7 Fix recently introduced bug, which caused a compilation without large
8 file support, where vol_id does not recognize raid signatures at the end
11 Firewire disks now create both, by-id/scsi-* and by-id/ieee-* links.
12 Seems some kernel versions prevent the creation of the ieee-* links,
13 so people used the scsi-* link which disappeared now.
15 More libudev work. Almost all udevadm functionality comes from libudev
18 udevadm trigger has a new option --type, which allows to trigger events
19 for "devices", for "subsystems", or "failed" devices. The old option
20 --retry-failed" still works, but is no longer mentioned in the man page.
26 The udevadm info --device-id-of-file= output has changed to use
27 the obvious format. Possible current users should use the --export
28 option which is not affected.
30 The old udev commands symlinks to udevadm are not installed, if
31 these symlinks are used, a warning is printed.
37 Optical drive's media is no longer probed for raid signatures,
38 reading the end of the device causes some devices to malfunction.
39 Also the offset of the last session found is used now to probe
42 The volume_id library got a major version number update to 1,
43 some deprecated functions are removed.
45 A shared library "libudev" gets installed now to provide access
46 to udev device information. DeviceKit, the successor of HAL, will
47 need this library to access the udev database and search sysfs for
49 The library is currently in an experimental state, also the API is
50 expected to change, as long as the DeviceKit integration is not
55 We use ./configure now. See INSTALL for details. Current
58 "/usr" - prefix for man pages, include files
60 "" - the root filesystem, prefix for libs and binaries
64 "lib" - directory name for libraries, not a path name
65 multilib 64bit systems may use "lib64" instead of "lib"
67 compile-in verbose debug messages
69 disable all logging and compile-out all log strings
71 link against SELInux libraries, to set the expected context
74 In the default rules, the group "disk" gets permissions 0660 instead
75 of 0640. One small step closer to unify distro rules. Some day, all
76 distros hopefully end up with the same set of rules.
78 No symlinks to udevadm are installed anymore, if they are still needed,
79 they should be provided by the package.
85 Default udev rules, which are not supposed to be edited by the user, should
86 be placed in /lib/udev/rules.d/ now, to make it clear that they are private to
87 the udev package and will be replaced with an update. Udev will pick up rule
89 /lib/udev/rules.d/ - default installed rules
90 /etc/udev/rules.d/ - user rules + on-the-fly generated rules
91 /dev/.udev/rules.d/ - temporary non-persistent rules created after bootup
92 It does not matter in which directory a rule file lives, all files are sorted
95 To help creating /dev/root, we have now:
96 $ udevadm info --export --export-prefix="ROOT_" --device-id-of-file=/
99 In case the current --device-id-of-file is already used, please switch to
100 the --export format version, it saves the output parsing and the old
101 format will be changed to use ':' as a separator, like the format in the
106 Fix cdrom_id to properly recognize blank media.
112 Tape drive id-data is queried from /dev/bsg/* instead of the tape
113 nodes. This avoids rewinding tapes on open().
119 The symlinks udevcontrol and udevtrigger are no longer installed by
122 The scsi_id program does not depend on sysfs anymore. It can speak
123 SGv4 now, so /dev/bsg/* device nodes can be used, to query SCSI device
124 data, which should solve some old problems with tape devices, where
125 we better do not open all tape device nodes to identify the device.
131 The cdrom_id program is replaced by an advanced version, which can
132 detect most common device types, and also properties of the inserted
133 media. This is part of moving some basic functionality from HAL into
134 udev (and the kernel).
140 The last WAIT_FOR_SYSFS rule is removed from the default rules.
142 The symlinks to udevadm for the debugging tools: udevmonitor and
143 udevtest are no longer created.
145 The symlinks to the udevadm man page for the old tool names are
148 Abstract namespace sockets paths in RUN+="socket:@<path>" rules,
149 should be prefixed with '@' to indicate that the path is not a
160 Udevstart is removed from the tree, it did not get installed for
161 a long time now, and is long replaced by trigger and settle.
167 All udev tools are merged into a single binary called udevadm.
168 The old names of the tools are built-in commands in udevadm now.
169 Symlinks to udevadm, with the names of the old tools, provide
170 the same functionality as the standalone tools. There is also
171 only a single udevadm.8 man page left for all tools.
