On 10/18/2017 06:18 PM, nirinA wrote:
not really a rtlwifi oops then. maybe issue with udev as that also disabled the
usb mouse.
but i have no problem since. i will report if i will catch it again.
thanks,
--
nirinA
James Cameron wrote:
On Thu, Oct 19, 2017 at 01:31:43AM +0300, nirinA wrote:
hello there,
i got the oops below with a rtl8192cu:0bda:8178 and kernel 4.13.6, but
cannot reproduce it.
i use this device since 4.3 or so without noticing any issue.
nirinA
[ 239.338040] usb 2-1.3: new high-speed USB device number 4 using ehci-pci
[ 239.417728] usb 2-1.3: New USB device found, idVendor=0bda,
idProduct=8178
[ 239.417730] usb 2-1.3: New USB device strings: Mfr=1, Product=2,
SerialNumber=3
[ 239.417731] usb 2-1.3: Product: 802.11n WLAN Adapter
[ 239.417732] usb 2-1.3: Manufacturer: Realtek
[ 239.417733] usb 2-1.3: SerialNumber: 00e04c000001
[ 239.578100] rtl8192cu: Chip version 0x11
[ 239.678225] usb 2-1-port3: disabled by hub (EMI?), re-enabling...
Just prior to the oops, your USB hub disabled the port being used by
the wireless device.
While the response of the driver seems wrong, it is a difficult
condition to reproduce; one must either force or forge the disabling
by the hub.
[ 239.678230] usb 2-1.3: USB disconnect, device number 4
[ 239.679128] rtl_usb: reg 0x30, usbctrl_vendorreq TimeOut!
status:0xffffffed value=0x0
This sequence of events is unusual. A driver should be able to trust that the
platform is behaving correctly and normally.
In reviewing the USB probe routine, the only thing I could see is that one
returned value is not being checked. If the problem happens again, please try
the following patch:
diff --git a/drivers/net/wireless/realtek/rtlwifi/usb.c
b/drivers/net/wireless/realtek/rtlwifi/usb.c
index 5590d07d0918..092cd2da15f6 100644
--- a/drivers/net/wireless/realtek/rtlwifi/usb.c
+++ b/drivers/net/wireless/realtek/rtlwifi/usb.c
@@ -1082,6 +1082,8 @@ int rtl_usb_probe(struct usb_interface *intf,
init_completion(&rtlpriv->firmware_loading_complete);
SET_IEEE80211_DEV(hw, &intf->dev);
udev = interface_to_usbdev(intf);
+ if (!udev)
+ return -ENODEV;
usb_get_dev(udev);
usb_priv = rtl_usbpriv(hw);
memset(usb_priv, 0, sizeof(*usb_priv));
I will test and push this patch because that value should be checked, but I'm
not sure it would correct your problem.
Larry
