<div dir="auto"><div style="color:rgb(80,0,80);font-family:sans-serif;font-size:12.8px" dir="auto">On Sun, Jul 12, 2020 at 3:08 AM Aleksander Morgado <<a href="mailto:aleksander@aleksander.es" style="text-decoration-line:none;color:rgb(66,133,244)">aleksander@aleksander.es</a>> wrote: > > Hey! > > > I've attached a patch which uses the "custom regex" feature of the > > existing parser in order to detect PPP and say that things are good. > > A few things to note: > > > > * I created the regex based on data I saw output from my modem. At > > best it is correct, at worst it won't do any harm. > > I find this regex is extremely specific to your usecase. As you said > it may be a fix, or it may not do anything, that isn't a complete > solution, and I think we should try to find some other way to handle > this if possible. > > I find it very very weird that the modem doesn't reply a CONNECT > response after the ATD, that would be a very weird firmware bug! Maybe > the TTY we're using for connection isn't the correct one? @Daniele > Palmas @Carlo Lobrano Do you guys have any idea of why this may be > happening in the ME310? </div>There are two ports which accept AT commands, and both behave in the same fashion. Here's some more information on the modem: Output of AT#SWPKGV: 37.00.210-B038-P0C.210000 M0C.200000-B038 P0C.210000 A0C.210000 ATI0: 332 ATI1: 0 ATI2: ATI3: Telit ATI4: ME310G1-WW ATI5: DOB v.2.0 I would not be surprised if the "-WW" part has different functionality. I issued the AT commands in picocom as well to confirm that CONNECT was never sent.<div style="color:rgb(80,0,80);font-family:sans-serif;font-size:12.8px" dir="auto"> > > * Ideally I would only use this for the ME310 modem, but I didn't see > > a way to do that in the plugin. I'm certainly open to suggestions. > > If we needed a quirk like this one, we could flag via udev with a new udev tag. > > > This particular device has a second port configuration (1bc7:1102), > > which which looks like this: > > > > /: Bus 01.Port 1: Dev 1, Class=root_hub, Driver=atmel-ehci/3p, 480M > > |__ Port 2: Dev 7, If 0, Class=Vendor Specific Class, Driver=option, 480M > > |__ Port 2: Dev 7, If 1, Class=Vendor Specific Class, Driver=option, 480M > > |__ Port 2: Dev 7, If 2, Class=Vendor Specific Class, Driver=option, 480M > > |__ Port 2: Dev 7, If 3, Class=Communications, Driver=cdc_ether, 480M > > |__ Port 2: Dev 7, If 4, Class=CDC Data, Driver=cdc_ether, 480M > > > > I've not been able to get the cdc_ether portion working yet, although > > I am hopeful. If both ECM and "option" serial interfaces are > > available, how does ModemManager choose between them? > > MM doesn't have any implementation yet to handle ECM interfaces in > Telit modems, but that would definitely be preferred over PPP in a > TTY. </div>I would also prefer to use ECM over pppd, but I've not found a way to make this work yet.<div style="color:rgb(80,0,80);font-family:sans-serif;font-size:12.8px" dir="auto"> > Regarding the patch itself, you shouldn't add the regex to the primary > port. You should run mm_base_modem_peek_data_ports() and add the regex > to all data ports that are TTYs (MM_IS_PORT_SERIAL()). But as said, I > would totally prefer to have this solved somehow without needing to > detect the PPP stream, because it is pppd the only one who should be > touching that (MM yields the control of the TTY to pppd as soon as the > port is in data mode). </div>This is easy enough to do, but I will focus on getting ECM working first. Thanks for the response, Aleksander.<div style="color:rgb(136,136,136);font-family:sans-serif;font-size:12.8px" dir="auto"> -M</div></div>