3 Add additional check to volume_id detection of via_raid, cause
4 some company decided to put a matching pattern all over the empty
5 storage area of their music players.
9 Fix path_id for SAS devices.
13 The udev daemon can be started with --debug-trace now, which will
14 execute all events serialized to get a chance to catch a possible
15 action that crashes the box.
17 A warning is logged, if PHYSDEV* keys, the "device" link, or a parent
18 device attribute like $attr{../file} is used, only WAIT_FOR_SYSFS rules
19 are excluded from the warning. Referencing parent attributes directly
20 may break when something in the kernel driver model changes. Udev will
21 just find the attribute by walking up the parent chain.
23 Udevtrigger now sorts the list of devices depending on the device
24 dependency, so a "usb" device is triggered after the parent "pci"
29 Revert persistent-storage ata-serial '_' '-' replacement.
35 Udevtrigger can now filter the list of devices to be triggered. Matches
36 for subsystems or sysfs attributes can be specified.
38 The entries in /dev/.udev/queue and /dev/.udev/failed have changed to
39 zero-sized files to avoid pointing to /sys and confuse broken tools which
40 scan the /dev directory. To retry failed events, udevtrigger --retry-failed
43 The rules and scripts to create udev rules for persistent network
44 devices and optical drives are in the extras/rules_generator directory
45 now. If you use something similar, please consider replacing your own
46 version with this, to share the support effort. The rule_generator
47 installs its own rules into /etc/udev/rules.d.
49 The cdrom_id tool installs its own rule now in /etc/udev/rules.d, cause
50 the rule_generator depends on cdrom_id to be called in an earlier rule.
56 Renaming of some key names (the old names still work):
57 BUS -> SUBSYSTEMS, ID -> KERNELS, SYSFS -> ATTRS, DRIVER -> DRIVERS.
58 (The behavior of the key DRIVER will change soon in one of the next
59 releases, to match only the event device, please switch to DRIVERS
60 instead. If DRIVER is used, it will behave like DRIVERS, but an error
62 With the new key names, we have a more consistent and simpler scheme.
63 We can match the properties of the event device only, with: KERNEL,
64 SUBSYSTEM, ATTR, DRIVER. Or include all the parent devices in the match,
65 with: KERNELS, SUBSYSTEMS, ATTRS, DRIVERS. ID, BUS, SYSFS, DRIVER are no
66 longer mentioned in the man page and should be switched in the rule
69 ATTR{file}="value" can be used now, to write to a sysfs file of the
70 event device. Instead of:
71 ..., SYSFS{type}=="0|7|14", RUN+="/bin/sh -c 'echo 60 > /sys$$DEVPATH/timeout'"
73 ..., ATTR{type}=="0|7|14", ATTR{timeout}="60"
75 All the PHYSDEV* keys are deprecated and will be removed from a
77 PHYDEVPATH - is the path of a parent device and should not be
79 PHYSDEVBUS - is just a SUBSYSTEM value of a parent, and can be
80 matched with SUBSYSTEMS==
81 PHYSDEVDRIVER - for bus devices it is available as ENV{DRIVER}.
82 Newer kernels will have DRIVER in the environment,
83 for older kernels udev puts in. Class device will
84 no longer carry this property of a parent and
85 DRIVERS== can be used to match such a parent value.
86 Note that ENV{DRIVER} is only available for a few bus devices, where
87 the driver is already bound at device event time. On coldplug, the
88 events for a lot devices are already bound to a driver, and they will have
89 that value set. But on hotplug, at the time the kernel creates the device,
90 it can't know what driver may claim the device after that, therefore
91 in most cases it will be empty.
93 Failed events should now be re-triggered with:
94 udevtrigger --retry-failed.
95 Please switch to this command, so we keep the details of the /dev/.udev/failed/
96 files private to the udev tools. We may need to switch the current symlink
97 target, cause some obviously broken tools try to scan all files in /dev
98 including /dev/.udev/, find the links to /sys and end up stat()'ing sysfs files
99 million times. This takes ages on slow boxes.
101 The udevinfo attribute walk (-a) now works with giving a device node
102 name (-n) instead of a devpath (-p). The query now always works, also when
103 no database file was created by udev.
