Backport systemd to make uevents "sticky"

[hugh chao]

Base on lwn 837033 [0], Linux 4.12 introduced two new uevents "bind" and "unbind" to the Linux device model which resulted in a number of software issues.

So in this PR [1] which minimize issues resulting from this kernel change (but not avoid them entirely) by make uevent "sticky", meaning that once a tag is assigned to a device it will not be removed from the device again until the device itself is removed (i.e. unplugged).

An example causing by this issue, I had a usb printer (non-ippusbxd) which connected to my laptop, but after unplugging it from my laptop, the .device unit were still present, if a different printer gets plugged to the same USB port then, it won't be configured correctly.

[0] https://lwn.net/Articles/837033/
[1] https://github.com/systemd/systemd/pull/16853/files

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Dan Streetman (ddstreet)]

I'm hesitant to backport, because the change causes significant issues on upgrade, e.g.:
https://github.com/systemd/systemd/issues/17605

if we do backport the 'sticky' udev tags, we'll need to make sure to also pull back the proper commits to handle the new udev format across daemon-reexec, e.g.:
https://github.com/systemd/systemd/pull/17622

[Dan Streetman (ddstreet)]

As I mentioned in my last comment, I'm skeptical that the benefit of backporting would outweigh the potential problems it would introduce to existing stable releases.

I'm marking this as wont-fix for that reason. I don't think we should consider this unless you can provide specific examples of significant, serious problems that would be fixed by backporting the change.

[Dan Streetman (ddstreet)]

for reference, the PR linked in the description is included upstream in v247 so this is fixed in h, i'll mark the bug as such

[hugh chao (hugh712)]

@Dan

I see, thank you anyway :)