@ -69,14 +71,14 @@ parameter. Only debug entries with a level that is lower or equal
than the actual level are written to the log. This means, when
writing events, high priority log entries should have a low level
value whereas low priority entries should have a high one.
The actual debug level can be changed with the help of the proc-filesystem
through writing a number string "x" to the 'level' proc file which is
The actual debug level can be changed with the help of the debugfs-filesystem
through writing a number string "x" to the 'level' debugfs file which is
provided for every debug log. Debugging can be switched off completely
by using "-" on the 'level' proc file.
by using "-" on the 'level' debugfs file.
Example:
> echo "-" > /proc/s390dbf/dasd/level
> echo "-" > /sys/kernel/debug/s390dbf/dasd/level
It is also possible to deactivate the debug feature globally for every
debug log. You can change the behavior using 2 sysctl parameters in
@ -99,11 +101,11 @@ Kernel Interfaces:
------------------
----------------------------------------------------------------------------
debug_info_t *debug_register(char *name, int pages_index, int nr_areas,
debug_info_t *debug_register(char *name, int pages, int nr_areas,
int buf_size);
Parameter: name: Name of debug log (e.g. used for proc entry)
pages_index: 2^pages_index pages will be allocated per area
Parameter: name: Name of debug log (e.g. used for debugfs entry)
pages: number of pages, which will be allocated per area
nr_areas: number of debug areas
buf_size: size of data area in each debug entry
@ -134,7 +136,7 @@ Return Value: none
Description: Sets new actual debug level if new_level is valid.
---------------------------------------------------------------------------
+void debug_stop_all(void);
void debug_stop_all(void);
Parameter: none
@ -270,7 +272,7 @@ Parameter: id: handle for debug log
Return Value: 0 : ok
< 0: Error
Description: registers new debug view and creates proc dir entry
Description: registers new debug view and creates debugfs dir entry
---------------------------------------------------------------------------
int debug_unregister_view (debug_info_t * id, struct debug_view *view);
@ -281,7 +283,7 @@ Parameter: id: handle for debug log
Return Value: 0 : ok
< 0: Error
Description: unregisters debug view and removes proc dir entry
Description: unregisters debug view and removes debugfs dir entry
@ -308,7 +310,7 @@ static int init(void)
{
/* register 4 debug areas with one page each and 4 byte data field */
debug_info = debug_register ("test", 0, 4, 4 );
debug_info = debug_register ("test", 1, 4, 4 );
debug_register_view(debug_info,&debug_hex_ascii_view);
debug_register_view(debug_info,&debug_raw_view);
@ -343,7 +345,7 @@ static int init(void)
/* register 4 debug areas with one page each and data field for */
/* format string pointer + 2 varargs (= 3 * sizeof(long)) */
debug_info = debug_register ("test", 0, 4, sizeof(long) * 3);
debug_info = debug_register ("test", 1, 4, sizeof(long) * 3);
debug_register_view(debug_info,&debug_sprintf_view);
debug_sprintf_event(debug_info, 2 , "first event in %s:%i\n",__FILE__,__LINE__);
@ -362,16 +364,16 @@ module_exit(cleanup);
ProcFS Interface
Debugfs Interface
----------------
Views to the debug logs can be investigated through reading the corresponding
proc-files:
debugfs-files:
Example:
> ls /proc/s390dbf/dasd
flush hex_ascii level raw
> cat /proc/s390dbf/dasd/hex_ascii | sort +1
> ls /sys/kernel/debug/s390dbf/dasd
flush hex_ascii level pages raw
> cat /sys/kernel/debug/s390dbf/dasd/hex_ascii | sort +1
00 00974733272:680099 2 - 02 0006ad7e 07 ea 4a 90 | ....
00 00974733272:682210 2 - 02 0006ade6 46 52 45 45 | FREE
00 00974733272:682213 2 - 02 0006adf6 07 ea 4a 90 | ....
@ -391,25 +393,36 @@ Changing the debug level
Example:
> cat /proc/s390dbf/dasd/level
> cat /sys/kernel/debug/s390dbf/dasd/level
3
> echo "5" > /proc/s390dbf/dasd/level
> cat /proc/s390dbf/dasd/level
> echo "5" > /sys/kernel/debug/s390dbf/dasd/level
> cat /sys/kernel/debug/s390dbf/dasd/level
5
Flushing debug areas
--------------------
Debug areas can be flushed with piping the number of the desired
area (0...n) to the proc file "flush". When using "-" all debug areas
area (0...n) to the debugfs file "flush". When using "-" all debug areas
are flushed.
Examples:
1. Flush debug area 0:
> echo "0" > /proc/s390dbf/dasd/flush
> echo "0" > /sys/kernel/debug/s390dbf/dasd/flush
2. Flush all debug areas:
> echo "-" > /proc/s390dbf/dasd/flush
> echo "-" > /sys/kernel/debug/s390dbf/dasd/flush
Changing the size of debug areas
------------------------------------
It is possible the change the size of debug areas through piping
the number of pages to the debugfs file "pages". The resize request will
also flush the debug areas.
Example:
Define 4 pages for the debug areas of debug feature "dasd":
> echo "4" > /sys/kernel/debug/s390dbf/dasd/pages
Stooping the debug feature
--------------------------
@ -491,7 +504,7 @@ Defining views
--------------
Views are specified with the 'debug_view' structure. There are defined
callback functions which are used for reading and writing the proc files:
callback functions which are used for reading and writing the debugfs files:
struct debug_view {
char name[DEBUG_MAX_PROCF_LEN];
@ -525,7 +538,7 @@ typedef int (debug_input_proc_t) (debug_info_t* id,
The "private_data" member can be used as pointer to view specific data.
It is not used by the debug feature itself.
The output when reading a debug-proc file is structured like this:
The output when reading a debugfs file is structured like this:
"prolog_proc output"
@ -534,13 +547,13 @@ The output when reading a debug-proc file is structured like this:
"header_proc output 3" "format_proc output 3"
...
When a view is read from the proc fs, the Debug Feature calls the
When a view is read from the debugfs, the Debug Feature calls the
'prolog_proc' once for writing the prolog.
Then 'header_proc' and 'format_proc' are called for each
existing debug entry.
The input_proc can be used to implement functionality when it is written to
the view (e.g. like with 'echo "0" > /proc/s390dbf/dasd/level).
the view (e.g. like with 'echo "0" > /sys/kernel/debug/s390dbf/dasd/level).
For header_proc there can be used the default function
debug_dflt_header_fn() which is defined in in debug.h.
@ -602,7 +615,7 @@ debug_info = debug_register ("test", 0, 4, 4 ));
debug_register_view(debug_info, &debug_test_view);
for(i = 0; i < 10; i ++) debug_int_event(debug_info, 1, i);
> cat /proc/s390dbf/test/myview
> cat /sys/kernel/debug/s390dbf/test/myview
00 00964419734:611402 1 - 00 88042ca This error...........
00 00964419734:611405 1 - 00 88042ca That error...........
00 00964419734:611408 1 - 00 88042ca Problem..............

