Currently the sigreturn compat code is copied to an offset in the
vectors table. When using a BE kernel this data will be stored in the
wrong endianess so when returning from a signal on a 32-bit BE system,
arbitrary code will be executed.

Instead of declaring the code inside a struct and copying that, use
the assembler's .byte directives to store the code in the correct
endianess regardless of platform endianess.
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Matthew Leach <matthew.leach@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

The arm64 port contains wrappers for arm32 syscalls that pass 64-bit
values. These wrappers concatenate the two registers to hold a 64-bit
value in a single X register. On BE, however, the lower and higher
words are swapped.

Create a new assembler macro, regs_to_64, that when on BE systems
swaps the registers in the orr instruction.
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Matthew Leach <matthew.leach@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This patch adds support for BE8 AArch32 tasks to the compat layer.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

uname -m reports the machine field from the current utsname, which should
reflect the endianness of the system.

This patch reports ELF_PLATFORM for the field, so that everything appears
consistent from userspace.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This patch adds support for the aarch64_be ELF format to the AArch64 ELF
loader.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

For big-endian processors, we must include
linux/byteorder/big_endian.h to get the relevant definitions for
swabbing between CPU order and a defined endianness.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This patch adds big-endian support to the AArch64 top-level Makefile.
This currently just passes the relevant flags to the toolchain and is
predicated on a Kconfig option that will be introduced later on.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This patch adds support for using PSCI CPU_OFF calls for CPU hotplug.
With this code it is possible to hot unplug CPUs with "psci" as their
boot-method, as long as there's an appropriate cpu_off function id
specified in the psci node.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This patch adds the basic infrastructure necessary to support
CPU_HOTPLUG on arm64, based on the arm implementation. Actual hotplug
support will depend on an implementation's cpu_operations (e.g. PSCI).
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

With the advent of CPU_HOTPLUG, the enable-method property for CPU0 may
tells us something useful (i.e. how to hotplug it back on), so we must
read it along with all the enable-method for all the other CPUs.  Even
on UP the enable-method may tell us useful information (e.g. if a core
has some mechanism that might be usable for cpuidle), so we should
always read it.

This patch factors out the reading of the enable method, and ensures
that CPU0's enable method is read regardless of whether the kernel is
built with SMP support.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

The arm64 kernel has an internal holding pen, which is necessary for
some systems where we can't bring CPUs online individually and must hold
multiple CPUs in a safe area until the kernel is able to handle them.
The current SMP infrastructure for arm64 is closely coupled to this
holding pen, and alternative boot methods must launch CPUs into the pen,
where they sit before they are launched into the kernel proper.

With PSCI (and possibly other future boot methods), we can bring CPUs
online individually, and need not perform the secondary_holding_pen
dance. Instead, this patch factors the holding pen management code out
to the spin-table boot method code, as it is the only boot method
requiring the pen.

A new entry point for secondaries, secondary_entry is added for other
boot methods to use, which bypasses the holding pen and its associated
overhead when bringing CPUs online. The smp.pen.text section is also
removed, as the pen can live in head.text without problem.

The cpu_operations structure is extended with two new functions,
cpu_boot and cpu_postboot, for bringing a cpu into the kernel and
performing any post-boot cleanup required by a bootmethod (e.g.
resetting the secondary_holding_pen_release to INVALID_HWID).
Documentation is added for cpu_operations.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

For hotplug support, we're going to want a place to store operations
that do more than bring CPUs online, and it makes sense to group these
with our current smp_enable_ops. For cpuidle support, we'll want to
group additional functions, and we may want them even for UP kernels.

This patch renames smp_enable_ops to the more general cpu_operations,
and pulls the definitions out of smp code such that they can be used in
UP kernels. While we're at it, fix up instances of the cpu parameter to
be an unsigned int, drop the init markings and rename the *_cpu
functions to cpu_* to reduce future churn when cpu_operations is
extended.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

The functions in psci.c are only used from smp_psci.c, and smp_psci
cannot function without psci.c. Additionally psci.c is built when !SMP,
where it's expected that cpu_suspend may be useful.

This patch unifies the two files, removing pointless duplication and
paving the way for PSCI support in UP systems.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

There are a few points in the arm64 booting document which are unclear
(such as the initial state of secondary CPUs), and/or have not been
documented (PSCI is a supported mechanism for booting secondary CPUs).

