xHCI regression(mass storage devices fail) in stable 3.13.5(14.04)

[Adam Lee]

Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.252465] usb 3-2: reset SuperSpeed USB device number 2 using xhci_hcd
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.268395] xhci_hcd 0000:00:14.0: xHCI xhci_drop_endpoint called with disabled ep ffff880046f75980
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.268400] xhci_hcd 0000:00:14.0: xHCI xhci_drop_endpoint called with disabled ep ffff880046f759c0
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.540388] usb 3-2: reset SuperSpeed USB device number 2 using xhci_hcd
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.556687] xhci_hcd 0000:00:14.0: xHCI xhci_drop_endpoint called with disabled ep ffff880046f75980
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.556693] xhci_hcd 0000:00:14.0: xHCI xhci_drop_endpoint called with disabled ep ffff880046f759c0
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.557393] sd 3:0:0:0: [sdb] Unhandled error code
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.557399] sd 3:0:0:0: [sdb]
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.557401] Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.557403] sd 3:0:0:0: [sdb] CDB:
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.557405] Read(10): 28 00 00 00 35 a1 00 00 66 00
Mar 11 02:00:25 u-ThinkPad-X240s kernel: [ 299.557415] end_request: I/O error, dev sdb, sector 13729

http://www.spinics.net/lists/linux-usb/msg103757.html

should backport:
e2ed511400d41e0d136089d5a55ceab57c6a2426
469d417b68958a064c09e7875646c97c6e783dfc

[Adam Lee (adam8157)]

patches are submitted to kernel-team list.

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 3.13.0-18.38

---------------
linux (3.13.0-18.38) trusty; urgency=low

  [ Tim Gardner ]

  * Release Tracking Bug
    - LP: #1293725
  * [Config] Add hv_balloon to d-i virtio-modules
    - LP: #1292216

  [ Andy Whitcroft ]

  * [Config] d-i -- add virtio_scsi to virtio-modules
    - LP: #1288607

  [ Colin Ian King ]

  * SAUCE: intel_pstate: inform user that thermald is worth considering

  [ dann frazier ]

  * [Config] arm64: KVM=y

  [ Gerd Hoffmann ]

  * SAUCE: vmbus: add missing breaks
    - LP: #1287398
  * SAUCE: vmbus: use resource for hyperv mmio region
    - LP: #1287398
  * SAUCE: hyperv-fb: add support for generation 2 virtual machines.
    - LP: #1287398
  * SAUCE: hyperv-fb: kick off efifb early
    - LP: #1287398

  [ Haiyang Zhang ]

  * SAUCE: hyperv: Change the receive buffer size for legacy hosts
    - LP: #1290151

  [ K. Y. Srinivasan ]

  * SAUCE: Drivers: hv: vmbus: Extract the mmio information from DSDT
    - LP: #1287398

  [ Paolo Pisati ]

  * SAUCE: leds-gpio: of: introduce MODULE_DEVICE_TABLE for module
    autoloading
  * [Config] amhf: LEDS_TRIGGER_HEARTBEAT=y

  [ Upstream Kernel Changes ]

  * Revert "xhci 1.0: Limit arbitrarily-aligned scatter gather."
    - LP: #1293361
  * Revert "USBNET: ax88179_178a: enable tso if usb host supports sg dma"
    - LP: #1293361
  * powerpc/tm: Fix crash when forking inside a transaction
  * AX88179_178A: Add VID:DID for Lenovo OneLinkDock Gigabit LAN
    - LP: #1291890
  * drm/vmwgfx: Fix a surface reference corner-case in legacy emulation
    mode
  * Input: wacom - scale up touch width and height values for Intuos Pro
  * Input: wacom - make sure touch_max is set for touch devices
  * Input: wacom - add support for three new Intuos devices
  * Input: wacom - add reporting of SW_MUTE_DEVICE events
  * Input: wacom - fix wacom->shared guards for dual input devices
  * Input: wacom - add support for DTU-1031
  * net: fix for a race condition in the inet frag code
    - CVE-2014-0100
  * net: sctp: fix sctp_sf_do_5_1D_ce to verify if we/peer is AUTH capable
    - CVE-2014-0101
  * KEYS: Make the keyring cycle detector ignore other keyrings of the same
    name
    - CVE-2014-0102
  * ipv6: don't set DST_NOCOUNT for remotely added routes
    - CVE-2014-2309
 -- Andy Whitcroft <email address hidden> Thu, 13 Mar 2014 10:06:05 +0000

[Rachel Greenham (rachel-strangenoises)]

FYI this also seems to have fixed a problem I was having with a USB DVB-S2 receiver, which apparently caused the drivers to hang on load *only* on xHCI; on EHCI, same kernel, all was fine. So this may not have been purely a mass storage device issue.

[Adam Lee (adam8157)]

@Rachel,

Yes, xHCI port with USB3 device will trigger this issue. EHCI or USB2 devices works fine.

[Jens (jens-launchpad-net)]

I am still seeing this issue with "3.13.0-24-generic #47-Ubuntu" in the 14.04 release with all updates applied.

Connecting an USB 3 device to my computer's front (USB3) port and reading large files from it will trigger repeated USB bus resets.

Please reopen.

[Jens (jens-launchpad-net)]

Also see Bug #1284399.

[Adam Lee (adam8157)]

@Jens,

Sorry, I will not reopen, your issue is not the same as this, this only happens on xHCI port with USB3 device, your issue is USB2 device on xHCI. Please stick on that bug.