The mlx4_en driver allocates the number of objects for the CPU affinity
reverse-map based on the number of rx rings of the device. However,
mlx4_assign_eq() calls irq_cpu_rmap_add() as many times as IRQ's are
assigned to EQ's, which can be as large as mlx4_dev->caps.comp_pool. If
caps.comp_pool is larger than rx_ring_num we will eventually hit the
BUG_ON() in cpu_rmap_add().

Fix this problem by allocating space for the maximum number of CPU
affinity reverse-map objects we might want to add.
Signed-off-by: NKleber Sacilotto de Souza <klebers@linux.vnet.ibm.com>
Acked-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The memory to hold the network device tx_cq is not being allocated with
the correct size in mlx4_en_init_netdev(). It should use MAX_TX_RINGS
instead of MAX_RX_RINGS. This can cause problems if the number of tx
rings being used is greater than MAX_RX_RINGS.
Signed-off-by: NKleber Sacilotto de Souza <klebers@linux.vnet.ibm.com>
Acked-by: NAmir Vadai <amirv@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In fast open the sender unncessarily reduces the space available
for data in SYN by 12 bytes.  This is because in the sender
incorrectly reserves space for TS option twice in tcp_send_syn_data():
tcp_mtu_to_mss() already accounts for TS option space. But it further
reserves MAX_TCP_OPTION_SPACE when computing the payload space.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The mem cgroup socket limit is only used if the config option is
enabled. Found with sparse
Signed-off-by: NStephen Hemminger <stephen@networkplumber.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Dave Jones reported a lockdep splat occurring in IP defrag code.

commit 6d7b857d (net: use lib/percpu_counter API for
fragmentation mem accounting) added a possible deadlock.

Because percpu_counter_sum_positive() needs to acquire
a lock that can be used from softirq, we need to disable BH
in sum_frag_mem_limit()
Reported-by: NDave Jones <davej@redhat.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

commit 7f7d6c28 (net: fec: Ensure that initialization is done prior to
request_irq()) placed fec_ptp_init() into a point that ptp clock was not
available, which causes a division by zero in fec_ptp_start_cyclecounter():

[   17.895723] Division by zero in kernel.
[   17.899571] Backtrace:
[   17.902094] [<80012564>] (dump_backtrace+0x0/0x10c) from [<8056deec>]
(dump_stack+0x18/0x1c)
[   17.910539]  r6:bfba8500 r5:8075c950 r4:bfba8000 r3:bfbd0000
[   17.916284] [<8056ded4>] (dump_stack+0x0/0x1c) from [<80012688>]
(__div0+0x18/0x20)
[   17.923968] [<80012670>] (__div0+0x0/0x20) from [<802829c4>] (Ldiv0+0x8/0x10)
[   17.931140] [<80398534>] (fec_ptp_start_cyclecounter+0x0/0x110) from
[<80394f64>] (fec_restart+0x6c8/0x754)
[   17.940898] [<8039489c>] (fec_restart+0x0/0x754) from [<803969a0>]
(fec_enet_adjust_link+0xdc/0x108)
[   17.950046] [<803968c4>] (fec_enet_adjust_link+0x0/0x108) from [<80390bc4>]
(phy_state_machine+0x178/0x534)
...

Fix this by rearraging the code so that fec_ptp_init() is called only after
the clocks have been properly acquired.

Tested on both mx53 and mx6 platforms.
Reported-by: NJim Baxter <jim_baxter@mentor.com>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Smatch found a locking bug in netif_set_xps_queue in which we were not
releasing the lock in the case of an allocation failure.

This change corrects that so that we release the xps_map_mutex before
returning -ENOMEM in the case of an allocation failure.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Now we handle icmp errors in each transport protocol's err_handler,
for icmp protocols, that is ping_err. Since this handler only care
of those icmp errors triggered by echo request, errors triggered
by echo reply(which sent by kernel) are sliently ignored.

So wrap ping_err() with icmp_err() to deal with those icmp errors.
Signed-off-by: NLi Wei <lw@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It looks like its possible to open thousands of TCP IPv6
sessions on a server, all landing in a single slot of TCP hash
table. Incoming packets have to lookup sockets in a very
long list.

We should hash all bits from foreign IPv6 addresses, using
a salt and hash mix, not a simple XOR.

inet6_ehashfn() can also separately use the ports, instead
of xoring them.
Reported-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NJitendra Kalsaria <jitendra.kalsaria@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NManish Chopra <manish.chopra@qlogic.com>
Signed-off-by: NJitendra Kalsaria <jitendra.kalsaria@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

o Use correct function to clear loopback mode.
Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NJitendra Kalsaria <jitendra.kalsaria@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently request_irq() is called prior to fec_enet_init() and fec_ptp_init(),
which causes the following crash on a mx53qsb:

Unable to handle kernel NULL pointer dereference at virtual address 00000002
pgd = 80004000
[00000002] *pgd=00000000
Internal error: Oops: 5 [#1] SMP ARM
Modules linked in:
CPU: 0    Not tainted  (3.8.0-rc7-next-20130215+ #346)
PC is at fec_enet_interrupt+0xd0/0x348
LR is at fec_enet_interrupt+0xb8/0x348
pc : [<80372b7c>]    lr : [<80372b64>]    psr: 60000193
sp : df855c20  ip : df855c20  fp : df855c74
r10: 00000516  r9 : 1c000000  r8 : 00000000
r7 : 00000000  r6 : 00000000  r5 : 00000000  r4 : df9b7800
r3 : df9b7df4  r2 : 00000000  r1 : 00000000  r0 : df9b7d34

Ensure that such initialization functions are called prior to requesting the
interrupts, so that all necessary the data structures are in place when the
irqs occur.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We should get 'type' and 'code' from the outer ICMP header.
Signed-off-by: NLi Wei <lw@cn.fujitsu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pull networking update from David Miller:

 1) Checkpoint/restarted TCP sockets now can properly propagate the TCP
    timestamp offset.  From Andrey Vagin.

