X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?p=elogind.git;a=blobdiff_plain;f=NEWS;h=301e725269dd6201064300d10b0c3cd3462b3f1d;hp=2ce846b58674acd04a11ce5d48c2a48224434944;hb=d21781c4576e161ada21447e4d85d82b86b381b4;hpb=804ab68550c5779a0a27b1ad2636184520f2a1cb diff --git a/NEWS b/NEWS index 2ce846b58..301e72526 100644 --- a/NEWS +++ b/NEWS @@ -1,7 +1,367 @@ -udev 145 +udev 171 ======== Bugfixes. +The systemd service files require systemd version 28. The systemd +socket activation make it possible now to start 'udevd' and 'udevadm +trigger' in parallel. + +udev 170 +======== +Fix bug in control message handling, which can lead to a failing +udevadm control --exit. Thanks to Jürg Billeter for help tracking +it down. + +udev 169 +======== +Bugfixes. + +We require at least Linux kernel 2.6.32 now. Some platforms might +require a later kernel that supports accept4() and similar, or +need to backport the trivial syscall wiring to the older kernels. + +The hid2hci tool moved to the bluez package and was removed. + +Many of the extras can be --enable/--disabled at ./configure +time. The --disable-extras option was removed. Some extras have +been disabled by default. The current options and their defaults +can be checked with './configure --help'. + +udev 168 +======== +Bugfixes. + +Udev logs a warning now if /run is not writable at udevd +startup. It will still fall back to /dev/.udev, but this is +now considered a bug. + +The running udev daemon can now cleanly shut down with: + udevadm control --exit + +Udev in initramfs should clean the state of the udev database +with: udevadm info --cleanup-db which will remove all state left +behind from events/rules in initramfs. If initramfs uses +--cleanup-db and device-mapper/LVM, the rules in initramfs need +to add OPTIONS+="db_persist" for all dm devices. This will +prevent removal of the udev database for these devices. + +Spawned programs by PROGRAM/IMPORT/RUN now have a hard timeout of +120 seconds per process. If that timeout is reached the spawned +process will be killed. The event timeout can be overwritten with +udev rules. + +If systemd is used, udev gets now activated by netlink data. +Systemd will bind the netlink socket which will buffer all data. +If needed, such setup allows a seemless update of the udev daemon, +where no event can be lost during a udevd update/restart. +Packages need to make sure to: systemctl stop udev.socket udev.service +or 'mask' udev.service during the upgrade to prevent any unwanted +auto-spawning of udevd. +This version of udev conflicts with systemd version below 25. The +unchanged service files will not wirk correctly. + +udev 167 +======== +Bugfixes. + +The udev runtime data moved from /dev/.udev/ to /run/udev/. The +/run mountpoint is supposed to be a tmpfs mounted during early boot, +available and writable to for all tools at any time during bootup, +it replaces /var/run/, which should become a symlink some day. + +If /run does not exist, or is not writable, udev will fall back using +/dev/.udev/. + +On systemd systems with initramfs and LVM used, packagers must +make sure, that the systemd and initramfs versions match. The initramfs +needs to create the /run mountpoint for udev to store the data, and +mount this tmpfs to /run in the rootfs, so the that the udev database +is preserved for the udev version started in the rootfs. + +The command 'udevadm info --convert-db' is gone. The udev daemon +itself, at startup, converts any old database version if necessary. + +The systemd services files have been reorganized. The udev control +socket is bound by systemd and passed to the started udev daemon. +The udev-settle.service is no longer active by default. Services which +can not handle hotplug setups properly need to actively pull it in, to +act like a barrier. Alternatively the settle service can be unconditionally +'systemctl'enabled, and act like a barrier for basic.target. + +The fstab_import callout is no longer built or installed. Udev +should not be used to mount, does not watch changes to fstab, and +should not mirror fstab values in the udev database. + +udev 166 +======== +Bugfixes. + +New and updated keymaps. + +udev 165 +======== +Bugfixes. + +The udev database has changed, After installation of a new