3 Fix queue export, which wasn't correct for subsequent add/remove
4 events for the same device.
8 Fix cramfs detection on big endian.
10 Make WAIT_FOR_SYSFS usable in "normal" rules and silent if the whole
15 If BUS== and SYSFS{}== have been used in the same rule, the sysfs
16 attributes were only checked at the parent device that matched the
17 by BUS requested subsystem. Fix it to also look at the device we
18 received the event for.
20 Build variable CROSS has changed to CROSS_COMPILE to match the kernel
25 Fix a bug where NAME="" would prevent RUN from beeing executed.
27 RUN="/bin/program" does not longer automatically add the subsystem
28 as the first parameter. This is from the days of /sbin/hotplug
29 which is dead now and it's just confusing to need to add a space at
30 the end of the program name to prevent this.
31 If you use rules that need the subsystem as the first parameter,
32 like the old "udev_run_hotlugd" and "udev_run_devd", add the subsystem
33 to the key like RUN+="/bin/program $env{SUBSYSTEM}".
37 The udev man page has moved to udev(7) as it does not describe a command
38 anymore. The programs udev, udevstart and udevsend are no longer installed
39 by default and must be copied manually, if they should be installed or
40 included in a package.
42 Fix a bug where "ignore_device" could run earlier collected RUN keys before
43 the ignore rule was applied.
45 More preparation for future sysfs changes. usb_id and scsi_id no longer
46 depend on a magic order of devices in the /devices chain. Specific devices
47 should be requested by their subsytem.
49 This will always find the scsi parent device without depending on a specific
51 dev = sysfs_device_get(devpath);
52 dev_usb = sysfs_device_get_parent_with_subsystem(dev, "scsi");
54 The "device" link in the current sysfs layout will be automatically
55 _resolved_ as a parent and in the new sysfs layout it will just _be_ the
56 parent in the devpath. If a device is requested by it's symlink, like all
57 class devices in the new sysfs layout will look like, it gets automatically
58 resolved and substituted with the real devpath and not the symlink path.
61 A similar logic must be applied to _all_ sysfs users, including
62 scripts, that search along parent devices in sysfs. The explicit use of
63 the "device" link must be avoided. With the future sysfs layout all
64 DEVPATH's will start with /devices/ and have a "subsystem" symlink poiting
65 back to the "class" or the "bus". The layout of the parent devices in
66 /devices is not necessarily expected to be stable across kernel releases and
67 searching for parents by their subsystem should make sysfs users tolerant
68 for changed parent chains.
72 Prepare udev to work with the experimental kernel patch, that moves
73 /sys/class devices to /sys/devices and /sys/block to /sys/class/block.
75 Clarify BUS, ID, $id usage and fix $id behavior. This prepares for
76 moving the class devices to /sys/devices.
78 Thanks again to Marco for help finding a hopefully nice compromise
79 to make %b simpler and working again.
83 Complete removal of libsysfs, replaced by simple helper functions
84 which are much simpler and a bit faster. The udev daemon operatesentirely
85 on event parameters and does not use sysfs for simple rules anymore.
86 Please report any new bugs/problems, that may be caused by this big
87 change. They will be fixed immediately.
89 The enumeration format character '%e' is deprecated and will be
90 removed sometimes from a future udev version. It never worked correctly
91 outside of udevstart, so we can't use it with the new parallel
92 coldplug. A simple enumeration is as useless as the devfs naming
93 scheme, just get rid of both if you still use it.
95 MODALIAS and $modalias is not needed and will be removed from one of
96 the next udev versions, replace it in all rules with ENV{MODALIAS} or
97 the sysfs "modalias" value.
99 Thanks a lot to Marco for all his help on finding and fixing bugs.
103 Let scsi_id request libata drive serial numbers from page 0x80.
105 Renamed etc/udev/persistent.rules to persistent-disk.rules and
106 added /dev/disk/by-name/* for device mapper device names.
108 Removed %e from the man page. It never worked reliably outside
109 of udevstart and udevstart is no longer recommended to use.