173 Tools like mkinitramfs should be checked, if they need to include
174 udevadm in the list of files.
184 The etc/udev/rules.d/ directory now contains a default set of basic
185 udev rules. This initial version is the result of a rules file merge
186 of Fedora and openSUSE. For these both distros only a few specific
187 rules are left in their own file, named after the distro. Rules which
188 are optionally installed, because they are only valid for a specific
189 architecture, or rules for subsystems which are not always used are
190 in etc/udev/packages/.
196 Dynamic rules can be created in /dev/.udev/rules.d/ to trigger
197 actions by dynamically created rules.
199 SYMLINK=="<value>" matches agains the entries in the list of
200 currently defined symlinks. The links are not created in the
201 filesystem at that point in time, but the values can be matched.
203 RUN{ignore_error}+="<program>" will ignore any exit code from the
204 program and not record as a failed event.
210 Final merge of patches/features from the Ubuntu package.
216 Control characters in filesystem label strings are no longer silenty
217 removed, but hex-encoded, to be able to uniquely identify the device
218 by its symlink in /dev/disk/by-label/.
219 If libvolume_id is used by mount(8), LABEL= will work as expected,
220 if slashes or other characters are used in the label string.
222 To test the existence of a file, TEST=="<file>" and TEST!="<file>"
223 can be specified now. The TEST key accepts an optional mode mask
224 TEST{0100}=="<is executable file>".
226 Scsi_id now supports a mode without expecting scsi-specific sysfs
227 entries to allow the extraction of cciss-device persistent properties.
233 In the future, we may see uuid's which are just simple character
234 strings (see the DDF Raid Specification). For that reason vol_id now
235 exports ID_FS_UUID_SAFE, just like ID_FS_LABEL_SAFE. For things like
236 the creation of symlinks, the *_SAFE values ensure, that no control
237 or whitespace characters are used in the filename.
239 Possible users of libvolume_id, please use the volume_id_get_* functions.
240 The public struct will go away in a future release of the library.
246 Removal of useless extras/eventrecorder.sh.
256 The directory multiplexer for dev.d/ and hotplug.d are finally removed
257 from the udev package.
263 Symlinks can have priorities now, the priority is assigned to the device
264 and specified with OPTIONS="link_priority=100". Devices with higher
265 priorities overwrite the symlinks of devices with lower priorities.
266 If the device that currently owns the link, goes away, the symlink
267 will be removed, and recreated, pointing to the next device with the
268 highest actual priority. This should make /dev/disk/by-{label,uuid,id}
269 more reliable, if multiple devices contain the same metadata and overwrite
272 The dasd_id program is removed from the udev tree, and dasdinfo, with the
273 needed rules, are part of the s390-tools now.
275 Please add KERNEL=="[0-9]*:[0-9]*" to the scsi wait-for-sysfs rule,
276 we may get the scsi sysfs mess fixed some day, and this will only catch
277 the devices we are looking for.
279 USB serial numbers for storage devices have the target:lun now appended,
280 to make it possibble to distinguish broken multi-lun devices with all
281 the same SCSI identifiers.
283 Note: The extra "run_directory" which searches and executes stuff in
284 /etc/hotplug.d/ and /etc/dev.d/ is long deprecated, and will be removed
285 with the next release. Make sure, that you don't use it anymore, or
286 provides your own implementation of that inefficient stuff.
287 We are tired of reports about a "slow udev", because these directories
288 contain stuff, that runs with _every_ event, instead of using rules,
289 that run programs only for the matching events.
299 DRIVER== will match only for devices that actually have a real
300 driver. DRIVERS== must be used, if parent devices should be
301 included in the match.
303 Libvolume_id's "linux_raid" detection needed another fix.
311 Add additional check to volume_id detection of via_raid, cause
312 some company decided to put a matching pattern all over the empty
313 storage area of their music players.
317 Fix path_id for SAS devices.
321 The udev daemon can be started with --debug-trace now, which will
322 execute all events serialized to get a chance to catch a possible
323 action that crashes the box.