15
Documentation/scsi/scsi_mid_low_api.txt
View File

@ -388,7 +388,6 @@ Summary:
scsi_remove_device - detach and remove a SCSI device
scsi_remove_host - detach and remove all SCSI devices owned by host
scsi_report_bus_reset - report scsi _bus_ reset observed
scsi_set_device - place device reference in host structure
scsi_track_queue_full - track successive QUEUE_FULL events
scsi_unblock_requests - allow further commands to be queued to given host
scsi_unregister - [calls scsi_host_put()]
@ -740,20 +739,6 @@ int scsi_remove_host(struct Scsi_Host *shost)
void scsi_report_bus_reset(struct Scsi_Host * shost, int channel)
/**
* scsi_set_device - place device reference in host structure
* @shost: a pointer to a scsi host instance
* @pdev: pointer to device instance to assign
*
* Returns nothing
*
* Might block: no
*
* Defined in: include/scsi/scsi_host.h .
**/
void scsi_set_device(struct Scsi_Host * shost, struct device * dev)
/**
* scsi_track_queue_full - track successive QUEUE_FULL events on given
* device to determine if and when there is a need

4
Documentation/serial/driver
View File

@ -107,8 +107,8 @@ hardware.
indicate that the signal is permanently active. If RI is
not available, the signal should not be indicated as active.
Locking: none.
Interrupts: caller dependent.
Locking: port->lock taken.
Interrupts: locally disabled.
This call must not sleep
stop_tx(port,tty_stop)

