One of the assertions in set_sub_devices_sync can be hit in the wild.
The crash-stack doesn't show which assertion is hit. My guess is that
the newly created aggregate-device may not be ready yet, to be used to
watch its own property-change, even if the created device is already in
the process-wide device list.
This patch tries to address BMO 1658982.
One of the assertions in
set_sub_devices_sync
can be hit in the wild. The crash-stack doesn't show which assertion is hit. My guess is that the newly created aggregate-device may not be ready yet, to be used to watch its own property-change, even if the created device is already in the process-wide device list.@kinetiknz : r?