325 A warning is logged, if PHYSDEV* keys, the "device" link, or a parent
326 device attribute like $attr{../file} is used, only WAIT_FOR_SYSFS rules
327 are excluded from the warning. Referencing parent attributes directly
328 may break when something in the kernel driver model changes. Udev will
329 just find the attribute by walking up the parent chain.
331 Udevtrigger now sorts the list of devices depending on the device
332 dependency, so a "usb" device is triggered after the parent "pci"
337 Revert persistent-storage ata-serial '_' '-' replacement.
343 Udevtrigger can now filter the list of devices to be triggered. Matches
344 for subsystems or sysfs attributes can be specified.
346 The entries in /dev/.udev/queue and /dev/.udev/failed have changed to
347 zero-sized files to avoid pointing to /sys and confuse broken tools which
348 scan the /dev directory. To retry failed events, udevtrigger --retry-failed
351 The rules and scripts to create udev rules for persistent network
352 devices and optical drives are in the extras/rules_generator directory
353 now. If you use something similar, please consider replacing your own
354 version with this, to share the support effort. The rule_generator
355 installs its own rules into /etc/udev/rules.d.
357 The cdrom_id tool installs its own rule now in /etc/udev/rules.d, cause
358 the rule_generator depends on cdrom_id to be called in an earlier rule.
364 Renaming of some key names (the old names still work):
365 BUS -> SUBSYSTEMS, ID -> KERNELS, SYSFS -> ATTRS, DRIVER -> DRIVERS.
366 (The behavior of the key DRIVER will change soon in one of the next
367 releases, to match only the event device, please switch to DRIVERS
368 instead. If DRIVER is used, it will behave like DRIVERS, but an error
370 With the new key names, we have a more consistent and simpler scheme.
371 We can match the properties of the event device only, with: KERNEL,
372 SUBSYSTEM, ATTR, DRIVER. Or include all the parent devices in the match,
373 with: KERNELS, SUBSYSTEMS, ATTRS, DRIVERS. ID, BUS, SYSFS, DRIVER are no
374 longer mentioned in the man page and should be switched in the rule
377 ATTR{file}="value" can be used now, to write to a sysfs file of the
378 event device. Instead of:
379 ..., SYSFS{type}=="0|7|14", RUN+="/bin/sh -c 'echo 60 > /sys$$DEVPATH/timeout'"
381 ..., ATTR{type}=="0|7|14", ATTR{timeout}="60"
383 All the PHYSDEV* keys are deprecated and will be removed from a
385 PHYDEVPATH - is the path of a parent device and should not be
387 PHYSDEVBUS - is just a SUBSYSTEM value of a parent, and can be
388 matched with SUBSYSTEMS==
389 PHYSDEVDRIVER - for bus devices it is available as ENV{DRIVER}.
390 Newer kernels will have DRIVER in the environment,
391 for older kernels udev puts in. Class device will
392 no longer carry this property of a parent and
393 DRIVERS== can be used to match such a parent value.
394 Note that ENV{DRIVER} is only available for a few bus devices, where
395 the driver is already bound at device event time. On coldplug, the
396 events for a lot devices are already bound to a driver, and they will have
397 that value set. But on hotplug, at the time the kernel creates the device,
398 it can't know what driver may claim the device after that, therefore
399 in most cases it will be empty.
401 Failed events should now be re-triggered with:
402 udevtrigger --retry-failed.
403 Please switch to this command, so we keep the details of the /dev/.udev/failed/
404 files private to the udev tools. We may need to switch the current symlink
405 target, cause some obviously broken tools try to scan all files in /dev
406 including /dev/.udev/, find the links to /sys and end up stat()'ing sysfs files
407 million times. This takes ages on slow boxes.
409 The udevinfo attribute walk (-a) now works with giving a device node
410 name (-n) instead of a devpath (-p). The query now always works, also when
411 no database file was created by udev.
413 The built-in /etc/passwd /etc/group parser is removed, we always depend on
414 getpwnam() and getgrnam() now. One of the next releases will depend on
415 fnmatch() and may use getopt_long().
419 Bugfixes and small improvements.
423 Fix path_id for recent kernels.
429 Added support for swapping network interface names, by temporarily
430 renaming the device and wait for the target name to become free.
