2.6.6-rc3 still oops on unplugging usb bluetooth bcm203x dongle
Hi...
I still get:
usb 2-1: USB disconnect, address 3
Oops: kernel access of bad area, sig: 11 [#1] NIP: C02134B4 LR: F205D414 SP: EFE87DD0 REGS: efe87d20 TRAP: 0600 Not tainted
MSR: 00009032 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
DAR: 6B6B6BB7, DSISR: 00000120
TASK = effa4030[5] 'khubd' THREAD: efe86000Last syscall: -1
GPR00: FFFF0001 EFE87DD0 EFFA4030 EE77C828 6B6B6B6B 00000000 EB8EE83C 00000000
GPR08: 00001388 EF0EE858 00010C00 C0213480 82008022 00000000 00000000 00000000
GPR16: 00000000 00000000 00000000 00000000 00000000 00220000 00230000 00000000
GPR24: 00000000 C0400000 00000001 6B6B6B6B 6B6B6BB7 EF07B8A0 EE77C828 EE77C6FC NIP [c02134b4] class_device_del+0x34/0x140
LR [f205d414] hci_unregister_sysfs+0x14/0x24 [bluetooth]
Call trace: [f205d414] hci_unregister_sysfs+0x14/0x24 [bluetooth] [f205876c] hci_unregister_dev+0x18/0xb0 [bluetooth] [f204cd94] hci_usb_disconnect+0x48/0x90 [hci_usb] [c0277a24] usb_unbind_interface+0x88/0x8c [c02125a4] device_release_driver+0x84/0x88 [c0212744] bus_remove_device+0x74/0xd0 [c0211120] device_del+0xa8/0x114 [c02111a4] device_unregister+0x18/0x30 [c027e248] usb_disable_device+0x9c/0xd8 [c0278768] usb_disconnect+0x9c/0x134 [c027ab14] hub_port_connect_change+0x294/0x298 [c027adec] hub_events+0x2d4/0x354 [c027aea8] hub_thread+0x3c/0xf0 [c00090b0] kernel_thread+0x44/0x60
Sometimes it helps to hciconfig hci0 down the bluetooth dongle + stop all programs + rmmod them... However also the
rmmod hci_usb rfcomm bluetooth firmware_class bcm203x l2cap
is very likely to cause the same oops...
Yes, this is on powerpc but quite a number of people have the same issue on x86 (with 2.6.5 at least... and thats around the time this oops on unplugging appeared).
Any ideas ?
Soeren
Quoted 1 line · Soeren Sonnenburg
Hi...
Maybe that additional information helps...
That is when I one by one rmmod:
history | grep rmmod
5703 rmmod bcm203x 5704 rmmod firmware_class 5705 rmmod rfcomm 5706 rmmod l2cap 5707 rmmod bluetooth 5709 rmmod hci_usb
that is the lsmod before I rmmod hci_usb
lsmod
Module Size Used by hci_usb 13824 0 bluetooth 55932 1 hci_usb
oops occurs after the rmmod hci_usb. looks to me as if the bluetooth module is not happy with that.
usbcore: deregistering driver hci_usb
Oops: kernel access of bad area, sig: 11 [#1] NIP: C02134B4 LR: F2064414 SP: C5C2DD70 REGS: c5c2dcc0 TRAP: 0600 Not tainted
MSR: 00009032 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
DAR: 6B6B6BB7, DSISR: 00000120
TASK = e1c54cf0[3957] 'rmmod' THREAD: c5c2c000Last syscall: 129
GPR00: FFFF0001 C5C2DD70 E1C54CF0 EE5A4244 EDFCF1DC 00000000 C12F09FC 00000000
GPR08: FFFFFFF7 C113B2FC 00000001 C0213480 22000222 10019FF0 00000000 100C0000
GPR16: 00000000 00000000 10204108 102044A8 10203068 101D81E8 10001400 00000000
GPR24: 10000000 00000000 00000000 6B6B6B6B 6B6B6BB7 C8D0561C EE5A4244 EE5A4118 NIP [c02134b4] class_device_del+0x34/0x140
LR [f2064414] hci_unregister_sysfs+0x14/0x24 [bluetooth]
Call trace: [f2064414] hci_unregister_sysfs+0x14/0x24 [bluetooth] [f205f76c] hci_unregister_dev+0x18/0xb0 [bluetooth] [f204ed94] hci_usb_disconnect+0x48/0x90 [hci_usb] [c0277a24] usb_unbind_interface+0x88/0x8c [c02125a4] device_release_driver+0x84/0x88 [c02125d4] driver_detach+0x2c/0x50 [c02128d0] bus_remove_driver+0x50/0xa8 [c0212cb8] driver_unregister+0x18/0x78 [c0277b20] usb_deregister+0x38/0x50 [f204edf4] hci_usb_exit+0x18/0x4cc [hci_usb] [c003501c] sys_delete_module+0x1a4/0x224 [c0005d60] ret_from_syscall+0x0/0x44
Quoted 42 lines
I still get:
usb 2-1: USB disconnect, address 3
Oops: kernel access of bad area, sig: 11 [#1] NIP: C02134B4 LR: F205D414 SP: EFE87DD0 REGS: efe87d20 TRAP: 0600 Not taintedMSR: 00009032 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
DAR: 6B6B6BB7, DSISR: 00000120
TASK = effa4030[5] 'khubd' THREAD: efe86000Last syscall: -1
GPR00: FFFF0001 EFE87DD0 EFFA4030 EE77C828 6B6B6B6B 00000000 EB8EE83C 00000000
GPR08: 00001388 EF0EE858 00010C00 C0213480 82008022 00000000 00000000 00000000
GPR16: 00000000 00000000 00000000 00000000 00000000 00220000 00230000 00000000
GPR24: 00000000 C0400000 00000001 6B6B6B6B 6B6B6BB7 EF07B8A0 EE77C828 EE77C6FC NIP [c02134b4]class_device_del+0x34/0x140
LR [f205d414]hci_unregister_sysfs+0x14/0x24[bluetooth]Call trace: [f205d414] hci_unregister_sysfs+0x14/0x24 [bluetooth] [f205876c] hci_unregister_dev+0x18/0xb0 [bluetooth] [f204cd94] hci_usb_disconnect+0x48/0x90 [hci_usb] [c0277a24] usb_unbind_interface+0x88/0x8c [c02125a4] device_release_driver+0x84/0x88 [c0212744] bus_remove_device+0x74/0xd0 [c0211120] device_del+0xa8/0x114 [c02111a4] device_unregister+0x18/0x30 [c027e248] usb_disable_device+0x9c/0xd8 [c0278768] usb_disconnect+0x9c/0x134 [c027ab14] hub_port_connect_change+0x294/0x298 [c027adec] hub_events+0x2d4/0x354 [c027aea8] hub_thread+0x3c/0xf0 [c00090b0] kernel_thread+0x44/0x60Sometimes it helps to hciconfig hci0 down the bluetooth dongle + stop all programs + rmmod them... However also the
rmmodhci_usbrfcomm bluetoothfirmware_classbcm203x l2capis very likely to cause the same oops...