This patch amends the arm64 boot document to better express the
(existing) requirements, and to describe PSCI as a supported booting
mechanism.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Martin <dave.martin@arm.com>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Fu Wei <tekkamanninja@gmail.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This function may be called from loadable modules, so it needs
exporting.
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reported-by: Loc Ho <lho@apm.com>

This patch introduces cmpxchg64_relaxed for arm64 using the existing
cmpxchg_local macro, which performs a cmpxchg operation (up to 64 bits)
without barrier semantics.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

Our spinlocks are only 32-bit (2x16-bit tickets) and our cmpxchg can
deal with 8-bytes (as one would hope!).

This patch wires up the cmpxchg-based lockless lockref implementation
for arm64.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

This patch introduces a ticket lock implementation for arm64, along the
same lines as the implementation for arch/arm/.
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

In ftrace_syscall_enter(),
    syscall_get_arguments(..., 0, n, ...)
        if (i == 0) { <handle orig_x0> ...; n--;}
        memcpy(..., n * sizeof(args[0]));
If 'number of arguments(n)' is zero and 'argument index(i)' is also zero in
syscall_get_arguments(), none of arguments should be copied by memcpy().
Otherwise 'n--' can be a big positive number and unexpected amount of data
will be copied. Tracing system calls which take no argument, say sync(void),
may hit this case and eventually make the system corrupted.
This patch fixes the issue both in syscall_get_arguments() and
syscall_set_arguments().
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

Modify the code to use current_euid(), and in_egroup_p, as in done
in fs/proc/proc_sysctl.c:test_perm()

Cc: stable@vger.kernel.org
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Reported-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull SCSI target fixes from Nicholas Bellinger:
 "Here are the outstanding target fixes queued up for v3.12-rc4 code.

  The highlights include:

   - Make vhost/scsi tag percpu_ida_alloc() use GFP_ATOMIC
   - Allow sess_cmd_map allocation failure fallback to use vzalloc
   - Fix COMPARE_AND_WRITE se_cmd->data_length bug with FILEIO backends
   - Fixes for COMPARE_AND_WRITE callback recursive failure OOPs + non
     zero scsi_status bug
   - Make iscsi-target do acknowledgement tag release from RX context
   - Setup iscsi-target with extra (cmdsn_depth / 2) percpu_ida tags

  Also included is a iscsi-target patch CC'ed for v3.10+ that avoids
  legacy wait_for_task=true release during fast-past StatSN
  acknowledgement, and two other SRP target related patches that address
  long-standing issues that are CC'ed for v3.3+.

  Extra thanks to Thomas Glanzmann for his testing feedback with
  COMPARE_AND_WRITE + EXTENDED_COPY VAAI logic"

* git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:
  iscsi-target; Allow an extra tag_num / 2 number of percpu_ida tags
  iscsi-target: Perform release of acknowledged tags from RX context
  iscsi-target: Only perform wait_for_tasks when performing shutdown
  target: Fail on non zero scsi_status in compare_and_write_callback
  target: Fix recursive COMPARE_AND_WRITE callback failure
  target: Reset data_length for COMPARE_AND_WRITE to NoLB * block_size
  ib_srpt: always set response for task management
  target: Fall back to vzalloc upon ->sess_cmd_map kzalloc failure
  vhost/scsi: Use GFP_ATOMIC with percpu_ida_alloc for obtaining tag
  ib_srpt: Destroy cm_id before destroying QP.
  target: Fix xop->dbl assignment in target_xcopy_parse_segdesc_02

Pull slave-dmaengine fixes from Vinod Koul:
 "Here is the slave dmanegine fixes.  We have the fix for deadlock issue
  on imx-dma by Michael and Josh's edma config fix along with author
  change"

* 'fixes' of git://git.infradead.org/users/vkoul/slave-dma:
  dmaengine: imx-dma: fix callback path in tasklet
  dmaengine: imx-dma: fix lockdep issue between irqhandler and tasklet
  dmaengine: imx-dma: fix slow path issue in prep_dma_cyclic
  dma/Kconfig: Make TI_EDMA select TI_PRIV_EDMA
  edma: Update author email address

Pull btrfs fixes from Chris Mason:
 "This is a small collection of fixes, including a regression fix from
  Liu Bo that solves rare crashes with compression on.

