Hello, Le mercredi 13 février 2008 à 10:58 +0530, pramod gurav a écrit : > Feb 13 09:52:43 localhost kernel: hci_scodata_packet: hci0 SCO packet > for unknown connection handle 1 > > Feb 13 09:52:43 localhost kernel: hci_scodata_packet: hci0 SCO packet > for unknown connection handle 1 Which kernel do you use now ? You may have the same issue as me with the Motorola HT820 headset. I have exactly the same type of messages, at the same time. > [root@localhost Pramod]# /usr/local/sbin/hciconfig -a > hci0: Type: USB > BD Address: 00:02:72:CE:5B:80 ACL MTU: 1017:8 SCO MTU: 64:0 > UP RUNNING PSCAN ISCAN > RX bytes:5974443 acl:279 sco:116978 events:188 errors:0 > TX bytes:4222 acl:168 sco:0 commands:90 errors:0 > Features: 0xff 0xff 0x8d > 0xfe It seems like me, your bluetooth dongle supports eSCO, but not your headset. (from http://support.zenwalk.org/index.php?topic=2060.0;wap2) Device Name: HBH-600 LMP Version: 1.1 (0x1) LMP Subversion: 0x9240 Manufacturer: Philips Semiconductors (37) Features: 0xbc 0x28 0x01 0x00 0x00 0x00 0x00 0x00 First, you can try a old bluetooth dongle =< 1.2 and not patch anything if you have one. Alternatively, you can try a slightly older kernel or patch your kernel to revert "patch-2.6.23-mh1", that will remove eSCO support, but also a lot of other things. I have an older kernel without this patch included ( 2.6.22.18 from mandriva ) that works fine with both of my headsets. Or if you feel adventurous, you can try the attached patch of mine (you have to know i am not a kernel dev but just a bluez user), which may solve your problem with *this headset*. Unfortunately, it will probably cause others with eSCO capable headsets, because a SCO link is always used the first time the headset connects itself, as the headset features are not known yet (mine, PROF PBH-6W, works only since the second try with this patch). Regards, Guillaume B.