105 The built-in /etc/passwd /etc/group parser is removed, we always depend on
106 getpwnam() and getgrnam() now. One of the next releases will depend on
107 fnmatch() and may use getopt_long().
111 Bugfixes and small improvements.
115 Fix path_id for recent kernels.
121 Added support for swapping network interface names, by temporarily
122 renaming the device and wait for the target name to become free.
126 The built-in MODALIAS key and substitution is removed.
130 The binary firmware helper is replaced by the usual simple
131 shell script. Udevsend is removed from the tree.
139 Some more keys require the correct use of '==' and '=' depending
140 on the kind of operation beeing an assignment or a match. Rules
141 with invalid operations are skipped and logged to syslog. Please
142 test with udevtest if the parsing of your rules throws errors and
143 fix possibly broken rules.
147 Provide "udevsettle" to wait for all current udev events to finish.
148 It also watches the current kernel netlink queue by comparing the
149 even sequence number to make sure that there are no current pending
150 events that have not already arrived in the daemon.
154 Fix rule to skip persistent rules for removable IDE devices, which
155 also skipped optical IDE drives.
157 All *_id program are installed in /lib/udev/ by default now.
159 No binary is stripped anymore as this should be done in the
160 packaging process and not at build time.
162 libvolume_id is provided as a shared library now and vol_id is
163 linked against it. Also one of the next HAL versions will require
164 this library, and the HAL build process will also require the
165 header file to be installed. The copy of the same code in HAL will
166 be removed to have only a single copy left on the system.
170 Add persistent links for SCSI tapes. The rules file is renamed
171 to 60-persistent-storage.rules.
173 Create persistent path for usb devices. Can be used for all sorts
174 of devices that can't be distinguished by other properties like
175 multiple identical keyboards and mice connected to the same box.
177 Provide "udevtrigger" program to request events on coldplug. The
178 shell script is much too slow with thousends of devices.
182 Fix persistent disk rules to exclude removable IDE drives.
184 Warn if %e, $modalias or MODALIAS is used.
188 Fix queue export, which wasn't correct for subsequent add/remove
189 events for the same device.
193 Fix cramfs detection on big endian.
195 Make WAIT_FOR_SYSFS usable in "normal" rules and silent if the whole
200 If BUS== and SYSFS{}== have been used in the same rule, the sysfs
201 attributes were only checked at the parent device that matched the
202 by BUS requested subsystem. Fix it to also look at the device we
203 received the event for.
205 Build variable CROSS has changed to CROSS_COMPILE to match the kernel
210 Fix a bug where NAME="" would prevent RUN from beeing executed.
212 RUN="/bin/program" does not longer automatically add the subsystem
213 as the first parameter. This is from the days of /sbin/hotplug
214 which is dead now and it's just confusing to need to add a space at
215 the end of the program name to prevent this.
216 If you use rules that need the subsystem as the first parameter,
217 like the old "udev_run_hotlugd" and "udev_run_devd", add the subsystem
218 to the key like RUN+="/bin/program $env{SUBSYSTEM}".
222 The udev man page has moved to udev(7) as it does not describe a command
223 anymore. The programs udev, udevstart and udevsend are no longer installed
224 by default and must be copied manually, if they should be installed or
225 included in a package.
227 Fix a bug where "ignore_device" could run earlier collected RUN keys before
228 the ignore rule was applied.
230 More preparation for future sysfs changes. usb_id and scsi_id no longer
231 depend on a magic order of devices in the /devices chain. Specific devices
232 should be requested by their subsytem.
234 This will always find the scsi parent device without depending on a specific
236 dev = sysfs_device_get(devpath);
237 dev_usb = sysfs_device_get_parent_with_subsystem(dev, "scsi");
239 The "device" link in the current sysfs layout will be automatically
240 _resolved_ as a parent and in the new sysfs layout it will just _be_ the
241 parent in the devpath. If a device is requested by it's symlink, like all
242 class devices in the new sysfs layout will look like, it gets automatically
243 resolved and substituted with the real devpath and not the symlink path.
246 A similar logic must be applied to _all_ sysfs users, including
247 scripts, that search along parent devices in sysfs. The explicit use of
248 the "device" link must be avoided. With the future sysfs layout all
249 DEVPATH's will start with /devices/ and have a "subsystem" symlink poiting
250 back to the "class" or the "bus". The layout of the parent devices in
251 /devices is not necessarily expected to be stable across kernel releases and
252 searching for parents by their subsystem should make sysfs users tolerant
253 for changed parent chains.