434 The built-in MODALIAS key and substitution is removed.
438 The binary firmware helper is replaced by the usual simple
439 shell script. Udevsend is removed from the tree.
447 Some more keys require the correct use of '==' and '=' depending
448 on the kind of operation beeing an assignment or a match. Rules
449 with invalid operations are skipped and logged to syslog. Please
450 test with udevtest if the parsing of your rules throws errors and
451 fix possibly broken rules.
455 Provide "udevsettle" to wait for all current udev events to finish.
456 It also watches the current kernel netlink queue by comparing the
457 even sequence number to make sure that there are no current pending
458 events that have not already arrived in the daemon.
462 Fix rule to skip persistent rules for removable IDE devices, which
463 also skipped optical IDE drives.
465 All *_id program are installed in /lib/udev/ by default now.
467 No binary is stripped anymore as this should be done in the
468 packaging process and not at build time.
470 libvolume_id is provided as a shared library now and vol_id is
471 linked against it. Also one of the next HAL versions will require
472 this library, and the HAL build process will also require the
473 header file to be installed. The copy of the same code in HAL will
474 be removed to have only a single copy left on the system.
478 Add persistent links for SCSI tapes. The rules file is renamed
479 to 60-persistent-storage.rules.
481 Create persistent path for usb devices. Can be used for all sorts
482 of devices that can't be distinguished by other properties like
483 multiple identical keyboards and mice connected to the same box.
485 Provide "udevtrigger" program to request events on coldplug. The
486 shell script is much too slow with thousends of devices.
490 Fix persistent disk rules to exclude removable IDE drives.
492 Warn if %e, $modalias or MODALIAS is used.
496 Fix queue export, which wasn't correct for subsequent add/remove
497 events for the same device.
501 Fix cramfs detection on big endian.
503 Make WAIT_FOR_SYSFS usable in "normal" rules and silent if the whole
508 If BUS== and SYSFS{}== have been used in the same rule, the sysfs
509 attributes were only checked at the parent device that matched the
510 by BUS requested subsystem. Fix it to also look at the device we
511 received the event for.
513 Build variable CROSS has changed to CROSS_COMPILE to match the kernel
518 Fix a bug where NAME="" would prevent RUN from beeing executed.
520 RUN="/bin/program" does not longer automatically add the subsystem
521 as the first parameter. This is from the days of /sbin/hotplug
522 which is dead now and it's just confusing to need to add a space at
523 the end of the program name to prevent this.
524 If you use rules that need the subsystem as the first parameter,
525 like the old "udev_run_hotlugd" and "udev_run_devd", add the subsystem
526 to the key like RUN+="/bin/program $env{SUBSYSTEM}".
530 The udev man page has moved to udev(7) as it does not describe a command
531 anymore. The programs udev, udevstart and udevsend are no longer installed
532 by default and must be copied manually, if they should be installed or
533 included in a package.
535 Fix a bug where "ignore_device" could run earlier collected RUN keys before
536 the ignore rule was applied.
538 More preparation for future sysfs changes. usb_id and scsi_id no longer
539 depend on a magic order of devices in the /devices chain. Specific devices
540 should be requested by their subsytem.
542 This will always find the scsi parent device without depending on a specific
544 dev = sysfs_device_get(devpath);
545 dev_usb = sysfs_device_get_parent_with_subsystem(dev, "scsi");
547 The "device" link in the current sysfs layout will be automatically
548 _resolved_ as a parent and in the new sysfs layout it will just _be_ the
549 parent in the devpath. If a device is requested by it's symlink, like all
550 class devices in the new sysfs layout will look like, it gets automatically
551 resolved and substituted with the real devpath and not the symlink path.
554 A similar logic must be applied to _all_ sysfs users, including
555 scripts, that search along parent devices in sysfs. The explicit use of
556 the "device" link must be avoided. With the future sysfs layout all
557 DEVPATH's will start with /devices/ and have a "subsystem" symlink poiting
558 back to the "class" or the "bus". The layout of the parent devices in
559 /devices is not necessarily expected to be stable across kernel releases and
560 searching for parents by their subsystem should make sysfs users tolerant
561 for changed parent chains.
