2018-07-02 21:15:39 +00:00
|
|
|
<?xml version='1.0'?>
|
2019-03-14 13:40:58 +00:00
|
|
|
<!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.5//EN"
|
2023-12-25 14:48:33 +00:00
|
|
|
"http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd">
|
2020-11-09 04:23:58 +00:00
|
|
|
<!-- SPDX-License-Identifier: LGPL-2.1-or-later -->
|
2017-11-18 16:52:56 +00:00
|
|
|
|
2014-12-25 23:04:19 +00:00
|
|
|
<refentry id="hwdb" conditional="ENABLE_HWDB">
|
2014-12-15 23:48:24 +00:00
|
|
|
<refentryinfo>
|
|
|
|
<title>hwdb</title>
|
|
|
|
<productname>systemd</productname>
|
|
|
|
</refentryinfo>
|
|
|
|
|
|
|
|
<refmeta>
|
|
|
|
<refentrytitle>hwdb</refentrytitle>
|
|
|
|
<manvolnum>7</manvolnum>
|
|
|
|
</refmeta>
|
|
|
|
|
|
|
|
<refnamediv>
|
|
|
|
<refname>hwdb</refname>
|
|
|
|
<refpurpose>Hardware Database</refpurpose>
|
|
|
|
</refnamediv>
|
|
|
|
|
|
|
|
<refsect1><title>Description</title>
|
|
|
|
<para>The hardware database is a key-value store for associating modalias-like keys to
|
2014-08-03 05:11:37 +00:00
|
|
|
udev-property-like values. It is used primarily by udev to add the relevant properties
|
2014-12-15 23:48:24 +00:00
|
|
|
to matching devices, but it can also be queried directly.</para>
|
|
|
|
</refsect1>
|
|
|
|
|
|
|
|
<refsect1><title>Hardware Database Files</title>
|
|
|
|
<para>The hwdb files are read from the files located in the
|
2015-06-18 17:47:44 +00:00
|
|
|
system hwdb directory <filename>/usr/lib/udev/hwdb.d</filename> and
|
|
|
|
the local administration directory <filename>/etc/udev/hwdb.d</filename>.
|
2014-12-15 23:48:24 +00:00
|
|
|
All hwdb files are collectively sorted and processed in lexical order,
|
|
|
|
regardless of the directories in which they live. However, files with
|
2020-10-05 16:08:21 +00:00
|
|
|
identical filenames replace each other. Files in <filename>/etc/</filename>
|
2015-06-18 17:47:44 +00:00
|
|
|
have the highest priority and take precedence over files with the same
|
2020-10-05 16:08:21 +00:00
|
|
|
name in <filename>/usr/lib/</filename>. This can be used to override a
|
2015-06-18 17:47:44 +00:00
|
|
|
system-supplied hwdb file with a local file if needed;
|
2020-10-05 16:08:21 +00:00
|
|
|
a symlink in <filename>/etc/</filename> with the same name as a hwdb file in
|
|
|
|
<filename>/usr/lib/</filename>, pointing to <filename>/dev/null</filename>,
|
2016-11-30 03:29:31 +00:00
|
|
|
disables that hwdb file entirely. hwdb files must have the extension
|
2014-12-15 23:48:24 +00:00
|
|
|
<filename>.hwdb</filename>; other extensions are ignored.</para>
|
|
|
|
|
2016-11-30 03:29:31 +00:00
|
|
|
<para>Each hwdb file contains data records consisting of matches and associated
|
|
|
|
key-value pairs. Every record in the hwdb starts with one or more match strings,
|
|
|
|
specifying a shell glob to compare the lookup string against. Multiple match lines
|
|
|
|
are specified in consecutive lines. Every match line is compared individually, and
|
|
|
|
they are combined by OR. Every match line must start at the first character of the
|
|
|
|
line.</para>
|
2014-12-15 23:48:24 +00:00
|
|
|
|
2020-10-15 15:58:43 +00:00
|
|
|
<para>Match patterns consist of literal characters, and shell-style wildcards:</para>
|
|
|
|
<itemizedlist>
|
|
|
|