  I've merged my for-linus up to 3.12-rc3 because the top commit is only
  meant for 3.12.  The rest of the fixes are also available in my master
  branch on top of my last 3.11 based pull"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  btrfs: Fix crash due to not allocating integrity data for a bioset
  Btrfs: fix a use-after-free bug in btrfs_dev_replace_finishing
  Btrfs: eliminate races in worker stopping code
  Btrfs: fix crash of compressed writes
  Btrfs: fix transid verify errors when recovering log tree

Pull GPIO fixes from Linus Walleij:
 "Two patches for the OMAP driver, dealing with setting up IRQs properly
  on the device tree boot path"

* tag 'gpio-v3.12-2' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio:
  gpio/omap: auto-setup a GPIO when used as an IRQ
  gpio/omap: maintain GPIO and IRQ usage separately

Pull USB fixes from Greg KH:
 "Here are none fixes for various USB driver problems.  The majority are
  gadget/musb fixes, but there are some new device ids in here as well"

* tag 'usb-3.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  usb: chipidea: add Intel Clovertrail pci id
  usb: gadget: s3c-hsotg: fix can_write limit for non-periodic endpoints
  usb: gadget: f_fs: fix error handling
  usb: musb: dsps: do not bind to "musb-hdrc"
  USB: serial: option: Ignore card reader interface on Huawei E1750
  usb: musb: gadget: fix otg active status flag
  usb: phy: gpio-vbus: fix deferred probe from __init
  usb: gadget: pxa25x_udc: fix deferred probe from __init
  usb: musb: fix otg default state

Pull tty fixes from Greg KH:
 "Here are two tty driver fixes for 3.12-rc4.

  One fixes the reported regression in the n_tty code that a number of
  people found recently, and the other one fixes an issue with xen
  consoles that broke in 3.10"

* tag 'tty-3.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  xen/hvc: allow xenboot console to be used again
  tty: Fix pty master read() after slave closes

Pull staging fixes from Greg KH:
 "Here are 4 tiny staging and iio driver fixes for 3.12-rc4.  Nothing
  major, just some small fixes for reported issues"

* tag 'staging-3.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:
  staging: comedi: ni_65xx: (bug fix) confine insn_bits to one subdevice
  iio:magnetometer: Bugfix magnetometer default output registers
  iio: Remove debugfs entries in iio_device_unregister()
  iio: amplifiers: ad8366: Remove regulator_put

When btrfs creates a bioset, we must also allocate the integrity data pool.
Otherwise btrfs will crash when it tries to submit a bio to a checksumming
disk:

 BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
 IP: [<ffffffff8111e28a>] mempool_alloc+0x4a/0x150
 PGD 2305e4067 PUD 23063d067 PMD 0
 Oops: 0000 [#1] PREEMPT SMP
 Modules linked in: btrfs scsi_debug xfs ext4 jbd2 ext3 jbd mbcache
sch_fq_codel eeprom lpc_ich mfd_core nfsd exportfs auth_rpcgss af_packet
raid6_pq xor zlib_deflate libcrc32c [last unloaded: scsi_debug]
 CPU: 1 PID: 4486 Comm: mount Not tainted 3.12.0-rc1-mcsum #2
 Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
 task: ffff8802451c9720 ti: ffff880230698000 task.ti: ffff880230698000
 RIP: 0010:[<ffffffff8111e28a>]  [<ffffffff8111e28a>] mempool_alloc+0x4a/0x150
 RSP: 0018:ffff880230699688  EFLAGS: 00010286
 RAX: 0000000000000001 RBX: 0000000000000000 RCX: 00000000005f8445
 RDX: 0000000000000001 RSI: 0000000000000010 RDI: 0000000000000000
 RBP: ffff8802306996f8 R08: 0000000000011200 R09: 0000000000000008
 R10: 0000000000000020 R11: ffff88009d6e8000 R12: 0000000000011210
 R13: 0000000000000030 R14: ffff8802306996b8 R15: ffff8802451c9720
 FS:  00007f25b8a16800(0000) GS:ffff88024fc80000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
 CR2: 0000000000000018 CR3: 0000000230576000 CR4: 00000000000007e0
 Stack:
  ffff8802451c9720 0000000000000002 ffffffff81a97100 0000000000281250
  ffffffff81a96480 ffff88024fc99150 ffff880228d18200 0000000000000000
  0000000000000000 0000000000000040 ffff880230e8c2e8 ffff8802459dc900
 Call Trace:
  [<ffffffff811b2208>] bio_integrity_alloc+0x48/0x1b0
  [<ffffffff811b26fc>] bio_integrity_prep+0xac/0x360
  [<ffffffff8111e298>] ? mempool_alloc+0x58/0x150
  [<ffffffffa03e8041>] ? alloc_extent_state+0x31/0x110 [btrfs]
  [<ffffffff81241579>] blk_queue_bio+0x1c9/0x460
  [<ffffffff8123e58a>] generic_make_request+0xca/0x100
  [<ffffffff8123e639>] submit_bio+0x79/0x160
  [<ffffffffa03f865e>] btrfs_map_bio+0x48e/0x5b0 [btrfs]
  [<ffffffffa03c821a>] btree_submit_bio_hook+0xda/0x110 [btrfs]
  [<ffffffffa03e7eba>] submit_one_bio+0x6a/0xa0 [btrfs]
  [<ffffffffa03ef450>] read_extent_buffer_pages+0x250/0x310 [btrfs]
  [<ffffffff8125eef6>] ? __radix_tree_preload+0x66/0xf0
  [<ffffffff8125f1c5>] ? radix_tree_insert+0x95/0x260
  [<ffffffffa03c66f6>] btree_read_extent_buffer_pages.constprop.128+0xb6/0x120
[btrfs]
  [<ffffffffa03c8c1a>] read_tree_block+0x3a/0x60 [btrfs]
  [<ffffffffa03caefd>] open_ctree+0x139d/0x2030 [btrfs]
  [<ffffffffa03a282a>] btrfs_mount+0x53a/0x7d0 [btrfs]
  [<ffffffff8113ab0b>] ? pcpu_alloc+0x8eb/0x9f0
  [<ffffffff81167305>] ? __kmalloc_track_caller+0x35/0x1e0
  [<ffffffff81176ba0>] mount_fs+0x20/0xd0
  [<ffffffff81191096>] vfs_kern_mount+0x76/0x120
  [<ffffffff81193320>] do_mount+0x200/0xa40
  [<ffffffff81135cdb>] ? strndup_user+0x5b/0x80
  [<ffffffff81193bf0>] SyS_mount+0x90/0xe0
  [<ffffffff8156d31d>] system_call_fastpath+0x1a/0x1f
 Code: 4c 8d 75 a8 4c 89 6d e8 45 89 e0 4c 8d 6f 30 48 89 5d d8 41 83 e0 af 48
89 fb 49 83 c6 18 4c 89 7d f8 65 4c 8b 3c 25 c0 b8 00 00 <48> 8b 73 18 44 89 c7
44 89 45 98 ff 53 20 48 85 c0 48 89 c2 74
 RIP  [<ffffffff8111e28a>] mempool_alloc+0x4a/0x150
  RSP <ffff880230699688>
 CR2: 0000000000000018
 ---[ end trace 7a96042017ed21e2 ]---
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Pull CIFS fixes from Steve French:
 "Small set of cifs fixes.  Most important is Jeff's fix that works
  around disconnection problems which can be caused by simultaneous use
  of user space tools (starting a long running smbclient backup then
  doing a cifs kernel mount) or multiple cifs mounts through a NAT, and
  Jim's fix to deal with reexport of cifs share.

  I expect to send two more cifs fixes next week (being tested now) -
  fixes to address an SMB2 unmount hang when server dies and a fix for
  cifs symlink handling of Windows "NFS" symlinks"

* 'for-linus' of git://git.samba.org/sfrench/cifs-2.6:
  [CIFS] update cifs.ko version
  [CIFS] Remove ext2 flags that have been moved to fs.h
  [CIFS] Provide sane values for nlink
  cifs: stop trying to use virtual circuits
  CIFS: FS-Cache: Uncache unread pages in cifs_readpages() before freeing them

Pull PCI fix from Bjorn Helgaas:
 "We merged what was intended to be an MMCONFIG cleanup, but in fact,
  for systems without _CBA (which is almost everything), it broke
  extended config space for domain 0 and it broke all config space for
  other domains.