Yes, this is on powerpc but quite a number of people have the same issue on x86 (with 2.6.5 at least... and thats around the time this oops on unplugging appeared).
Any ideas ?
Soeren
Quoted 16 lines · Soeren Sonnenburg
Hi...
I still get:
usb 2-1: USB disconnect, address 3
Oops: kernel access of bad area, sig: 11 [#1] NIP: C02134B4 LR: F205D414 SP: EFE87DD0 REGS: efe87d20 TRAP: 0600 Not taintedMSR: 00009032 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
DAR: 6B6B6BB7, DSISR: 00000120
TASK = effa4030[5] 'khubd' THREAD: efe86000Last syscall: -1
GPR00: FFFF0001 EFE87DD0 EFFA4030 EE77C828 6B6B6B6B 00000000 EB8EE83C 00000000
GPR08: 00001388 EF0EE858 00010C00 C0213480 82008022 00000000 00000000 00000000
GPR16: 00000000 00000000 00000000 00000000 00000000 00220000 00230000 00000000
GPR24: 00000000 C0400000 00000001 6B6B6B6B 6B6B6BB7 EF07B8A0 EE77C828 EE77C6FC NIP [c02134b4]class_device_del+0x34/0x140
LR [f205d414]hci_unregister_sysfs+0x14/0x24[bluetooth]
Quoted 4 lines
Call trace: [f205d414] hci_unregister_sysfs+0x14/0x24 [bluetooth] [f205876c] hci_unregister_dev+0x18/0xb0 [bluetooth] [f204cd94] hci_usb_disconnect+0x48/0x90 [hci_usb]
Marcel Holtman is working on this; some other people have reported the same problem. Have you been in touch with him? It appears to be a problem in the Bluetooth driver, not in the USB stack.
Alan STern
Quoted 23 lines · Alan Stern
Quoted 16 lines · Soeren Sonnenburg
Hi...
I still get:
usb 2-1: USB disconnect, address 3
Oops: kernel access of bad area, sig: 11 [#1] NIP: C02134B4 LR: F205D414 SP: EFE87DD0 REGS: efe87d20 TRAP: 0600 Not taintedMSR: 00009032 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
DAR: 6B6B6BB7, DSISR: 00000120
TASK = effa4030[5] 'khubd' THREAD: efe86000Last syscall: -1
GPR00: FFFF0001 EFE87DD0 EFFA4030 EE77C828 6B6B6B6B 00000000 EB8EE83C 00000000
GPR08: 00001388 EF0EE858 00010C00 C0213480 82008022 00000000 00000000 00000000
GPR16: 00000000 00000000 00000000 00000000 00000000 00220000 00230000 00000000
GPR24: 00000000 C0400000 00000001 6B6B6B6B 6B6B6BB7 EF07B8A0 EE77C828 EE77C6FC NIP [c02134b4]class_device_del+0x34/0x140
LR [f205d414]hci_unregister_sysfs+0x14/0x24[bluetooth]Quoted 4 lines
Call trace: [f205d414] hci_unregister_sysfs+0x14/0x24 [bluetooth] [f205876c] hci_unregister_dev+0x18/0xb0 [bluetooth] [f204cd94] hci_usb_disconnect+0x48/0x90 [hci_usb]
Quoted 3 lines
Marcel Holtman is working on this; some other people have reported the same problem. Have you been in touch with him? It appears to be a problem in the Bluetooth driver, not in the USB stack.
Are there any patches available? The most recent patch I could find is for 2.6.5, and does not change this problem.
I would like to try the latest version of the bluez-patch, if this can help to find the bug.
LLAP, Martin
Quoted 11 lines · Martin Hermanowski
Quoted 4 lines · Alan Stern
Marcel Holtman is working on this; some other people have reported the same problem. Have you been in touch with him? It appears to be a problem in the Bluetooth driver, not in the USB stack.
Are there any patches available? The most recent patch I could find is for 2.6.5, and does not change this problem.
I would like to try the latest version of the bluez-patch, if this can help to find the bug.
As I said, Marcel is working on it. You should ask him there are any patches available (but I don't think he's gotten that far yet).
Alan Stern