We have found a race in the optimisation used in the dm cache
writethrough implementation.  Currently, dm core sends the cache target
two bios, one for the origin device and one for the cache device and
these are processed in parallel.  This patch avoids the race by
changing the code back to a simpler (slower) implementation which
processes the two writes in series, one after the other, until we can
develop a complete fix for the problem.

When the cache is in writethrough mode it needs to send WRITE bios to
both the origin and cache devices.

Previously we've been implementing this by having dm core query the
cache target on every write to find out how many copies of the bio it
wants.  The cache will ask for two bios if the block is in the cache,
and one otherwise.

Then main problem with this is it's racey.  At the time this check is
made the bio hasn't yet been submitted and so isn't being taken into
account when quiescing a block for migration (promotion or demotion).
This means a single bio may be submitted when two were needed because
the block has since been promoted to the cache (catastrophic), or two
bios where only one is needed (harmless).

I really don't want to start entering bios into the quiescing system
(deferred_set) in the get_num_write_bios callback.  Instead this patch
simplifies things; only one bio is submitted by the core, this is
first written to the origin and then the cache device in series.
Obviously this will have a latency impact.

deferred_writethrough_bios is introduced to record bios that must be
later issued to the cache device from the worker thread.  This deferred
submission, after the origin bio completes, is required given that we're
in interrupt context (writethrough_endio).
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

When writing the dirty bitset to the metadata device on a clean
shutdown, clear the dirty bits.  Previously they were left indicating
the cache was dirty. This led to confusion about whether there really
was dirty data in the cache or not.  (This was a harmless bug.)
Reported-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

If the cache policy's config values are not able to be set we must
set the policy to NULL after destroying it in create_cache_policy()
so we don't attempt to destroy it a second time later.
Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

Return error if cache_create() fails.

A missing return check made cache_ctr continue even after an error in
cache_create() resulting in the cache object being destroyed.  So a
simple failure like an odd number of cache policy config value arguments
would result in an oops.
Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

Squash various 32bit link errors.

  >> on i386:
  >> drivers/built-in.o: In function `is_discarded_oblock':
  >> dm-cache-target.c:(.text+0x1ea28e): undefined reference to `__udivdi3'
  ...
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

A deadlock was found in the prefetch code in the dm verity map
function.  This patch fixes this by transferring the prefetch
to a worker thread and skipping it completely if kmalloc fails.

If generic_make_request is called recursively, it queues the I/O
request on the current->bio_list without making the I/O request
and returns. The routine making the recursive call cannot wait
for the I/O to complete.

The deadlock occurs when one thread grabs the bufio_client
mutex and waits for an I/O to complete but the I/O is queued
on another thread's current->bio_list and is waiting to get
the mutex held by the first thread.

The fix recognises that prefetching is not essential.  If memory
can be allocated, it queues the prefetch request to the worker thread,
but if not, it does nothing.
Signed-off-by: Paul Taysom <taysom@chromium.org>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Cc: stable@kernel.org

Fix a discard granularity calculation to work for non power of 2 block sizes.

In order for thinp to passdown discard bios to the underlying data
device, the data device must have a discard granularity that is a
factor of the thinp block size.  Originally this check was done by
using bitops since the block_size was known to be a power of two.

Introduced by commit f13945d7
("dm thin: support a non power of 2 discard_granularity").
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

Fix a bug in dm_btree_remove that could leave leaf values with incorrect
reference counts.  The effect of this was that removal of a shared block
could result in the space maps thinking the block was no longer used.
More concretely, if you have a thin device and a snapshot of it, sending
a discard to a shared region of the thin could corrupt the snapshot.

Thinp uses a 2-level nested btree to store it's mappings.  This first
level is indexed by thin device, and the second level by logical
block.

Often when we're removing an entry in this mapping tree we need to
rebalance nodes, which can involve shadowing them, possibly creating a
copy if the block is shared.  If we do create a copy then children of
that node need to have their reference counts incremented.  In this
way reference counts percolate down the tree as shared trees diverge.

