Earlier versions of Tree SRCU were subject to a counter overflow bug that
could theoretically result in too-short grace periods.  This commit
eliminates this problem by adding an update-side memory barrier.
The short explanation is that if the updater sums the unlock counts
too late to see a given __srcu_read_unlock() increment, that CPU's
next __srcu_read_lock() must see the new value of ->srcu_idx, thus
incrementing the other bank of counters.  This eliminates the possibility
of destructive counter overflow as long as the srcu_read_lock() nesting
level does not exceed floor(ULONG_MAX/NR_CPUS/2), which should be an
eminently reasonable nesting limit, especially on 64-bit systems.
Reported-by: Lance Roy <ldr709@gmail.com>
Suggested-by: Lance Roy <ldr709@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

A number of the rcutorture test scenarios were not using the desired
Kconfig options because dependencies were preventing the selections in the
Kconfig-fragment files from being honored.  This commit therefore updates
the Kconfig-fragment files to account for these changes in dependencies.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

The rcutorture scripting handles the CONFIG_*_TORTURE_TEST Kconfig
options specially, and therefore greps them out of the Kconfig-fragment
files.  Unfortunately, a poor choice of grep pattern means that the
CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP, CONFIG_RCU_TORTURE_TEST_SLOW_INIT,
and CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT Kconfig options are also grepped
out, preventing rcutorture from using them.  This commit therefore fixes
the offending grep pattern to focus only on the CONFIG_*_TORTURE_TEST
Kconfig options.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

Currently rcu_barrier() uses call_rcu() to enqueue new callbacks
on each CPU with a non-empty callback list.  This works, but means
that rcu_barrier() forces grace periods that are not otherwise needed.
The key point is that rcu_barrier() never needs to wait for a grace
period, but instead only for all pre-existing callbacks to be invoked.
This means that rcu_barrier()'s new callbacks should be placed in
the callback-list segment containing the last pre-existing callback.

This commit makes this change using the new rcu_segcblist_entrain()
function.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

A robust combination of paranoia and cowardice has resulted in retaining
Classic SRCU (CONFIG_CLASSIC_SRCU) as a backup for the shiny new Tiny
and Tree SRCU implementations.  If it is to be a viable backup, it of
course needs to be tested.  This commit therefore adds an rcutorture
scenario named SRCU-C for Classic SRCU.  This commit also adds this
scenario to the set that are run by default.

Once sufficient good experience has accumulated for Tiny and Tree SRCU,
this test will be removed, along with the Classic SRCU implementation
itself.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

This commit adds an SRCU-t rcutorture scenario for the new Tiny SRCU
implementation, removing the need to pass the --bootargs parameter to
kvm.sh to run Tiny SRCU tests.  This commit also adds SRCU-t to the set
of scenarios that are run by default.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

Kconfig "select" clauses can defeat Kconfig-fragment file attempts to
clear a given Kconfig variable, and dependencies can defeat attempts to
set a given Kconfig variable.  Because "select" clauses and dependencies
can be added at any time, there needs to be a way to verify that the
Kconfig-fragment file's requests were honored.  And there is, except
that it is buggy.  This commit therefore provides the needed fix.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

This commit adds a test for a three-level srcu_node tree for Tree SRCU
in the existing SRCU-P scenario.  This requires enabling CONFIG_RCU_EXPERT,
so the CONFIG_RCU_EXPERT=n scenario is now SRCU-N.  The reason for using
SRCU-P for the tall tree is that preemption raises the possibility of
locating more bugs than does the non-preemptive SRCU-N.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

Back when SRCU was simpler, there wasn't much need for lockdep.
However, with Tree SRCU, it is needed.  This commit therefore adds
CONFIG_PROVE_LOCKING to the SRCU-P scenario.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

Linu Cherian reported a WARN in cleanup_srcu_struct() when shutting
down a guest running iperf on a VFIO assigned device.  This happens
because irqfd_wakeup() calls srcu_read_lock(&kvm->irq_srcu) in interrupt
context, while a worker thread does the same inside kvm_set_irq().  If the
interrupt happens while the worker thread is executing __srcu_read_lock(),
updates to the Classic SRCU ->lock_count[] field or the Tree SRCU
->srcu_lock_count[] field can be lost.