113 Symlinks are now exported to the event environment. Hopefully it's no
114 longer needed to run udevinfo from an event process, like it was
115 mentioned on the hotplug list:
116 UDEV [1134776873.702967] add@/block/sdb
119 DEVLINKS=/dev/disk/by-id/usb-IBM_Memory_Key_0218B301030027E8 /dev/disk/by-path/usb-0218B301030027E8:0:0:0
123 Fix a problem if udevsend is used as the hotplug handler and tries to use
124 syslog, which causes a "vc" event loop. 2.6.15 will make udevsend obsolete
125 and this kind of problems will hopefully go away soon.
129 All built-in logic to work around bad sysfs timing is removed with this
130 version. The need to wait for sysfs files is almost fixed with a kernel
131 version that doesn't work with this udev version anyway. Until we fix
132 the timing of the "bus" link creation, the former integrated logic should
133 be emulated by a rule placed before all other rules:
134 ACTION=="add", DEVPATH=="/devices/*", ENV{PHYSDEVBUS}=="?*", WAIT_FOR_SYSFS="bus"
136 The option "udev_db" does no longer exist. All udev state will be in
137 /$udev_root/.udev/ now, there is no longer an option to set this
139 If the init script or something else used this value, just depend on
140 this hardcoded path. But remember _all_content_ of this directory is
141 still private to udev and can change at any time.
143 Default location for rule sripts and helper programs is now: /lib/udev/.
144 Everything that is not useful on the commandline should go into this
145 directory. Some of the helpers in the extras folder are installed there
146 now. The rules need to be changed, to find the helpers there.
148 Also /lib/udev/devices is recommended as a directory where packages or
149 the user can place real device nodes, which get copied over to /dev at
150 every boot. This should replace the various solutions with custom config
153 Udevsend does no longer start the udev daemon. This must be done with
154 the init script that prepares /dev on tmpfs and creates the initial nodes,
155 before starting the daemon.
159 Silent a too verbose error logging for the old hotplug.d/ dev.d/
162 The copy of klibc is removed. A systemwide installed version of klibc
163 should be used to build a klibc udev now.
167 NAME="" will not create any nodes, but execute RUN keys. To completely
168 ignore an event the OPTION "ignore_device" should be used.
170 After removal of the reorder queue, events with a TIMEOUT can be executed
171 without any queuing now.
175 Fixed bug in udevd, if inotify is not available. We depend on netlink
176 uevents now, kernels without that event source will not work with that
177 version of udev anymore.
181 The rule parsing happens now in the daemon once at startup, all udev
182 event processes inherit the already parsed rules from the daemon.
183 It is shipped with SUSE10.0 and reduces heavily the system load at
184 startup. The option to save precompiled rules and let the udev process
185 pick the them up is removed, as it's no longer needed.
187 Kernel 2.6.15 will have symlinks at /class/input pointing to the real
188 device. Libsysfs is changed to "translate" the requested link into the
189 real device path, as it would happen with the hotplug event. Otherwise
190 device removal and the udev database will not work.
192 Using 'make STRIPCMD=' will leave the binaries unstripped for debugging
195 A few improvements for vol_id, the filesytem probing code.
199 Fix a stupid typo in extras/run_directory for "make install".
201 scsi_id creates the temporary devnode now in /dev for usage with a
202 non-writable /tmp directory.
204 The uevent kernel socket buffer can carry app. 50.000 events now,
205 let's see who can break this again. :)
207 The upcoming kernel will have a new input driver core integration.
208 Some class devices are now symlinks to the real device. libsysfs
209 needs a fix for this to work correctly. Udevstart of older udev
210 versions will _not_ create these devices!
214 Fix a 'install' target in the Makefile, that prevents EXTRAS from
219 A bunch of mostly trivial bugfixes. From now on no node name or
220 symlink name can contain any character than plain whitelisted ascii
221 characters or validated utf8 byte-streams. This is needed for the
222 /dev/disk/by-label/* links, cause we import untrusted data and
223 export it to the filesystem.
227 More bugfixes. If udevd was started from the kernel, we don't
228 have stdin/stdout/stderr, which broke the forked tools in some
233 Bugfix. udevstart event ordering was broken for a long time.
234 The new run_program() uncovered it, cause /dev/null was not
235 available while we try to run external programs.
