I have a Sony Vaio Z 2010 model (VPCZ1). The internal Bluetooth adapter is, I believe, part of this USB device: Bus 001 Device 006: ID 0489:e00f Foxconn / Hon Hai I'll attach an lsusb -v dump. With both a mostly vanilla upstream kernel 2.6.36-rc5 (actually Jesse Barnes' edp-fixes tree which has various fixes for the graphics on the system, but the Bluetooth stuff should be unchanged), and Fedora kernel 2.6.36-0.20.rc3.git4.fc15 , I can't reliably use a Bluetooth mouse with the system. From the user POV what happens is I pair the mouse, using GNOME's Bluetooth stuff, and it comes up working right away. I think it works as long as I keep it active, but I'm not 100% sure. Shortly after - I think when the mouse goes idle - it stops working. Sometimes, if I move it or switch it off and then on again, I get an authorization request from the mouse via gnome-bluetooth; if I permit the access, the mouse still doesn't work. From /var/log/messages , it seems like the Bluetooth bus gets reset after some 'hci0 urb failed to resubmit' messages, and after it comes up again the mouse never works. I'll attach the /var/log/messages . I pair the device at Oct 10 15:13 and it stops working at 15:17:54, with the 'hcio0 urb failed to resubmit' messages. The authorization request is I think the 'link_key_request' stuff. The dbus errors may be interesting, but they definitely occur *after* the mouse stops working.
Created attachment 33152 [details] /var/log/messages extract; see report for relevant timestamps
Created attachment 33162 [details] lsusb -v output
This is fixed in 2.6.37-rc4 (maybe earlier, didn't run anything between 2.6.36 and 2.6.37rc4). In fact it seems to have been a general 'Bluetooth is screwed on this system' bug, because Bluetooth DUN also failed with 2.6.36 but worked with 2.6.37. Yay. Of course, 2.6.37 managed to break sound, so now I can pick between working sound and working Bluetooth =)