565 Prepare udev to work with the experimental kernel patch, that moves
566 /sys/class devices to /sys/devices and /sys/block to /sys/class/block.
568 Clarify BUS, ID, $id usage and fix $id behavior. This prepares for
569 moving the class devices to /sys/devices.
571 Thanks again to Marco for help finding a hopefully nice compromise
572 to make %b simpler and working again.
576 Complete removal of libsysfs, replaced by simple helper functions
577 which are much simpler and a bit faster. The udev daemon operatesentirely
578 on event parameters and does not use sysfs for simple rules anymore.
579 Please report any new bugs/problems, that may be caused by this big
580 change. They will be fixed immediately.
582 The enumeration format character '%e' is deprecated and will be
583 removed sometimes from a future udev version. It never worked correctly
584 outside of udevstart, so we can't use it with the new parallel
585 coldplug. A simple enumeration is as useless as the devfs naming
586 scheme, just get rid of both if you still use it.
588 MODALIAS and $modalias is not needed and will be removed from one of
589 the next udev versions, replace it in all rules with ENV{MODALIAS} or
590 the sysfs "modalias" value.
592 Thanks a lot to Marco for all his help on finding and fixing bugs.
596 Let scsi_id request libata drive serial numbers from page 0x80.
598 Renamed etc/udev/persistent.rules to persistent-disk.rules and
599 added /dev/disk/by-name/* for device mapper device names.
601 Removed %e from the man page. It never worked reliably outside
602 of udevstart and udevstart is no longer recommended to use.
606 Symlinks are now exported to the event environment. Hopefully it's no
607 longer needed to run udevinfo from an event process, like it was
608 mentioned on the hotplug list:
609 UDEV [1134776873.702967] add@/block/sdb
612 DEVLINKS=/dev/disk/by-id/usb-IBM_Memory_Key_0218B301030027E8 /dev/disk/by-path/usb-0218B301030027E8:0:0:0
616 Fix a problem if udevsend is used as the hotplug handler and tries to use
617 syslog, which causes a "vc" event loop. 2.6.15 will make udevsend obsolete
618 and this kind of problems will hopefully go away soon.
622 All built-in logic to work around bad sysfs timing is removed with this
623 version. The need to wait for sysfs files is almost fixed with a kernel
624 version that doesn't work with this udev version anyway. Until we fix
625 the timing of the "bus" link creation, the former integrated logic should
626 be emulated by a rule placed before all other rules:
627 ACTION=="add", DEVPATH=="/devices/*", ENV{PHYSDEVBUS}=="?*", WAIT_FOR_SYSFS="bus"
629 The option "udev_db" does no longer exist. All udev state will be in
630 /$udev_root/.udev/ now, there is no longer an option to set this
632 If the init script or something else used this value, just depend on
633 this hardcoded path. But remember _all_content_ of this directory is
634 still private to udev and can change at any time.
636 Default location for rule sripts and helper programs is now: /lib/udev/.
637 Everything that is not useful on the commandline should go into this
638 directory. Some of the helpers in the extras folder are installed there
639 now. The rules need to be changed, to find the helpers there.
641 Also /lib/udev/devices is recommended as a directory where packages or
642 the user can place real device nodes, which get copied over to /dev at
643 every boot. This should replace the various solutions with custom config
646 Udevsend does no longer start the udev daemon. This must be done with
647 the init script that prepares /dev on tmpfs and creates the initial nodes,
648 before starting the daemon.
652 Silent a too verbose error logging for the old hotplug.d/ dev.d/
655 The copy of klibc is removed. A systemwide installed version of klibc
656 should be used to build a klibc udev now.
660 NAME="" will not create any nodes, but execute RUN keys. To completely
661 ignore an event the OPTION "ignore_device" should be used.
663 After removal of the reorder queue, events with a TIMEOUT can be executed
664 without any queuing now.
668 Fixed bug in udevd, if inotify is not available. We depend on netlink
669 uevents now, kernels without that event source will not work with that
670 version of udev anymore.
674 The rule parsing happens now in the daemon once at startup, all udev
675 event processes inherit the already parsed rules from the daemon.
