5 Symlinks can have priorities now, the priority is assigned to the device
6 and specified with OPTIONS="link_priority=100". Devices with higher
7 priorities overwrite the symlinks of devices with lower priorities.
8 If the device that currently owns the link, goes away, the symlink
9 will be removed, and recreated, pointing to the next device with the
10 highest actual priority. This should make /dev/disk/by-{label,uuid,id}
11 more reliable, if multiple devices contain the same metadata and overwrite
14 The dasd_id program is removed from the udev tree, and dasdinfo, with the
15 needed rules, are part of the s390-tools now.
17 Please add KERNEL=="[0-9]*:[0-9]*" to the scsi wait-for-sysfs rule,
18 we may get the scsi sysfs mess fixed some day, and this will only catch
19 the devices we are looking for.
21 USB serial numbers for storage devices have the target:lun now appended,
22 to make it possibble to distinguish broken multi-lun devices with all
23 the same SCSI identifiers.
25 Note: The extra "run_directory" which searches and executes stuff in
26 /etc/hotplug.d/ and /etc/dev.d/ is long deprecated, and will be removed
27 with the next release. Make sure, that you don't use it anymore, or
28 provides your own implementation of that inefficient stuff.
29 We are tired of reports about a "slow udev", because these directories
30 contain stuff, that runs with _every_ event, instead of using rules,
31 that run programs only for the matching events.
41 DRIVER== will match only for devices that actually have a real
42 driver. DRIVERS== must be used, if parent devices should be
43 included in the match.
45 Libvolume_id's "linux_raid" detection needed another fix.
53 Add additional check to volume_id detection of via_raid, cause
54 some company decided to put a matching pattern all over the empty
55 storage area of their music players.
59 Fix path_id for SAS devices.
63 The udev daemon can be started with --debug-trace now, which will
64 execute all events serialized to get a chance to catch a possible
65 action that crashes the box.
67 A warning is logged, if PHYSDEV* keys, the "device" link, or a parent
68 device attribute like $attr{../file} is used, only WAIT_FOR_SYSFS rules
69 are excluded from the warning. Referencing parent attributes directly
70 may break when something in the kernel driver model changes. Udev will
71 just find the attribute by walking up the parent chain.
73 Udevtrigger now sorts the list of devices depending on the device
74 dependency, so a "usb" device is triggered after the parent "pci"
79 Revert persistent-storage ata-serial '_' '-' replacement.
85 Udevtrigger can now filter the list of devices to be triggered. Matches
86 for subsystems or sysfs attributes can be specified.
88 The entries in /dev/.udev/queue and /dev/.udev/failed have changed to
89 zero-sized files to avoid pointing to /sys and confuse broken tools which
90 scan the /dev directory. To retry failed events, udevtrigger --retry-failed
93 The rules and scripts to create udev rules for persistent network
94 devices and optical drives are in the extras/rules_generator directory
95 now. If you use something similar, please consider replacing your own
96 version with this, to share the support effort. The rule_generator
97 installs its own rules into /etc/udev/rules.d.
99 The cdrom_id tool installs its own rule now in /etc/udev/rules.d, cause
100 the rule_generator depends on cdrom_id to be called in an earlier rule.
106 Renaming of some key names (the old names still work):
107 BUS -> SUBSYSTEMS, ID -> KERNELS, SYSFS -> ATTRS, DRIVER -> DRIVERS.
108 (The behavior of the key DRIVER will change soon in one of the next
109 releases, to match only the event device, please switch to DRIVERS
110 instead. If DRIVER is used, it will behave like DRIVERS, but an error
112 With the new key names, we have a more consistent and simpler scheme.
113 We can match the properties of the event device only, with: KERNEL,
114 SUBSYSTEM, ATTR, DRIVER. Or include all the parent devices in the match,
115 with: KERNELS, SUBSYSTEMS, ATTRS, DRIVERS. ID, BUS, SYSFS, DRIVER are no
116 longer mentioned in the man page and should be switched in the rule
119 ATTR{file}="value" can be used now, to write to a sysfs file of the
120 event device. Instead of:
121 ..., SYSFS{type}=="0|7|14", RUN+="/bin/sh -c 'echo 60 > /sys$$DEVPATH/timeout'"
123 ..., ATTR{type}=="0|7|14", ATTR{timeout}="60"
125 All the PHYSDEV* keys are deprecated and will be removed from a
127 PHYDEVPATH - is the path of a parent device and should not be
129 PHYSDEVBUS - is just a SUBSYSTEM value of a parent, and can be
130 matched with SUBSYSTEMS==
131 PHYSDEVDRIVER - for bus devices it is available as ENV{DRIVER}.
132 Newer kernels will have DRIVER in the environment,
133 for older kernels udev puts in. Class device will
134 no longer carry this property of a parent and
135 DRIVERS== can be used to match such a parent value.
136 Note that ENV{DRIVER} is only available for a few bus devices, where
137 the driver is already bound at device event time. On coldplug, the
138 events for a lot devices are already bound to a driver, and they will have
139 that value set. But on hotplug, at the time the kernel creates the device,
140 it can't know what driver may claim the device after that, therefore
141 in most cases it will be empty.
