3 NAME="" will not create any nodes, but execute RUN keys. To completely
4 ignore an event the OPTION "ignore_device" should be used.
6 After removal of the reorder queue, events with a TIMEOUT can be executed
7 without any queuing now.
11 Fixed bug in udevd, if inotify is not available. We depend on netlink
12 uevents now, kernels without that event source will not work with that
13 version of udev anymore.
17 The rule parsing happens now in the daemon once at startup, all udev
18 event processes inherit the already parsed rules from the daemon.
19 It is shipped with SUSE10.0 and reduces heavily the system load at
20 startup. The option to save precompiled rules and let the udev process
21 pick the them up is removed, as it's no longer needed.
23 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
24 device. Libsysfs is changed to "translate" the requested link into the
25 real device path, as it would happen with the hotplug event. Otherwise
26 device removal and the udev database will not work.
28 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
31 A few improvements for vol_id, the filesytem probing code.
35 Fix a stupid typo in extras/run_directory for "make install".
37 scsi_id creates the temporary devnode now in /dev for usage with a
38 non-writable /tmp directory.
40 The uevent kernel socket buffer can carry app. 50.000 events now,
41 let's see who can break this again. :)
43 The upcoming kernel will have a new input driver core integration.
44 Some class devices are now symlinks to the real device. libsysfs
45 needs a fix for this to work correctly. Udevstart of older udev
46 versions will _not_ create these devices!
50 Fix a 'install' target in the Makefile, that prevents EXTRAS from
55 A bunch of mostly trivial bugfixes. From now on no node name or
56 symlink name can contain any character than plain whitelisted ascii
57 characters or validated utf8 byte-streams. This is needed for the
58 /dev/disk/by-label/* links, cause we import untrusted data and
59 export it to the filesystem.
63 More bugfixes. If udevd was started from the kernel, we don't
64 have stdin/stdout/stderr, which broke the forked tools in some
69 Bugfix. udevstart event ordering was broken for a long time.
70 The new run_program() uncovered it, cause /dev/null was not
71 available while we try to run external programs.
72 Now udevstart should create it before we run anything.
76 Minor bugfixes and some distro rules updates. If you don't have the
77 persistent disk rules in /dev/disk/by-*/* on your distro, just
82 We can use socket communication now to pass events from udev to
84 RUN+="socket:/org/freedesktop/hal/udev_event"
85 will pass the whole udev event to the HAL daemon without the need
86 for a forked helper. (See ChangeLog for udevmonitor, as an example)
90 Mostly bugfixes and see ChangeLog.
92 The test for the existence of an environment value should be
94 ENV{KEY}=="*" to ENV{KEY}=="?*"
95 cause "*" will not fail anymore, if the key does not exist or
100 Bugfixes and a few tweaks described in the ChangeLog.
104 Mostly a Bugfix release.
106 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
107 timing with custom rules.
111 We changed the internal rule storage format. Our large rule files took
112 2 MB of RAM, with the change we are down to 99kB.
114 If the device-node has been created with default name and no symlink or
115 options are to remenber, it is not longer stored in the udevdb. HAL will
116 need to be updated to work correctly with that change.
118 To overrride optimization flags, OPTFLAGS may be used now.
126 Major changes happened with this release. The goal is to take over the
127 complete kernel-event handling and provide a more efficient way to dispatch
128 kernel events. Replacing most of the current shell script logic and the
129 kernel forked helper with a netlink-daemon and a rule-based event handling.
131 o udevd listens to netlink events now. The first valid netlink event
132 will make udevd ignore any message from udevsend that contains a
133 SEQNUM, to avoid duplicate events. The forked events can be disabled
135 echo "" > /proc/sys/kernel/hotplug
136 For full support, the broken input-subsytem needs to be fixed, not to
137 bypass the driver core.
139 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
140 removed from udev itself and must be emulated by calling small
141 helper binaries provided in the extras folder:
142 make EXTRAS=extras/run_directory/
143 will build udev_run_devd and udev_run_hotplugd, which can be called
144 from a rule if needed:
145 RUN+="/sbin/udev_run_hotplugd"
146 The recommended way to handle this is to convert all the calls from
147 the directories to explicit udev rules and get completely rid of the
148 multiplexing. (To catch a ttyUSB event, you now no longer need to
149 fork and exit 300 tty script instances you are not interested in, it
150 is just one rule that matches exactly the device.)