257 Prepare udev to work with the experimental kernel patch, that moves
258 /sys/class devices to /sys/devices and /sys/block to /sys/class/block.
260 Clarify BUS, ID, $id usage and fix $id behavior. This prepares for
261 moving the class devices to /sys/devices.
263 Thanks again to Marco for help finding a hopefully nice compromise
264 to make %b simpler and working again.
268 Complete removal of libsysfs, replaced by simple helper functions
269 which are much simpler and a bit faster. The udev daemon operatesentirely
270 on event parameters and does not use sysfs for simple rules anymore.
271 Please report any new bugs/problems, that may be caused by this big
272 change. They will be fixed immediately.
274 The enumeration format character '%e' is deprecated and will be
275 removed sometimes from a future udev version. It never worked correctly
276 outside of udevstart, so we can't use it with the new parallel
277 coldplug. A simple enumeration is as useless as the devfs naming
278 scheme, just get rid of both if you still use it.
280 MODALIAS and $modalias is not needed and will be removed from one of
281 the next udev versions, replace it in all rules with ENV{MODALIAS} or
282 the sysfs "modalias" value.
284 Thanks a lot to Marco for all his help on finding and fixing bugs.
288 Let scsi_id request libata drive serial numbers from page 0x80.
290 Renamed etc/udev/persistent.rules to persistent-disk.rules and
291 added /dev/disk/by-name/* for device mapper device names.
293 Removed %e from the man page. It never worked reliably outside
294 of udevstart and udevstart is no longer recommended to use.
298 Symlinks are now exported to the event environment. Hopefully it's no
299 longer needed to run udevinfo from an event process, like it was
300 mentioned on the hotplug list:
301 UDEV [1134776873.702967] add@/block/sdb
304 DEVLINKS=/dev/disk/by-id/usb-IBM_Memory_Key_0218B301030027E8 /dev/disk/by-path/usb-0218B301030027E8:0:0:0
308 Fix a problem if udevsend is used as the hotplug handler and tries to use
309 syslog, which causes a "vc" event loop. 2.6.15 will make udevsend obsolete
310 and this kind of problems will hopefully go away soon.
314 All built-in logic to work around bad sysfs timing is removed with this
315 version. The need to wait for sysfs files is almost fixed with a kernel
316 version that doesn't work with this udev version anyway. Until we fix
317 the timing of the "bus" link creation, the former integrated logic should
318 be emulated by a rule placed before all other rules:
319 ACTION=="add", DEVPATH=="/devices/*", ENV{PHYSDEVBUS}=="?*", WAIT_FOR_SYSFS="bus"
321 The option "udev_db" does no longer exist. All udev state will be in
322 /$udev_root/.udev/ now, there is no longer an option to set this
324 If the init script or something else used this value, just depend on
325 this hardcoded path. But remember _all_content_ of this directory is
326 still private to udev and can change at any time.
328 Default location for rule sripts and helper programs is now: /lib/udev/.
329 Everything that is not useful on the commandline should go into this
330 directory. Some of the helpers in the extras folder are installed there
331 now. The rules need to be changed, to find the helpers there.
333 Also /lib/udev/devices is recommended as a directory where packages or
334 the user can place real device nodes, which get copied over to /dev at
335 every boot. This should replace the various solutions with custom config
338 Udevsend does no longer start the udev daemon. This must be done with
339 the init script that prepares /dev on tmpfs and creates the initial nodes,
340 before starting the daemon.
344 Silent a too verbose error logging for the old hotplug.d/ dev.d/
347 The copy of klibc is removed. A systemwide installed version of klibc
348 should be used to build a klibc udev now.
352 NAME="" will not create any nodes, but execute RUN keys. To completely
353 ignore an event the OPTION "ignore_device" should be used.
355 After removal of the reorder queue, events with a TIMEOUT can be executed
356 without any queuing now.
360 Fixed bug in udevd, if inotify is not available. We depend on netlink
361 uevents now, kernels without that event source will not work with that
362 version of udev anymore.