143 Failed events should now be re-triggered with:
144 udevtrigger --retry-failed.
145 Please switch to this command, so we keep the details of the /dev/.udev/failed/
146 files private to the udev tools. We may need to switch the current symlink
147 target, cause some obviously broken tools try to scan all files in /dev
148 including /dev/.udev/, find the links to /sys and end up stat()'ing sysfs files
149 million times. This takes ages on slow boxes.
151 The udevinfo attribute walk (-a) now works with giving a device node
152 name (-n) instead of a devpath (-p). The query now always works, also when
153 no database file was created by udev.
155 The built-in /etc/passwd /etc/group parser is removed, we always depend on
156 getpwnam() and getgrnam() now. One of the next releases will depend on
157 fnmatch() and may use getopt_long().
161 Bugfixes and small improvements.
165 Fix path_id for recent kernels.
171 Added support for swapping network interface names, by temporarily
172 renaming the device and wait for the target name to become free.
176 The built-in MODALIAS key and substitution is removed.
180 The binary firmware helper is replaced by the usual simple
181 shell script. Udevsend is removed from the tree.
189 Some more keys require the correct use of '==' and '=' depending
190 on the kind of operation beeing an assignment or a match. Rules
191 with invalid operations are skipped and logged to syslog. Please
192 test with udevtest if the parsing of your rules throws errors and
193 fix possibly broken rules.
197 Provide "udevsettle" to wait for all current udev events to finish.
198 It also watches the current kernel netlink queue by comparing the
199 even sequence number to make sure that there are no current pending
200 events that have not already arrived in the daemon.
204 Fix rule to skip persistent rules for removable IDE devices, which
205 also skipped optical IDE drives.
207 All *_id program are installed in /lib/udev/ by default now.
209 No binary is stripped anymore as this should be done in the
210 packaging process and not at build time.
212 libvolume_id is provided as a shared library now and vol_id is
213 linked against it. Also one of the next HAL versions will require
214 this library, and the HAL build process will also require the
215 header file to be installed. The copy of the same code in HAL will
216 be removed to have only a single copy left on the system.
220 Add persistent links for SCSI tapes. The rules file is renamed
221 to 60-persistent-storage.rules.
223 Create persistent path for usb devices. Can be used for all sorts
224 of devices that can't be distinguished by other properties like
225 multiple identical keyboards and mice connected to the same box.
227 Provide "udevtrigger" program to request events on coldplug. The
228 shell script is much too slow with thousends of devices.
232 Fix persistent disk rules to exclude removable IDE drives.
234 Warn if %e, $modalias or MODALIAS is used.
238 Fix queue export, which wasn't correct for subsequent add/remove
239 events for the same device.
243 Fix cramfs detection on big endian.
245 Make WAIT_FOR_SYSFS usable in "normal" rules and silent if the whole
250 If BUS== and SYSFS{}== have been used in the same rule, the sysfs
251 attributes were only checked at the parent device that matched the
252 by BUS requested subsystem. Fix it to also look at the device we
253 received the event for.
255 Build variable CROSS has changed to CROSS_COMPILE to match the kernel
260 Fix a bug where NAME="" would prevent RUN from beeing executed.
262 RUN="/bin/program" does not longer automatically add the subsystem
263 as the first parameter. This is from the days of /sbin/hotplug
264 which is dead now and it's just confusing to need to add a space at
265 the end of the program name to prevent this.
266 If you use rules that need the subsystem as the first parameter,
267 like the old "udev_run_hotlugd" and "udev_run_devd", add the subsystem
268 to the key like RUN+="/bin/program $env{SUBSYSTEM}".
272 The udev man page has moved to udev(7) as it does not describe a command
273 anymore. The programs udev, udevstart and udevsend are no longer installed
274 by default and must be copied manually, if they should be installed or
275 included in a package.
277 Fix a bug where "ignore_device" could run earlier collected RUN keys before
278 the ignore rule was applied.