udev +version, 'udevadm info --convert-db' should be called, to let the new +udev/libudev version read the already stored data. + +udevadm now supports quoting of property values, and prefixing of +key names: + $ udevadm info --export --export-prefix=MY_ --query=property -n sda + MY_MAJOR='259' + MY_MINOR='0' + MY_DEVNAME='/dev/sda' + MY_DEVTYPE='disk' + ... + +libudev now supports: + udev_device_get_is_initialized() + udev_enumerate_add_match_is_initialized() +to be able to skip devices the kernel has created , but udev has +not already handled. + +libudev now supports: + udev_device_get_usec_since_initialized() +to retrieve the "age" of a udev device record. + +GUdev supports a more generic GUdevEnumerator class, udev TAG +handling, device initialization and timestamp now. + +The counterpart of /sys/dev/{char,block}/$major:$minor, +/dev/{char,block}/$major:$minor symlinks are now unconditionally +created, even when no rule files exist. + +New and updated keymaps. + +udev 164 +======== +Bugfixes. + +GUdev moved from /usr to /. + +udev 163 +======== +Bugfixes. + +udev 162 +======== +Bugfixes. + +Persistent network naming rules are disabled inside of Qemu/KVM now. + +New and updated keymaps. + +Udev gets unconditionally enabled on systemd installations now. There +is no longer the need to to run 'systemctl enable udev.service'. + +udev 161 +======== +Bugfixes. + +udev 160 +======== +Bugfixes. + +udev 159 +======== +Bugfixes. + +New and fixed keymaps. + +Install systemd service files if applicable. + +udev 158 +======== +Bugfixes. + +All distribution specific rules are removed from the udev source tree, +most of them are no longer needed. The Gentoo rules which allow to support +older kernel versions, which are not covered by the default rules anymore +has moved to rules/misc/30-kernel-compat.rules. + +udev 157 +======== +Bugfixes. + +The option --debug-trace and the environemnt variable UDEVD_MAX_CHILDS= +was removed from udevd. + +Udevd now checks the kernel commandline for the following variables: + udev.log-priority= + udev.children-max= + udev.exec-delay= +to help debuging coldplug setups where the loading of a kernel +module crashes the system. + +The subdirectory in the source tree rules/packages has been renamed to +rules/arch, anc contains only architecture specific rules now. + +udev 156 +======== +Bugfixes. + +udev 155 +======== +Bugfixes. + +Now the udev daemon itself, does on startup: + - copy the content of /lib/udev/devices to /dev + - create the standard symlinks like /dev/std{in,out,err}, + /dev/core, /dev/fd, ... + - use static node information provided by kernel modules + and creates these nodes to allow module on-demand loading + - possibly apply permissions to all ststic nodes from udev + rules which are annotated to match a static node + +The default mode for a device node is 0600 now to match the kernel +created devtmpfs defaults. If GROUP= is specified and no MODE= is +given the default will be 0660. + +udev 154 +======== +Bugfixes. + +Udev now gradually starts to pass control over the primary device nodes +and their names to the kernel, and will in the end only manage the +permissions of the node, and possibly create additional symlinks. +As a first step NAME="" will be ignored, and NAME= setings with names +other than the kernel provided name will result in a logged warning. +Kernels that don't provide device names, or devtmpfs is not used, will +still work as they did before, but it is strongly recommended to use +only the same names for the primary device node as the recent kernel +provides for all devices. + +udev 153 +======== +Fix broken firmware loader search path. + +udev 152 +======== +Bugfixes. + +"udevadm trigger" defaults to "change" events now instead of "add" +events. The "udev boot script" might need to add "--action=add" to +the trigger command if not already there, in case the initial coldplug +events are expected as "add" events. + +The option "all_partitons" was removed from udev. This should not be +needed for usual hardware. Udev can not safely make assumptions +about non-existing partition major/minor numbers, and therefore no +longer provide this unreliable and unsafe option. + +The option "ignore_remove" was removed from udev. With devtmpfs +udev passed