The rebalance functions were incrementing the children at the
appropriate time, but they were always assuming the children were
internal nodes.  This meant the leaf values (in our case packed
block/flags entries) were not being incremented.

Cc: stable@vger.kernel.org
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>

Commit 1d9d8639 ("perf,x86: fix kernel crash with PEBS/BTS after
suspend/resume") introduces a link failure since
perf_restore_debug_store() is only defined for CONFIG_CPU_SUP_INTEL:

	arch/x86/power/built-in.o: In function `restore_processor_state':
	(.text+0x45c): undefined reference to `perf_restore_debug_store'

Fix it by defining the dummy function appropriately.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Commit 1d9d8639 ("perf,x86: fix kernel crash with PEBS/BTS after
suspend/resume") fixed a crash when doing PEBS performance profiling
after resuming, but in using init_debug_store_on_cpu() to restore the
DS_AREA mtrr it also resulted in a new WARN_ON() triggering.

init_debug_store_on_cpu() uses "wrmsr_on_cpu()", which in turn uses CPU
cross-calls to do the MSR update.  Which is not really valid at the
early resume stage, and the warning is quite reasonable.  Now, it all
happens to _work_, for the simple reason that smp_call_function_single()
ends up just doing the call directly on the CPU when the CPU number
matches, but we really should just do the wrmsr() directly instead.

This duplicates the wrmsr() logic, but hopefully we can just remove the
wrmsr_on_cpu() version eventually.
Reported-and-tested-by: Parag Warudkar <parag.lkml@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull btrfs fixes from Chris Mason:
 "Eric's rcu barrier patch fixes a long standing problem with our
  unmount code hanging on to devices in workqueue helpers.  Liu Bo
  nailed down a difficult assertion for in-memory extent mappings."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: fix warning of free_extent_map
  Btrfs: fix warning when creating snapshots
  Btrfs: return as soon as possible when edquot happens
  Btrfs: return EIO if we have extent tree corruption
  btrfs: use rcu_barrier() to wait for bdev puts at unmount
  Btrfs: remove btrfs_try_spin_lock
  Btrfs: get better concurrency for snapshot-aware defrag work

Users report that an extent map's list is still linked when it's actually
going to be freed from cache.

The story is that

a) when we're going to drop an extent map and may split this large one into
smaller ems, and if this large one is flagged as EXTENT_FLAG_LOGGING which means
that it's on the list to be logged, then the smaller ems split from it will also
be flagged as EXTENT_FLAG_LOGGING, and this is _not_ expected.

b) we'll keep ems from unlinking the list and freeing when they are flagged with
EXTENT_FLAG_LOGGING, because the log code holds one reference.

The end result is the warning, but the truth is that we set the flag
EXTENT_FLAG_LOGGING only during fsync.

So clear flag EXTENT_FLAG_LOGGING for extent maps split from a large one.
Reported-by: Johannes Hirte <johannes.hirte@fem.tu-ilmenau.de>
Reported-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Pull kbuild fix from Michal Marek:
 "One fix for for make headers_install/headers_check to not require make
  3.81.  The requirement has been accidentally introduced in 3.7."

* 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
  kbuild: fix make headers_check with make 3.80

Pull OpenRISC bug fixes from Jonas Bonn:

 - The GPIO descriptor work has exposed how broken the non-GPIOLIB bits
   for OpenRISC were.  We now require GPIOLIB as this is the preferred
   way forward.

 - The system.h split introduced a bug in llist.h for arches using
   asm-generic/cmpxchg.h directly, which is currently only OpenRISC.
   The patch here moves two defines from asm-generic/atomic.h to
   asm-generic/cmpxchg.h to make things work as they should.

 - The VIRT_TO_BUS selector was added for OpenRISC, but OpenRISC does
   not have the virt_to_bus methods, so there's a patch to remove it
   again.