The docs say you are not supposed to call srcu_read_lock() and
srcu_read_unlock() from irq context, but KVM interrupt injection happens
from (host) interrupt context and it would be nice if SRCU supported the
use case.  KVM is using SRCU here not really for the "sleepable" part,
but rather due to its IPI-free fast detection of grace periods.  It is
therefore not desirable to switch back to RCU, which would effectively
revert commit 719d93cd ("kvm/irqchip: Speed up KVM_SET_GSI_ROUTING",
2014-01-16).

However, the docs are overly conservative.  You can have an SRCU instance
only has users in irq context, and you can mix process and irq context
as long as process context users disable interrupts.  In addition,
__srcu_read_unlock() actually uses this_cpu_dec() on both Tree SRCU and
Classic SRCU.  For those two implementations, only srcu_read_lock()
is unsafe.

When Classic SRCU's __srcu_read_unlock() was changed to use this_cpu_dec(),
in commit 5a41344a ("srcu: Simplify __srcu_read_unlock() via
this_cpu_dec()", 2012-11-29), __srcu_read_lock() did two increments.
Therefore it kept __this_cpu_inc(), with preempt_disable/enable in
the caller.  Tree SRCU however only does one increment, so on most
architectures it is more efficient for __srcu_read_lock() to use
this_cpu_inc(), and any performance differences appear to be down in
the noise.

Cc: stable@vger.kernel.org
Fixes: 719d93cd ("kvm/irqchip: Speed up KVM_SET_GSI_ROUTING")
Reported-by: Linu Cherian <linuc.decode@gmail.com>
Suggested-by: Linu Cherian <linuc.decode@gmail.com>
Cc: kvm@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

Linu Cherian reported a WARN in cleanup_srcu_struct() when shutting
down a guest running iperf on a VFIO assigned device.  This happens
because irqfd_wakeup() calls srcu_read_lock(&kvm->irq_srcu) in interrupt
context, while a worker thread does the same inside kvm_set_irq().  If the
interrupt happens while the worker thread is executing __srcu_read_lock(),
updates to the Classic SRCU ->lock_count[] field or the Tree SRCU
->srcu_lock_count[] field can be lost.

The docs say you are not supposed to call srcu_read_lock() and
srcu_read_unlock() from irq context, but KVM interrupt injection happens
from (host) interrupt context and it would be nice if SRCU supported the
use case.  KVM is using SRCU here not really for the "sleepable" part,
but rather due to its IPI-free fast detection of grace periods.  It is
therefore not desirable to switch back to RCU, which would effectively
revert commit 719d93cd ("kvm/irqchip: Speed up KVM_SET_GSI_ROUTING",
2014-01-16).

However, the docs are overly conservative.  You can have an SRCU instance
only has users in irq context, and you can mix process and irq context
as long as process context users disable interrupts.  In addition,
__srcu_read_unlock() actually uses this_cpu_dec() on both Tree SRCU and
Classic SRCU.  For those two implementations, only srcu_read_lock()
is unsafe.

When Classic SRCU's __srcu_read_unlock() was changed to use this_cpu_dec(),
in commit 5a41344a ("srcu: Simplify __srcu_read_unlock() via
this_cpu_dec()", 2012-11-29), __srcu_read_lock() did two increments.
Therefore it kept __this_cpu_inc(), with preempt_disable/enable in
the caller.  Tree SRCU however only does one increment, so on most
architectures it is more efficient for __srcu_read_lock() to use
this_cpu_inc(), and any performance differences appear to be down in
the noise.

Unlike Classic and Tree SRCU, Tiny SRCU does increments and decrements on
a single variable.  Therefore, as Peter Zijlstra pointed out, Tiny SRCU's
implementation already supports mixed-context use of srcu_read_lock()
and srcu_read_unlock(), at least as long as uses of srcu_read_lock()
and srcu_read_unlock() in each handler are nested and paired properly.
In other words, it is still illegal to (say) invoke srcu_read_lock()
in an interrupt handler and to invoke the matching srcu_read_unlock()
in a softirq handler.  Therefore, the only change required for Tiny SRCU
is to its comments.

Fixes: 719d93cd ("kvm/irqchip: Speed up KVM_SET_GSI_ROUTING")
Reported-by: Linu Cherian <linuc.decode@gmail.com>
Suggested-by: Linu Cherian <linuc.decode@gmail.com>
Cc: kvm@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Paolo Bonzini <pbonzini@redhat.com>

This fixes a problem with reading files larger than 2GB from a UFS-2
file system:

    https://bugzilla.kernel.org/show_bug.cgi?id=195721

The incorrect UFS s_maxsize limit became a problem as of commit
c2a9737f ("vfs,mm: fix a dead loop in truncate_inode_pages_range()")
which started using s_maxbytes to avoid a page index overflow in
do_generic_file_read().