366 The rule parsing happens now in the daemon once at startup, all udev
367 event processes inherit the already parsed rules from the daemon.
368 It is shipped with SUSE10.0 and reduces heavily the system load at
369 startup. The option to save precompiled rules and let the udev process
370 pick the them up is removed, as it's no longer needed.
372 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
373 device. Libsysfs is changed to "translate" the requested link into the
374 real device path, as it would happen with the hotplug event. Otherwise
375 device removal and the udev database will not work.
377 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
380 A few improvements for vol_id, the filesytem probing code.
384 Fix a stupid typo in extras/run_directory for "make install".
386 scsi_id creates the temporary devnode now in /dev for usage with a
387 non-writable /tmp directory.
389 The uevent kernel socket buffer can carry app. 50.000 events now,
390 let's see who can break this again. :)
392 The upcoming kernel will have a new input driver core integration.
393 Some class devices are now symlinks to the real device. libsysfs
394 needs a fix for this to work correctly. Udevstart of older udev
395 versions will _not_ create these devices!
399 Fix a 'install' target in the Makefile, that prevents EXTRAS from
404 A bunch of mostly trivial bugfixes. From now on no node name or
405 symlink name can contain any character than plain whitelisted ascii
406 characters or validated utf8 byte-streams. This is needed for the
407 /dev/disk/by-label/* links, because we import untrusted data and
408 export it to the filesystem.
412 More bugfixes. If udevd was started from the kernel, we don't
413 have stdin/stdout/stderr, which broke the forked tools in some
418 Bugfix. udevstart event ordering was broken for a long time.
419 The new run_program() uncovered it, because /dev/null was not
420 available while we try to run external programs.
421 Now udevstart should create it before we run anything.
425 Minor bugfixes and some distro rules updates. If you don't have the
426 persistent disk rules in /dev/disk/by-*/* on your distro, just
427 grab it from here. :)
431 We can use socket communication now to pass events from udev to
433 RUN+="socket:/org/freedesktop/hal/udev_event"
434 will pass the whole udev event to the HAL daemon without the need
435 for a forked helper. (See ChangeLog for udevmonitor, as an example)
439 Mostly bugfixes and see ChangeLog.
441 The test for the existence of an environment value should be
443 ENV{KEY}=="*" to ENV{KEY}=="?*"
444 because "*" will not fail anymore, if the key does not exist or
449 Bugfixes and a few tweaks described in the ChangeLog.
453 Mostly a Bugfix release.
455 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
456 timing with custom rules.
460 We changed the internal rule storage format. Our large rule files took
461 2 MB of RAM, with the change we are down to 99kB.
463 If the device-node has been created with default name and no symlink or
464 options are to remenber, it is not longer stored in the udevdb. HAL will
465 need to be updated to work correctly with that change.
467 To overrride optimization flags, OPTFLAGS may be used now.
475 Major changes happened with this release. The goal is to take over the
476 complete kernel-event handling and provide a more efficient way to dispatch
477 kernel events. Replacing most of the current shell script logic and the
478 kernel forked helper with a netlink-daemon and a rule-based event handling.
480 o udevd listens to netlink events now. The first valid netlink event
481 will make udevd ignore any message from udevsend that contains a
482 SEQNUM, to avoid duplicate events. The forked events can be disabled
484 echo "" > /proc/sys/kernel/hotplug
485 For full support, the broken input-subsytem needs to be fixed, not to
486 bypass the driver core.
488 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
489 removed from udev itself and must be emulated by calling small
490 helper binaries provided in the extras folder:
491 make EXTRAS=extras/run_directory/
492 will build udev_run_devd and udev_run_hotplugd, which can be called
493 from a rule if needed:
494 RUN+="/sbin/udev_run_hotplugd"
495 The recommended way to handle this is to convert all the calls from
496 the directories to explicit udev rules and get completely rid of the
497 multiplexing. (To catch a ttyUSB event, you now no longer need to
498 fork and exit 300 tty script instances you are not interested in, it
499 is just one rule that matches exactly the device.)
501 o udev handles now _all_ events not just events for class and block
502 devices, this way it is possible to control the complete event
503 behavior with udev rules. Especially useful for rules like:
504 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
506 o As used in the modalias rule, udev supports now textual
507 substitution placeholder along with the usual format chars. This
508 needs to be documented, for now it's only visible in udev_rules_parse.c.