* tag 'for-3.9-rc3' of git://openrisc.net/jonas/linux:
  openrisc: remove HAVE_VIRT_TO_BUS
  asm-generic: move cmpxchg*_local defs to cmpxchg.h
  openrisc: require gpiolib

Pull char/misc fixes from Greg Kroah-Hartman:
 "Here are some tiny fixes for the w1 drivers and the final removal
  patch for getting rid of CONFIG_EXPERIMENTAL (all users of it are now
  gone from your tree, this just drops the Kconfig item itself.)

  All have been in the linux-next tree for a while"

* tag 'char-misc-3.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
  final removal of CONFIG_EXPERIMENTAL
  w1: fix oops when w1_search is called from netlink connector
  w1-gpio: fix unused variable warning
  w1-gpio: remove erroneous __exit and __exit_p()
  ARM: w1-gpio: fix erroneous gpio requests

Pull sound fixes from Takashi Iwai:
 "A collection of small fixes, as expected for the middle rc:
   - A couple of fixes for potential NULL dereferences and out-of-range
     array accesses revealed by static code parsers
   - A fix for the wrong error handling detected by trinity
   - A regression fix for missing audio on some MacBooks
   - CA0132 DSP loader fixes
   - Fix for EAPD control of IDT codecs on machines w/o speaker
   - Fix a regression in the HD-audio widget list parser code
   - Workaround for the NuForce UDH-100 USB audio"

* tag 'sound-3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: hda - Fix missing EAPD/GPIO setup for Cirrus codecs
  sound: sequencer: cap array index in seq_chn_common_event()
  ALSA: hda/ca0132 - Remove extra setting of dsp_state.
  ALSA: hda/ca0132 - Check download state of DSP.
  ALSA: hda/ca0132 - Check if dspload_image succeeded.
  ALSA: hda - Disable IDT eapd_switch if there are no internal speakers
  ALSA: hda - Fix snd_hda_get_num_raw_conns() to return a correct value
  ALSA: usb-audio: add a workaround for the NuForce UDH-100
  ALSA: asihpi - fix potential NULL pointer dereference
  ALSA: seq: Fix missing error handling in snd_seq_timer_open()

Pull DMA-mapping fix from Marek Szyprowski:
 "An important fix for all ARM architectures which use ZONE_DMA.
  Without it dma_alloc_* calls with GFP_ATOMIC flag might have allocated
  buffers outsize DMA zone."

* 'fixes-for-3.9' of git://git.linaro.org/people/mszyprowski/linux-dma-mapping:
  ARM: DMA-mapping: add missing GFP_DMA flag for atomic buffer allocation

Pull MFD fixes from Samuel Ortiz:
 "This is the first batch of MFD fixes for 3.9.

  With this one we have:

   - An ab8500 build failure fix.
   - An ab8500 device tree parsing fix.
   - A fix for twl4030_madc remove routine to work properly (when
     built-in).
   - A fix for properly registering palmas interrupt handler.
   - A fix for omap-usb init routine to actually write into the
     hostconfig register.
   - A couple of warning fixes for ab8500-gpadc and tps65912"

* tag 'mfd-fixes-3.9-1' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-fixes:
  mfd: twl4030-madc: Remove __exit_p annotation
  mfd: ab8500: Kill "reg" property from binding
  mfd: ab8500-gpadc: Complain if we fail to enable vtvout LDO
  mfd: wm831x: Don't forward declare enum wm831x_auxadc
  mfd: twl4030-audio: Fix argument type for twl4030_audio_disable_resource()
  mfd: tps65912: Declare and use tps65912_irq_exit()
  mfd: palmas: Provide irq flags through DT/platform data
  mfd: Make AB8500_CORE select POWER_SUPPLY to fix build error
  mfd: omap-usb-host: Actually update hostconfig

Pull hwmon fixes from Guenter Roeck:
 "Bug fixes for pmbus, ltc2978, and lineage-pem drivers

  Added specific maintainer for some hwmon drivers"

* tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (pmbus/ltc2978) Fix temperature reporting
  hwmon: (pmbus) Fix krealloc() misuse in pmbus_add_attribute()
  hwmon: (lineage-pem) Add missing terminating entry for pem_[input|fan]_attributes
  MAINTAINERS: Add maintainer for MAX6697, INA209, and INA2XX drivers

This patch fixes a kernel crash when using precise sampling (PEBS)
after a suspend/resume. Turns out the CPU notifier code is not invoked
on CPU0 (BP). Therefore, the DS_AREA (used by PEBS) is not restored properly
by the kernel and keeps it power-on/resume value of 0 causing any PEBS
measurement to crash when running on CPU0.

The workaround is to add a hook in the actual resume code to restore
the DS Area MSR value. It is invoked for all CPUS. So for all but CPU0,
the DS_AREA will be restored twice but this is harmless.
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

During the transition to the generic parser, the hook to the codec
specific automute function was forgotten.  This resulted in the silent
output on some MacBooks.
Signed-off-by: Takashi Iwai <tiwai@suse.de>

"chn" here is a number between 0 and 255, but ->chn_info[] only has
16 elements so there is a potential write beyond the end of the
array.

If the seq_mode isn't SEQ_2 then we let the individual drivers
(either opl3.c or midi_synth.c) handle it.  Those functions all
do a bounds check on "chn" so I haven't changed anything here.
The opl3.c driver has up to 18 channels and not 16.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

4740f73f "mfd: remove use of __devexit" removed the __devexit annotation
on the twl4030_madc_remove function, but left an __exit_p() present on the
pointer to this function. Using __exit_p was as wrong with the devexit in
place as it is now, but now we get a gcc warning about an unused function.

In order for the twl4030_madc_remove to work correctly in built-in code, we
have to remove the __exit_p.

Cc: Bill Pemberton <wfp5p@virginia.edu>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

spec->dsp_state is initialized to DSP_DOWNLOAD_INIT, no need to reset
and check it in ca0132_download_dsp().
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

Instead of using the dspload_is_loaded() function, check the dsp_state
that is kept in the spec.  The dspload_is_loaded() function returns
true if the DSP transfer was never started.  This false-positive leads
to multiple second delays when ca0132_setup_efaults() times out on
each write.
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

If dspload_image() fails, it was ignored and dspload_wait_loaded() was
still called.  dsp_loaded should never be set to true in this case,
skip it.  The check in dspload_wait_loaded() return true if the DSP is
loaded or if it never started.
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

The vm_flags introduced in 6d7825b1 ("mm/fremap.c: fix oops on error
path") is supposed to avoid a compiler warning about unitialized
vm_flags without changing the generated code.

However I am concerned that this is going to be very brittle, and fail
with some compiler versions. The failure could be either of:

- compiler could actually load vma->vm_flags before checking for the
  !vma condition, thus reintroducing the oops

- compiler could optimize out the !vma check, since the pointer just got
  dereferenced shortly before (so the compiler knows it can't be NULL!)

I propose reversing this part of the change and initializing vm_flags to 0
just to avoid the bogus uninitialized use warning.
Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Tommi Rantala <tt.rantala@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull fix for hlist_entry_safe() regression from Paul McKenney:
 "This contains a single commit that fixes a regression in
  hlist_entry_safe().  This macro references its argument twice, which
  can cause NULL-pointer errors.  This commit applies a gcc statement
  expression, creating a temporary variable to avoid the double
  reference.  This has been posted to LKML at

    https://lkml.org/lkml/2013/3/9/75.

  Kudos to CAI Qian, whose testing uncovered this, to Eric Dumazet, who
  spotted root cause, and to Li Zefan, who tested this commit."

