pci.ids.man

.TH pci.ids 5 "@TODAY@" "@VERSION@" "The PCI Utilities"

.SH NAME
pci.ids \- list of known identifiers related to PCI devices

.SH INTRODUCTION
Devices on the PCI bus are identified by a combination
of a vendor ID (assigned by the PCI SIG) and device ID
(assigned by the vendor). Both IDs are 16-bit integers
and the device itself provides no translation to a human-readable
string.

In addition to the vendor and device, devices also report several
other identifiers:

.IP \(bu 3
Device class and subclass (two 8-bit numbers)

.IP \(bu 3
Programming interface (8-bit number, meaning specific for the subclass)

.IP \(bu 3
Subsystem, which identifies the assembly in which the device is contained.
A typical example is an Ethernet add-in card: the device is the Ethernet
controller chip, while the card plays the role of the subsystem. Subsystems
have their vendor ID (from the same namespace as device vendors) and subsystem
ID. Generally, the meaning of the subsystem ID depends on the device, but there
are cases in which a single subsystem ID is used for many devices - e.g.,
laptop motherboards.

The PCI utilities use the
.B pci.ids
file to translate all these numeric IDs to strings.

.SH KEEPING THE LIST UP-TO-DATE
The
.B pci.ids
file is generated from the PCI ID database, which is maintained at
.UR https://pci-ids.ucw.cz/
.UE .
If you find any IDs missing from the list, please contribute them
to the database.

You can use the
.B update-pciids
command to download the current version of the list.

Alternatively, you can use
.B lspci -q
to query the database online.

.SH FILE FORMAT
The pci.ids file is a text file in plain ASCII, interpreted line by line.
Lines starting with the hash sign are treated as comments are ignored.
Comments regarding a specific entry are written immediately before the entry.

Vendor entries start with a 4-digit hexadecimal vendor ID, followed by one
or more spaces, and the name of the vendor extending to the end of the line.

Device entries are placed below the vendor entry. Each device entry consists
of a single TAB character, a 4-digit hexadecimal device ID, followed by one or more
spaces, and the name of the device extending to the end of the line.

Subsystem entries are placed below the device entry. They start with two TAB
characters, a 4-digit hexadecimal vendor ID (which must be defined elsewhere in
the list), a single space, a 4-digit hexadecimal subsystem ID, one or more
spaces, and the name of the subsystem extending to the end of the line.

Class entries consist of "C", one space, 2-digit hexadecimal class ID, one or
more spaces, and the name of the class. Subclasses are placed below the
corresponding class, indented by a single TAB, followed by a 2-digit
hexadecimal subclass ID, one or more spaces, and the name of the subclass.
Programming interfaces are below the subclass, indented by two TABs, followed
by a 2-digit hexadecimal prog-if ID, one or more spaces, and the name.

There can be device-independent subsystem IDs, although the web interface of
the database does not support them yet. They start with a subsystem vendor line
consisting of "S", one space, and a 4-digit hexadecimal vendor ID (which must
correspond to an already listed vendor). Subsystems follow on subsequent lines,
each indented by one TAB, followed by a 4-digit hexadecimal subsystem ID, one
or more spaces, and the name of the subsystem.

To ensure extensibility of the format, lines starting with an unrecognized letter
followed by a single space are ignored and so are all following TAB-indented lines.

.SH SEE ALSO
.BR lspci (8),
.BR update-pciids (8),
.BR pcilib (7)

.SH AUTHOR
The PCI Utilities are maintained by Martin Mares <mj@ucw.cz>.