<listitem><para>Asterisk <literal>*</literal> matches any number of characters
|
|
|
|
</para></listitem>
|
|
|
|
<listitem><para>Question mark <literal>?</literal> matches a single character
|
|
|
|
</para></listitem>
|
|
|
|
<listitem><para>Character list <literal>[<replaceable>chars</replaceable>]</literal> matches one of
|
|
|
|
the characters <replaceable>chars</replaceable> listed between <literal>[</literal> and
|
|
|
|
<literal>]</literal>. A range may be specified as with a dash as
|
|
|
|
<literal>[<replaceable>first</replaceable>-<replaceable>last</replaceable>]</literal>. The match may
|
|
|
|
be inverted with a caret <literal>[^…]</literal>.</para></listitem>
|
|
|
|
</itemizedlist>
|
|
|
|
|
2016-11-30 03:29:31 +00:00
|
|
|
<para>The match lines are followed by one or more key-value pair lines, which are
|
|
|
|
recognized by a leading space character. The key name and value are separated by
|
|
|
|
<literal>=</literal>. An empty line signifies the end of a record. Lines beginning
|
|
|
|
with <literal>#</literal> are ignored.</para>
|
|
|
|
|
|
|
|
<para>In case multiple records match a given lookup string, the key-value pairs
|
|
|
|
from all records are combined. If a key is specified multiple times, the value
|
|
|
|
from the record with the highest priority is used (each key can have only a single
|
|
|
|
value). The priority is higher when the record is in a file that sorts later
|
|
|
|
lexicographically, and in case of records in the same file, later records have
|
|
|
|
higher priority.</para>
|
2014-12-15 23:48:24 +00:00
|
|
|
|
|
|
|
<para>The content of all hwdb files is read by
|
|
|
|
<citerefentry><refentrytitle>systemd-hwdb</refentrytitle><manvolnum>8</manvolnum></citerefentry>
|
|
|
|
and compiled to a binary database located at <filename>/etc/udev/hwdb.bin</filename>,
|
2016-11-30 03:29:31 +00:00
|
|
|
or alternatively <filename>/usr/lib/udev/hwdb.bin</filename> if you want ship the
|
|
|
|
compiled database in an immutable image. During runtime, only the binary database
|
|
|
|
is used.</para>
|
|
|
|
</refsect1>
|
|
|
|
|
|
|
|
<refsect1>
|
|
|
|
<title>Examples</title>
|
|
|
|
|
|
|
|
<example>
|
|
|
|
<title>General syntax of hwdb files</title>
|
|
|
|
|
|
|
|
<programlisting># /usr/lib/udev/hwdb.d/example.hwdb
|
|
|
|
# Comments can be placed before any records. This is a good spot
|
|
|
|
# to describe what that file is used for, what kind of properties
|
|
|
|
# it defines, and the ordering convention.
|
|
|
|
|
|
|
|
# A record with three matches and one property
|
hwdb: add trailing ":*" everywhere
No functional change is intended.
The general pattern of changes:
-usb:v04F3p2B7C*
+usb:v04F3p2B7C:*
This is mostly a clarification, to make the part that makes the usb vXXXXpYYYY
part visually separated. It would only make a difference if we added further
keys with a different number of digits, which is unlikely.
-usb:v0627p0001:*QEMU USB Keyboard*
-usb:v0627p0001:*QEMU USB Mouse*
-usb:v0627p0001:*QEMU USB Tablet*
+usb:v0627p0001:*QEMU USB Keyboard*:*
+usb:v0627p0001:*QEMU USB Mouse*:*
+usb:v0627p0001:*QEMU USB Tablet*:*
Again, only a clarification. We know that ":" will appear somewhere later in
the match key, so anything that matches "…Keyboard*" will also match "…Keyboard*:*".