* 'rcu/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu:
  list: Fix double fetch of pointer in hlist_entry_safe()

The current version of hlist_entry_safe() fetches the pointer twice,
once to test for NULL and the other to compute the offset back to the
enclosing structure.  This is OK for normal lock-based use because in
that case, the pointer cannot change.  However, when the pointer is
protected by RCU (as in "rcu_dereference(p)"), then the pointer can
change at any time.  This use case can result in the following sequence
of events:

1.	CPU 0 invokes hlist_entry_safe(), fetches the RCU-protected
	pointer as sees that it is non-NULL.

2.	CPU 1 invokes hlist_del_rcu(), deleting the entry that CPU 0
	just fetched a pointer to.  Because this is the last entry
	in the list, the pointer fetched by CPU 0 is now NULL.

3.	CPU 0 refetches the pointer, obtains NULL, and then gets a
	NULL-pointer crash.

This commit therefore applies gcc's "({ })" statement expression to
create a temporary variable so that the specified pointer is fetched
only once, avoiding the above sequence of events.  Please note that
it is the caller's responsibility to use rcu_dereference() as needed.
This allows RCU-protected uses to work correctly without imposing
any additional overhead on the non-RCU case.

Many thanks to Eric Dumazet for spotting root cause!
Reported-by: CAI Qian <caiqian@redhat.com>
Reported-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Li Zefan <lizefan@huawei.com>

Pull ext2, ext3, reiserfs, quota fixes from Jan Kara:
 "A fix for regression in ext2, and a format string issue in ext3.  The
  rest isn't too serious."

* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  ext2: Fix BUG_ON in evict() on inode deletion
  reiserfs: Use kstrdup instead of kmalloc/strcpy
  ext3: Fix format string issues
  quota: add missing use of dq_data_lock in __dquot_initialize

Creating snapshot passes extent_root to commit its transaction,
but it can lead to the warning of checking root for quota in
the __btrfs_end_transaction() when someone else is committing
the current transaction.  Since we've recorded the needed root
in trans_handle, just use it to get rid of the warning.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

If one of qgroup fails to reserve firstly, we should return immediately,
it is unnecessary to continue check.
Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

The callers of lookup_inline_extent_info all handle getting an error back
properly, so return an error if we have corruption instead of being a jerk and
panicing.  Still WARN_ON() since this is kind of crucial and I've been seeing it
a bit too much recently for my taste, I think we're doing something wrong
somewhere.  Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Doing this would reliably fail with -EBUSY for me:

# mount /dev/sdb2 /mnt/scratch; umount /mnt/scratch; mkfs.btrfs -f /dev/sdb2
...
unable to open /dev/sdb2: Device or resource busy

because mkfs.btrfs tries to open the device O_EXCL, and somebody still has it.

Using systemtap to track bdev gets & puts shows a kworker thread doing a
blkdev put after mkfs attempts a get; this is left over from the unmount
path:

btrfs_close_devices
	__btrfs_close_devices
		call_rcu(&device->rcu, free_device);
			free_device
				INIT_WORK(&device->rcu_work, __free_device);
				schedule_work(&device->rcu_work);

so unmount might complete before __free_device fires & does its blkdev_put.

Adding an rcu_barrier() to btrfs_close_devices() causes unmount to wait
until all blkdev_put()s are done, and the device is truly free once
unmount completes.

Cc: stable@vger.kernel.org
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Remove a useless function declaration
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Using spinning case instead of blocking will result in better concurrency
overall.
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

On LTC2978, only READ_TEMPERATURE is supported. It reports
the internal junction temperature. This register is unpaged.

On LTC3880, READ_TEMPERATURE and READ_TEMPERATURE2 are supported.
READ_TEMPERATURE is paged and reports external temperatures.
READ_TEMPERATURE2 is unpaged and reports the internal junction
temperature.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Cc: stable@vger.kernel.org # 3.2+
Acked-by: Jean Delvare <khali@linux-fr.org>

If there are no internal speakers, we should not turn the eapd switch
off, because it might be necessary to keep high for Headphone.

BugLink: https://bugs.launchpad.net/bugs/1155016Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

If krealloc() returns NULL, it *doesn't* free the original. So any code
of the form 'foo = krealloc(foo, …);' is almost certainly a bug.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>