280 More preparation for future sysfs changes. usb_id and scsi_id no longer
281 depend on a magic order of devices in the /devices chain. Specific devices
282 should be requested by their subsytem.
284 This will always find the scsi parent device without depending on a specific
286 dev = sysfs_device_get(devpath);
287 dev_usb = sysfs_device_get_parent_with_subsystem(dev, "scsi");
289 The "device" link in the current sysfs layout will be automatically
290 _resolved_ as a parent and in the new sysfs layout it will just _be_ the
291 parent in the devpath. If a device is requested by it's symlink, like all
292 class devices in the new sysfs layout will look like, it gets automatically
293 resolved and substituted with the real devpath and not the symlink path.
296 A similar logic must be applied to _all_ sysfs users, including
297 scripts, that search along parent devices in sysfs. The explicit use of
298 the "device" link must be avoided. With the future sysfs layout all
299 DEVPATH's will start with /devices/ and have a "subsystem" symlink poiting
300 back to the "class" or the "bus". The layout of the parent devices in
301 /devices is not necessarily expected to be stable across kernel releases and
302 searching for parents by their subsystem should make sysfs users tolerant
303 for changed parent chains.
307 Prepare udev to work with the experimental kernel patch, that moves
308 /sys/class devices to /sys/devices and /sys/block to /sys/class/block.
310 Clarify BUS, ID, $id usage and fix $id behavior. This prepares for
311 moving the class devices to /sys/devices.
313 Thanks again to Marco for help finding a hopefully nice compromise
314 to make %b simpler and working again.
318 Complete removal of libsysfs, replaced by simple helper functions
319 which are much simpler and a bit faster. The udev daemon operatesentirely
320 on event parameters and does not use sysfs for simple rules anymore.
321 Please report any new bugs/problems, that may be caused by this big
322 change. They will be fixed immediately.
324 The enumeration format character '%e' is deprecated and will be
325 removed sometimes from a future udev version. It never worked correctly
326 outside of udevstart, so we can't use it with the new parallel
327 coldplug. A simple enumeration is as useless as the devfs naming
328 scheme, just get rid of both if you still use it.
330 MODALIAS and $modalias is not needed and will be removed from one of
331 the next udev versions, replace it in all rules with ENV{MODALIAS} or
332 the sysfs "modalias" value.
334 Thanks a lot to Marco for all his help on finding and fixing bugs.
338 Let scsi_id request libata drive serial numbers from page 0x80.
340 Renamed etc/udev/persistent.rules to persistent-disk.rules and
341 added /dev/disk/by-name/* for device mapper device names.
343 Removed %e from the man page. It never worked reliably outside
344 of udevstart and udevstart is no longer recommended to use.
348 Symlinks are now exported to the event environment. Hopefully it's no
349 longer needed to run udevinfo from an event process, like it was
350 mentioned on the hotplug list:
351 UDEV [1134776873.702967] add@/block/sdb
354 DEVLINKS=/dev/disk/by-id/usb-IBM_Memory_Key_0218B301030027E8 /dev/disk/by-path/usb-0218B301030027E8:0:0:0
358 Fix a problem if udevsend is used as the hotplug handler and tries to use
359 syslog, which causes a "vc" event loop. 2.6.15 will make udevsend obsolete
360 and this kind of problems will hopefully go away soon.
364 All built-in logic to work around bad sysfs timing is removed with this
365 version. The need to wait for sysfs files is almost fixed with a kernel
366 version that doesn't work with this udev version anyway. Until we fix
367 the timing of the "bus" link creation, the former integrated logic should
368 be emulated by a rule placed before all other rules:
369 ACTION=="add", DEVPATH=="/devices/*", ENV{PHYSDEVBUS}=="?*", WAIT_FOR_SYSFS="bus"
371 The option "udev_db" does no longer exist. All udev state will be in
372 /$udev_root/.udev/ now, there is no longer an option to set this
374 If the init script or something else used this value, just depend on
375 this hardcoded path. But remember _all_content_ of this directory is
376 still private to udev and can change at any time.
378 Default location for rule sripts and helper programs is now: /lib/udev/.
379 Everything that is not useful on the commandline should go into this
380 directory. Some of the helpers in the extras folder are installed there
381 now. The rules need to be changed, to find the helpers there.
383 Also /lib/udev/devices is recommended as a directory where packages or
384 the user can place real device nodes, which get copied over to /dev at
385 every boot. This should replace the various solutions with custom config
388 Udevsend does no longer start the udev daemon. This must be done with
389 the init script that prepares /dev on tmpfs and creates the initial nodes,
390 before starting the daemon.