510 o The rule keys support now more operations. This is documented in the
511 man page. It is possible to add values to list-keys like the SYMLINK
512 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
513 Also "final"-assignments are supported by using KEY:="value", which will
514 prevent changing the key by any later rule.
516 o kernel 2.6.12 has the "detached_state" attribute removed from
517 sysfs, which was used to recognize sysfs population. We switched that
518 to wait for the "bus" link, which is only available in kernels after 2.6.11.
519 Running this udev version on older kernels may cause a short delay for
522 o To provide infrastructure for persistent device naming, the id programs:
523 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
524 to export the probed data in environment key format:
525 pim:~ # /sbin/ata_id --export /dev/hda
526 ID_MODEL=HTS726060M9AT00
527 ID_SERIAL=MRH401M4G6UM9B
531 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
532 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
535 kay@pim:~> tree /dev/disk
538 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
539 | `-- IBM-Memory_Key -> ../../sda
541 | |-- swap -> ../../hda1
542 | |-- date -> ../../sda1
543 | `-- home -> ../../hda3
545 |-- 2E08712B0870F2E7 -> ../../hda3
546 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
547 |-- E845-7A89 -> ../../sda1
548 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
550 The IMPORT= operation will import these keys in the environment and make
551 it available for later PROGRAM= and RUN= executed programs. The keys are
552 also stored in the udevdb and can be queried from there with one of the
555 o A few binaries are silently added to the repository, which can be used
556 to replay kernel events from initramfs instead of using coldplug. udevd
557 can be instructed now to queue-up events while the stored events from
558 initramfs are filled into the udevd-queue. This code is still under
559 development and there is no documentation now besides the code itself.
560 The additional binaries get compiled, but are not installed by default.
562 o There is also a temporary fix for a performance problem where too many
563 events happen in parallel and every event needs to parse the rules.
564 udev can now read precompiled rules stored on disk. This is likely to be
565 replaced by a more elegant solution in a future udev version.
569 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
570 Fix for libsysfs not to expect this file was added.
574 All rules are applied now, but only the first matching rule with a NAME-key
575 will be applied. All later rules with NAME-key are completely ignored. This
576 way system supplied symlinks or permissions gets applied to user-defined
580 Please check your rules setup, if you may need to add OPTIONS="last_rule"
581 to some rules, to keep the old behavior.
583 The rules are read on "remove"-events too. That makes is possible to match
584 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
585 instruct udev to ignore an event (OPTIONS="ignore_device").
586 The new ACTION-key may be used to let a rule act only at a "remove"-event.
588 The new RUN-key supports rule-based execution of programs after device-node
589 handling. This is meant as a general replacement for the dev.d/-directories
590 to give fine grained control over the execution of programs.
592 The %s{}-sysfs format char replacement values are searched at any of the
593 devices in the device chain now, not only at the class-device.
595 We support log priority levels now. The value udev_log in udev.conf is used
596 to determine what is printed to syslog. This makes it possible to
597 run a version with compiled-in debug messages in a production environment
598 which is sometimes needed to find a bug.
599 It is still possible to supress the inclusion of _any_ syslog usage with
600 USE_LOG=false to create the smallest possible binaries if needed.
601 The configured udev_log value can be overridden with the environment variable
606 Possible use of a system-wide klibc:
607 make USE_KLIBC=true KLCC=/usr/bin/klcc all
608 will link against an external klibc and our own version will be ignored.
612 We support an unlimited count of symlinks now.
614 If USE_STATIC=true is passed to a glibc build, we link statically and use
615 a built-in userdb parser to resolve user and group names.
617 The PLACE= key is gone. It can be replaced by an ID= for a long time, because
618 we walk up the chain of physical devices to find a match.
620 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
621 easy to skip certain attribute matches without composing rules with weird
622 character class negations like:
623 KERNEL="[!s][!c][!d]*"
624 this can now be replaced with:
626 The current simple '=' is still supported, and should work as it does today,
627 but existing rules should be converted if possible, to be better readable.
629 We have new ENV{}== key now, to match against a maximum of 5 environment
632 udevstart is its own binary again, because we don't need co carry this araound
633 with every forked event.