 2) VMWARE VM VSOCK layer, from Andy King.

 3) Much improved support for virtual functions and SR-IOV in bnx2x,
    from Ariel ELior.

 4) All protocols on ipv4 and ipv6 are now network namespace aware, and
    all the compatability checks for initial-namespace-only protocols is
    removed.  Thanks to Tom Parkin for helping deal with the last major
    holdout, L2TP.

 5) IPV6 support in netpoll and network namespace support in pktgen,
    from Cong Wang.

 6) Multiple Registration Protocol (MRP) and Multiple VLAN Registration
    Protocol (MVRP) support, from David Ward.

 7) Compute packet lengths more accurately in the packet scheduler, from
    Eric Dumazet.

 8) Use per-task page fragment allocator in skb_append_datato_frags(),
    also from Eric Dumazet.

 9) Add support for connection tracking labels in netfilter, from
    Florian Westphal.

10) Fix default multicast group joining on ipv6, and add anti-spoofing
    checks to 6to4 and 6rd.  From Hannes Frederic Sowa.

11) Make ipv4/ipv6 fragmentation memory limits more reasonable in modern
    times, rearrange inet frag datastructures for better cacheline
    locality, and move more operations outside of locking.  From Jesper
    Dangaard Brouer.

12) Instead of strict master <--> slave relationships, allow arbitrary
    scenerios with "upper device lists".  From Jiri Pirko.

13) Improve rate limiting accuracy in TBF and act_police, also from Jiri
    Pirko.

14) Add a BPF filter netfilter match target, from Willem de Bruijn.

15) Orphan and delete a bunch of pre-historic networking drivers from
    Paul Gortmaker.

16) Add TSO support for GRE tunnels, from Pravin B SHelar.  Although
    this still needs some minor bug fixing before it's %100 correct in
    all cases.

17) Handle unresolved IPSEC states like ARP, with a resolution packet
    queue.  From Steffen Klassert.

18) Remove TCP Appropriate Byte Count support (ABC), from Stephen
    Hemminger.  This was long overdue.

19) Support SO_REUSEPORT, from Tom Herbert.

20) Allow locking a socket BPF filter, so that it cannot change after a
    process drops capabilities.

21) Add VLAN filtering to bridge, from Vlad Yasevich.

22) Bring ipv6 on-par with ipv4 and do not cache neighbour entries in
    the ipv6 routes, from YOSHIFUJI Hideaki.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1538 commits)
  ipv6: fix race condition regarding dst->expires and dst->from.
  net: fix a wrong assignment in skb_split()
  ip_gre: remove an extra dst_release()
  ppp: set qdisc_tx_busylock to avoid LOCKDEP splat
  atl1c: restore buffer state
  net: fix a build failure when !CONFIG_PROC_FS
  net: ipv4: fix waring -Wunused-variable
  net: proc: fix build failed when procfs is not configured
  Revert "xen: netback: remove redundant xenvif_put"
  net: move procfs code to net/core/net-procfs.c
  qmi_wwan, cdc-ether: add ADU960S
  bonding: set sysfs device_type to 'bond'
  bonding: fix bond_release_all inconsistencies
  b44: use netdev_alloc_skb_ip_align()
  xen: netback: remove redundant xenvif_put
  net: fec: Do a sanity check on the gpio number
  ip_gre: propogate target device GSO capability to the tunnel device
  ip_gre: allow CSUM capable devices to handle packets
  bonding: Fix initialize after use for 3ad machine state spinlock
  bonding: Fix race condition between bond_enslave() and bond_3ad_update_lacp_rate()
  ...

Pull sparc updates from David Miller:
 "Mostly more sparc64 THP bug fixes, and a refactoring of SMP bootup on
  sparc32 from Sam Ravnborg."

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
  sparc32: refactor smp boot
  sparc64: Fix huge PMD to PTE translation for sun4u in TLB miss handler.
  sparc64: Fix tsb_grow() in atomic context.
  sparc64: Handle hugepage TSB being NULL.
  sparc64: Fix gfp_flags setting in tsb_grow().

Pull arm64 patches from Catalin Marinas:

 - SMP support for the PSCI booting protocol (power state coordination
   interface).

 - Simple earlyprintk support.

 - Platform devices populated by default from the DT (SoC-agnostic).

 - CONTEXTIDR support (used by external trace tools).

* tag 'arm64-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/cmarinas/linux-aarch64:
  arm64: mm: update CONTEXTIDR register to contain PID of current process
  arm64: atomics: fix grossly inconsistent asm constraints for exclusives
  arm64: compat: use compat_uptr_t type for compat_ucontext.uc_link
  arm64: Select ARCH_WANT_FRAME_POINTERS
  arm64: Add kvm_para.h and xor.h generic headers
  arm64: SMP: enable PSCI boot method
  arm64: psci: add support for PSCI invocations from the kernel
  arm64: SMP: rework the SMP code to be enabling method agnostic
  arm64: perf: add guest vs host discrimination
  arm64: add COMPAT_PSR_*_BIT flags
  arm64: Add simple earlyprintk support
  arm64: Populate the platform devices

Pull ARM updates (part two) from Russell King:

 - breakpoint and perf updates from Will Deacon.