236 Now udevstart should create it before we run anything.
240 Minor bugfixes and some distro rules updates. If you don't have the
241 persistent disk rules in /dev/disk/by-*/* on your distro, just
242 grab it from here. :)
246 We can use socket communication now to pass events from udev to
248 RUN+="socket:/org/freedesktop/hal/udev_event"
249 will pass the whole udev event to the HAL daemon without the need
250 for a forked helper. (See ChangeLog for udevmonitor, as an example)
254 Mostly bugfixes and see ChangeLog.
256 The test for the existence of an environment value should be
258 ENV{KEY}=="*" to ENV{KEY}=="?*"
259 cause "*" will not fail anymore, if the key does not exist or
264 Bugfixes and a few tweaks described in the ChangeLog.
268 Mostly a Bugfix release.
270 Added WAIT_FOR_SYSFS="<attribute>" to be able to fight against the sysfs
271 timing with custom rules.
275 We changed the internal rule storage format. Our large rule files took
276 2 MB of RAM, with the change we are down to 99kB.
278 If the device-node has been created with default name and no symlink or
279 options are to remenber, it is not longer stored in the udevdb. HAL will
280 need to be updated to work correctly with that change.
282 To overrride optimization flags, OPTFLAGS may be used now.
290 Major changes happened with this release. The goal is to take over the
291 complete kernel-event handling and provide a more efficient way to dispatch
292 kernel events. Replacing most of the current shell script logic and the
293 kernel forked helper with a netlink-daemon and a rule-based event handling.
295 o udevd listens to netlink events now. The first valid netlink event
296 will make udevd ignore any message from udevsend that contains a
297 SEQNUM, to avoid duplicate events. The forked events can be disabled
299 echo "" > /proc/sys/kernel/hotplug
300 For full support, the broken input-subsytem needs to be fixed, not to
301 bypass the driver core.
303 o /etc/dev.d/ + /etc/hotplug.d/ directory multiplexing is completely
304 removed from udev itself and must be emulated by calling small
305 helper binaries provided in the extras folder:
306 make EXTRAS=extras/run_directory/
307 will build udev_run_devd and udev_run_hotplugd, which can be called
308 from a rule if needed:
309 RUN+="/sbin/udev_run_hotplugd"
310 The recommended way to handle this is to convert all the calls from
311 the directories to explicit udev rules and get completely rid of the
312 multiplexing. (To catch a ttyUSB event, you now no longer need to
313 fork and exit 300 tty script instances you are not interested in, it
314 is just one rule that matches exactly the device.)
316 o udev handles now _all_ events not just events for class and block
317 devices, this way it is possible to control the complete event
318 behavior with udev rules. Especially useful for rules like:
319 ACTION="add", DEVPATH="/devices/*", MODALIAS=="?*", RUN+="/sbin/modprobe $modalias"
321 o As used in the modalias rule, udev supports now textual
322 substitution placeholder along with the usual format chars. This
323 needs to be documented, for now it's only visible in udev_rules_parse.c.
325 o The rule keys support now more operations. This is documented in the
326 man page. It is possible to add values to list-keys like the SYMLINK
327 and RUN list with KEY+="value" and to clear the list by assigning KEY="".
328 Also "final"-assignments are supported by using KEY:="value", which will
329 prevent changing the key by any later rule.
331 o kernel 2.6.12 has the "detached_state" attribute removed from
332 sysfs, which was used to recognize sysfs population. We switched that
333 to wait for the "bus" link, which is only available in kernels after 2.6.11.
