linux/drivers/usb
Douglas Anderson ed62ca2f4f USB: core: Avoid race of async_completed() w/ usbdev_release()
While running reboot tests w/ a specific set of USB devices (and
slub_debug enabled), I found that once every few hours my device would
be crashed with a stack that looked like this:

[   14.012445] BUG: spinlock bad magic on CPU#0, modprobe/2091
[   14.012460]  lock: 0xffffffc0cb055978, .magic: ffffffc0, .owner: cryption contexts: %lu/%lu
[   14.012460] /1025536097, .owner_cpu: 0
[   14.012466] CPU: 0 PID: 2091 Comm: modprobe Not tainted 4.4.79 #352
[   14.012468] Hardware name: Google Kevin (DT)
[   14.012471] Call trace:
[   14.012483] [<....>] dump_backtrace+0x0/0x160
[   14.012487] [<....>] show_stack+0x20/0x28
[   14.012494] [<....>] dump_stack+0xb4/0xf0
[   14.012500] [<....>] spin_dump+0x8c/0x98
[   14.012504] [<....>] spin_bug+0x30/0x3c
[   14.012508] [<....>] do_raw_spin_lock+0x40/0x164
[   14.012515] [<....>] _raw_spin_lock_irqsave+0x64/0x74
[   14.012521] [<....>] __wake_up+0x2c/0x60
[   14.012528] [<....>] async_completed+0x2d0/0x300
[   14.012534] [<....>] __usb_hcd_giveback_urb+0xc4/0x138
[   14.012538] [<....>] usb_hcd_giveback_urb+0x54/0xf0
[   14.012544] [<....>] xhci_irq+0x1314/0x1348
[   14.012548] [<....>] usb_hcd_irq+0x40/0x50
[   14.012553] [<....>] handle_irq_event_percpu+0x1b4/0x3f0
[   14.012556] [<....>] handle_irq_event+0x4c/0x7c
[   14.012561] [<....>] handle_fasteoi_irq+0x158/0x1c8
[   14.012564] [<....>] generic_handle_irq+0x30/0x44
[   14.012568] [<....>] __handle_domain_irq+0x90/0xbc
[   14.012572] [<....>] gic_handle_irq+0xcc/0x18c

Investigation using kgdb() found that the wait queue that was passed
into wake_up() had been freed (it was filled with slub_debug poison).

I analyzed and instrumented the code and reproduced.  My current
belief is that this is happening:

1. async_completed() is called (from IRQ).  Moves "as" onto the
   completed list.
2. On another CPU, proc_reapurbnonblock_compat() calls
   async_getcompleted().  Blocks on spinlock.
3. async_completed() releases the lock; keeps running; gets blocked
   midway through wake_up().
4. proc_reapurbnonblock_compat() => async_getcompleted() gets the
   lock; removes "as" from completed list and frees it.
5. usbdev_release() is called.  Frees "ps".
6. async_completed() finally continues running wake_up().  ...but
   wake_up() has a pointer to the freed "ps".

The instrumentation that led me to believe this was based on adding
some trace_printk() calls in a select few functions and then using
kdb's "ftdump" at crash time.  The trace follows (NOTE: in the trace
below I cheated a little bit and added a udelay(1000) in
async_completed() after releasing the spinlock because I wanted it to
trigger quicker):

<...>-2104   0d.h2 13759034us!: async_completed at start: as=ffffffc0cc638200
mtpd-2055    3.... 13759356us : async_getcompleted before spin_lock_irqsave
mtpd-2055    3d..1 13759362us : async_getcompleted after list_del_init: as=ffffffc0cc638200
mtpd-2055    3.... 13759371us+: proc_reapurbnonblock_compat: free_async(ffffffc0cc638200)
mtpd-2055    3.... 13759422us+: async_getcompleted before spin_lock_irqsave
mtpd-2055    3.... 13759479us : usbdev_release at start: ps=ffffffc0cc042080
mtpd-2055    3.... 13759487us : async_getcompleted before spin_lock_irqsave
mtpd-2055    3.... 13759497us!: usbdev_release after kfree(ps): ps=ffffffc0cc042080
<...>-2104   0d.h2 13760294us : async_completed before wake_up(): as=ffffffc0cc638200