 - hypervisor boot mode updates from Will.

 - support for Power State Coordination Interface via the Hypervisor

 - core ARM support for KVM

* 'for-linus-2' of git://git.linaro.org/people/rmk/linux-arm: (32 commits)
  KVM: ARM: Add maintainer entry for KVM/ARM
  KVM: ARM: Power State Coordination Interface implementation
  KVM: ARM: Handle I/O aborts
  KVM: ARM: Handle guest faults in KVM
  KVM: ARM: VFP userspace interface
  KVM: ARM: Demux CCSIDR in the userspace API
  KVM: ARM: User space API for getting/setting co-proc registers
  KVM: ARM: Emulation framework and CP15 emulation
  KVM: ARM: World-switch implementation
  KVM: ARM: Inject IRQs and FIQs from userspace
  KVM: ARM: Memory virtualization setup
  KVM: ARM: Hypervisor initialization
  KVM: ARM: Initial skeleton to compile KVM support
  ARM: Section based HYP idmap
  ARM: Add page table and page defines needed by KVM
  ARM: perf: simplify __hw_perf_event_init err handling
  ARM: perf: remove unnecessary checks for idx < 0
  ARM: perf: handle armpmu_register failing
  ARM: perf: don't pretend to support counting of L1I writes
  ARM: perf: remove redundant NULL check on cpu_pmu
  ...

Pull ARM updates (part one) from Russell King:

 - MMC patches from Ulf Hansson and Pawel Moll.  These add support for
   DDR mode and the latest variant found on ARM Versatile Express, as
   well as a number of cleanups.

 - A fix for to improve the behaviour of ARMs sched_clock()

 - Changes to the ARM ioremap() code.  I'm not convinced with the
   primary arguments for this, but it's been around for a while, and
   people seem happy with it - and the "other" justification for this is
   at

      http://lkml.org/lkml/2012/12/6/184

 - Add SCHED_HRTICK to ARMs Kconfig

 - Making the ARM SHA/AES code Thumb-2 compatible

 - A collection of other small updates.

* 'for-linus' of git://git.linaro.org/people/rmk/linux-arm: (26 commits)
  ARM: add SCHED_HRTICK config option
  ARM: 7650/1: mm: replace direct access to mm->context.id with new macro
  ARM: 7649/1: mm: mm->context.id fix for big-endian
  ARM: 7648/1: pci: Allow passing per-controller private data
  ARM: 7647/1: pci: Keep pci_common_init() around after init
  ARM: fix warnings introduced by previous patch
  ARM: 7646/1: mm: use static_vm for managing static mapped areas
  ARM: 7645/1: ioremap: introduce an infrastructure for static mapped area
  ARM: 7644/1: vmregion: remove vmregion code entirely
  MAINTAINERS: Re-assert MMCI driver maintainer status
  MAINTAINERS: add additional file for MMCI driver
  MAINTAINERS: add maintainer entry for AMBA serial drivers
  ARM: 7637/1: memory: use SZ_ constants for defining the virtual memory layout
  ARM: 7643/1: sched: correct update_sched_clock()
  ARM: 7635/1: versatile: fix the PCI IRQ regression
  ARM: 7639/1: cache-l2x0: add missed dummy outer_resume entry
  ARM: 7630/1: mmc: mmci: Fixup and cleanup code for DMA handling
  ARM: 7632/1: spinlock: avoid exclusive accesses on unlock() path
  ARM: 7631/1: mmc: mmci: Add new VE MMCI variant
  ARM: 7623/1: mmc: mmci: Fixup clock gating when freq is 0 for ST-variants
  ...

Pull m68k update from Geert Uytterhoeven.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k:
  m68k: Sort out !CONFIG_MMU_SUN3 vs. CONFIG_HAS_DMA
  swim: Add missing spinlock init

Introduce a common smp_callin() function to call
from trampoline_32.S.
Add platform specific functions to handle the
platform details.

This is in preparation for a patch that will
unify the smp boot stuff for all architectures.
sparc32 was significantly different to warrant
this patch in preparation.
Signed-off-by: NSam Ravnborg <sam@ravnborg.org>
Acked-by: NDavid S. Miller <davem@davemloft.net>
Cc: "Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com>
Acked-by: NSrivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When we set the sun4u version of the PTE execute bit, it's:

	or	REG, _PAGE_EXEC_4U, REG

_PAGE_EXEC_4U is 0x1000, unfortunately the immedate field of the
'or' instruction is a signed 13-bit value.  So the above actually
assembles into:

	or	REG, -4096, REG

completely corrupting the final PTE value.

Set it with a:

	sethi	%hi(_PAGE_EXEC_4U), TMP
	or	REG, TMP, REG

sequence instead.

This fixes "git gc" crashes on sun4u machines.
Reported-by: NMeelis Roos <mroos@linux.ee>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Eric Dumazet wrote:
| Some strange crashes happen in rt6_check_expired(), with access
| to random addresses.
|
| At first glance, it looks like the RTF_EXPIRES and
| stuff added in commit 1716a961
| (ipv6: fix problem with expired dst cache)
| are racy : same dst could be manipulated at the same time
| on different cpus.
|
| At some point, our stack believes rt->dst.from contains a dst pointer,
| while its really a jiffie value (as rt->dst.expires shares the same area
| of memory)
|
| rt6_update_expires() should be fixed, or am I missing something ?
|
| CC Neil because of https://bugzilla.redhat.com/show_bug.cgi?id=892060

Because we do not have any locks for dst_entry, we cannot change
essential structure in the entry; e.g., we cannot change reference
to other entity.