171
Documentation/sound/alsa/ALSA-Configuration.txt
View File

@ -615,9 +615,11 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
Module snd-hda-intel
--------------------
Module for Intel HD Audio (ICH6, ICH6M, ICH7)
Module for Intel HD Audio (ICH6, ICH6M, ICH7), ATI SB450,
VIA VT8251/VT8237A
model - force the model name
position_fix - Fix DMA pointer (0 = FIFO size, 1 = none, 2 = POSBUF)
Module supports up to 8 cards.
@ -634,7 +636,16 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
3stack-digout 3-jack in back, a HP out and a SPDIF out
5stack 5-jack in back, 2-jack in front
5stack-digout 5-jack in back, 2-jack in front, a SPDIF out
6stack 6-jack in back, 2-jack in front
6stack-digout 6-jack with a SPDIF out
w810 3-jack
z71v 3-jack (HP shared SPDIF)
asus 3-jack
uniwill 3-jack
F1734 2-jack
test for testing/debugging purpose, almost all controls can be
adjusted. Appearing only when compiled with
$CONFIG_SND_DEBUG=y
CMI9880
minimal 3-jack in back
@ -642,6 +653,15 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
full 6-jack in back, 2-jack in front
full_dig 6-jack in back, 2-jack in front, SPDIF I/O
allout 5-jack in back, 2-jack in front, SPDIF out
auto auto-config reading BIOS (default)
Note 2: If you get click noises on output, try the module option
position_fix=1 or 2. position_fix=1 will use the SD_LPIB
register value without FIFO size correction as the current
DMA pointer. position_fix=2 will make the driver to use
the position buffer instead of reading SD_LPIB register.
(Usually SD_LPLIB register is more accurate than the
position buffer.)
Module snd-hdsp
---------------
@ -660,7 +680,19 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
module did formerly. It will allocate the buffers in advance
when any HDSP cards are found. To make the buffer
allocation sure, load snd-page-alloc module in the early
stage of boot sequence.
stage of boot sequence. See "Early Buffer Allocation"
section.
Module snd-hdspm
----------------
Module for RME HDSP MADI board.
precise_ptr - Enable precise pointer, or disable.
line_outs_monitor - Send playback streams to analog outs by default.
enable_monitor - Enable Analog Out on Channel 63/64 by default.
See hdspm.txt for details.
Module snd-ice1712
------------------
@ -677,15 +709,19 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
* TerraTec EWS 88D
* TerraTec EWX 24/96
* TerraTec DMX 6Fire
* TerraTec Phase 88
* Hoontech SoundTrack DSP 24
* Hoontech SoundTrack DSP 24 Value
* Hoontech SoundTrack DSP 24 Media 7.1
* Event Electronics, EZ8
* Digigram VX442
* Lionstracs, Mediastaton
model - Use the given board model, one of the following:
delta1010, dio2496, delta66, delta44, audiophile, delta410,
delta1010lt, vx442, ewx2496, ews88mt, ews88mt_new, ews88d,
dmx6fire, dsp24, dsp24_value, dsp24_71, ez8
dmx6fire, dsp24, dsp24_value, dsp24_71, ez8,
phase88, mediastation
omni - Omni I/O support for MidiMan M-Audio Delta44/66
cs8427_timeout - reset timeout for the CS8427 chip (S/PDIF transciever)
in msec resolution, default value is 500 (0.5 sec)
@ -694,20 +730,46 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
is not used with all Envy24 based cards (for example in the MidiMan Delta
serie).
Note: The supported board is detected by reading EEPROM or PCI
SSID (if EEPROM isn't available). You can override the
model by passing "model" module option in case that the
driver isn't configured properly or you want to try another
type for testing.
Module snd-ice1724
------------------
Module for Envy24HT (VT/ICE1724) based PCI sound cards.
Module for Envy24HT (VT/ICE1724), Envy24PT (VT1720) based PCI sound cards.
* MidiMan M Audio Revolution 7.1
* AMP Ltd AUDIO2000
* TerraTec Aureon Sky-5.1, Space-7.1
* TerraTec Aureon 5.1 Sky
* TerraTec Aureon 7.1 Space
* TerraTec Aureon 7.1 Universe
* TerraTec Phase 22
* TerraTec Phase 28
* AudioTrak Prodigy 7.1
* AudioTrak Prodigy 192
* Pontis MS300
* Albatron K8X800 Pro II
* Chaintech ZNF3-150
* Chaintech ZNF3-250
* Chaintech 9CJS
* Chaintech AV-710
* Shuttle SN25P
model - Use the given board model, one of the following:
revo71, amp2000, prodigy71, aureon51, aureon71,
k8x800
revo71, amp2000, prodigy71, prodigy192, aureon51,
aureon71, universe, k8x800, phase22, phase28, ms300,
av710
Module supports up to 8 cards and autoprobe.
Note: The supported board is detected by reading EEPROM or PCI
SSID (if EEPROM isn't available). You can override the
model by passing "model" module option in case that the
driver isn't configured properly or you want to try another
type for testing.
Module snd-intel8x0
-------------------
@ -997,6 +1059,13 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
The power-management is supported.
Module snd-pxa2xx-ac97 (on arm only)
------------------------------------
Module for AC97 driver for the Intel PXA2xx chip
For ARM architecture only.
Module snd-rme32
----------------
@ -1026,7 +1095,8 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
module did formerly. It will allocate the buffers in advance
when any RME9652 cards are found. To make the buffer
allocation sure, load snd-page-alloc module in the early