That caused files to be truncated on UFS-2 file systems because the
default maximum file size is 2GB (MAX_NON_LFS) and UFS didn't update it.

Here I simply increase the default to a common value used by other file
systems.
Signed-off-by: Richard Narron <comet.berkeley@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Will B <will.brokenbourgh2877@gmail.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: <stable@vger.kernel.org> # v4.9 and backports of c2a9737fSigned-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull NFS client bugfixes from Trond Myklebust:
 "Bugfixes include:

   - Fix a typo in commit e0926934 ("NFS append COMMIT after
     synchronous COPY") that breaks copy offload

   - Fix the connect error propagation in xs_tcp_setup_socket()

   - Fix a lock leak in nfs40_walk_client_list

   - Verify that pNFS requests lie within the offset range of the layout
     segment"

* tag 'nfs-for-4.12-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  nfs: Mark unnecessarily extern functions as static
  SUNRPC: ensure correct error is reported by xs_tcp_setup_socket()
  NFSv4.0: Fix a lock leak in nfs40_walk_client_list
  pnfs: Fix the check for requests in range of layout segment
  xprtrdma: Delete an error message for a failed memory allocation in xprt_rdma_bc_setup()
  pNFS/flexfiles: missing error code in ff_layout_alloc_lseg()
  NFS fix COMMIT after COPY

Pull tty fix from Greg KH:
 "Here is a single tty core fix for 4.12-rc4. It reverts a patch that a
  lot of people reported as causing lockdep and other warnings.

  Right after I reverted this in my tree, it seems like another
  "correct" fix might have shown up, but it's too late in the release
  cycle to be messing with tty core locking, so let's just revert this
  for now to go back how things always have been and try it again for
  4.13.

  This has not been in linux-next as I only reverted it a few hours ago"

* tag 'tty-4.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  Revert "tty: fix port buffer locking"

Pull input subsystem fixes from Dmitry Torokhov:

 - a couple of regression fixes in synaptics and axp20x-pek drivers

 - try to ease transition from PS/2 to RMI for Synaptics touchpad users
   by ensuring we do not try to activate RMI mode when RMI SMBus support
   is not enabled, and nag users a bit to enable it

 - plus a couple of other changes that seemed worthwhile for this
   release

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: axp20x-pek - switch to acpi_dev_present and check for ACPI0011 too
  Input: axp20x-pek - only check for "INTCFD9" ACPI device on Cherry Trail
  Input: tm2-touchkey - use LEN_ON as boolean value instead of LED_FULL
  Input: synaptics - tell users to report when they should be using rmi-smbus
  Input: synaptics - warn the users when there is a better mode
  Input: synaptics - keep PS/2 around when RMI4_SMB is not enabled
  Input: synaptics - clear device info before filling in
  Input: silead - disable interrupt during suspend

Pull RTC fixlet from Alexandre Belloni:
 "A single patch, not really a fix but I don't think there is any reason
  to delay it.

  Change the mailing list address"

* tag 'rtc-4.12-2' of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:
  MAINTAINERS: update RTC mailing list

Pull SCSI fixes from James Bottomley:
 "This is nine fixes, seven of which are for the qedi driver (new as of
  4.10) the other two are a use after free in the cxgbi drivers and a
  potential NULL dereference in the rdac device handler"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: libcxgbi: fix skb use after free
  scsi: qedi: Fix endpoint NULL panic during recovery.
  scsi: qedi: set max_fin_rt default value
  scsi: qedi: Set firmware tcp msl timer value.
  scsi: qedi: Fix endpoint NULL panic in qedi_set_path.
  scsi: qedi: Set dma_boundary to 0xfff.
  scsi: qedi: Correctly set firmware max supported BDs.
  scsi: qedi: Fix bad pte call trace when iscsiuio is stopped.
  scsi: scsi_dh_rdac: Use ctlr directly in rdac_failover_get()

Pull rdma fixes from Doug Ledford:
 "For the most part this is just a minor -rc cycle for the rdma
  subsystem. Even given that this is all of the -rc patches since the
  merge window closed, it's still only about 25 patches:

   - Multiple i40iw, nes, iw_cxgb4, hfi1, qib, mlx4, mlx5 fixes

   - A few upper layer protocol fixes (IPoIB, iSER, SRP)

   - A modest number of core fixes"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (26 commits)
  RDMA/SA: Fix kernel panic in CMA request handler flow
  RDMA/umem: Fix missing mmap_sem in get umem ODP call
  RDMA/core: not to set page dirty bit if it's already set.
  RDMA/uverbs: Declare local function static and add brackets to sizeof
  RDMA/netlink: Reduce exposure of RDMA netlink functions
  RDMA/srp: Fix NULL deref at srp_destroy_qp()
  RDMA/IPoIB: Limit the ipoib_dev_uninit_default scope
  RDMA/IPoIB: Replace netdev_priv with ipoib_priv for ipoib_get_link_ksettings
  RDMA/qedr: add null check before pointer dereference
  RDMA/mlx5: set UMR wqe fence according to HCA cap
  net/mlx5: Define interface bits for fencing UMR wqe
  RDMA/mlx4: Fix MAD tunneling when SRIOV is enabled
  RDMA/qib,hfi1: Fix MR reference count leak on write with immediate
  RDMA/hfi1: Defer setting VL15 credits to link-up interrupt
  RDMA/hfi1: change PCI bar addr assignments to Linux API functions
  RDMA/hfi1: fix array termination by appending NULL to attr array
  RDMA/iw_cxgb4: fix the calculation of ipv6 header size
  RDMA/iw_cxgb4: calculate t4_eq_status_entries properly
  RDMA/iw_cxgb4: Avoid touch after free error in ARP failure handlers
  RDMA/nes: ACK MPA Reply frame
  ...

This reverts commit 925bb1ce.

It causes lots of warnings and problems so for now, let's just revert
it.

Reported-by: <valdis.kletnieks@vt.edu>
Reported-by: Russell King <linux@armlinux.org.uk>
Reported-by: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reported-by: Jiri Slaby <jslaby@suse.cz>
Reported-by: Andrey Konovalov <andreyknvl@google.com>
Acked-by: Vegard Nossum <vegard.nossum@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

nfs_initialise_sb() and nfs_clone_super() are declared as extern even
though they are used only in fs/nfs/super.c. Mark them as static.

Also remove explicit 'inline' directive from nfs_initialise_sb() and
leave it upto compiler to decide whether inlining is worth it.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>

Merge tag 'hwmon-for-linus-v4.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging

Pull hwmon fixes from Guenter Roeck:
 "A couple of patches for the aspeed pwm fan driver"

* tag 'hwmon-for-linus-v4.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
  hwmon: (aspeed-pwm-tacho) make fan/pwm names start with index 1
  hwmon: (aspeed-pwm-tacho) Call of_node_put() on a node not claimed
  hwmon: (aspeed-pwm-tacho) On read failure return -ETIMEDOUT
  hwmon: (aspeed-pwm-tacho) Select REGMAP

Pull MTD fixes from Brian Norris:
 "NAND updates from Boris:

  tango fixes:
   - Add missing MODULE_DEVICE_TABLE() in tango_nand.c
   - Update the number of corrected bitflips

  core fixes:
   - Fix a long standing memory leak in nand_scan_tail()
   - Fix several bugs introduced by the per-vendor init/detection
     infrastructure (introduced in 4.12)
   - Add a static specifier to nand_ooblayout_lp_hamming_ops definition"

* tag 'for-linus-20170602' of git://git.infradead.org/linux-mtd:
  mtd: nand: make nand_ooblayout_lp_hamming_ops static
  mtd: nand: tango: Update ecc_stats.corrected
  mtd: nand: tango: Export OF device ID table as module aliases
  mtd: nand: samsung: warn about un-parseable ECC info
  mtd: nand: free vendor-specific resources in init failure paths
  mtd: nand: drop unneeded module.h include
  mtd: nand: don't leak buffers when ->scan_bbt() fails

Make fan and pwm names in sysfs start with index 1 in accordance to
Documentation/hwmon/sysfs-interface conventions.