control over device nodes to the kernel. This option +should not be needed, or can not work as advertised. Neither +udev nor the kernel will remove device nodes which are copied from +the /lib/udev/devices/ directory. + +All "add|change" matches are replaced by "!remove" in the rules and +in the udev logic. All types of events will update possible symlinks +and permissions, only "remove" is handled special now. + +The modem modeswitch extra was removed and the external usb_modeswitch +program should be used instead. + +New and fixed keymaps. + +udev 151 +======== +Bugfixes. + +udev 150 +======== +Bugfixes. + +Kernels with SYSFS_DEPRECATED=y are not supported since a while. Many users +depend on the current sysfs layout and the information not available in the +deprecated layout. All remaining support for the deprecated sysfs layout is +removed now. + +udev 149 +======== +Fix for a possible endless loop in the new input_id program. + +udev 148 +======== +Bugfixes. + +The option "ignore_device" does no longer exist. There is no way to +ignore an event, as libudev events can not be suppressed by rules. +It only prevented RUN keys from being executed, which results in an +inconsistent behavior in current setups. + +BUS=, SYSFS{}=, ID= are long deprecated and should be SUBSYSTEM(S)=, +ATTR(S){}=, KERNEL(S)=. It will cause a warning once for every rule +file from now on. + +The support for the deprecated IDE devices has been removed from the +default set of rules. Distros who still care about non-libata drivers +need to add the rules to the compat rules file. + +The ID_CLASS property on input devices has been replaced by the more accurate +set of flags ID_INPUT_{KEYBOARD,KEY,MOUSE,TOUCHPAD,TABLET,JOYSTICK}. These are +determined by the new "input_id" prober now. Some devices, such as touchpads, +can have several classes. So if you previously had custom udev rules which e. g. +checked for ENV{ID_CLASS}=="kbd", you need to replace this with +ENV{ID_INPUT_KEYBOARD}=="?*". + +udev 147 +======== +Bugfixes. + +To support DEVPATH strings larger than the maximum file name length, the +private udev database format has changed. If some software still reads the +private files in /dev/.udev/, which it shouldn't, now it's time to fix it. +Please do not port anything to the new format again, everything in /dev/.udev +is and always was private to udev, and may and will change any time without +prior notice. + +Multiple devices claiming the same names in /dev are limited to symlinks +only now. Mixing identical symlink names and node names is not supported. +This reduces the amount of data in the database significantly. + +NAME="%k" causes a warning now. It's is and always was completely superfluous. +It will break kernel supplied DEVNAMEs and therefore it needs to be removed +from all rules. + +Most NAME= instructions got removed. Kernel 2.6.31 supplies the needed names +if they are not the default. To support older kernels, the NAME= rules need to +be added to the compat rules file. + +Symlinks to udevadm with the old command names are no longer resolved to +the udevadm commands. + +The udev-acl tool got adopted to changes in ConsoleKit. Version 0.4.1 is +required now. + +The option "last_rule" does no longer exist. Its use breaks too many +things which expect to be run from independent later rules, and is an idication +that something needs to be fixed properly instead. + +The gudev API is no longer marked as experimental, +G_UDEV_API_IS_SUBJECT_TO_CHANGE is no longer needed. The gudev introspection +is enabled by default now. Various projects already depend on introspection +information to bind dynamic languages to the gudev interfaces. + +udev 146 +======== +Bugfixes. + +The udevadm trigger "--retry-failed" option, which is replaced since quite +a while by "--type=failed" is removed. + +The failed tracking was not working at all for a few releases. The RUN +option "ignore_error" is replaced by a "fail_event_on_error" option, and the +default is not to track any failing RUN executions. + +New keymaps, new modem, hid2hci updated. + +udev 145 +======== +Fix possible crash in udevd when worker processes are busy, rules are +changed at the same time, and workers get killed to reload the rules. + udev 144 ======== Bugfixes.