stage of boot sequence.
stage of boot sequence. See "Early Buffer Allocation"
section.
Module snd-sa11xx-uda1341 (on arm only)
---------------------------------------
@ -1115,6 +1185,13 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
Module supports up to 8 cards.
Module snd-sun-dbri (on sparc only)
-----------------------------------
Module for DBRI sound chips found on Sparcs.
Module supports up to 8 cards.
Module snd-wavefront
--------------------
@ -1211,16 +1288,18 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
------------------
Module for AC'97 motherboards based on VIA 82C686A/686B, 8233,
8233A, 8233C, 8235 (south) bridge.
8233A, 8233C, 8235, 8237 (south) bridge.
mpu_port - 0x300,0x310,0x320,0x330, otherwise obtain BIOS setup
[VIA686A/686B only]
joystick - Enable joystick (default off) [VIA686A/686B only]
ac97_clock - AC'97 codec clock base (default 48000Hz)
dxs_support - support DXS channels,
0 = auto (defalut), 1 = enable, 2 = disable,
3 = 48k only, 4 = no VRA
[VIA8233/C,8235 only]
0 = auto (default), 1 = enable, 2 = disable,
3 = 48k only, 4 = no VRA, 5 = enable any sample
rate and different sample rates on different
channels
[VIA8233/C, 8235, 8237 only]
ac97_quirk - AC'97 workaround for strange hardware
See the description of intel8x0 module for details.
@ -1232,18 +1311,23 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
default value 1.4. Then the interrupt number will be
assigned under 15. You might also upgrade your BIOS.
Note: VIA8233/5 (not VIA8233A) can support DXS (direct sound)
Note: VIA8233/5/7 (not VIA8233A) can support DXS (direct sound)
channels as the first PCM. On these channels, up to 4
streams can be played at the same time.
streams can be played at the same time, and the controller
can perform sample rate conversion with separate rates for
each channel.
As default (dxs_support = 0), 48k fixed rate is chosen
except for the known devices since the output is often
noisy except for 48k on some mother boards due to the
bug of BIOS.
Please try once dxs_support=1 and if it works on other
Please try once dxs_support=5 and if it works on other
sample rates (e.g. 44.1kHz of mp3 playback), please let us
know the PCI subsystem vendor/device id's (output of
"lspci -nv").
If it doesn't work, try dxs_support=4. If it still doesn't
If dxs_support=5 does not work, try dxs_support=4; if it
doesn't work too, try dxs_support=1. (dxs_support=1 is
usually for old motherboards. The correct implementated
board should work with 4 or 5.) If it still doesn't
work and the default setting is ok, dxs_support=3 is the
right choice. If the default setting doesn't work at all,
try dxs_support=2 to disable the DXS channels.
@ -1306,7 +1390,7 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
Module snd-vxpocket
-------------------
Module for Digigram VX-Pocket VX2 PCMCIA card.
Module for Digigram VX-Pocket VX2 and 440 PCMCIA cards.
ibl - Capture IBL size. (default = 0, minimum size)
@ -1326,29 +1410,6 @@ Prior to version 0.9.0rc4 options had a 'snd_' prefix. This was removed.
Note: the driver is build only when CONFIG_ISA is set.
Module snd-vxp440
-----------------
Module for Digigram VX-Pocket 440 PCMCIA card.
ibl - Capture IBL size. (default = 0, minimum size)
Module supports up to 8 cards. The module is compiled only when
PCMCIA is supported on kernel.
To activate the driver via the card manager, you'll need to set
up /etc/pcmcia/vxp440.conf. See the sound/pcmcia/vx/vxp440.c.
When the driver is compiled as a module and the hotplug firmware
is supported, the firmware data is loaded via hotplug automatically.
Install the necessary firmware files in alsa-firmware package.
When no hotplug fw loader is available, you need to load the
firmware via vxloader utility in alsa-tools package.
About capture IBL, see the description of snd-vx222 module.
Note: the driver is build only when CONFIG_ISA is set.
Module snd-ymfpci
-----------------
@ -1497,6 +1558,36 @@ Proc interfaces (/proc/asound)
echo "rvplayer 0 0 block" > /proc/asound/card0/pcm0p/oss
Early Buffer Allocation
=======================
Some drivers (e.g. hdsp) require the large contiguous buffers, and
sometimes it's too late to find such spaces when the driver module is
actually loaded due to memory fragmentation. You can pre-allocate the
PCM buffers by loading snd-page-alloc module and write commands to its
proc file in prior, for example, in the early boot stage like
/etc/init.d/*.local scripts.
Reading the proc file /proc/drivers/snd-page-alloc shows the current
usage of page allocation. In writing, you can send the following
commands to the snd-page-alloc driver:
- add VENDOR DEVICE MASK SIZE BUFFERS
VENDOR and DEVICE are PCI vendor and device IDs. They take
integer numbers (0x prefix is needed for the hex).
MASK is the PCI DMA mask. Pass 0 if not restricted.
SIZE is the size of each buffer to allocate. You can pass
k and m suffix for KB and MB. The max number is 16MB.
BUFFERS is the number of buffers to allocate. It must be greater
than 0. The max number is 4.
- erase
This will erase the all pre-allocated buffers which are not in
use.
Links
=====

