3 Fix a problem if udevsend is used as the hotplug handler and tries to use
4 syslog, which causes a "vc" event loop. 2.6.15 will make udevsend obsolete
5 and this kind of problems will hopefully go away soon.
9 All built-in logic to work around bad sysfs timing is removed with this
10 version. The need to wait for sysfs files is almost fixed with a kernel
11 version that doesn't work with this udev version anyway. Until we fix
12 the timing of the "bus" link creation, the former integrated logic should
13 be emulated by a rule placed before all other rules:
14 ACTION=="add", DEVPATH=="/devices/*", ENV{PHYSDEVBUS}=="?*", WAIT_FOR_SYSFS="bus"
16 The option "udev_db" does no longer exist. All udev state will be in
17 /$udev_root/.udev/ now, there is no longer an option to set this
19 If the init script or something else used this value, just depend on
20 this hardcoded path. But remember _all_content_ of this directory is
21 still private to udev and can change at any time.
23 Default location for rule sripts and helper programs is now: /lib/udev/.
24 Everything that is not useful on the commandline should go into this
25 directory. Some of the helpers in the extras folder are installed there
26 now. The rules need to be changed, to find the helpers there.
28 Also /lib/udev/devices is recommended as a directory where packages or
29 the user can place real device nodes, which get copied over to /dev at
30 every boot. This should replace the various solutions with custom config
33 Udevsend does no longer start the udev daemon. This must be done with
34 the init script that prepares /dev on tmpfs and creates the initial nodes,
35 before starting the daemon.
39 Silent a too verbose error logging for the old hotplug.d/ dev.d/
42 The copy of klibc is removed. A systemwide installed version of klibc
43 should be used to build a klibc udev now.
47 NAME="" will not create any nodes, but execute RUN keys. To completely
48 ignore an event the OPTION "ignore_device" should be used.
50 After removal of the reorder queue, events with a TIMEOUT can be executed
51 without any queuing now.
55 Fixed bug in udevd, if inotify is not available. We depend on netlink
56 uevents now, kernels without that event source will not work with that
57 version of udev anymore.
61 The rule parsing happens now in the daemon once at startup, all udev
62 event processes inherit the already parsed rules from the daemon.
63 It is shipped with SUSE10.0 and reduces heavily the system load at
64 startup. The option to save precompiled rules and let the udev process
65 pick the them up is removed, as it's no longer needed.
67 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
68 device. Libsysfs is changed to "translate" the requested link into the
69 real device path, as it would happen with the hotplug event. Otherwise
70 device removal and the udev database will not work.
72 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
75 A few improvements for vol_id, the filesytem probing code.
79 Fix a stupid typo in extras/run_directory for "make install".
81 scsi_id creates the temporary devnode now in /dev for usage with a
82 non-writable /tmp directory.
84 The uevent kernel socket buffer can carry app. 50.000 events now,
85 let's see who can break this again. :)
87 The upcoming kernel will have a new input driver core integration.
88 Some class devices are now symlinks to the real device. libsysfs
89 needs a fix for this to work correctly. Udevstart of older udev
90 versions will _not_ create these devices!
94 Fix a 'install' target in the Makefile, that prevents EXTRAS from
99 A bunch of mostly trivial bugfixes. From now on no node name or
100 symlink name can contain any character than plain whitelisted ascii
101 characters or validated utf8 byte-streams. This is needed for the
102 /dev/disk/by-label/* links, cause we import untrusted data and
103 export it to the filesystem.
107 More bugfixes. If udevd was started from the kernel, we don't
108 have stdin/stdout/stderr, which broke the forked tools in some
113 Bugfix. udevstart event ordering was broken for a long time.
114 The new run_program() uncovered it, cause /dev/null was not
115 available while we try to run external programs.
116 Now udevstart should create it before we run anything.
120 Minor bugfixes and some distro rules updates. If you don't have the
121 persistent disk rules in /dev/disk/by-*/* on your distro, just
122 grab it from here. :)
126 We can use socket communication now to pass events from udev to
128 RUN+="socket:/org/freedesktop/hal/udev_event"
129 will pass the whole udev event to the HAL daemon without the need
130 for a forked helper. (See ChangeLog for udevmonitor, as an example)
134 Mostly bugfixes and see ChangeLog.
136 The test for the existence of an environment value should be
138 ENV{KEY}=="*" to ENV{KEY}=="?*"
139 cause "*" will not fail anymore, if the key does not exist or
144 Bugfixes and a few tweaks described in the ChangeLog.
148 Mostly a Bugfix release.
150 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
151 timing with custom rules.
155 We changed the internal rule storage format. Our large rule files took
156 2 MB of RAM, with the change we are down to 99kB.
158 If the device-node has been created with default name and no symlink or
159 options are to remenber, it is not longer stored in the udevdb. HAL will
160 need to be updated to work correctly with that change.
162 To overrride optimization flags, OPTFLAGS may be used now.
170 Major changes happened with this release. The goal is to take over the
171 complete kernel-event handling and provide a more efficient way to dispatch
172 kernel events. Replacing most of the current shell script logic and the
173 kernel forked helper with a netlink-daemon and a rule-based event handling.