To fix this issue, split 'from' and 'expires' field in dst_entry
out of union.  Once it is 'from' is assigned in the constructor,
keep the reference until the very last stage of the life time of
the object.

Of course, it is unsafe to change 'from', so make rt6_set_from simple
just for fresh entries.
Reported-by: NEric Dumazet <eric.dumazet@gmail.com>
Reported-by: NNeil Horman <nhorman@tuxdriver.com>
CC: Gao Feng <gaofeng@cn.fujitsu.com>
Signed-off-by: NYOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Reviewed-by: NEric Dumazet <edumazet@google.com>
Reported-by: NSteinar H. Gunderson <sesse@google.com>
Reviewed-by: NNeil Horman <nhorman@tuxdriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

commit c9af6db4 (net: Fix possible wrong checksum generation)
has a suspicous piece:

	-       skb_shinfo(skb1)->gso_type = skb_shinfo(skb)->gso_type;
	-
	+       skb_shinfo(skb)->tx_flags = skb_shinfo(skb1)->tx_flags & SKBTX_SHARED_FRAG;

skb1 is the new skb, therefore should be on the left side of the assignment.
This patch fixes it.

Cc: Pravin B Shelar <pshelar@nicira.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: NCong Wang <amwang@redhat.com>
Acked-by: NPravin B Shelar <pshelar@nicira.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pull EDAC updates from Borislav Petkov:
 "Mostly AMD's side of EDAC.  It is basically a new family enablement
  stuff: AMD F16h MCE decoding enablement from Jacob Shin.  The rest is
  trivial cleanups."

* tag 'edac_3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/bp/bp:
  mpc85xx_edac: Fix typo
  EDAC, MCE, AMD: Remove unneeded exports
  EDAC, MCE, AMD: Add MCE decoding support for Family 16h
  EDAC, MCE, AMD: Make MC2 decoding per-family
  amd64_edac: Remove dead code

Pull ACPI and power management updates from Rafael Wysocki:

 - Rework of the ACPI namespace scanning code from Rafael J.  Wysocki
   with contributions from Bjorn Helgaas, Jiang Liu, Mika Westerberg,
   Toshi Kani, and Yinghai Lu.

 - ACPI power resources handling and ACPI device PM update from Rafael
   J Wysocki.

 - ACPICA update to version 20130117 from Bob Moore and Lv Zheng with
   contributions from Aaron Lu, Chao Guan, Jesper Juhl, and Tim Gardner.

 - Support for Intel Lynxpoint LPSS from Mika Westerberg.

 - cpuidle update from Len Brown including Intel Haswell support, C1
   state for intel_idle, removal of global pm_idle.

 - cpuidle fixes and cleanups from Daniel Lezcano.

 - cpufreq fixes and cleanups from Viresh Kumar and Fabio Baltieri with
   contributions from Stratos Karafotis and Rickard Andersson.

 - Intel P-states driver for Sandy Bridge processors from Dirk
   Brandewie.

 - cpufreq driver for Marvell Kirkwood SoCs from Andrew Lunn.

 - cpufreq fixes related to ordering issues between acpi-cpufreq and
   powernow-k8 from Borislav Petkov and Matthew Garrett.

 - cpufreq support for Calxeda Highbank processors from Mark Langsdorf
   and Rob Herring.

 - cpufreq driver for the Freescale i.MX6Q SoC and cpufreq-cpu0 update
   from Shawn Guo.

 - cpufreq Exynos fixes and cleanups from Jonghwan Choi, Sachin Kamat,
   and Inderpal Singh.

 - Support for "lightweight suspend" from Zhang Rui.

 - Removal of the deprecated power trace API from Paul Gortmaker.

 - Assorted updates from Andreas Fleig, Colin Ian King, Davidlohr Bueso,
   Joseph Salisbury, Kees Cook, Li Fei, Nishanth Menon, ShuoX Liu,
   Srinivas Pandruvada, Tejun Heo, Thomas Renninger, and Yasuaki
   Ishimatsu.

* tag 'pm+acpi-3.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (267 commits)
  PM idle: remove global declaration of pm_idle
  unicore32 idle: delete stray pm_idle comment
  openrisc idle: delete pm_idle
  mn10300 idle: delete pm_idle
  microblaze idle: delete pm_idle
  m32r idle: delete pm_idle, and other dead idle code
  ia64 idle: delete pm_idle
  cris idle: delete idle and pm_idle
  ARM64 idle: delete pm_idle
  ARM idle: delete pm_idle
  blackfin idle: delete pm_idle
  sparc idle: rename pm_idle to sparc_idle
  sh idle: rename global pm_idle to static sh_idle
  x86 idle: rename global pm_idle to static x86_idle
  APM idle: register apm_cpu_idle via cpuidle
  cpufreq / intel_pstate: Add kernel command line option disable intel_pstate.
  cpufreq / intel_pstate: Change to disallow module build
  tools/power turbostat: display SMI count by default
  intel_idle: export both C1 and C1E
  ACPI / hotplug: Fix concurrency issues and memory leaks
  ...