41
Documentation/sound/alsa/CMIPCI.txt
View File

@ -89,19 +89,22 @@ and use the interleaved 4 channel data.
There are some control switchs affecting to the speaker connections:
"Line-In As Rear" - As mentioned above, the line-in jack is used
for the rear (3th and 4th channels) output.
"Line-In As Bass" - The line-in jack is used for the bass (5th
and 6th channels) output.
"Mic As Center/LFE" - The mic jack is used for the bass output.
If this switch is on, you cannot use a microphone as a capture
source, of course.
"Line-In Mode" - an enum control to change the behavior of line-in
jack. Either "Line-In", "Rear Output" or "Bass Output" can
be selected. The last item is available only with model 039
or newer.
When "Rear Output" is chosen, the surround channels 3 and 4
are output to line-in jack.
"Mic-In Mode" - an enum control to change the behavior of mic-in
jack. Either "Mic-In" or "Center/LFE Output" can be
selected.
When "Center/LFE Output" is chosen, the center and bass
channels (channels 5 and 6) are output to mic-in jack.
Digital I/O
-----------
The CM8x38 provides the excellent SPDIF capability with very chip
The CM8x38 provides the excellent SPDIF capability with very cheap
price (yes, that's the reason I bought the card :)
The SPDIF playback and capture are done via the third PCM device
@ -122,8 +125,9 @@ respectively, so you cannot playback both analog and digital streams
simultaneously.
To enable SPDIF output, you need to turn on "IEC958 Output Switch"
control via mixer or alsactl. Then you'll see the red light on from
the card so you know that's working obviously :)
control via mixer or alsactl ("IEC958" is the official name of
so-called S/PDIF). Then you'll see the red light on from the card so
you know that's working obviously :)
The SPDIF input is always enabled, so you can hear SPDIF input data
from line-out with "IEC958 In Monitor" switch at any time (see
below).
@ -205,9 +209,10 @@ In addition to the standard SB mixer, CM8x38 provides more functions.
MIDI CONTROLLER
---------------
The MPU401-UART interface is enabled as default only for the first
(CMIPCI) card. You need to set module option "midi_port" properly
for the 2nd (CMIPCI) card.
The MPU401-UART interface is disabled as default. You need to set
module option "mpu_port" with a valid I/O port address to enable the
MIDI support. The valid I/O ports are 0x300, 0x310, 0x320 and 0x330.
Choose the value which doesn't conflict with other cards.
There is _no_ hardware wavetable function on this chip (except for
OPL3 synth below).
@ -229,9 +234,11 @@ I don't know why..
Joystick and Modem
------------------
The joystick and modem should be available by enabling the control
switch "Joystick" and "Modem" respectively. But I myself have never
tested them yet.
The legacy joystick is supported. To enable the joystick support, pass
joystick_port=1 module option. The value 1 means the auto-detection.
If the auto-detection fails, try to pass the exact I/O address.
The modem is enabled dynamically via a card control switch "Modem".
Debugging Information

