Return-Path: Message-ID: <4AB1EF78.4000708@koala.ie> Date: Thu, 17 Sep 2009 09:12:40 +0100 From: Simon Kenyon MIME-Version: 1.0 CC: BlueZ devel list Subject: Re: hci0 is invisible References: <200909161204.13479.gene.heskett@verizon.net> <20090916202224.3eef8bd3@strolchi.home.s3e.de> In-Reply-To: <20090916202224.3eef8bd3@strolchi.home.s3e.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed To: unlisted-recipients:; (no To-header on input) Sender: linux-bluetooth-owner@vger.kernel.org List-ID: Stefan Seyfried wrote: > Hi Gene, > > On Wed, 16 Sep 2009 12:04:13 -0400 > Gene Heskett wrote: > > >> Greetings Marcel; >> Yet another failed pass at getting these &*^$ dongles from Conwise >> Tech to work as a simple rs-232 link. >> >> I have simlinked the contents of /etc/bluetooth, /usr/etc/bluetooth, >> and /usr/local/etc/bluetooth so that regardless of where it might >> look for config files, it will find something. >> > > It should actually work without lots of config files. > > >> I've put the device on the other end of the path back into the >> non-paired state. It is an eb101, according to test-discovery, is: >> [root@coyote test]# ./test-discovery >> [ 00:0C:84:00:86:F8 ] >> Name = eb101 >> Paired = 0 >> LegacyPairing = 1 >> Alias = eb101 >> Address = 00:0C:84:00:86:F8 >> RSSI = 0 >> Class = 0x001f00 >> > > Very good, it means your adapter and the other side is up and running. > > >> [root@coyote tools]# hcitool inq >> Inquiring ... >> 00:0C:84:00:86:F8 clock offset: 0x5711 class: >> 0x001f00 [root@coyote tools]# hcitool cc 00:0C:84:00:86:F8 >> This last command can be repeated, with no errors reported. And no >> device can be created either in /dev, or in the link the messages >> file reports when the dongle is plugged in: >> > > This is a misunderstanding on your side. There is no such thing as a > device in /dev for the bluetooth adapter. Think of it like an ethernet > interface - you also don't have /dev/eth0 and still it works. > > >> Sep 16 11:39:04 coyote kernel: [564988.826049] usb 2-5.1: new full >> speed USB device using ohci_hcd and address 9 >> Sep 16 11:39:05 coyote kernel: [564988.919053] usb 2-5.1: New USB >> device found, idVendor=0e5e, idProduct=6622 >> Sep 16 11:39:05 coyote kernel: [564988.919057] usb 2-5.1: New USB >> device strings: Mfr=0, Product=0, SerialNumber=0 >> Sep 16 11:39:05 coyote kernel: [564988.919145] usb 2-5.1: >> configuration #1 chosen from 1 choice >> Sep 16 11:39:05 coyote bluetoothd[892]: HCI dev 0 registered >> Sep 16 11:39:05 coyote bluetoothd[892]: HCI dev 0 up >> Sep 16 11:39:05 coyote bluetoothd[892]: Starting security manager 0 >> Sep 16 11:39:05 coyote bluetoothd[892]: Parsing >> /usr/local/etc/bluetooth/serial.conf failed: Key file does not start >> with a group >> Sep 16 11:39:05 coyote bluetoothd[892]: Adapter /org/bluez/892/hci0 >> has been enabled >> > > >> Q: What defines this missing "group" in the serial.conf file?, which >> is now 100% commented. Copied from the 4.51 serial tree verbatum >> since it wasn't installed by a make install. >> > > It is not needed, since the defaults are just fine. > > >> Firing up bluetooth-wizard, the eb101 is displayed, and can be >> selected, but the pairing attempt fails. Pin on both ends is 0000. >> >> What can I do to make minicom find and use this hci0 device as a >> modem circuit? /dev/hci0 doesn't exist, and minicom can't find >> /org/bluez/892/hci0. >> > > You need to either create a rfcomm device with "rfcomm bind > " or use something like test-serial to set up the rfcomm > device for you. > > This rfcomm device (usually "/dev/rfcomm0") then is the "serial port" > that you hook up screen, or minicom or pppd or whatever. > > I have written something up about rfcomm in a former life on > http://en.opensuse.org/Bluetooth > (it is not suse specific at all, but it might be outdated), for your > case http://en.opensuse.org/Bluetooth/rfcomm might be even more > specific. > > Nowadays, you might also get good results with "test-serial". > At least on my phone "test-serial " gets me an /dev/rfcomm0 > which I can use to talk to it with AT commands. Be prepared that all > those test-* commands will cancel the connection after 1000 seconds, > but increasing the timeout is easy as it is a simple python script ;) > > The /org/bluez/892/hci0 is no filesystem path but a pointer to where > the device lives on the famous D-Bus ;) > > oh i so wish there was some "user level" documentation for bluez!