On Mon, Jan 02, 2017 at 11:25:14PM +0100, B&A Consultants wrote: > Hello, > > Kernels 4.1.35 and 4.4.39, on Gentoo boxes. > > We are trying to set up Yubikey-compatible (made by Plugup) U2F usb keys > on our systems, and the keys seems to be randomly detected by the usbhid > driver. Since it's for authentication, this seemingly random behaviour > is bothersome... > > Sometimes, the key is detected properly. It may be almost instantly (ie > plug the key and it's detected) or after a 5-second delay as below : > > 11:17:35 kernel: usb 3-2: new full-speed USB device number 9 using xhci_hcd > 11:17:35 kernel: usb 3-2: New USB device found, idVendor=2581, > idProduct=f1d0 > 11:17:35 kernel: usb 3-2: New USB device strings: Mfr=1, Product=1, > SerialNumber=1 > 11:17:35 kernel: usb 3-2: Product: Plug-up > 11:17:35 kernel: usb 3-2: Manufacturer: Plug-up > 11:17:35 kernel: usb 3-2: SerialNumber: Plug-up > 11:17:40 kernel: hid-generic 0003:2581:F1D0.0004: hiddev0,hidraw3: USB > HID v1.01 Device [Plug-up Plug-up] on usb-0000:00:14.0-2/input0 > > More often than not, the key isn't detected at all and we get an error > after a 20-second delay : > > 11:17:52 kernel: usb 3-2: new full-speed USB device number 10 using xhci_hcd > 11:17:53 kernel: usb 3-2: New USB device found, idVendor=2581, > idProduct=f1d0 > 11:17:53 kernel: usb 3-2: New USB device strings: Mfr=1, Product=1, > SerialNumber=1 > 11:17:53 kernel: usb 3-2: Product: Plug-up > 11:17:53 kernel: usb 3-2: Manufacturer: Plug-up > 11:17:53 kernel: usb 3-2: SerialNumber: Plug-up > 11:18:13 kernel: usbhid 3-2:1.0: can't add hid device: -110 > 11:18:13 kernel: usbhid: probe of 3-2:1.0 failed with error -110 > > This happens on all USB ports (USB2 or USB3), on several computers. > > When the key is properly detected, usb-devices says > > T: Bus=03 Lev=02 Prnt=13 Port=02 Cnt=03 Dev#= 16 Spd=12 MxCh= 0 > D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=16 #Cfgs= 1 > P: Vendor=2581 ProdID=f1d0 Rev=00.01 > S: Manufacturer=Plug-up > S: Product=Plug-up > S: SerialNumber=Plug-up > C: #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=100mA > I: If#= 0 Alt= 0 #EPs= 2 Cls=03(HID ) Sub=00 Prot=00 Driver=usbhid > > And, when it's not, we get > > T: Bus=01 Lev=02 Prnt=02 Port=03 Cnt=03 Dev#= 8 Spd=12 MxCh= 0 > D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=16 #Cfgs= 1 > P: Vendor=2581 ProdID=f1d0 Rev=00.01 > S: Manufacturer=Plug-up > S: Product=Plug-up > S: SerialNumber=Plug-up > C: #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=100mA > I: If#= 0 Alt= 0 #EPs= 2 Cls=03(HID ) Sub=00 Prot=00 Driver=(none) > > (note the 'Driver=(none)' on the last line). > > After activating dynamic_debug on the kernel, here is what we get when > the key is detected : > > 09:06:13 kernel: usb 3-6: new full-speed USB device number 13 using xhci_hcd > 09:06:13 kernel: usb 3-6: New USB device found, idVendor=2581, > idProduct=f1d0 > 09:06:13 kernel: usb 3-6: New USB device strings: Mfr=1, Product=1, > SerialNumber=1 > 09:06:13 kernel: usb 3-6: Product: Plug-up > 09:06:13 kernel: usb 3-6: Manufacturer: Plug-up > 09:06:13 kernel: usb 3-6: SerialNumber: Plug-up > 09:06:13 kernel: usbhid_restart_ctrl_queue:238: hid-generic > 0003:2581:F1D0.0008: Kicking head 1 tail 0 > 09:06:13 kernel: hid-generic 0003:2581:F1D0.0008: hiddev0,hidraw0: USB > HID v1.01 Device [Plug-up Plug-up] on usb-0000:00:14.0-6/input0 > > And when it is not : > > 09:08:02 kernel: usb 3-6: new full-speed USB device number 14 using xhci_hcd > 09:08:02 kernel: usb 3-6: New USB device found, idVendor=2581, > idProduct=f1d0 > 09:08:02 kernel: usb 3-6: New USB device strings: Mfr=1, Product=1, > SerialNumber=1 > 09:08:02 kernel: usb 3-6: Product: Plug-up > 09:08:02 kernel: usb 3-6: Manufacturer: Plug-up > 09:08:02 kernel: usb 3-6: SerialNumber: Plug-up > 09:08:22 kernel: usbhid 3-6:1.0: can't add hid device: -110 > 09:08:22 kernel: usbhid: probe of 3-6:1.0 failed with error -110 > > The 'usbhid_restart_ctrl_queue' line is present when everything is ok, > absent otherwise. > > During the 20-second delay, no other USB activity will be taken into > account (meaning, if we insert a USB storage key, it will be detected > after the 'blocking-delay' will be over). > > We configured dynamic_debug for a more verbosity, but the output is a > little too long (around 1000 lines for the ok and nok cases) for this > mailing list. We will send it directly if requested. > > Has anyone some idea of what's going on and, more importantly, how to > correct this ? > > If needed, we can test patches and perform any test you wizards may find > useful to pinpoint (and hopefully correct) this really annoying behaviour. Does this happen on all systems you try these devices in, or only on one type of system? And can you try a more "modern" kernel version? Lots of xhci changes and fixes have happened in the years since the kernels you are testing were released. thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