  This reverts the change"

* tag 'pci-v3.12-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  Revert "x86/PCI: MMCONFIG: Check earlier for MMCONFIG region at address zero"

This reverts commit 07f9b61c.

07f9b61c was intended to be a cleanup that didn't change anything, but in
fact, for systems without _CBA (which is almost everything), it broke
extended config space for domain 0 and all config space for other domains.

Reference: http://lkml.kernel.org/r/20131004011806.GE20450@dangermouse.emea.sgi.comReported-by: Hedi Berriche <hedi@sgi.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

Pull ACPI and power management fixes from Rafael Wysocki:

 - The resume part of user space driven hibernation (s2disk) is now
   broken after the change that moved the creation of memory bitmaps to
   after the freezing of tasks, because I forgot that the resume utility
   loaded the image before freezing tasks and needed the bitmaps for
   that.  The fix adds special handling for that case.

 - One of recent commits changed the export of acpi_bus_get_device() to
   EXPORT_SYMBOL_GPL(), which was technically correct but broke existing
   binary modules using that function including one in particularly
   widespread use.  Change it back to EXPORT_SYMBOL().

 - The intel_pstate driver sometimes fails to disable turbo if its
   no_turbo sysfs attribute is set.  Fix from Srinivas Pandruvada.

 - One of recent cpufreq fixes forgot to update a check in cpufreq-cpu0
   which still (incorrectly) treats non-NULL as non-error.  Fix from
   Philipp Zabel.

 - The SPEAr cpufreq driver uses a wrong variable type in one place
   preventing it from catching errors returned by one of the functions
   called by it.  Fix from Sachin Kamat.

* tag 'pm+acpi-3.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI: Use EXPORT_SYMBOL() for acpi_bus_get_device()
  intel_pstate: fix no_turbo
  cpufreq: cpufreq-cpu0: NULL is a valid regulator, part 2
  cpufreq: SPEAr: Fix incorrect variable type
  PM / hibernate: Fix user space driven resume regression

Pull xfs bugfixes from Ben Myers:
 "There are lockdep annotations for project quotas, a fix for dirent
  dtype support on v4 filesystems, a fix for a memory leak in recovery,
  and a fix for the build error that resulted from it.  D'oh"

* tag 'xfs-for-linus-v3.12-rc4' of git://oss.sgi.com/xfs/xfs:
  xfs: Use kmem_free() instead of free()
  xfs: fix memory leak in xlog_recover_add_to_trans
  xfs: dirent dtype presence is dependent on directory magic numbers
  xfs: lockdep needs to know about 3 dquot-deep nesting

Now avc_audit() has no more users with that parameter. Remove it.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

.. so get rid of it.  The only indirect users were all the
avc_has_perm() callers which just expanded to have a zero flags
argument.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

free_device rcu callback, scheduled from btrfs_rm_dev_replace_srcdev,
can be processed before btrfs_scratch_superblock is called, which would
result in a use-after-free on btrfs_device contents.  Fix this by
zeroing the superblock before the rcu callback is registered.

Cc: Stefan Behrens <sbehrens@giantdisaster.de>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>

The current implementation of worker threads in Btrfs has races in
worker stopping code, which cause all kinds of panics and lockups when
running btrfs/011 xfstest in a loop.  The problem is that
btrfs_stop_workers is unsynchronized with respect to check_idle_worker,
check_busy_worker and __btrfs_start_workers.

E.g., check_idle_worker race flow:

       btrfs_stop_workers():            check_idle_worker(aworker):
- grabs the lock
- splices the idle list into the
  working list
- removes the first worker from the
  working list
- releases the lock to wait for
  its kthread's completion
                                  - grabs the lock
                                  - if aworker is on the working list,
                                    moves aworker from the working list
                                    to the idle list
                                  - releases the lock
- grabs the lock
- puts the worker
- removes the second worker from the
  working list
                              ......
        btrfs_stop_workers returns, aworker is on the idle list
                 FS is umounted, memory is freed
                              ......
              aworker is waken up, fireworks ensue

With this applied, I wasn't able to trigger the problem in 48 hours,
whereas previously I could reliably reproduce at least one of these
races within an hour.
Reported-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>