6
Documentation/sound/alsa/DocBook/writing-an-alsa-driver.tmpl
View File

@ -371,7 +371,7 @@
<listitem><para>create <function>probe()</function> callback.</para></listitem>
<listitem><para>create <function>remove()</function> callback.</para></listitem>
<listitem><para>create pci_driver table which contains the three pointers above.</para></listitem>
<listitem><para>create <function>init()</function> function just calling <function>pci_module_init()</function> to register the pci_driver table defined above.</para></listitem>
<listitem><para>create <function>init()</function> function just calling <function>pci_register_driver()</function> to register the pci_driver table defined above.</para></listitem>
<listitem><para>create <function>exit()</function> function to call <function>pci_unregister_driver()</function> function.</para></listitem>
</itemizedlist>
</para>
@ -1198,7 +1198,7 @@
/* initialization of the module */
static int __init alsa_card_mychip_init(void)
{
return pci_module_init(&driver);
return pci_register_driver(&driver);
}
/* clean up the module */
@ -1654,7 +1654,7 @@
<![CDATA[
static int __init alsa_card_mychip_init(void)
{
return pci_module_init(&driver);
return pci_register_driver(&driver);
}
static void __exit alsa_card_mychip_exit(void)

74
Documentation/sound/alsa/emu10k1-jack.txt Normal file
View File

@ -0,0 +1,74 @@
This document is a guide to using the emu10k1 based devices with JACK for low
latency, multichannel recording functionality. All of my recent work to allow
Linux users to use the full capabilities of their hardware has been inspired
by the kX Project. Without their work I never would have discovered the true
power of this hardware.
http://www.kxproject.com
- Lee Revell, 2005.03.30
Low latency, multichannel audio with JACK and the emu10k1/emu10k2
-----------------------------------------------------------------
Until recently, emu10k1 users on Linux did not have access to the same low
latency, multichannel features offered by the "kX ASIO" feature of their
Windows driver. As of ALSA 1.0.9 this is no more!
For those unfamiliar with kX ASIO, this consists of 16 capture and 16 playback
channels. With a post 2.6.9 Linux kernel, latencies down to 64 (1.33 ms) or
even 32 (0.66ms) frames should work well.
The configuration is slightly more involved than on Windows, as you have to
select the correct device for JACK to use. Actually, for qjackctl users it's
fairly self explanatory - select Duplex, then for capture and playback select
the multichannel devices, set the in and out channels to 16, and the sample
rate to 48000Hz. The command line looks like this:
/usr/local/bin/jackd -R -dalsa -r48000 -p64 -n2 -D -Chw:0,2 -Phw:0,3 -S
This will give you 16 input ports and 16 output ports.
The 16 output ports map onto the 16 FX buses (or the first 16 of 64, for the
Audigy). The mapping from FX bus to physical output is described in
SB-Live-mixer.txt (or Audigy-mixer.txt).
The 16 input ports are connected to the 16 physical inputs. Contrary to
popular belief, all emu10k1 cards are multichannel cards. Which of these
input channels have physical inputs connected to them depends on the card
model. Trial and error is highly recommended; the pinout diagrams
for the card have been reverse engineered by some enterprising kX users and are
available on the internet. Meterbridge is helpful here, and the kX forums are
packed with useful information.
Each input port will either correspond to a digital (SPDIF) input, an analog
input, or nothing. The one exception is the SBLive! 5.1. On these devices,
the second and third input ports are wired to the center/LFE output. You will
still see 16 capture channels, but only 14 are available for recording inputs.
This chart, borrowed from kxfxlib/da_asio51.cpp, describes the mapping of JACK
ports to FXBUS2 (multitrack recording input) and EXTOUT (physical output)
channels.
/*JACK (& ASIO) mappings on 10k1 5.1 SBLive cards:
--------------------------------------------
JACK Epilog FXBUS2(nr)
--------------------------------------------
capture_1 asio14 FXBUS2(0xe)
capture_2 asio15 FXBUS2(0xf)
capture_3 asio0 FXBUS2(0x0)
~capture_4 Center EXTOUT(0x11) // mapped to by Center
~capture_5 LFE EXTOUT(0x12) // mapped to by LFE
capture_6 asio3 FXBUS2(0x3)
capture_7 asio4 FXBUS2(0x4)
capture_8 asio5 FXBUS2(0x5)
capture_9 asio6 FXBUS2(0x6)
capture_10 asio7 FXBUS2(0x7)
capture_11 asio8 FXBUS2(0x8)
capture_12 asio9 FXBUS2(0x9)
capture_13 asio10 FXBUS2(0xa)
capture_14 asio11 FXBUS2(0xb)
capture_15 asio12 FXBUS2(0xc)
capture_16 asio13 FXBUS2(0xd)
*/
TODO: describe use of ld10k1/qlo10k1 in conjunction with JACK