-evdev:name:ETPS/2 Elantech Touchpad:dmi:*svnASUSTeKComputerInc.:pnN53SV*
+evdev:name:ETPS/2 Elantech Touchpad:dmi:*svnASUSTeKComputerInc.:pnN53SV:*
This makes the match narrower. Previously we would match product "N53SV"
and "N53SV2", "N53SV3", and others. Here we are saying that the ':pn' part must
match exactly. Most of the changes in this patch match this pattern. I made a few
judgement calls and used "pn…*:*" when I wasn't sure if the full pn is included:
-evdev:name:Dell WMI hotkeys:dmi:bvn*:bvr*:bd*:svnDell*:pnPrecision*
+evdev:name:Dell WMI hotkeys:dmi:bvn*:bvr*:bd*:svnDell*:pnPrecision*:*
-evdev:name:Cypress APA Trackpad ?cyapa?:dmi:*:svnHewlett-Packard*:pnFalco*:
+evdev:name:Cypress APA Trackpad ?cyapa?:dmi:*:svnHewlett-Packard*:pnFalco*:*
This more like the "QEMU" example above, since all dmi strings end in ":", so
anything which matches the old version will also match the new version.
-evdev:atkbd:dmi:bvn*:bvr*:bd*:svnGateway*:pnA0A1*:pvr*
+evdev:atkbd:dmi:bvn*:bvr*:bd*:svnGateway*:pnA0A1*:*
I replaced trailing ":pvr*" by ":*". This makes no functional difference because
we expect "pvr" to always appear in the dmi string. This makes patterns shorter.
-evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*
+evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*:*
OTOH, ":pn*" is kept. This is because almost patterns include ":pn*", and if we
skip it, we should make it clear that this is on purpose, that we really want to
match any product name.
The python script to generate autosuspend rules is updated to use ":*" too.
Inspired by https://github.com/systemd/systemd/pull/17281#discussion_r501489750.
2020-10-15 07:23:30 +00:00
|
|
|
mouse:*:name:*Trackball*:*
|
|
|
|
mouse:*:name:*trackball*:*
|
|
|
|
mouse:*:name:*TrackBall*:*
|
2016-11-30 03:29:31 +00:00
|
|
|
ID_INPUT_TRACKBALL=1
|
|
|
|
|
2020-10-15 15:58:43 +00:00
|
|
|
# The rule above could be also be written in a form that
|
|
|
|
# matches Tb, tb, TB, tB:
|
|
|
|
mouse:*:name:*[tT]rack[bB]all*:*
|
|
|
|
ID_INPUT_TRACKBALL=1
|
|
|
|
|
2016-11-30 03:29:31 +00:00
|
|
|
# A record with a single match and five properties
|
hwdb: add trailing ":*" everywhere
No functional change is intended.
The general pattern of changes:
-usb:v04F3p2B7C*
+usb:v04F3p2B7C:*
This is mostly a clarification, to make the part that makes the usb vXXXXpYYYY
part visually separated. It would only make a difference if we added further
keys with a different number of digits, which is unlikely.
-usb:v0627p0001:*QEMU USB Keyboard*
-usb:v0627p0001:*QEMU USB Mouse*
-usb:v0627p0001:*QEMU USB Tablet*
+usb:v0627p0001:*QEMU USB Keyboard*:*
+usb:v0627p0001:*QEMU USB Mouse*:*
+usb:v0627p0001:*QEMU USB Tablet*:*
Again, only a clarification. We know that ":" will appear somewhere later in
the match key, so anything that matches "…Keyboard*" will also match "…Keyboard*:*".
