3 Prepare udev to work with the experimental kernel patch, that moves
4 /sys/class devices to /sys/devices and /sys/block to /sys/class/block.
6 Clarify BUS, ID, $id usage and fix $id behavior. This prepares for
7 moving the class devices to /sys/devices.
9 Thanks again to Marco for help finding a hopefully nice compromise
10 to make %b simpler and working again.
14 Complete removal of libsysfs, replaced by simple helper functions
15 which are much simpler and a bit faster. The udev daemon operatesentirely
16 on event parameters and does not use sysfs for simple rules anymore.
17 Please report any new bugs/problems, that may be caused by this big
18 change. They will be fixed immediately.
20 The enumeration format character '%e' is deprecated and will be
21 removed sometimes from a future udev version. It never worked correctly
22 outside of udevstart, so we can't use it with the new parallel
23 coldplug. A simple enumeration is as useless as the devfs naming
24 scheme, just get rid of both if you still use it.
26 MODALIAS and $modalias is not needed and will be removed from one of
27 the next udev versions, replace it in all rules with ENV{MODALIAS} or
28 the sysfs "modalias" value.
30 Thanks a lot to Marco for all his help on finding and fixing bugs.
34 Let scsi_id request libata drive serial numbers from page 0x80.
36 Renamed etc/udev/persistent.rules to persistent-disk.rules and
37 added /dev/disk/by-name/* for device mapper device names.
39 Removed %e from the man page. It never worked reliably outside
40 of udevstart and udevstart is no longer recommended to use.
44 Symlinks are now exported to the event environment. Hopefully it's no
45 longer needed to run udevinfo from an event process, like it was
46 mentioned on the hotplug list:
47 UDEV [1134776873.702967] add@/block/sdb
50 DEVLINKS=/dev/disk/by-id/usb-IBM_Memory_Key_0218B301030027E8 /dev/disk/by-path/usb-0218B301030027E8:0:0:0
54 Fix a problem if udevsend is used as the hotplug handler and tries to use
55 syslog, which causes a "vc" event loop. 2.6.15 will make udevsend obsolete
56 and this kind of problems will hopefully go away soon.
60 All built-in logic to work around bad sysfs timing is removed with this
61 version. The need to wait for sysfs files is almost fixed with a kernel
62 version that doesn't work with this udev version anyway. Until we fix
63 the timing of the "bus" link creation, the former integrated logic should
64 be emulated by a rule placed before all other rules:
65 ACTION=="add", DEVPATH=="/devices/*", ENV{PHYSDEVBUS}=="?*", WAIT_FOR_SYSFS="bus"
67 The option "udev_db" does no longer exist. All udev state will be in
68 /$udev_root/.udev/ now, there is no longer an option to set this
70 If the init script or something else used this value, just depend on
71 this hardcoded path. But remember _all_content_ of this directory is
72 still private to udev and can change at any time.
74 Default location for rule sripts and helper programs is now: /lib/udev/.
75 Everything that is not useful on the commandline should go into this
76 directory. Some of the helpers in the extras folder are installed there
77 now. The rules need to be changed, to find the helpers there.
79 Also /lib/udev/devices is recommended as a directory where packages or
80 the user can place real device nodes, which get copied over to /dev at
81 every boot. This should replace the various solutions with custom config
84 Udevsend does no longer start the udev daemon. This must be done with
85 the init script that prepares /dev on tmpfs and creates the initial nodes,
86 before starting the daemon.
90 Silent a too verbose error logging for the old hotplug.d/ dev.d/
93 The copy of klibc is removed. A systemwide installed version of klibc
94 should be used to build a klibc udev now.
98 NAME="" will not create any nodes, but execute RUN keys. To completely
99 ignore an event the OPTION "ignore_device" should be used.
101 After removal of the reorder queue, events with a TIMEOUT can be executed
102 without any queuing now.
106 Fixed bug in udevd, if inotify is not available. We depend on netlink
107 uevents now, kernels without that event source will not work with that
108 version of udev anymore.
112 The rule parsing happens now in the daemon once at startup, all udev
113 event processes inherit the already parsed rules from the daemon.