676 It is shipped with SUSE10.0 and reduces heavily the system load at
677 startup. The option to save precompiled rules and let the udev process
678 pick the them up is removed, as it's no longer needed.
680 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
681 device. Libsysfs is changed to "translate" the requested link into the
682 real device path, as it would happen with the hotplug event. Otherwise
683 device removal and the udev database will not work.
685 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
688 A few improvements for vol_id, the filesytem probing code.
692 Fix a stupid typo in extras/run_directory for "make install".
694 scsi_id creates the temporary devnode now in /dev for usage with a
695 non-writable /tmp directory.
697 The uevent kernel socket buffer can carry app. 50.000 events now,
698 let's see who can break this again. :)
700 The upcoming kernel will have a new input driver core integration.
701 Some class devices are now symlinks to the real device. libsysfs
702 needs a fix for this to work correctly. Udevstart of older udev
703 versions will _not_ create these devices!
707 Fix a 'install' target in the Makefile, that prevents EXTRAS from
712 A bunch of mostly trivial bugfixes. From now on no node name or
713 symlink name can contain any character than plain whitelisted ascii
714 characters or validated utf8 byte-streams. This is needed for the
715 /dev/disk/by-label/* links, because we import untrusted data and
716 export it to the filesystem.
720 More bugfixes. If udevd was started from the kernel, we don't
721 have stdin/stdout/stderr, which broke the forked tools in some
726 Bugfix. udevstart event ordering was broken for a long time.
727 The new run_program() uncovered it, because /dev/null was not
728 available while we try to run external programs.
729 Now udevstart should create it before we run anything.
733 Minor bugfixes and some distro rules updates. If you don't have the
734 persistent disk rules in /dev/disk/by-*/* on your distro, just
735 grab it from here. :)
739 We can use socket communication now to pass events from udev to
741 RUN+="socket:/org/freedesktop/hal/udev_event"
742 will pass the whole udev event to the HAL daemon without the need
743 for a forked helper. (See ChangeLog for udevmonitor, as an example)
747 Mostly bugfixes and see ChangeLog.
749 The test for the existence of an environment value should be
751 ENV{KEY}=="*" to ENV{KEY}=="?*"
752 because "*" will not fail anymore, if the key does not exist or
757 Bugfixes and a few tweaks described in the ChangeLog.
761 Mostly a Bugfix release.
763 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
764 timing with custom rules.
768 We changed the internal rule storage format. Our large rule files took
769 2 MB of RAM, with the change we are down to 99kB.
771 If the device-node has been created with default name and no symlink or
772 options are to remenber, it is not longer stored in the udevdb. HAL will
773 need to be updated to work correctly with that change.
775 To overrride optimization flags, OPTFLAGS may be used now.
783 Major changes happened with this release. The goal is to take over the
784 complete kernel-event handling and provide a more efficient way to dispatch
785 kernel events. Replacing most of the current shell script logic and the
786 kernel forked helper with a netlink-daemon and a rule-based event handling.
788 o udevd listens to netlink events now. The first valid netlink event
789 will make udevd ignore any message from udevsend that contains a
790 SEQNUM, to avoid duplicate events. The forked events can be disabled
792 echo "" > /proc/sys/kernel/hotplug
793 For full support, the broken input-subsytem needs to be fixed, not to
794 bypass the driver core.
796 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
797 removed from udev itself and must be emulated by calling small
798 helper binaries provided in the extras folder:
799 make EXTRAS=extras/run_directory/
800 will build udev_run_devd and udev_run_hotplugd, which can be called
801 from a rule if needed:
802 RUN+="/sbin/udev_run_hotplugd"
803 The recommended way to handle this is to convert all the calls from
804 the directories to explicit udev rules and get completely rid of the
805 multiplexing. (To catch a ttyUSB event, you now no longer need to
806 fork and exit 300 tty script instances you are not interested in, it
807 is just one rule that matches exactly the device.)
809 o udev handles now _all_ events not just events for class and block
810 devices, this way it is possible to control the complete event
811 behavior with udev rules. Especially useful for rules like:
812 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
814 o As used in the modalias rule, udev supports now textual
815 substitution placeholder along with the usual format chars. This
816 needs to be documented, for now it's only visible in udev_rules_parse.c.