-evdev:name:ETPS/2 Elantech Touchpad:dmi:*svnASUSTeKComputerInc.:pnN53SV*
+evdev:name:ETPS/2 Elantech Touchpad:dmi:*svnASUSTeKComputerInc.:pnN53SV:*
This makes the match narrower. Previously we would match product "N53SV"
and "N53SV2", "N53SV3", and others. Here we are saying that the ':pn' part must
match exactly. Most of the changes in this patch match this pattern. I made a few
judgement calls and used "pn…*:*" when I wasn't sure if the full pn is included:
-evdev:name:Dell WMI hotkeys:dmi:bvn*:bvr*:bd*:svnDell*:pnPrecision*
+evdev:name:Dell WMI hotkeys:dmi:bvn*:bvr*:bd*:svnDell*:pnPrecision*:*
-evdev:name:Cypress APA Trackpad ?cyapa?:dmi:*:svnHewlett-Packard*:pnFalco*:
+evdev:name:Cypress APA Trackpad ?cyapa?:dmi:*:svnHewlett-Packard*:pnFalco*:*
This more like the "QEMU" example above, since all dmi strings end in ":", so
anything which matches the old version will also match the new version.
-evdev:atkbd:dmi:bvn*:bvr*:bd*:svnGateway*:pnA0A1*:pvr*
+evdev:atkbd:dmi:bvn*:bvr*:bd*:svnGateway*:pnA0A1*:*
I replaced trailing ":pvr*" by ":*". This makes no functional difference because
we expect "pvr" to always appear in the dmi string. This makes patterns shorter.
-evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*
+evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*:*
OTOH, ":pn*" is kept. This is because almost patterns include ":pn*", and if we
skip it, we should make it clear that this is on purpose, that we really want to
match any product name.
The python script to generate autosuspend rules is updated to use ":*" too.
Inspired by https://github.com/systemd/systemd/pull/17281#discussion_r501489750.
2020-10-15 07:23:30 +00:00
|
|
|
mouse:usb:v046dp4041:name:Logitech MX Master:*
|
2016-11-30 03:29:31 +00:00
|
|
|
MOUSE_DPI=1000@166
|
|
|
|
MOUSE_WHEEL_CLICK_ANGLE=15
|
|
|
|
MOUSE_WHEEL_CLICK_ANGLE_HORIZONTAL=26
|
|
|
|
MOUSE_WHEEL_CLICK_COUNT=24
|
|
|
|
MOUSE_WHEEL_CLICK_COUNT_HORIZONTAL=14
|
|
|
|
</programlisting>
|
|
|
|
</example>
|
|
|
|
|
|
|
|
<example>
|
|
|
|
<title>Overriding of properties</title>
|
|
|
|
|
|
|
|
<programlisting># /usr/lib/udev/hwdb.d/60-keyboard.hwdb
|
hwdb: add trailing ":*" everywhere
No functional change is intended.
The general pattern of changes:
-usb:v04F3p2B7C*
+usb:v04F3p2B7C:*
This is mostly a clarification, to make the part that makes the usb vXXXXpYYYY
part visually separated. It would only make a difference if we added further
keys with a different number of digits, which is unlikely.
-usb:v0627p0001:*QEMU USB Keyboard*
-usb:v0627p0001:*QEMU USB Mouse*
-usb:v0627p0001:*QEMU USB Tablet*
+usb:v0627p0001:*QEMU USB Keyboard*:*
+usb:v0627p0001:*QEMU USB Mouse*:*
+usb:v0627p0001:*QEMU USB Tablet*:*
Again, only a clarification. We know that ":" will appear somewhere later in
the match key, so anything that matches "…Keyboard*" will also match "…Keyboard*:*".
