3 The rule parsing happens now in the daemon once at startup, all udev
4 event processes inherit the already parsed rules from the daemon.
5 It is shipped with SUSE10.0 and reduces heavily the system load at
6 startup. The option to save precompiled rules and let the udev process
7 pick the them up is removed, as it's no longer needed.
9 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
10 device. Libsysfs is changed to "translate" the requested link into the
11 real device path, as it would happen with the hotplug event. Otherwise
12 device removal and the udev database will not work.
14 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
17 A few improvements for vol_id, the filesytem probing code.
21 Fix a stupid typo in extras/run_directory for "make install".
23 scsi_id creates the temporary devnode now in /dev for usage with a
24 non-writable /tmp directory.
26 The uevent kernel socket buffer can carry app. 50.000 events now,
27 let's see who can break this again. :)
29 The upcoming kernel will have a new input driver core integration.
30 Some class devices are now symlinks to the real device. libsysfs
31 needs a fix for this to work correctly. Udevstart of older udev
32 versions will _not_ create these devices!
36 Fix a 'install' target in the Makefile, that prevents EXTRAS from
41 A bunch of mostly trivial bugfixes. From now on no node name or
42 symlink name can contain any character than plain whitelisted ascii
43 characters or validated utf8 byte-streams. This is needed for the
44 /dev/disk/by-label/* links, cause we import untrusted data and
45 export it to the filesystem.
49 More bugfixes. If udevd was started from the kernel, we don't
50 have stdin/stdout/stderr, which broke the forked tools in some
55 Bugfix. udevstart event ordering was broken for a long time.
56 The new run_program() uncovered it, cause /dev/null was not
57 available while we try to run external programs.
58 Now udevstart should create it before we run anything.
62 Minor bugfixes and some distro rules updates. If you don't have the
63 persistent disk rules in /dev/disk/by-*/* on your distro, just
68 We can use socket communication now to pass events from udev to
70 RUN+="socket:/org/freedesktop/hal/udev_event"
71 will pass the whole udev event to the HAL daemon without the need
72 for a forked helper. (See ChangeLog for udevmonitor, as an example)
76 Mostly bugfixes and see ChangeLog.
78 The test for the existence of an environment value should be
80 ENV{KEY}=="*" to ENV{KEY}=="?*"
81 cause "*" will not fail anymore, if the key does not exist or
86 Bugfixes and a few tweaks described in the ChangeLog.
90 Mostly a Bugfix release.
92 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
93 timing with custom rules.
97 We changed the internal rule storage format. Our large rule files took
98 2 MB of RAM, with the change we are down to 99kB.
100 If the device-node has been created with default name and no symlink or
101 options are to remenber, it is not longer stored in the udevdb. HAL will
102 need to be updated to work correctly with that change.
104 To overrride optimization flags, OPTFLAGS may be used now.
112 Major changes happened with this release. The goal is to take over the
113 complete kernel-event handling and provide a more efficient way to dispatch
114 kernel events. Replacing most of the current shell script logic and the
115 kernel forked helper with a netlink-daemon and a rule-based event handling.
117 o udevd listens to netlink events now. The first valid netlink event
118 will make udevd ignore any message from udevsend that contains a
119 SEQNUM, to avoid duplicate events. The forked events can be disabled
121 echo "" > /proc/sys/kernel/hotplug
122 For full support, the broken input-subsytem needs to be fixed, not to
123 bypass the driver core.
125 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
126 removed from udev itself and must be emulated by calling small
127 helper binaries provided in the extras folder:
128 make EXTRAS=extras/run_directory/
129 will build udev_run_devd and udev_run_hotplugd, which can be called
130 from a rule if needed:
131 RUN+="/sbin/udev_run_hotplugd"
132 The recommended way to handle this is to convert all the calls from
133 the directories to explicit udev rules and get completely rid of the
134 multiplexing. (To catch a ttyUSB event, you now no longer need to
135 fork and exit 300 tty script instances you are not interested in, it
136 is just one rule that matches exactly the device.)
138 o udev handles now _all_ events not just events for class and block
139 devices, this way it is possible to control the complete event
140 behavior with udev rules. Especially useful for rules like:
141 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
143 o As used in the modalias rule, udev supports now textual
144 substitution placeholder along with the usual format chars. This
145 needs to be documented, for now it's only visible in udev_rules_parse.c.