818 o The rule keys support now more operations. This is documented in the
819 man page. It is possible to add values to list-keys like the SYMLINK
820 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
821 Also "final"-assignments are supported by using KEY:="value", which will
822 prevent changing the key by any later rule.
824 o kernel 2.6.12 has the "detached_state" attribute removed from
825 sysfs, which was used to recognize sysfs population. We switched that
826 to wait for the "bus" link, which is only available in kernels after 2.6.11.
827 Running this udev version on older kernels may cause a short delay for
830 o To provide infrastructure for persistent device naming, the id programs:
831 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
832 to export the probed data in environment key format:
833 pim:~ # /sbin/ata_id --export /dev/hda
834 ID_MODEL=HTS726060M9AT00
835 ID_SERIAL=MRH401M4G6UM9B
839 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
840 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
843 kay@pim:~> tree /dev/disk
846 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
847 | `-- IBM-Memory_Key -> ../../sda
849 | |-- swap -> ../../hda1
850 | |-- date -> ../../sda1
851 | `-- home -> ../../hda3
853 |-- 2E08712B0870F2E7 -> ../../hda3
854 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
855 |-- E845-7A89 -> ../../sda1
856 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
858 The IMPORT= operation will import these keys in the environment and make
859 it available for later PROGRAM= and RUN= executed programs. The keys are
860 also stored in the udevdb and can be queried from there with one of the
863 o A few binaries are silently added to the repository, which can be used
864 to replay kernel events from initramfs instead of using coldplug. udevd
865 can be instructed now to queue-up events while the stored events from
866 initramfs are filled into the udevd-queue. This code is still under
867 development and there is no documentation now besides the code itself.
868 The additional binaries get compiled, but are not installed by default.
870 o There is also a temporary fix for a performance problem where too many
871 events happen in parallel and every event needs to parse the rules.
872 udev can now read precompiled rules stored on disk. This is likely to be
873 replaced by a more elegant solution in a future udev version.
877 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
878 Fix for libsysfs not to expect this file was added.
882 All rules are applied now, but only the first matching rule with a NAME-key
883 will be applied. All later rules with NAME-key are completely ignored. This
884 way system supplied symlinks or permissions gets applied to user-defined
888 Please check your rules setup, if you may need to add OPTIONS="last_rule"
889 to some rules, to keep the old behavior.
891 The rules are read on "remove"-events too. That makes is possible to match
892 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
893 instruct udev to ignore an event (OPTIONS="ignore_device").
894 The new ACTION-key may be used to let a rule act only at a "remove"-event.
896 The new RUN-key supports rule-based execution of programs after device-node
897 handling. This is meant as a general replacement for the dev.d/-directories
898 to give fine grained control over the execution of programs.
900 The %s{}-sysfs format char replacement values are searched at any of the
901 devices in the device chain now, not only at the class-device.
903 We support log priority levels now. The value udev_log in udev.conf is used
904 to determine what is printed to syslog. This makes it possible to
905 run a version with compiled-in debug messages in a production environment
906 which is sometimes needed to find a bug.
907 It is still possible to supress the inclusion of _any_ syslog usage with
908 USE_LOG=false to create the smallest possible binaries if needed.
909 The configured udev_log value can be overridden with the environment variable
914 Possible use of a system-wide klibc:
915 make USE_KLIBC=true KLCC=/usr/bin/klcc all
916 will link against an external klibc and our own version will be ignored.
920 We support an unlimited count of symlinks now.
922 If USE_STATIC=true is passed to a glibc build, we link statically and use
923 a built-in userdb parser to resolve user and group names.
925 The PLACE= key is gone. It can be replaced by an ID= for a long time, because
926 we walk up the chain of physical devices to find a match.
928 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
929 easy to skip certain attribute matches without composing rules with weird
930 character class negations like:
931 KERNEL="[!s][!c][!d]*"
932 this can now be replaced with:
934 The current simple '=' is still supported, and should work as it does today,
935 but existing rules should be converted if possible, to be better readable.
937 We have new ENV{}== key now, to match against a maximum of 5 environment
940 udevstart is its own binary again, because we don't need co carry this araound
941 with every forked event.