114 It is shipped with SUSE10.0 and reduces heavily the system load at
115 startup. The option to save precompiled rules and let the udev process
116 pick the them up is removed, as it's no longer needed.
118 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
119 device. Libsysfs is changed to "translate" the requested link into the
120 real device path, as it would happen with the hotplug event. Otherwise
121 device removal and the udev database will not work.
123 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
126 A few improvements for vol_id, the filesytem probing code.
130 Fix a stupid typo in extras/run_directory for "make install".
132 scsi_id creates the temporary devnode now in /dev for usage with a
133 non-writable /tmp directory.
135 The uevent kernel socket buffer can carry app. 50.000 events now,
136 let's see who can break this again. :)
138 The upcoming kernel will have a new input driver core integration.
139 Some class devices are now symlinks to the real device. libsysfs
140 needs a fix for this to work correctly. Udevstart of older udev
141 versions will _not_ create these devices!
145 Fix a 'install' target in the Makefile, that prevents EXTRAS from
150 A bunch of mostly trivial bugfixes. From now on no node name or
151 symlink name can contain any character than plain whitelisted ascii
152 characters or validated utf8 byte-streams. This is needed for the
153 /dev/disk/by-label/* links, cause we import untrusted data and
154 export it to the filesystem.
158 More bugfixes. If udevd was started from the kernel, we don't
159 have stdin/stdout/stderr, which broke the forked tools in some
164 Bugfix. udevstart event ordering was broken for a long time.
165 The new run_program() uncovered it, cause /dev/null was not
166 available while we try to run external programs.
167 Now udevstart should create it before we run anything.
171 Minor bugfixes and some distro rules updates. If you don't have the
172 persistent disk rules in /dev/disk/by-*/* on your distro, just
173 grab it from here. :)
177 We can use socket communication now to pass events from udev to
179 RUN+="socket:/org/freedesktop/hal/udev_event"
180 will pass the whole udev event to the HAL daemon without the need
181 for a forked helper. (See ChangeLog for udevmonitor, as an example)
185 Mostly bugfixes and see ChangeLog.
187 The test for the existence of an environment value should be
189 ENV{KEY}=="*" to ENV{KEY}=="?*"
190 cause "*" will not fail anymore, if the key does not exist or
195 Bugfixes and a few tweaks described in the ChangeLog.
199 Mostly a Bugfix release.
201 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
202 timing with custom rules.
206 We changed the internal rule storage format. Our large rule files took
207 2 MB of RAM, with the change we are down to 99kB.
209 If the device-node has been created with default name and no symlink or
210 options are to remenber, it is not longer stored in the udevdb. HAL will
211 need to be updated to work correctly with that change.
213 To overrride optimization flags, OPTFLAGS may be used now.
221 Major changes happened with this release. The goal is to take over the
222 complete kernel-event handling and provide a more efficient way to dispatch
223 kernel events. Replacing most of the current shell script logic and the
224 kernel forked helper with a netlink-daemon and a rule-based event handling.
226 o udevd listens to netlink events now. The first valid netlink event
227 will make udevd ignore any message from udevsend that contains a
228 SEQNUM, to avoid duplicate events. The forked events can be disabled
230 echo "" > /proc/sys/kernel/hotplug
231 For full support, the broken input-subsytem needs to be fixed, not to
232 bypass the driver core.
234 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
235 removed from udev itself and must be emulated by calling small
236 helper binaries provided in the extras folder:
237 make EXTRAS=extras/run_directory/
238 will build udev_run_devd and udev_run_hotplugd, which can be called
239 from a rule if needed:
240 RUN+="/sbin/udev_run_hotplugd"
241 The recommended way to handle this is to convert all the calls from
242 the directories to explicit udev rules and get completely rid of the
243 multiplexing. (To catch a ttyUSB event, you now no longer need to
244 fork and exit 300 tty script instances you are not interested in, it
245 is just one rule that matches exactly the device.)
247 o udev handles now _all_ events not just events for class and block
248 devices, this way it is possible to control the complete event
249 behavior with udev rules. Especially useful for rules like:
250 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
252 o As used in the modalias rule, udev supports now textual
253 substitution placeholder along with the usual format chars. This
254 needs to be documented, for now it's only visible in udev_rules_parse.c.