175 o udevd listens to netlink events now. The first valid netlink event
176 will make udevd ignore any message from udevsend that contains a
177 SEQNUM, to avoid duplicate events. The forked events can be disabled
179 echo "" > /proc/sys/kernel/hotplug
180 For full support, the broken input-subsytem needs to be fixed, not to
181 bypass the driver core.
183 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
184 removed from udev itself and must be emulated by calling small
185 helper binaries provided in the extras folder:
186 make EXTRAS=extras/run_directory/
187 will build udev_run_devd and udev_run_hotplugd, which can be called
188 from a rule if needed:
189 RUN+="/sbin/udev_run_hotplugd"
190 The recommended way to handle this is to convert all the calls from
191 the directories to explicit udev rules and get completely rid of the
192 multiplexing. (To catch a ttyUSB event, you now no longer need to
193 fork and exit 300 tty script instances you are not interested in, it
194 is just one rule that matches exactly the device.)
196 o udev handles now _all_ events not just events for class and block
197 devices, this way it is possible to control the complete event
198 behavior with udev rules. Especially useful for rules like:
199 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
201 o As used in the modalias rule, udev supports now textual
202 substitution placeholder along with the usual format chars. This
203 needs to be documented, for now it's only visible in udev_rules_parse.c.
205 o The rule keys support now more operations. This is documented in the
206 man page. It is possible to add values to list-keys like the SYMLINK
207 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
208 Also "final"-assignments are supported by using KEY:="value", which will
209 prevent changing the key by any later rule.
211 o kernel 2.6.12 has the "detached_state" attribute removed from
212 sysfs, which was used to recognize sysfs population. We switched that
213 to wait for the "bus" link, which is only available in kernels after 2.6.11.
214 Running this udev version on older kernels may cause a short delay for
217 o To provide infrastructure for persistent device naming, the id programs:
218 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
219 to export the probed data in environment key format:
220 pim:~ # /sbin/ata_id --export /dev/hda
221 ID_MODEL=HTS726060M9AT00
222 ID_SERIAL=MRH401M4G6UM9B
226 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
227 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
230 kay@pim:~> tree /dev/disk
233 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
234 | `-- IBM-Memory_Key -> ../../sda
236 | |-- swap -> ../../hda1
237 | |-- date -> ../../sda1
238 | `-- home -> ../../hda3
240 |-- 2E08712B0870F2E7 -> ../../hda3
241 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
242 |-- E845-7A89 -> ../../sda1
243 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
245 The IMPORT= operation will import these keys in the environment and make
246 it available for later PROGRAM= and RUN= executed programs. The keys are
247 also stored in the udevdb and can be queried from there with one of the
250 o A few binaries are silently added to the repository, which can be used
251 to replay kernel events from initramfs instead of using coldplug. udevd
252 can be instructed now to queue-up events while the stored events from
253 initramfs are filled into the udevd-queue. This code is still under
254 development and there is no documentation now besides the code itself.
255 The additional binaries get compiled, but are not installed by default.
257 o There is also a temporary fix for a performance problem where too many
258 events happen in parallel and every event needs to parse the rules.
259 udev can now read precompiled rules stored on disk. This is likely to be
260 replaced by a more elegant solution in a future udev version.
264 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
265 Fix for libsysfs not to expect this file was added.
269 All rules are applied now, but only the first matching rule with a NAME-key
270 will be applied. All later rules with NAME-key are completely ignored. This
271 way system supplied symlinks or permissions gets applied to user-defined
275 Please check your rules setup, if you may need to add OPTIONS="last_rule"
276 to some rules, to keep the old behavior.
278 The rules are read on "remove"-events too. That makes is possible to match
279 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
280 instruct udev to ignore an event (OPTIONS="ignore_device").
281 The new ACTION-key may be used to let a rule act only at a "remove"-event.
283 The new RUN-key supports rule-based execution of programs after device-node
284 handling. This is meant as a general replacement for the dev.d/-directories
285 to give fine grained control over the execution of programs.
287 The %s{}-sysfs format char replacement values are searched at any of the
288 devices in the device chain now, not only at the class-device.
290 We support log priority levels now. The value udev_log in udev.conf is used
291 to determine what is printed to syslog. This makes it possible to
292 run a version with compiled-in debug messages in a production environment
293 which is sometimes needed to find a bug.
294 It is still possible to supress the inclusion of _any_ syslog usage with
295 USE_LOG=false to create the smallest possible binaries if needed.
296 The configured udev_log value can be overridden with the environment variable
301 Possible use of a system-wide klibc:
302 make USE_KLIBC=true KLCC=/usr/bin/klcc all
303 will link against an external klibc and our own version will be ignored.
307 We support an unlimited count of symlinks now.
309 If USE_STATIC=true is passed to a glibc build, we link statically and use
310 a built-in userdb parser to resolve user and group names.
312 The PLACE= key is gone. It can be replaced by an ID= for a long time, cause
313 we walk up the chain of physical devices to find a match.
315 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
316 easy to skip certain attribute matches without composing rules with weird
317 character class negations like:
318 KERNEL="[!s][!c][!d]*"
319 this can now be replaced with:
321 The current simple '=' is still supported, and should work as it does today,
322 but existing rules should be converted if possible, to be better readable.
324 We have new ENV{}== key now, to match against a maximum of 5 environment
327 udevstart is its own binary again, cause we don't need co carry this araound
328 with every forked event.