362
Documentation/sound/alsa/hdspm.txt Normal file
View File

@ -0,0 +1,362 @@
Software Interface ALSA-DSP MADI Driver
(translated from German, so no good English ;-),
2004 - winfried ritsch
Full functionality has been added to the driver. Since some of
the Controls and startup-options are ALSA-Standard and only the
special Controls are described and discussed below.
hardware functionality:
Audio transmission:
number of channels -- depends on transmission mode
The number of channels chosen is from 1..Nmax. The reason to
use for a lower number of channels is only resource allocation,
since unused DMA channels are disabled and less memory is
allocated. So also the throughput of the PCI system can be
scaled. (Only important for low performance boards).
Single Speed -- 1..64 channels
(Note: Choosing the 56channel mode for transmission or as
receiver, only 56 are transmitted/received over the MADI, but
all 64 channels are available for the mixer, so channel count
for the driver)
Double Speed -- 1..32 channels
Note: Choosing the 56-channel mode for
transmission/receive-mode , only 28 are transmitted/received
over the MADI, but all 32 channels are available for the mixer,
so channel count for the driver
Quad Speed -- 1..16 channels
Note: Choosing the 56-channel mode for
transmission/receive-mode , only 14 are transmitted/received
over the MADI, but all 16 channels are available for the mixer,
so channel count for the driver
Format -- signed 32 Bit Little Endian (SNDRV_PCM_FMTBIT_S32_LE)
Sample Rates --
Single Speed -- 32000, 44100, 48000
Double Speed -- 64000, 88200, 96000 (untested)
Quad Speed -- 128000, 176400, 192000 (untested)
access-mode -- MMAP (memory mapped), Not interleaved
(PCM_NON-INTERLEAVED)
buffer-sizes -- 64,128,256,512,1024,2048,8192 Samples
fragments -- 2
Hardware-pointer -- 2 Modi
The Card supports the readout of the actual Buffer-pointer,
where DMA reads/writes. Since of the bulk mode of PCI it is only
64 Byte accurate. SO it is not really usable for the
ALSA-mid-level functions (here the buffer-ID gives a better
result), but if MMAP is used by the application. Therefore it
can be configured at load-time with the parameter
precise-pointer.
(Hint: Experimenting I found that the pointer is maximum 64 to
large never to small. So if you subtract 64 you always have a
safe pointer for writing, which is used on this mode inside
ALSA. In theory now you can get now a latency as low as 16
Samples, which is a quarter of the interrupt possibilities.)
Precise Pointer -- off
interrupt used for pointer-calculation
Precise Pointer -- on
hardware pointer used.
Controller:
Since DSP-MADI-Mixer has 8152 Fader, it does not make sense to
use the standard mixer-controls, since this would break most of
(especially graphic) ALSA-Mixer GUIs. So Mixer control has be
provided by a 2-dimensional controller using the
hwdep-interface.
Also all 128+256 Peak and RMS-Meter can be accessed via the
hwdep-interface. Since it could be a performance problem always
copying and converting Peak and RMS-Levels even if you just need
one, I decided to export the hardware structure, so that of
needed some driver-guru can implement a memory-mapping of mixer
or peak-meters over ioctl, or also to do only copying and no
conversion. A test-application shows the usage of the controller.
Latency Controls --- not implemented !!!
Note: Within the windows-driver the latency is accessible of a
control-panel, but buffer-sizes are controlled with ALSA from
hwparams-calls and should not be changed in run-state, I did not
implement it here.
System Clock -- suspended !!!!
Name -- "System Clock Mode"
Access -- Read Write
Values -- "Master" "Slave"
!!!! This is a hardware-function but is in conflict with the
Clock-source controller, which is a kind of ALSA-standard. I
makes sense to set the card to a special mode (master at some
frequency or slave), since even not using an Audio-application
a studio should have working synchronisations setup. So use
Clock-source-controller instead !!!!
Clock Source
Name -- "Sample Clock Source"
Access -- Read Write
Values -- "AutoSync", "Internal 32.0 kHz", "Internal 44.1 kHz",
"Internal 48.0 kHz", "Internal 64.0 kHz", "Internal 88.2 kHz",
"Internal 96.0 kHz"
Choose between Master at a specific Frequency and so also the