Pull device tree changes from Grant Likely:
 "All around device tree changes destined for v3.8.  Aside from the
  documentation updates the highlights in this branch include:

   - Kbuild changes for using CPP with .dts files
   - locking fix from preempt_rt patchset
   - include DT alias names in device uevent
   - Selftest bugfixes and improvements
   - New function for counting phandles stanzas in a property
   - constify argument to of_node_full_name()
   - Various bug fixes

  This tree did also contain a commit to use platform_device_add instead
  of open-coding the device add code, but it caused problems with amba
  devices and needed to be reverted."

* tag 'dt-for-linus' of git://git.secretlab.ca/git/linux: (23 commits)
  Revert "of: use platform_device_add"
  kbuild: limit dtc+cpp include path
  gpio: Make of_count_named_gpios() use new of_count_phandle_with_args()
  of: Create function for counting number of phandles in a property
  of/base: Clean up exit paths for of_parse_phandle_with_args()
  of/selftest: Use selftest() macro throughout
  of/selftest: Fix GPIOs selftest to cover the 7th case
  of: fix recursive locking in of_get_next_available_child()
  documentation/devicetree: Fix a typo in exynos-dw-mshc.txt
  OF: convert devtree lock from rw_lock to raw spinlock
  of/exynos_g2d: Add Bindings for exynos G2D driver
  kbuild: create a rule to run the pre-processor on *.dts files
  input: Extend matrix-keypad device tree binding
  devicetree: Move NS2 LEDs binding into LEDs directory
  of: use platform_device_add
  powerpc/5200: Fix size to request_mem_region() call
  documentation/devicetree: Fix typos
  of: add 'const' to of_node_full_name parameter
  of: Output devicetree alias names in uevent
  DT: add vendor prefixes for Renesas and Toshiba
  ...

Pull SPI changes from Grant Likely:
 "Changes to both core spi code and spi device drivers.  The driver
  changes are the usual set of bug fixes and platform enablement.

  Core code changes include:

   - More intelligent assignment of SPI bus numbers when using DT

   - Common mechanism for using gpios as CS lines

   - Pull checks for bits_per_word and transfer speed out of drivers and
     into core code

   - Ensure temporary DMA buffers are DMA safe"

* tag 'spi-for-linus' of git://git.secretlab.ca/git/linux: (50 commits)
  spi: Document cs_gpios and cs_gpio in kernel-doc
  spi/of: Fix initialization of cs_gpios array
  spi/pxa2xx: add support for Lynxpoint SPI controllers
  spi/pxa2xx: add support for Intel Low Power Subsystem SPI
  spi/pxa2xx: add support for SPI_LOOP
  spi/pxa2xx: add support for runtime PM
  spi/pxa2xx: add support for DMA engine
  spi/pxa2xx: break out the private DMA API usage into a separate file
  spi/ath79: add shutdown handler
  spi/mips-lantiq: set SPI_MASTER_HALF_DUPLEX flag
  spi/mips-lantiq: make use of spi_finalize_current_message
  spi/bcm63xx: work around inability to keep CS up
  spi/davinci: use request_threaded_irq() to fix deadlock
  spi/orion: Use module_platform_driver()
  spi/bcm63xx: reject transfers unable to transfer
  spi: Ensure memory used for spi_write_then_read() is DMA safe
  spi/spi-mpc512x-psc: init mode bits supported by the driver
  spi/mpc512x-psc: don't use obsolet cell-index property
  spi: Remove erroneous __init, __exit and __exit_p() references in drivers
  spi/s3c64xx: fix checkpatch warnings and error
  ...

Pull clock framework update from Michael Turquette:
 "The common clock framework changes for 3.9 are almost entirely fixes.

  None are dire enough to be Cc'd to stable which may be interpreted to
  mean that users of the framework are reaching stability.  Lots of new
  adoption of this framework is via DeviceTree data and that comes
  through the respective architecture and platform trees instead of
  through the clk framework tree.

  Two new features are improved debugfs output and an improvement to how
  DT clocks are initialized by reusing a common method."

* tag 'clk-for-linus' of git://git.linaro.org/people/mturquette/linux: (25 commits)
  clk: sunxi: remove stale Makefile entry
  clk: vexpress: Use common of_clk_init() function
  clk: zynq: Use common of_clk_init() function
  clk: vt8500: Use common of_clk_init() function
  clk: highbank: Use common of_clk_init() function
  clk: sunxi: Use common of_clk_init() function
  clk: add common of_clk_init() function
  clk: Deduplicate exit code in clk_set_rate
  clk: beautify Makefile
  clk-divider: fix macros
  clk: prima2: enable dt-binding clkdev mapping
  clk: mxs: Index is always positive
  clk: max77686: Avoid double free at remove time
  clk: remove exported function from __init section
  clk: vt8500: Add support for WM8750/WM8850 PLL clocks
  clk: vt8500: Fix division-by-0 when requested rate=0
  clk: vt8500: Fix device clock divisor calculations
  clk: vt8500: Fix error in PLL calculations on non-exact match.
  clk: max77686: Remove unnecessary NULL checking for container_of()
  clk: JSON debugfs clock tree summary
  ...

Pull input updates from Dmitry Torokhov:
 "Two new touchpad drivers - Cypress APA I2C Trackpad and Cypress PS/2
  touchpad and a big update to ALPS driver from Kevin Cernekee that adds
  support for "Rushmore" touchpads and paves way for adding support for
  "Dolphin" touchpads.

  There is also a new input driver for Goldfish emulator and also
  Android keyreset driver was folded into SysRq code.