334 Running this udev version on older kernels may cause a short delay for
337 o To provide infrastructure for persistent device naming, the id programs:
338 scsi_id, vol_id (former udev_volume_id), and ata_id (new) are able now
339 to export the probed data in environment key format:
340 pim:~ # /sbin/ata_id --export /dev/hda
341 ID_MODEL=HTS726060M9AT00
342 ID_SERIAL=MRH401M4G6UM9B
346 KERNEL="hd*[!0-9]", IMPORT="/sbin/ata_id --export $tempnode"
347 KERNEL="hd*[!0-9]", ENV{ID_SERIAL}=="?*", SYMLINK+="disk/by-id/$env{ID_MODEL}_$env{ID_SERIAL}"
350 kay@pim:~> tree /dev/disk
353 | |-- HTS726060M9AT00_MRH401M4G6UM9B -> ../../hda
354 | `-- IBM-Memory_Key -> ../../sda
356 | |-- swap -> ../../hda1
357 | |-- date -> ../../sda1
358 | `-- home -> ../../hda3
360 |-- 2E08712B0870F2E7 -> ../../hda3
361 |-- 9352cfef-7687-47bc-a2a3-34cf136f72e1 -> ../../hda1
362 |-- E845-7A89 -> ../../sda1
363 `-- b2a61681-3812-4f13-a4ff-920d70604299 -> ../../hda2
365 The IMPORT= operation will import these keys in the environment and make
366 it available for later PROGRAM= and RUN= executed programs. The keys are
367 also stored in the udevdb and can be queried from there with one of the
370 o A few binaries are silently added to the repository, which can be used
371 to replay kernel events from initramfs instead of using coldplug. udevd
372 can be instructed now to queue-up events while the stored events from
373 initramfs are filled into the udevd-queue. This code is still under
374 development and there is no documentation now besides the code itself.
375 The additional binaries get compiled, but are not installed by default.
377 o There is also a temporary fix for a performance problem where too many
378 events happen in parallel and every event needs to parse the rules.
379 udev can now read precompiled rules stored on disk. This is likely to be
380 replaced by a more elegant solution in a future udev version.
384 With kernel version 2.6.12, the sysfs file "detached_state" was removed.
385 Fix for libsysfs not to expect this file was added.
389 All rules are applied now, but only the first matching rule with a NAME-key
390 will be applied. All later rules with NAME-key are completely ignored. This
391 way system supplied symlinks or permissions gets applied to user-defined
395 Please check your rules setup, if you may need to add OPTIONS="last_rule"
396 to some rules, to keep the old behavior.
398 The rules are read on "remove"-events too. That makes is possible to match
399 with keys that are available on remove (KERNEL, SUBSYSTEM, ID, ENV, ...) to
400 instruct udev to ignore an event (OPTIONS="ignore_device").
401 The new ACTION-key may be used to let a rule act only at a "remove"-event.
403 The new RUN-key supports rule-based execution of programs after device-node
404 handling. This is meant as a general replacement for the dev.d/-directories
405 to give fine grained control over the execution of programs.
407 The %s{}-sysfs format char replacement values are searched at any of the
408 devices in the device chain now, not only at the class-device.
410 We support log priority levels now. The value udev_log in udev.conf is used
411 to determine what is printed to syslog. This makes it possible to
412 run a version with compiled-in debug messages in a production environment
413 which is sometimes needed to find a bug.
414 It is still possible to supress the inclusion of _any_ syslog usage with
415 USE_LOG=false to create the smallest possible binaries if needed.
416 The configured udev_log value can be overridden with the environment variable
421 Possible use of a system-wide klibc:
422 make USE_KLIBC=true KLCC=/usr/bin/klcc all
423 will link against an external klibc and our own version will be ignored.
427 We support an unlimited count of symlinks now.
429 If USE_STATIC=true is passed to a glibc build, we link statically and use
430 a built-in userdb parser to resolve user and group names.
432 The PLACE= key is gone. It can be replaced by an ID= for a long time, cause
433 we walk up the chain of physical devices to find a match.
435 The KEY="<value>" format supports '=', '==', '!=,' , '+=' now. This makes it
436 easy to skip certain attribute matches without composing rules with weird
437 character class negations like:
438 KERNEL="[!s][!c][!d]*"
439 this can now be replaced with:
441 The current simple '=' is still supported, and should work as it does today,
442 but existing rules should be converted if possible, to be better readable.
444 We have new ENV{}== key now, to match against a maximum of 5 environment
447 udevstart is its own binary again, cause we don't need co carry this araound
448 with every forked event.