To fix this problem we can just move the wake_up() under the ps->lock.
There should be no issues there that I'm aware of.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-08-28 11:17:57 +02:00
..
atm usb: atm: ueagle-atm: constify attribute_group structures. 2017-08-10 11:31:28 -07:00
c67x00 c67x00-hcd: constify hc_driver structures 2017-07-30 07:26:52 -07:00
chipidea Chipidea changes for v4.14-rc1 2017-08-28 10:50:22 +02:00
class usb: usbtmc: constify attribute_group structures. 2017-08-10 11:31:28 -07:00
common usb: make device_type const 2017-08-28 11:17:57 +02:00
core USB: core: Avoid race of async_completed() w/ usbdev_release() 2017-08-28 11:17:57 +02:00
dwc2 usb: dwc2: skip L2 state of hcd if controller work in device mode 2017-08-15 14:18:55 +03:00
dwc3 usb: changes for v4.14 merge window 2017-08-22 13:16:06 -07:00
early usb/early: Remove trace_printk() callers in xhci-dbc 2017-06-13 10:54:40 +02:00
gadget usb: changes for v4.14 merge window 2017-08-22 13:16:06 -07:00
host xhci: rework bus_resume and check ports are suspended before resuming them. 2017-08-16 15:26:26 -07:00
image USB: microtek: remove unneeded DRIVER_VERSION macro 2017-07-22 15:56:53 +02:00
isp1760 usb: isp1760: compress return logic into one line 2017-07-17 13:13:44 +02:00
misc usb: usbsevseg: constify attribute_group structures. 2017-08-10 11:31:27 -07:00
mon sched/headers: Prepare to move signal wakeup & sigpending methods from <linux/sched.h> into <linux/sched/signal.h> 2017-03-02 08:42:32 +01:00
mtu3 usb: changes for v4.14 merge window 2017-08-22 13:16:06 -07:00
musb USB: musb: dsps: add explicit runtime resume at suspend 2017-08-28 10:51:56 +02:00
phy usb: changes for v4.14 merge window 2017-08-22 13:16:06 -07:00
renesas_usbhs usb: changes for v4.14 merge window 2017-08-22 13:16:06 -07:00
serial USB: serial: pl2303: add new ATEN device id 2017-08-10 11:55:00 -07:00
storage Merge 4.13-rc5 into usb-next 2017-08-14 14:50:58 -07:00
typec usb: typec: include linux/device.h in ucsi.h 2017-07-17 13:11:56 +02:00
usbip USB: usbip: remove unneeded MODULE_VERSION() usage 2017-07-22 15:56:53 +02:00
wusbcore usb: wusbcore: dev-sysfs: constify attribute_group structures. 2017-08-10 11:31:27 -07:00
Kconfig usb: USB Type-C connector class 2017-03-23 13:48:44 +01:00
Makefile USB patches for 4.12-rc1 2017-05-04 18:03:51 -07:00
README
usb-skeleton.c USB: usb-skeleton: refactor endpoint retrieval 2017-03-23 13:54:08 +01:00

To understand all the Linux-USB framework, you'll use these resources:

    * This source code.  This is necessarily an evolving work, and
      includes kerneldoc that should help you get a current overview.
      ("make pdfdocs", and then look at "usb.pdf" for host side and
      "gadget.pdf" for peripheral side.)  Also, Documentation/usb has
      more information.

    * The USB 2.0 specification (from www.usb.org), with supplements
      such as those for USB OTG and the various device classes.
      The USB specification has a good overview chapter, and USB
      peripherals conform to the widely known "Chapter 9".

    * Chip specifications for USB controllers.  Examples include
      host controllers (on PCs, servers, and more); peripheral
      controllers (in devices with Linux firmware, like printers or
      cell phones); and hard-wired peripherals like Ethernet adapters.

    * Specifications for other protocols implemented by USB peripheral
      functions.  Some are vendor-specific; others are vendor-neutral
      but just standardized outside of the www.usb.org team.

Here is a list of what each subdirectory here is, and what is contained in
them.

core/		- This is for the core USB host code, including the
		  usbfs files and the hub class driver ("hub_wq").

host/		- This is for USB host controller drivers.  This
		  includes UHCI, OHCI, EHCI, and others that might
		  be used with more specialized "embedded" systems.

gadget/		- This is for USB peripheral controller drivers and
		  the various gadget drivers which talk to them.


Individual USB driver directories.  A new driver should be added to the
first subdirectory in the list below that it fits into.

image/		- This is for still image drivers, like scanners or
		  digital cameras.
../input/	- This is for any driver that uses the input subsystem,
		  like keyboard, mice, touchscreens, tablets, etc.
../media/	- This is for multimedia drivers, like video cameras,
		  radios, and any other drivers that talk to the v4l
		  subsystem.
../net/		- This is for network drivers.
serial/		- This is for USB to serial drivers.
storage/	- This is for USB mass-storage drivers.
class/		- This is for all USB device drivers that do not fit
		  into any of the above categories, and work for a range
		  of USB Class specified devices. 
misc/		- This is for all USB device drivers that do not fit
		  into any of the above categories.