  A few more drivers were updated with device tree bindings and others
  got some small cleanups and fixes."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: (55 commits)
  Input: cyttsp-spi - remove duplicate MODULE_ALIAS()
  Input: tsc2005 - add MODULE_ALIAS
  Input: tegra-kbc - require CONFIG_OF, remove platform data
  Input: synaptics - initialize pointer emulation usage
  Input: MT - do not apply filtering on emulated events
  Input: bma150 - make some defines public and fix some comments
  Input: bma150 - fix checking pm_runtime_get_sync() return value
  Input: ALPS - enable trackstick on Rushmore touchpads
  Input: ALPS - add support for "Rushmore" touchpads
  Input: ALPS - make the V3 packet field decoder "pluggable"
  Input: ALPS - move pixel and bitmap info into alps_data struct
  Input: ALPS - fix command mode check
  Input: ALPS - rework detection of Pinnacle AGx touchpads
  Input: ALPS - move {addr,nibble}_command settings into alps_set_defaults()
  Input: ALPS - use function pointers for different protocol handlers
  Input: ALPS - rework detection sequence
  Input: ALPS - introduce helper function for repeated commands
  Input: ALPS - move alps_get_model() down below hw_init code
  Input: ALPS - copy "model" info into alps_data struct
  Input: ALPS - document the alps.h data structures
  ...

Pull battery updates from Anton Vorontsov:
 "Four new drivers:

   - goldfish_battery:

     This is Android Emulator battery driver.  Originally from Google,
     but Intel folks reshaped it for mainline

   - pm2301_charger:

     A new driver for ST-Ericsson 2301 Power Management chip, uses
     AB8500 battery management core

   - qnap-poweroff:

     The driver adds poweroff functionality for QNAP NAS boxes

   - restart-poweroff:

     A generic driver that implements 'power off by restarting'.  The
     actual poweroff functionality is implemented through a bootloader,
     so Linux' task is just to restart the box.  The driver is useful on
     Buffalo Linkstation LS-XHL and LS-CHLv2 boards.  Andrew Lunn worked
     on submitting the driver (as well as qnap-poweroff above).

  Additionally:

   - A lot of fixes for ab8500 drivers.  This is a part of efforts of
     syncing internal ST-Ericsson development tree with the mainline.
     Lee Jones @ Linaro worked on compilation and reshaping these
     series.

   - New health properties for the power supplies: "Watchdog timer
     expire" and "Safety timer expire"

   - As usual, a bunch of fixes/cleanups here and there"

* tag 'for-v3.9' of git://git.infradead.org/battery-2.6: (81 commits)
  bq2415x_charger: Add support for offline and 100mA mode
  generic-adc-battery: Fix forever loop in gab_remove()
  goldfish_battery: Add missing GENERIC_HARDIRQS dependency
  da9030_battery: Include notifier.h
  bq27x00_battery: Fix reporting battery temperature
  power/reset: Remove newly introduced __dev* annotations
  lp8727_charger: Small cleanup in naming
  ab8500_btemp: Demote initcall sequence
  ds2782_battery: Add power_supply_changed() calls for proper uevent support
  power: Add battery driver for goldfish emulator
  u8500-charger: Delay for USB enumeration
  ab8500-bm: Remove individual [charger|btemp|fg|chargalg] pdata structures
  ab8500-charger: Do not touch VBUSOVV bits
  ab8500-fg: Use correct battery charge full design
  pm2301: LPN mode control support
  pm2301: Enable vbat low monitoring
  ab8500-bm: Flush all work queues before suspending
  ab8500-fg: Go to INIT_RECOVERY when charger removed
  ab8500-charger: Add support for autopower on AB8505 and AB9540
  abx500-chargalg: Add new sysfs interface to get current charge status
  ...

Fix up fairly straightforward conflicts in the ab8500 driver.  But since
it seems to be ARM-specific, I can't even compile-test the result..

If our first THP installation for an MM is via the set_pmd_at() done
during khugepaged's collapsing we'll end up in tsb_grow() trying to do
a GFP_KERNEL allocation with several locks held.

Simply using GFP_ATOMIC in this situation is not the best option
because we really can't have this fail, so we'd really like to keep
this an order 0 GFP_KERNEL allocation if possible.

Also, doing the TSB allocation from khugepaged is a really bad idea
because we'll allocate it potentially from the wrong NUMA node in that
context.

So what we do is defer the hugepage TSB allocation until the first TLB
miss we take on a hugepage.  This is slightly tricky because we have
to handle two unusual cases:

1) Taking the first hugepage TLB miss in the window trap handler.
   We'll call the winfix_trampoline when that is detected.

2) An initial TSB allocation via TLB miss races with a hugetlb
   fault on another cpu running the same MM.  We handle this by
   unconditionally loading the TSB we see into the current cpu
   even if it's non-NULL at hugetlb_setup time.
Reported-by: NMeelis Roos <mroos@ut.ee>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Accomodate the possibility that the TSB might be NULL at
the point that update_mmu_cache() is invoked.  This is
necessary because we will sometimes need to defer the TSB
allocation to the first fault that happens in the 'mm'.

Seperate out the hugepage PTE test into a seperate function
so that the logic is clearer.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We should "|= more_flags" rather than "= more_flags".
Reported-by: NDavid Rientjes <rientjes@google.com>
Acked-by: NDavid Rientjes <rientjes@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pull hwmon updates from Guenter Roeck:

 - New drivers for MAX6697 and compatibles and for INA209.