147 o The rule keys support now more operations. This is documented in the
148 man page. It is possible to add values to list-keys like the SYMLINK
149 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
150 Also "final"-assignments are supported by using KEY:="value", which will
151 prevent changing the key by any later rule.
153 o kernel 2.6.12 has the "detached_state" attribute removed from
154 sysfs, which was used to recognize sysfs population. We switched that
155 to wait for the "bus" link, which is only available in kernels after 2.6.11.
156 Running this udev version on older kernels may cause a short delay for
159 o To provide infrastructure for persistent device naming, the id programs:
160 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
161 to export the probed data in environment key format:
162 pim:~ # /sbin/ata_id --export /dev/hda
163 ID_MODEL=HTS726060M9AT00
164 ID_SERIAL=MRH401M4G6UM9B
168 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
169 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
172 kay@pim:~> tree /dev/disk
175 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
176 | `-- IBM-Memory_Key -> ../../sda
178 | |-- swap -> ../../hda1
179 | |-- date -> ../../sda1
180 | `-- home -> ../../hda3
182 |-- 2E08712B0870F2E7 -> ../../hda3
183 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
184 |-- E845-7A89 -> ../../sda1
185 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
187 The IMPORT= operation will import these keys in the environment and make
188 it available for later PROGRAM= and RUN= executed programs. The keys are
189 also stored in the udevdb and can be queried from there with one of the
192 o A few binaries are silently added to the repository, which can be used
193 to replay kernel events from initramfs instead of using coldplug. udevd
194 can be instructed now to queue-up events while the stored events from
195 initramfs are filled into the udevd-queue. This code is still under
196 development and there is no documentation now besides the code itself.
197 The additional binaries get compiled, but are not installed by default.
199 o There is also a temporary fix for a performance problem where too many
200 events happen in parallel and every event needs to parse the rules.
201 udev can now read precompiled rules stored on disk. This is likely to be
202 replaced by a more elegant solution in a future udev version.
206 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
207 Fix for libsysfs not to expect this file was added.
211 All rules are applied now, but only the first matching rule with a NAME-key
212 will be applied. All later rules with NAME-key are completely ignored. This
213 way system supplied symlinks or permissions gets applied to user-defined
217 Please check your rules setup, if you may need to add OPTIONS="last_rule"
218 to some rules, to keep the old behavior.
220 The rules are read on "remove"-events too. That makes is possible to match
221 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
222 instruct udev to ignore an event (OPTIONS="ignore_device").
223 The new ACTION-key may be used to let a rule act only at a "remove"-event.
225 The new RUN-key supports rule-based execution of programs after device-node
226 handling. This is meant as a general replacement for the dev.d/-directories
227 to give fine grained control over the execution of programs.
229 The %s{}-sysfs format char replacement values are searched at any of the
230 devices in the device chain now, not only at the class-device.
232 We support log priority levels now. The value udev_log in udev.conf is used
233 to determine what is printed to syslog. This makes it possible to
234 run a version with compiled-in debug messages in a production environment
235 which is sometimes needed to find a bug.
236 It is still possible to supress the inclusion of _any_ syslog usage with
237 USE_LOG=false to create the smallest possible binaries if needed.
238 The configured udev_log value can be overridden with the environment variable
243 Possible use of a system-wide klibc:
244 make USE_KLIBC=true KLCC=/usr/bin/klcc all
245 will link against an external klibc and our own version will be ignored.
249 We support an unlimited count of symlinks now.
251 If USE_STATIC=true is passed to a glibc build, we link statically and use
252 a built-in userdb parser to resolve user and group names.
254 The PLACE= key is gone. It can be replaced by an ID= for a long time, cause
255 we walk up the chain of physical devices to find a match.
257 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
258 easy to skip certain attribute matches without composing rules with weird
259 character class negations like:
260 KERNEL="[!s][!c][!d]*"
261 this can now be replaced with:
263 The current simple '=' is still supported, and should work as it does today,
264 but existing rules should be converted if possible, to be better readable.
266 We have new ENV{}== key now, to match against a maximum of 5 environment
269 udevstart is its own binary again, cause we don't need co carry this araound
270 with every forked event.