Udev-154

[Matthew Burgess]

New version. Release announcement at ​http://www.spinics.net/lists/hotplug/msg03684.html.

[Matthew Burgess]

The only bit of the announcement that seems to apply to our bootscript & rules is the following:

"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.

[Bryan Kadzban]

Bootscripts fixed in r9256.

However, the rule that calls setclock might be an issue here. If (a future version of) udev makes it hard to do "add" rules properly somehow, under the assumption that everyone uses devtmpfs and so add rules are "wrong" somehow, we'll need to change this to a "!remove" rule. (Or change it to an "add|change" rule, though it doesn't look like that's preferred anymore either.)

And that will rerun the bootscript if udev does happen to get a change event for the rtc device node(s). Which is horribly wrong, if you're using NTP. :-(

We can't really save any state away either (to make setclock only do its work once per boot), since this script runs during udev, and before mountfs. And it won't work to compare the hardware clock time to the system clock time, since if the hwclock is five hours off, you don't want to break the system on every change event.

We might be able to save something in the udev database though; hmm. Maybe:

SUBSYSTEM=="rtc", ACTION=="!remove", ENV{CLOCK_SET}!="1" MODE="0644", \
    RUN+="/etc/rc.d/init.d/setclock start", ENV{CLOCK_SET}="1"
KERNEL=="rtc", ACTION=="!remove", ENV{CLOCK_SET}!="1" MODE="0644", \
    RUN+="/etc/rc.d/init.d/setclock start", ENV{CLOCK_SET}="1"

Opinions? Better ideas?

(Also, I should note that udev-152 has a bug: ​http://www.spinics.net/lists/hotplug/msg03698.html -- we probably want to wait for -153.)

[Matthew Burgess]

Now 153. Release announcement at ​http://www.spinics.net/lists/hotplug/msg03699.html.

[Gilles Espinasse]

udev-153 has a broken empty FIRMWARE_PATH instead of /lib/firmware/update:/lib/firmware

I have a e100 netcard and firmware fail to load.

Probably wait for 154 or borrow the patch from

​http://marc.info/?l=linux-hotplug&m=127205071208335&w=2

[Steffen Pankratz]

Udev-154 was released yesterday:

​http://www.spinics.net/lists/hotplug/msg03766.html

[Matthew Burgess]

Fixed in r9271.