Speed-mode or Slave (Autosync). Also see "Preferred Sync Ref"
!!!! This is no pure hardware function but was implemented by
ALSA by some ALSA-drivers before, so I use it also. !!!
Preferred Sync Ref
Name -- "Preferred Sync Reference"
Access -- Read Write
Values -- "Word" "MADI"
Within the Auto-sync-Mode the preferred Sync Source can be
chosen. If it is not available another is used if possible.
Note: Since MADI has a much higher bit-rate than word-clock, the
card should synchronise better in MADI Mode. But since the
RME-PLL is very good, there are almost no problems with
word-clock too. I never found a difference.
TX 64 channel ---
Name -- "TX 64 channels mode"
Access -- Read Write
Values -- 0 1
Using 64-channel-modus (1) or 56-channel-modus for
MADI-transmission (0).
Note: This control is for output only. Input-mode is detected
automatically from hardware sending MADI.
Clear TMS ---
Name -- "Clear Track Marker"
Access -- Read Write
Values -- 0 1
Don't use to lower 5 Audio-bits on AES as additional Bits.
Safe Mode oder Auto Input ---
Name -- "Safe Mode"
Access -- Read Write
Values -- 0 1
(default on)
If on (1), then if either the optical or coaxial connection
has a failure, there is a takeover to the working one, with no
sample failure. Its only useful if you use the second as a
backup connection.
Input ---
Name -- "Input Select"
Access -- Read Write
Values -- optical coaxial
Choosing the Input, optical or coaxial. If Safe-mode is active,
this is the preferred Input.
-------------- Mixer ----------------------
Mixer
Name -- "Mixer"
Access -- Read Write
Values - <channel-number 0-127> <Value 0-65535>
Here as a first value the channel-index is taken to get/set the
corresponding mixer channel, where 0-63 are the input to output
fader and 64-127 the playback to outputs fader. Value 0
is channel muted 0 and 32768 an amplification of 1.
Chn 1-64
fast mixer for the ALSA-mixer utils. The diagonal of the
mixer-matrix is implemented from playback to output.
Line Out
Name -- "Line Out"
Access -- Read Write
Values -- 0 1
Switching on and off the analog out, which has nothing to do
with mixing or routing. the analog outs reflects channel 63,64.
--- information (only read access):
Sample Rate
Name -- "System Sample Rate"
Access -- Read-only
getting the sample rate.
External Rate measured
Name -- "External Rate"
Access -- Read only
Should be "Autosync Rate", but Name used is
ALSA-Scheme. External Sample frequency liked used on Autosync is
reported.
MADI Sync Status
Name -- "MADI Sync Lock Status"
Access -- Read
Values -- 0,1,2
MADI-Input is 0=Unlocked, 1=Locked, or 2=Synced.
Word Clock Sync Status
Name -- "Word Clock Lock Status"
Access -- Read
Values -- 0,1,2
Word Clock Input is 0=Unlocked, 1=Locked, or 2=Synced.
AutoSync
Name -- "AutoSync Reference"
Access -- Read
Values -- "WordClock", "MADI", "None"
Sync-Reference is either "WordClock", "MADI" or none.
RX 64ch --- noch nicht implementiert
MADI-Receiver is in 64 channel mode oder 56 channel mode.
AB_inp --- not tested
Used input for Auto-Input.
actual Buffer Position --- not implemented
!!! this is a ALSA internal function, so no control is used !!!
Calling Parameter:
index int array (min = 1, max = 8),
"Index value for RME HDSPM interface." card-index within ALSA
note: ALSA-standard
id string array (min = 1, max = 8),
"ID string for RME HDSPM interface."
note: ALSA-standard
enable int array (min = 1, max = 8),
"Enable/disable specific HDSPM sound-cards."
note: ALSA-standard
precise_ptr int array (min = 1, max = 8),
"Enable precise pointer, or disable."
note: Use only when the application supports this (which is a special case).
line_outs_monitor int array (min = 1, max = 8),
"Send playback streams to analog outs by default."
note: each playback channel is mixed to the same numbered output
channel (routed). This is against the ALSA-convention, where all
channels have to be muted on after loading the driver, but was
used before on other cards, so i historically use it again)
enable_monitor int array (min = 1, max = 8),
"Enable Analog Out on Channel 63/64 by default."
note: here the analog output is enabled (but not routed).

Some files were not shown because too many files have changed in this diff Show More