Re: iwd needs restart after suspending

[Denis Kenzior <denkenz@gmail.com>]

Hi Joachim,

> I recorded the output of iwd -d on 5.15 and 5.19, see attachement. The
> relevant bit seems to be
> 

We should add this to our debug output, but what driver / hardware are you using?

> src/station.c:station_enter_state() Old State: connected, new state: disconnected
> src/station.c:station_enter_state() Old State: disconnected, new state: autoconnect_quick
> src/wiphy.c:wiphy_radio_work_insert() Inserting work item 3
> src/wiphy.c:wiphy_radio_work_next() Starting work item 3
> src/wiphy.c:wiphy_reg_notify() Notification of command Wiphy Reg Change(113)
> src/wiphy.c:wiphy_update_reg_domain() New reg domain country code for phy0 is XX
> src/scan.c:scan_notify() Scan notification Trigger Scan(33)
> src/netconfig.c:netconfig_ifaddr_deleted() ifaddr 192.168.178.39/24
> src/scan.c:scan_request_triggered() Active scan triggered for wdev 6
> src/station.c:station_quick_scan_triggered() Quick scan triggered for wlan0
> src/scan.c:scan_notify() Scan notification New Scan Results(34)
> src/netdev.c:netdev_link_notify() event 16 on ifindex 5
> src/scan.c:get_scan_callback() get_scan_callback
> src/scan.c:scan_parse_bss_information_elements() Load: 31/255
> src/scan.c:get_scan_callback() get_scan_callback
> src/scan.c:get_scan_callback() get_scan_callback
> 
>                             

So with 5.15, the underlying netdev is disconnected.  We get notified of that 
and start scanning.

> with 5.15 and
> 
> src/station.c:station_enter_state() Old State: connected, new state:
> disconnected
> src/station.c:station_enter_state() Old State: disconnected, new state:
> autoconnect_quick
> src/wiphy.c:wiphy_radio_work_insert() Inserting work item 3
> src/wiphy.c:wiphy_radio_work_next() Starting work item 3
> src/netdev.c:netdev_link_notify() event 16 on ifindex 3

Here, we receive what looks to be an IFDOWN event.

> src/station.c:station_free()
> src/netconfig.c:netconfig_destroy()
> src/scan.c:scan_cancel() Trying to cancel scan id 3 for wdev 3
> src/scan.c:scan_cancel() Scan is already started
> src/wiphy.c:wiphy_radio_work_done() Work item 3 done
> src/station.c:station_roam_state_clear() 3
> src/rrm.c:rrm_station_watch_destroy()
> 

So the question is why is the kernel changing this behavior?  iwd is not aware 
of resume/suspend, so I'm not sure we can tell whether the ifdown event came 
from the user manually bringing down the device or due to suspend / resume.

What does rfkill look like on 5.15 & 5.19 after resume?

Regards,
-Denis