-evdev:name:ETPS/2 Elantech Touchpad:dmi:*svnASUSTeKComputerInc.:pnN53SV*
+evdev:name:ETPS/2 Elantech Touchpad:dmi:*svnASUSTeKComputerInc.:pnN53SV:*
This makes the match narrower. Previously we would match product "N53SV"
and "N53SV2", "N53SV3", and others. Here we are saying that the ':pn' part must
match exactly. Most of the changes in this patch match this pattern. I made a few
judgement calls and used "pn…*:*" when I wasn't sure if the full pn is included:
-evdev:name:Dell WMI hotkeys:dmi:bvn*:bvr*:bd*:svnDell*:pnPrecision*
+evdev:name:Dell WMI hotkeys:dmi:bvn*:bvr*:bd*:svnDell*:pnPrecision*:*
-evdev:name:Cypress APA Trackpad ?cyapa?:dmi:*:svnHewlett-Packard*:pnFalco*:
+evdev:name:Cypress APA Trackpad ?cyapa?:dmi:*:svnHewlett-Packard*:pnFalco*:*
This more like the "QEMU" example above, since all dmi strings end in ":", so
anything which matches the old version will also match the new version.
-evdev:atkbd:dmi:bvn*:bvr*:bd*:svnGateway*:pnA0A1*:pvr*
+evdev:atkbd:dmi:bvn*:bvr*:bd*:svnGateway*:pnA0A1*:*
I replaced trailing ":pvr*" by ":*". This makes no functional difference because
we expect "pvr" to always appear in the dmi string. This makes patterns shorter.
-evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*
+evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*:*
OTOH, ":pn*" is kept. This is because almost patterns include ":pn*", and if we
skip it, we should make it clear that this is on purpose, that we really want to
match any product name.
The python script to generate autosuspend rules is updated to use ":*" too.
Inspired by https://github.com/systemd/systemd/pull/17281#discussion_r501489750.
2020-10-15 07:23:30 +00:00
|
|
|
evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer*:pn*:*
|
2016-11-30 03:29:31 +00:00
|
|
|
KEYBOARD_KEY_a1=help
|
|
|
|
KEYBOARD_KEY_a2=setup
|
|
|
|
KEYBOARD_KEY_a3=battery
|
|
|
|
|
2020-10-15 15:58:43 +00:00
|
|
|
# Match vendor name "Acer" and any product name starting with "X123"
|
|
|
|
evdev:atkbd:dmi:bvn*:bvr*:bd*:svnAcer:pnX123*:*
|
2016-11-30 03:29:31 +00:00
|
|
|
KEYBOARD_KEY_a2=wlan
|
|
|
|
|
|
|
|
# /etc/udev/hwdb.d/70-keyboard.hwdb
|
|
|
|
# disable wlan key on all at keyboards
|
|
|
|
evdev:atkbd:*
|
2020-10-15 15:58:43 +00:00
|
|
|
KEYBOARD_KEY_a2=reserved
|
|
|
|
PROPERTY_WITH_SPACES=some string</programlisting>
|
2016-11-30 03:29:31 +00:00
|
|
|
|
|
|
|
<para>If the hwdb consists of those two files, a keyboard with the lookup string
|
2023-12-05 12:53:32 +00:00
|
|
|
<literal>evdev:atkbd:dmi:bvnAcer:bvr:bdXXXXX:bd08/05/2010:svnAcer:pnX123:</literal>
|
2016-11-30 03:29:31 +00:00
|
|
|
will match all three records, and end up with the following properties:</para>
|
|
|
|
|
|
|
|
<programlisting>KEYBOARD_KEY_a1=help
|
|
|
|
KEYBOARD_KEY_a2=reserved
|
2020-10-15 15:58:43 +00:00
|
|
|
KEYBOARD_KEY_a3=battery
|
|
|
|
PROPERTY_WITH_SPACES=some string</programlisting>
|
|
|
|
|
2016-11-30 03:29:31 +00:00
|
|
|
</example>
|
2014-12-15 23:48:24 +00:00
|
|
|
</refsect1>
|
|
|
|
|
|
|
|
<refsect1>
|
|
|
|
<title>See Also</title>
|
|
|
|
<para>
|
|
|
|
<citerefentry>
|
|
|
|
<refentrytitle>systemd-hwdb</refentrytitle><manvolnum>8</manvolnum>
|
|
|
|
</citerefentry>
|
|
|
|
</para>
|
|
|
|
</refsect1>
|
|
|
|
</refentry>
|