394 Silent a too verbose error logging for the old hotplug.d/ dev.d/
397 The copy of klibc is removed. A systemwide installed version of klibc
398 should be used to build a klibc udev now.
402 NAME="" will not create any nodes, but execute RUN keys. To completely
403 ignore an event the OPTION "ignore_device" should be used.
405 After removal of the reorder queue, events with a TIMEOUT can be executed
406 without any queuing now.
410 Fixed bug in udevd, if inotify is not available. We depend on netlink
411 uevents now, kernels without that event source will not work with that
412 version of udev anymore.
416 The rule parsing happens now in the daemon once at startup, all udev
417 event processes inherit the already parsed rules from the daemon.
418 It is shipped with SUSE10.0 and reduces heavily the system load at
419 startup. The option to save precompiled rules and let the udev process
420 pick the them up is removed, as it's no longer needed.
422 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
423 device. Libsysfs is changed to "translate" the requested link into the
424 real device path, as it would happen with the hotplug event. Otherwise
425 device removal and the udev database will not work.
427 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
430 A few improvements for vol_id, the filesytem probing code.
434 Fix a stupid typo in extras/run_directory for "make install".
436 scsi_id creates the temporary devnode now in /dev for usage with a
437 non-writable /tmp directory.
439 The uevent kernel socket buffer can carry app. 50.000 events now,
440 let's see who can break this again. :)
442 The upcoming kernel will have a new input driver core integration.
443 Some class devices are now symlinks to the real device. libsysfs
444 needs a fix for this to work correctly. Udevstart of older udev
445 versions will _not_ create these devices!
449 Fix a 'install' target in the Makefile, that prevents EXTRAS from
454 A bunch of mostly trivial bugfixes. From now on no node name or
455 symlink name can contain any character than plain whitelisted ascii
456 characters or validated utf8 byte-streams. This is needed for the
457 /dev/disk/by-label/* links, because we import untrusted data and
458 export it to the filesystem.
462 More bugfixes. If udevd was started from the kernel, we don't
463 have stdin/stdout/stderr, which broke the forked tools in some
468 Bugfix. udevstart event ordering was broken for a long time.
469 The new run_program() uncovered it, because /dev/null was not
470 available while we try to run external programs.
471 Now udevstart should create it before we run anything.
475 Minor bugfixes and some distro rules updates. If you don't have the
476 persistent disk rules in /dev/disk/by-*/* on your distro, just
477 grab it from here. :)
481 We can use socket communication now to pass events from udev to
483 RUN+="socket:/org/freedesktop/hal/udev_event"
484 will pass the whole udev event to the HAL daemon without the need
485 for a forked helper. (See ChangeLog for udevmonitor, as an example)
489 Mostly bugfixes and see ChangeLog.
491 The test for the existence of an environment value should be
493 ENV{KEY}=="*" to ENV{KEY}=="?*"
494 because "*" will not fail anymore, if the key does not exist or
499 Bugfixes and a few tweaks described in the ChangeLog.
503 Mostly a Bugfix release.
505 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
506 timing with custom rules.
510 We changed the internal rule storage format. Our large rule files took
511 2 MB of RAM, with the change we are down to 99kB.
513 If the device-node has been created with default name and no symlink or
514 options are to remenber, it is not longer stored in the udevdb. HAL will
515 need to be updated to work correctly with that change.
517 To overrride optimization flags, OPTFLAGS may be used now.
525 Major changes happened with this release. The goal is to take over the
526 complete kernel-event handling and provide a more efficient way to dispatch
527 kernel events. Replacing most of the current shell script logic and the
528 kernel forked helper with a netlink-daemon and a rule-based event handling.
530 o udevd listens to netlink events now. The first valid netlink event
531 will make udevd ignore any message from udevsend that contains a
532 SEQNUM, to avoid duplicate events. The forked events can be disabled
534 echo "" > /proc/sys/kernel/hotplug
535 For full support, the broken input-subsytem needs to be fixed, not to
536 bypass the driver core.
538 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
539 removed from udev itself and must be emulated by calling small
540 helper binaries provided in the extras folder:
541 make EXTRAS=extras/run_directory/
542 will build udev_run_devd and udev_run_hotplugd, which can be called
543 from a rule if needed:
544 RUN+="/sbin/udev_run_hotplugd"
545 The recommended way to handle this is to convert all the calls from
546 the directories to explicit udev rules and get completely rid of the
547 multiplexing. (To catch a ttyUSB event, you now no longer need to
548 fork and exit 300 tty script instances you are not interested in, it
549 is just one rule that matches exactly the device.)
