For a long time over many different kernels (currently on hardened-sources-3.11.2 with grsecurity disabled), I've been sometimes experiencing a crash with a USB device is plugged into the USB3 port on my Dell XPS 13 laptop. Such devices include a USB3 HD, a wifi interface, and a tethered Galaxy Nexus cell phone. Here are the logs leading up to the latest crash (enp2s0u1u3 is the LAN interface and easytether0 is the WAN): Nov 08 23:01:16 [kernel] [37544.759542] asix 3-1.3:1.0 enp2s0u1u3: link up, 1000Mbps, full-duplex, lpa 0xC5E1 Nov 08 23:01:23 [kernel] [37552.050212] asix 3-1.3:1.0 enp2s0u1u3: link up, 1000Mbps, full-duplex, lpa 0xC5E1 Nov 08 23:01:26 [kernel] [37554.992168] asix 3-1.3:1.0 enp2s0u1u3: link up, 1000Mbps, full-duplex, lpa 0xC5E1 Nov 08 23:01:35 [kernel] [37563.945543] asix 3-1.3:1.0 enp2s0u1u3: link up, 1000Mbps, full-duplex, lpa 0xC5E1 Nov 08 23:08:43 [dhcpcd] easytether0: carrier lost Nov 08 23:08:43 [dhcpcd] del_address: No such device Nov 08 23:08:43 [kernel] [37991.797043] usb 3-1.4: USB disconnect, device number 6 Nov 08 23:08:43 [dhcpcd] manage_link: No such device or address Nov 08 23:08:43 [dhcpcd] easytether0: removing interface The crash seems to happen when the USB connection starts to fail to the device attached to the USB3 port. For example, the device was physically moved or the USB3 HD starts making sounds. When it crashes I get the typical frozen white text on a black background. The text always references xhci_hcd. What can I submit to make this a better bug report?
(In reply to A. Person from comment #0) > For a long time over many different kernels Hardened kernels only or also different types of kernels? > When it crashes I get the typical frozen white text on a black background. > The text always references xhci_hcd. > What can I submit to make this a better bug report? That screen would help, can you take a picture of it?
(In reply to Tom Wijsman (TomWij) from comment #1) > (In reply to A. Person from comment #0) > > For a long time over many different kernels > > Hardened kernels only or also different types of kernels? > > > When it crashes I get the typical frozen white text on a black background. > > The text always references xhci_hcd. > > What can I submit to make this a better bug report? > > That screen would help, can you take a picture of it? The two questions I always ask are: 1) does it happen on the vanilla with approximately the smae config? 2) did you test the very latest grsec/pax patches. I'm cc-ing upstream.
The crash does happen on at least pf-sources as well. The latest version I've tested is 3.11.2. I will take a photo the next time I see the crash.
(In reply to A. Person from comment #3) > The crash does happen on at least pf-sources as well. The latest version > I've tested is 3.11.2. I will take a photo the next time I see the crash. If its in pf-sources and hardened-sources, its probably in vanilla. Can you test that? Also post your config file.
Created attachment 364602 [details] screenshot
Created attachment 364604 [details] config
This crashed for me again on hardened-sources after my USB hub and USB wifi adapter fell to the ground. I just switched to the vanilla kernel so I will be able to report if it crashes there. I've attached a photo of the crash screen and my .config. Here is the last chunk from the log before the crash: Dec 04 07:50:57 [kernel] [ 2381.230425] usb 3-1.4: USB disconnect, device number 4 Dec 04 07:50:57 [kernel] [ 2381.230682] wlp2s0u1u4: deauthenticating from (masked) by local choice (reason=3) Dec 04 07:50:57 [kernel] [ 2381.315406] ath: phy0: Failed to wakeup in 500us Dec 04 07:50:57 [kernel] [ 2381.417590] ath: phy0: RX failed to go idle in 10 ms RXSM=0xe050b70a Dec 04 07:50:57 [kernel] [ 2381.427671] ath: phy0: Failed to wakeup in 500us Dec 04 07:50:57 [dhcpcd] wlp2s0u1u4: carrier lost Dec 04 07:50:57 [kernel] [ 2381.732484] ath: phy0: RX failed to go idle in 10 ms RXSM=0xe050b70a Dec 04 07:50:57 [dhcpcd] del_address: No such device Dec 04 07:50:57 [kernel] [ 2381.743537] cfg80211: Calling CRDA for country: CN Dec 04 07:50:57 [kernel] [ 2381.752964] cfg80211: Calling CRDA to update world regulatory domain Dec 04 07:50:57 [kernel] [ 2381.752991] usb 3-1.4: ath9k_htc: USB layer deinitialized Dec 04 07:50:57 [kernel] [ 2381.755806] cfg80211: World regulatory domain updated: Dec 04 07:50:57 [kernel] [ 2381.755810] cfg80211: (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) Dec 04 07:50:57 [kernel] [ 2381.755812] cfg80211: (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.755814] cfg80211: (2457000 KHz - 2482000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.755815] cfg80211: (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.755817] cfg80211: (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.755818] cfg80211: (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.755830] cfg80211: Calling CRDA to update world regulatory domain Dec 04 07:50:57 [kernel] [ 2381.758075] cfg80211: World regulatory domain updated: Dec 04 07:50:57 [kernel] [ 2381.758080] cfg80211: (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) Dec 04 07:50:57 [kernel] [ 2381.758083] cfg80211: (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.758086] cfg80211: (2457000 KHz - 2482000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.758089] cfg80211: (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.758091] cfg80211: (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [kernel] [ 2381.758093] cfg80211: (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) Dec 04 07:50:57 [dhcpcd] manage_link: No such device or address Dec 04 07:50:57 [dhcpcd] wlp2s0u1u4: removing interface Dec 04 07:50:58 [kernel] [ 2381.920189] usb 3-1.4: new high-speed USB device number 6 using xhci_hcd Dec 04 07:50:59 [ntpd] Deleting interface #3 wlp2s0u1u4, (masked)#123, interface stats: received=132, sent=142, dropped=1, active_time=2295 secs Dec 04 07:50:59 [ntpd] (masked) interface (masked) -> (none) Dec 04 07:50:59 [ntpd] (masked) interface (masked) -> (none) Dec 04 07:50:59 [ntpd] (masked) interface (masked) -> (none) Dec 04 07:50:59 [ntpd] (masked) interface (masked) -> (none) Dec 04 07:50:59 [ntpd] peers refreshed
(In reply to A. Person from comment #7) > This crashed for me again on hardened-sources after my USB hub and USB wifi > adapter fell to the ground. I just switched to the vanilla kernel so I will > be able to report if it crashes there. I've attached a photo of the crash > screen and my .config. Here is the last chunk from the log before the crash: There is nothing specific in that log file that suggests that hardened is the issue. Let me know what happens with the vanilla sources.
Will do, thanks Anthony.
I'm very glad to report that this doesn't seem to happen any more on 3.12. I will re-open if necessary.