152 o udev handles now _all_ events not just events for class and block
153 devices, this way it is possible to control the complete event
154 behavior with udev rules. Especially useful for rules like:
155 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
157 o As used in the modalias rule, udev supports now textual
158 substitution placeholder along with the usual format chars. This
159 needs to be documented, for now it's only visible in udev_rules_parse.c.
161 o The rule keys support now more operations. This is documented in the
162 man page. It is possible to add values to list-keys like the SYMLINK
163 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
164 Also "final"-assignments are supported by using KEY:="value", which will
165 prevent changing the key by any later rule.
167 o kernel 2.6.12 has the "detached_state" attribute removed from
168 sysfs, which was used to recognize sysfs population. We switched that
169 to wait for the "bus" link, which is only available in kernels after 2.6.11.
170 Running this udev version on older kernels may cause a short delay for
173 o To provide infrastructure for persistent device naming, the id programs:
174 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
175 to export the probed data in environment key format:
176 pim:~ # /sbin/ata_id --export /dev/hda
177 ID_MODEL=HTS726060M9AT00
178 ID_SERIAL=MRH401M4G6UM9B
182 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
183 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
186 kay@pim:~> tree /dev/disk
189 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
190 | `-- IBM-Memory_Key -> ../../sda
192 | |-- swap -> ../../hda1
193 | |-- date -> ../../sda1
194 | `-- home -> ../../hda3
196 |-- 2E08712B0870F2E7 -> ../../hda3
197 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
198 |-- E845-7A89 -> ../../sda1
199 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
201 The IMPORT= operation will import these keys in the environment and make
202 it available for later PROGRAM= and RUN= executed programs. The keys are
203 also stored in the udevdb and can be queried from there with one of the
206 o A few binaries are silently added to the repository, which can be used
207 to replay kernel events from initramfs instead of using coldplug. udevd
208 can be instructed now to queue-up events while the stored events from
209 initramfs are filled into the udevd-queue. This code is still under
210 development and there is no documentation now besides the code itself.
211 The additional binaries get compiled, but are not installed by default.
213 o There is also a temporary fix for a performance problem where too many
214 events happen in parallel and every event needs to parse the rules.
215 udev can now read precompiled rules stored on disk. This is likely to be
216 replaced by a more elegant solution in a future udev version.
220 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
221 Fix for libsysfs not to expect this file was added.
225 All rules are applied now, but only the first matching rule with a NAME-key
226 will be applied. All later rules with NAME-key are completely ignored. This
227 way system supplied symlinks or permissions gets applied to user-defined
231 Please check your rules setup, if you may need to add OPTIONS="last_rule"
232 to some rules, to keep the old behavior.
234 The rules are read on "remove"-events too. That makes is possible to match
235 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
236 instruct udev to ignore an event (OPTIONS="ignore_device").
237 The new ACTION-key may be used to let a rule act only at a "remove"-event.
239 The new RUN-key supports rule-based execution of programs after device-node
240 handling. This is meant as a general replacement for the dev.d/-directories
241 to give fine grained control over the execution of programs.
243 The %s{}-sysfs format char replacement values are searched at any of the
244 devices in the device chain now, not only at the class-device.
246 We support log priority levels now. The value udev_log in udev.conf is used
247 to determine what is printed to syslog. This makes it possible to
248 run a version with compiled-in debug messages in a production environment
249 which is sometimes needed to find a bug.
250 It is still possible to supress the inclusion of _any_ syslog usage with
251 USE_LOG=false to create the smallest possible binaries if needed.
252 The configured udev_log value can be overridden with the environment variable
257 Possible use of a system-wide klibc:
258 make USE_KLIBC=true KLCC=/usr/bin/klcc all
259 will link against an external klibc and our own version will be ignored.
263 We support an unlimited count of symlinks now.
265 If USE_STATIC=true is passed to a glibc build, we link statically and use
266 a built-in userdb parser to resolve user and group names.
268 The PLACE= key is gone. It can be replaced by an ID= for a long time, cause
269 we walk up the chain of physical devices to find a match.
271 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
272 easy to skip certain attribute matches without composing rules with weird
273 character class negations like:
274 KERNEL="[!s][!c][!d]*"
275 this can now be replaced with:
277 The current simple '=' is still supported, and should work as it does today,
278 but existing rules should be converted if possible, to be better readable.
280 We have new ENV{}== key now, to match against a maximum of 5 environment
283 udevstart is its own binary again, cause we don't need co carry this araound
284 with every forked event.