> At this point the disconnection logic has finished, but the CGACT? > that got in the middle never gets a reply. After this, all commands we > attempt to send to ttyUSB1 end up timing out, and after 10 consecutive > timeouts, MM removes the modem from its logic because it cannot use > the primary control port. > > Now, in this specific case, given that the connection status command > is triggered *during* the disconnection process, I believe this could > be caused by MM. > I believe that we could either remove those connection checks also if > bearer state goes to "disconnecting", or otherwise keep them removed > only when "disconnected" but make sure they are a no-op while > disconnecting. I think the latter would be a bit safer, as we wouldn't > need to care for when the disconnection procedure fails. > > Let me prepare a patch to see if we can avoid this. >
Could you test with the following patch to see if the stability improves? https://gitlab.freedesktop.org/mobile-broadband/ModemManager/merge_requests/229 -- Aleksander https://aleksander.es _______________________________________________ ModemManager-devel mailing list ModemManager-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/modemmanager-devel
