Instead of using systemd oneshot services that have to be careful to not
toggle wakeups back on, use a udev rule to disable wakeups by device ID.
On a B550 Vision D, these do almost the same thing:
````
$ lspci -n | grep 1022:1483
00:01.1 0604: 1022:1483
00:01.2 0604: 1022:1483
00:03.1 0604: 1022:1483
$ cat /proc/acpi/wakeup
Device S-state Status Sysfs node
...
GPP0 S4 *disabled pci:0000:00:01.1
GPP8 S4 *disabled pci:0000:00:03.1
````
Two of the three devices with the PCI vendor/device ID specified by the
udev rule correspond to devices previously disabled via ACPI (if I
understand correctly disabling these via either /proc/acpi/wakeup or
udev device attribute has the same effect).
The third device is (like the other two) using the "pcieport" driver.
Using a device connected via that port as a wakeup device still works.
fixes b550 suspend bug by setting up systemd services that disable GPP0 and GPP8 in /proc/acpi/wakeup
Co-authored-by: Jörg Thalheim <Mic92@users.noreply.github.com>