 - Added support for IT8771E, IT8772E, MAX34460, MAX34461, MCP98244, and
   ADT7420 to existing drivers.

 - Added support for additional attributes to various drivers.

 - Replaced SENSORS_LIMIT with clamp_val; retire SENSORS_LIMIT;

 - Clean up PMBus code to reduce its size; clean up adt7410 driver.

 - A couple of minor bug fixes as well as documentation cleanup.

 - Out-of-tree change: Replace SENSORS_LIMIT with clamp_val in
   platform/x86/eeepc-laptop driver.

* tag 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging: (32 commits)
  hwmon: (ntc_thermistor): Fix sparse warnings
  hwmon: (adt7410) Add device table entry for the adt7420
  hwmon: (adt7410) Use I2C_ADDRS helper macro
  hwmon: (adt7410) Use the SIMPLE_DEV_PM_OPS helper macro
  hwmon: (adt7410) Let suspend/resume depend on CONFIG_PM_SLEEP
  hwmon: (adt7410) Clear unwanted bits in the config register
  hwmon: (jc42) Add support for MCP98244
  hwmon: (pmbus) Clean up for code size reduction
  hwmon: (pmbus/max34440) Add support for MAX34460 and MAX34461
  hwmon: (pmbus) Add support for word status register
  hwmon: (pmbus/zl6100) Add support for VMON/VDRV
  hwmon: (pmbus) Add function to clear sensor cache
  hwmon: (pmbus) Add support for additional voltage sensor
  hwmon: (pmbus) Use krealloc to allocate attribute memory
  hwmon: (pmbus) Simplify memory allocation for sensor attributes
  hwmon: (pmbus) Improve boolean handling
  hwmon: (pmbus) Simplify memory allocation for labels and booleans
  hwmon: (pmbus) Use dev variable to represent client->dev
  hwmon: (pmbus) Fix 'Macros with multiple statements' checkpatch error
  hwmon: (pmbus) Drop unnecessary error messages in probe error path
  ...

Pull pinctrl changes from Linus Walleij:
 "These are the main pinctrl changes for the v3.9 merge window.  The
  most interesting change by far is how the device core grabs pinctrl
  default handles avoiding the need to stick boilerplate into driver
  consumers.

   - Grabbing of default pinctrl handles from the device core.  These
     are the hunks hitting drivers/base.  All is ACKed by Greg, after a
     long discussion about different alternatives.

   - Some stuff also touches the MFD and ARM SoC trees, this has been
     coordinated and ACKed.

   - New drivers for:
     - The Tegra 114 sub-SoC
     - Allwinner sunxi
     - New ABx500 driver and sub-SoC drivers for AB8500, AB8505, AB9540
       and AB8540.

   - Make it possible for hogged pins to enter a sleep mode, and make it
     possible for drivers to control that mode.

   - Various clean-up, extensions and device tree support to various pin
     controllers."

* tag 'pinctrl-for-v3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl: (68 commits)
  pinctrl: tegra: add clfvs function to Tegra114 support
  pinctrl: generic: rename input schmitt disable
  pinctrl/pinconfig: add debug interface
  pinctrl: samsung: remove duplicated line
  ARM: ux500: use real AB8500 IRQ numbers instead of virtual ones
  ARM: ux500: remove irq_base property from platform_data
  pinctrl/abx500: use direct IRQ defines
  pinctrl/abx500: replace IRQ offsets with table read-in values
  pinctrl/abx500: move IRQ handling to ab8500-core
  pinctrl: exynos5440: remove erroneous __init
  pinctrl/abx500: adjust offset for get_mode()
  pinctrl/abx500: add Device Tree support
  pinctrl/abx500: align GPIO cluster boundaries
  pinctrl/abx500: prevent error path from corrupting returning error
  pinctrl: sunxi: add of_xlate function
  pinctrl/lantiq: fix pin number in ltq_pmx_gpio_request_enable
  pinctrl/lantiq: add functionality to falcon_pinconf_dbg_show
  pinctrl/lantiq: fix pinconfig parameters
  pinctrl/lantiq: one of the boot leds was defined incorrectly
  pinctrl/lantiq: only probe available pad controllers
  ...

Pull regulator updates from Mark Brown:
 "A fairly quiet release for the regulator API, the bulk of the changes
  being lots of small cleanups and API updates contributed by Axel Lin
  with just a small set of larger changes:

   - New driver for LP8755

   - DT support for S5M8767, TPS51632, TPS6507x and TPS65090

   - Support for writing a "commit changes" bit in the regmap helper
     functions."

* tag 'regulator-3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator: (60 commits)
  regulator: Fix memory garbage dev_err printout.
  regulator: max77686: Reuse rdev_get_id() function.
  regulator: tps51632: Use regulator_[get|set]_voltage_sel_regmap
  regulator: as3711: Fix checking if no platform initialization data
  regulator: s5m8767: Prevent possible NULL pointer dereference
  regulator: s5m8767: Fix dev argument for devm_kzalloc and of_get_regulator_init_data
  regulator: core: Optimize _regulator_do_set_voltage if voltage does not change
  regulator: max8998: Let regulator core handle the case selector == old_selector
  regulator: s5m8767: Use of_get_child_count()
  regulator: anatop: improve precision of delay time
  regulator: show state for GPIO-controlled regulators
  regulator: s5m8767: Fix build in non-DT case
  regulator: add device tree support for s5m8767
  regulator: palmas: Remove a redundant setting for warm_reset
  regulator: mc13xxx: Use of_get_child_count()
  regulator: max8997: Use of_get_child_count()
  regulator: tps65090: Fix using wrong dev argument for calling of_regulator_match
  regulators: anatop: add set_voltage_time_sel interface
  regulator: Add missing of_node_put()
  regulator: tps6507x: Fix using wrong dev argument for calling of_regulator_match
  ...