256 o The rule keys support now more operations. This is documented in the
257 man page. It is possible to add values to list-keys like the SYMLINK
258 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
259 Also "final"-assignments are supported by using KEY:="value", which will
260 prevent changing the key by any later rule.
262 o kernel 2.6.12 has the "detached_state" attribute removed from
263 sysfs, which was used to recognize sysfs population. We switched that
264 to wait for the "bus" link, which is only available in kernels after 2.6.11.
265 Running this udev version on older kernels may cause a short delay for
268 o To provide infrastructure for persistent device naming, the id programs:
269 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
270 to export the probed data in environment key format:
271 pim:~ # /sbin/ata_id --export /dev/hda
272 ID_MODEL=HTS726060M9AT00
273 ID_SERIAL=MRH401M4G6UM9B
277 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
278 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
281 kay@pim:~> tree /dev/disk
284 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
285 | `-- IBM-Memory_Key -> ../../sda
287 | |-- swap -> ../../hda1
288 | |-- date -> ../../sda1
289 | `-- home -> ../../hda3
291 |-- 2E08712B0870F2E7 -> ../../hda3
292 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
293 |-- E845-7A89 -> ../../sda1
294 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
296 The IMPORT= operation will import these keys in the environment and make
297 it available for later PROGRAM= and RUN= executed programs. The keys are
298 also stored in the udevdb and can be queried from there with one of the
301 o A few binaries are silently added to the repository, which can be used
302 to replay kernel events from initramfs instead of using coldplug. udevd
303 can be instructed now to queue-up events while the stored events from
304 initramfs are filled into the udevd-queue. This code is still under
305 development and there is no documentation now besides the code itself.
306 The additional binaries get compiled, but are not installed by default.
308 o There is also a temporary fix for a performance problem where too many
309 events happen in parallel and every event needs to parse the rules.
310 udev can now read precompiled rules stored on disk. This is likely to be
311 replaced by a more elegant solution in a future udev version.
315 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
316 Fix for libsysfs not to expect this file was added.
320 All rules are applied now, but only the first matching rule with a NAME-key
321 will be applied. All later rules with NAME-key are completely ignored. This
322 way system supplied symlinks or permissions gets applied to user-defined
326 Please check your rules setup, if you may need to add OPTIONS="last_rule"
327 to some rules, to keep the old behavior.
329 The rules are read on "remove"-events too. That makes is possible to match
330 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
331 instruct udev to ignore an event (OPTIONS="ignore_device").
332 The new ACTION-key may be used to let a rule act only at a "remove"-event.
334 The new RUN-key supports rule-based execution of programs after device-node
335 handling. This is meant as a general replacement for the dev.d/-directories
336 to give fine grained control over the execution of programs.
338 The %s{}-sysfs format char replacement values are searched at any of the
339 devices in the device chain now, not only at the class-device.
341 We support log priority levels now. The value udev_log in udev.conf is used
342 to determine what is printed to syslog. This makes it possible to
343 run a version with compiled-in debug messages in a production environment
344 which is sometimes needed to find a bug.
345 It is still possible to supress the inclusion of _any_ syslog usage with
346 USE_LOG=false to create the smallest possible binaries if needed.
347 The configured udev_log value can be overridden with the environment variable
352 Possible use of a system-wide klibc:
353 make USE_KLIBC=true KLCC=/usr/bin/klcc all
354 will link against an external klibc and our own version will be ignored.
358 We support an unlimited count of symlinks now.
360 If USE_STATIC=true is passed to a glibc build, we link statically and use
361 a built-in userdb parser to resolve user and group names.
363 The PLACE= key is gone. It can be replaced by an ID= for a long time, cause
364 we walk up the chain of physical devices to find a match.
366 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
367 easy to skip certain attribute matches without composing rules with weird
368 character class negations like:
369 KERNEL="[!s][!c][!d]*"
370 this can now be replaced with:
372 The current simple '=' is still supported, and should work as it does today,
373 but existing rules should be converted if possible, to be better readable.
375 We have new ENV{}== key now, to match against a maximum of 5 environment
378 udevstart is its own binary again, cause we don't need co carry this araound
379 with every forked event.