551 o udev handles now _all_ events not just events for class and block
552 devices, this way it is possible to control the complete event
553 behavior with udev rules. Especially useful for rules like:
554 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
556 o As used in the modalias rule, udev supports now textual
557 substitution placeholder along with the usual format chars. This
558 needs to be documented, for now it's only visible in udev_rules_parse.c.
560 o The rule keys support now more operations. This is documented in the
561 man page. It is possible to add values to list-keys like the SYMLINK
562 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
563 Also "final"-assignments are supported by using KEY:="value", which will
564 prevent changing the key by any later rule.
566 o kernel 2.6.12 has the "detached_state" attribute removed from
567 sysfs, which was used to recognize sysfs population. We switched that
568 to wait for the "bus" link, which is only available in kernels after 2.6.11.
569 Running this udev version on older kernels may cause a short delay for
572 o To provide infrastructure for persistent device naming, the id programs:
573 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
574 to export the probed data in environment key format:
575 pim:~ # /sbin/ata_id --export /dev/hda
576 ID_MODEL=HTS726060M9AT00
577 ID_SERIAL=MRH401M4G6UM9B
581 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
582 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
585 kay@pim:~> tree /dev/disk
588 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
589 | `-- IBM-Memory_Key -> ../../sda
591 | |-- swap -> ../../hda1
592 | |-- date -> ../../sda1
593 | `-- home -> ../../hda3
595 |-- 2E08712B0870F2E7 -> ../../hda3
596 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
597 |-- E845-7A89 -> ../../sda1
598 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
600 The IMPORT= operation will import these keys in the environment and make
601 it available for later PROGRAM= and RUN= executed programs. The keys are
602 also stored in the udevdb and can be queried from there with one of the
605 o A few binaries are silently added to the repository, which can be used
606 to replay kernel events from initramfs instead of using coldplug. udevd
607 can be instructed now to queue-up events while the stored events from
608 initramfs are filled into the udevd-queue. This code is still under
609 development and there is no documentation now besides the code itself.
610 The additional binaries get compiled, but are not installed by default.
612 o There is also a temporary fix for a performance problem where too many
613 events happen in parallel and every event needs to parse the rules.
614 udev can now read precompiled rules stored on disk. This is likely to be
615 replaced by a more elegant solution in a future udev version.
619 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
620 Fix for libsysfs not to expect this file was added.
624 All rules are applied now, but only the first matching rule with a NAME-key
625 will be applied. All later rules with NAME-key are completely ignored. This
626 way system supplied symlinks or permissions gets applied to user-defined
630 Please check your rules setup, if you may need to add OPTIONS="last_rule"
631 to some rules, to keep the old behavior.
633 The rules are read on "remove"-events too. That makes is possible to match
634 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
635 instruct udev to ignore an event (OPTIONS="ignore_device").
636 The new ACTION-key may be used to let a rule act only at a "remove"-event.
638 The new RUN-key supports rule-based execution of programs after device-node
639 handling. This is meant as a general replacement for the dev.d/-directories
640 to give fine grained control over the execution of programs.
642 The %s{}-sysfs format char replacement values are searched at any of the
643 devices in the device chain now, not only at the class-device.
645 We support log priority levels now. The value udev_log in udev.conf is used
646 to determine what is printed to syslog. This makes it possible to
647 run a version with compiled-in debug messages in a production environment
648 which is sometimes needed to find a bug.
649 It is still possible to supress the inclusion of _any_ syslog usage with
650 USE_LOG=false to create the smallest possible binaries if needed.
651 The configured udev_log value can be overridden with the environment variable
656 Possible use of a system-wide klibc:
657 make USE_KLIBC=true KLCC=/usr/bin/klcc all
658 will link against an external klibc and our own version will be ignored.
662 We support an unlimited count of symlinks now.
664 If USE_STATIC=true is passed to a glibc build, we link statically and use
665 a built-in userdb parser to resolve user and group names.
667 The PLACE= key is gone. It can be replaced by an ID= for a long time, because
668 we walk up the chain of physical devices to find a match.
670 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
671 easy to skip certain attribute matches without composing rules with weird
672 character class negations like:
673 KERNEL="[!s][!c][!d]*"
674 this can now be replaced with:
676 The current simple '=' is still supported, and should work as it does today,
677 but existing rules should be converted if possible, to be better readable.
679 We have new ENV{}== key now, to match against a maximum of 5 environment
682 udevstart is its own binary again, because we don't need co carry this araound
683 with every forked event.