Pull regmap updates from Mark Brown:
 "Several nice new features and performance improvements here,
  especially the first:

   - Support for using the cache infrastructure without the physical
     I/O, allowing devices which don't fit the physical model regmap has
     to take advantage of the cache infrastructure, contributed by
     Andrey Smirnov.

   - Several small improvements to the support for wake capable IRQs.

   - Support for asynchronous I/O, allowing us to come much closer to
     saturating fast buses like SPI.

   - Support for simple array caches, giving higher performance for use
     with MMIO devices.

   - Restoration of the use of bulk reads for handling interrupts,
     giving a performance improvement.

   - Support for 24 bit register addresses.

   - More performance improvements for debugfs."

* tag 'regmap-3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap: (24 commits)
  regmap: mmio: add register clock support
  regmap: debugfs: Factor out debugfs_tot_len calc into a function
  regmap: debugfs: Optimize seeking within blocks of registers
  regmap: debugfs: Add a `max_reg' member in struct regmap_debugfs_off_cache
  regmap: debugfs: Fix reading in register field units
  regmap: spi: Handle allocation failures gracefully
  regmap: Export regmap_async_complete()
  regmap: Export regmap_async_complete_cb
  regmap: include linux/sched.h to fix build
  regmap: spi: Support asynchronous I/O for SPI
  regmap: Add asynchronous I/O support
  regmap: Add "no-bus" option for regmap API
  regmap: regmap: avoid spurious warning in regmap_read_debugfs
  regmap: Add provisions to have user-defined write operation
  regmap: Add provisions to have user-defined read operation
  regmap: Add support for 24 bit wide register addresses
  mfd: wm5110: Mark wakes as inverted
  mfd: wm5102: Mark wakes as inverted
  regmap: irq: Support wake IRQ mask inversion
  regmap: irq: Fix sync of wake statuses to hardware
  ...

Pull cpuset changes from Tejun Heo:

 - Synchornization has seen a lot of changes with focus on decoupling
   cpuset synchronization from cgroup internal locking.

   After this change, there only remain a couple of mostly trivial
   dependencies on cgroup_lock outside cgroup core proper.  cgroup_lock
   is scheduled to be unexported in this devel cycle.

   This will finally remove the fragile locking order around cgroup
   (cgroup locking wants to / should be one of the outermost but yet has
   been acquired from deep inside individual controllers).

 - At this point, Li is most knowlegeable with cpuset and taking over
   the maintainership of cpuset.

* 'for-3.9-cpuset' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cpuset: drop spurious retval assignment in proc_cpuset_show()
  cpuset: fix RCU lockdep splat
  cpuset: update MAINTAINERS
  cpuset: remove cpuset->parent
  cpuset: replace cpuset->stack_list with cpuset_for_each_descendant_pre()
  cpuset: replace cgroup_mutex locking with cpuset internal locking
  cpuset: schedule hotplug propagation from cpuset_attach() if the cpuset is empty
  cpuset: pin down cpus and mems while a task is being attached
  cpuset: make CPU / memory hotplug propagation asynchronous
  cpuset: drop async_rebuild_sched_domains()
  cpuset: don't nest cgroup_mutex inside get_online_cpus()
  cpuset: reorganize CPU / memory hotplug handling
  cpuset: cleanup cpuset[_can]_attach()
  cpuset: introduce cpuset_for_each_child()
  cpuset: introduce CS_ONLINE
  cpuset: introduce ->css_on/offline()
  cpuset: remove fast exit path from remove_tasks_in_empty_cpuset()
  cpuset: remove unused cpuset_unlock()

Pull cgroup changes from Tejun Heo:
 "Nothing too drastic.

   - Removal of synchronize_rcu() from userland visible paths.

   - Various fixes and cleanups from Li.

   - cgroup_rightmost_descendant() added which will be used by cpuset
     changes (it will be a separate pull request)."

* 'for-3.9' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
  cgroup: fail if monitored file and event_control are in different cgroup
  cgroup: fix cgroup_rmdir() vs close(eventfd) race
  cpuset: fix cpuset_print_task_mems_allowed() vs rename() race
  cgroup: fix exit() vs rmdir() race
  cgroup: remove bogus comments in cgroup_diput()
  cgroup: remove synchronize_rcu() from cgroup_diput()
  cgroup: remove duplicate RCU free on struct cgroup
  sched: remove redundant NULL cgroup check in task_group_path()
  sched: split out css_online/css_offline from tg creation/destruction
  cgroup: initialize cgrp->dentry before css_alloc()
  cgroup: remove a NULL check in cgroup_exit()
  cgroup: fix bogus kernel warnings when cgroup_create() failed
  cgroup: remove synchronize_rcu() from rebind_subsystems()
  cgroup: remove synchronize_rcu() from cgroup_attach_{task|proc}()
  cgroup: use new hashtable implementation
  cgroups: fix cgroup_event_listener error handling
  cgroups: move cgroup_event_listener.c to tools/cgroup
  cgroup: implement cgroup_rightmost_descendant()
  cgroup: remove unused dummy cgroup_fork_callbacks()