Current implementation starts with index 0, making tools such as
sensors(1) skip the first fan.
Signed-off-by: Stefan Schaeckeler <sschaeck@cisco.com>
Fixes: 2d7a548a ("drivers: hwmon: Support for ASPEED PWM/Fan tach")
Signed-off-by: Guenter Roeck <linux@roeck-us.net>

Call of_node_put() on a node claimed with of_node_get() or by any other
means such as for_each_child_of_node().
Signed-off-by: Stefan Schaeckeler <sschaeck@cisco.com>
Fixes: 2d7a548a ("drivers: hwmon: Support for ASPEED PWM/Fan tach")
Signed-off-by: Guenter Roeck <linux@roeck-us.net>

acpi_dev_found checks that there is a matching ACPI node, but it
may be disabled (_STA method returns 0) in which case the
soc_button_array driver will not bind to it and axp20x-pek should
handle the power-button.

This commit switches from acpi_dev_found to acpi_dev_present to
avoid not registering an input-dev for the powerbutton when there
is a disabled PNP0C40 device.

The ACPI-6.0 standard defines a standard gpio button device using
the ACPI0011 HID replacing the custom PNP0C40 gpio device, many
newer devices define both PNP0C40 and ACPI0011 devices enabling one
or the other depending on whether the BIOS thinks it is going to boot
Android or Windows.

This commit adds a check for the ACPI0011 device, so that if
either device is present *and* enabled we don't register an input-dev
for the powerbutton.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Commit 9b13a4ca ("Input: axp20x-pek - do not register input device
on some systems") added a check for the INTCFD9 ACPI device which also
handles the powerbutton as on some systems the powerbutton is connected
to both the PMIC, handled by axp20x-pek, and to a gpio on the SoC, handled
by soc_button_array which attaches itself to the INTCFD9 ACPI device.

Testing + comparing DSDTs has shown that this only happens on Cherry
Trail devices with an AXP288 PMIC, the AXP288 PMIC is also used on
Bay Trail devices but there the power button is only connected to
the PMIC and not handled by soc_button_array.

This means that the INTCFD9 check has caused a regression on Bay Trail
devices, causing power-button presses to no longer be seen.

This commit fixes this by limiting the check to devices where the ACPI
node for the AXP288 contains a _HRV (hardware revision) attribute with
a value of 3 which indicates we are dealing with a Cherry Trail platform.

Fixes: 9b13a4ca ("Input: axp20x-pek - do not register input ...")
Reported-by: Сергей Трусов <t.rus76@ya.ru>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

Merge with mainline to get acpi_dev_present() needed by patches to
axp20x-pek driver.

Pull ACPI fixes from Rafael Wysocki:
 "These revert one more problematic commit related to the ACPI-based
  handling of laptop lids and make some unuseful error messages coming
  from ACPICA go away.

  Specifics:

   - Revert one more commit related to the ACPI-based handling of laptop
     lids that changed the default behavior on laptops that booted with
     closed lids and introduced a regression there (Benjamin Tissoires).

   - Add a missing acpi_put_table() to the code implementing the
     /sys/firmware/acpi/tables interface to prevent a counter in the
     ACPICA core from overflowing (Dan Williams).

   - Drop error messages printed by ACPICA on acpi_get_table() reference
     counting mismatches as they need not indicate real errors at this
     point (Lv Zheng)"

* tag 'acpi-4.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPICA: Tables: Fix regression introduced by a too early mechanism enabling
  Revert "ACPI / button: Change default behavior to lid_init_state=open"
  ACPI / sysfs: fix acpi_get_table() leak / acpi-sysfs denial of service

Pull power management fixes from Rafael Wysocki:
 "These fix two bugs in error code paths in the cpufreq core and in the
  kirkwood-cpufreq driver.

  Specifics:

   - Make cpufreq_register_driver() return an error if the ->init()
     calls fail for all CPUs to prevent non-functional drivers from
     hanging around for no reason (David Arcari).

   - Make kirkwood-cpufreq check the return value of
     clk_prepare_enable() (which may fail) as appropriate (Arvind
     Yadav)"

* tag 'pm-4.12-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  cpufreq: kirkwood-cpufreq:- Handle return value of clk_prepare_enable()
  cpufreq: cpufreq_register_driver() should return -ENODEV if init fails

Pull /dev/random bug fix from Ted Ts'o:
 "Fix a race on architectures with prioritized interrupts (such as m68k)
  which can causes crashes in drivers/char/random.c:get_reg()"

* tag 'random_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/random:
  fix race in drivers/char/random.c:get_reg()

Merge misc fixes from Andrew Morton:
 "15 fixes"

* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
  scripts/gdb: make lx-dmesg command work (reliably)
  mm: consider memblock reservations for deferred memory initialization sizing
  mm/hugetlb: report -EHWPOISON not -EFAULT when FOLL_HWPOISON is specified
  mlock: fix mlock count can not decrease in race condition
  mm/migrate: fix refcount handling when !hugepage_migration_supported()
  dax: fix race between colliding PMD & PTE entries
  mm: avoid spurious 'bad pmd' warning messages
  mm/page_alloc.c: make sure OOM victim can try allocations with no watermarks once
  pcmcia: remove left-over %Z format
  slub/memcg: cure the brainless abuse of sysfs attributes
  initramfs: fix disabling of initramfs (and its compression)
  mm: clarify why we want kmalloc before falling backto vmallock
  frv: declare jiffies to be located in the .data section
  include/linux/gfp.h: fix ___GFP_NOLOCKDEP value
  ksm: prevent crash after write_protect_page fails

lx-dmesg needs access to the log_buf symbol from printk.c.
Unfortunately, the symbol log_buf also exists in BPF's verifier.c and
hence gdb can pick one or the other.  If it happens to pick BPF's
log_buf, lx-dmesg doesn't work:

  (gdb) lx-dmesg
  Python Exception <class 'gdb.MemoryError'> Cannot access memory at address 0x0:
  Error occurred in Python command: Cannot access memory at address 0x0
  (gdb) p log_buf
  $15 = 0x0

Luckily, GDB has a way to deal with this, see
  https://sourceware.org/gdb/onlinedocs/gdb/Symbols.html

  (gdb) info variables ^log_buf$
  All variables matching regular expression "^log_buf$":

  File <linux.git>/kernel/bpf/verifier.c:
  static char *log_buf;

  File <linux.git>/kernel/printk/printk.c:
  static char *log_buf;
  (gdb) p 'verifier.c'::log_buf
  $1 = 0x0
  (gdb) p 'printk.c'::log_buf
  $2 = 0x811a6aa0 <__log_buf> ""
  (gdb) p &log_buf
  $3 = (char **) 0x8120fe40 <log_buf>
  (gdb) p &'verifier.c'::log_buf
  $4 = (char **) 0x8120fe40 <log_buf>
  (gdb) p &'printk.c'::log_buf
  $5 = (char **) 0x8048b7d0 <log_buf>

By being explicit about the location of the symbol, we can make lx-dmesg
work again.  While at it, do the same for the other symbols we need from
printk.c

Link: http://lkml.kernel.org/r/20170526112222.3414-1-git@andred.netSigned-off-by: André Draszik <git@andred.net>
Tested-by: Kieran Bingham <kieran@bingham.xyz>
Acked-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

We have seen an early OOM killer invocation on ppc64 systems with
crashkernel=4096M:

	kthreadd invoked oom-killer: gfp_mask=0x16040c0(GFP_KERNEL|__GFP_COMP|__GFP_NOTRACK), nodemask=7, order=0, oom_score_adj=0
	kthreadd cpuset=/ mems_allowed=7
	CPU: 0 PID: 2 Comm: kthreadd Not tainted 4.4.68-1.gd7fe927-default #1
	Call Trace:
	  dump_stack+0xb0/0xf0 (unreliable)
	  dump_header+0xb0/0x258
	  out_of_memory+0x5f0/0x640
	  __alloc_pages_nodemask+0xa8c/0xc80
	  kmem_getpages+0x84/0x1a0
	  fallback_alloc+0x2a4/0x320
	  kmem_cache_alloc_node+0xc0/0x2e0
	  copy_process.isra.25+0x260/0x1b30
	  _do_fork+0x94/0x470
	  kernel_thread+0x48/0x60
	  kthreadd+0x264/0x330
	  ret_from_kernel_thread+0x5c/0xa4

	Mem-Info:
	active_anon:0 inactive_anon:0 isolated_anon:0
	 active_file:0 inactive_file:0 isolated_file:0
	 unevictable:0 dirty:0 writeback:0 unstable:0
	 slab_reclaimable:5 slab_unreclaimable:73
	 mapped:0 shmem:0 pagetables:0 bounce:0
	 free:0 free_pcp:0 free_cma:0
	Node 7 DMA free:0kB min:0kB low:0kB high:0kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:52428800kB managed:110016kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:320kB slab_unreclaimable:4672kB kernel_stack:1152kB pagetables:0kB unstable:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? yes
	lowmem_reserve[]: 0 0 0 0
	Node 7 DMA: 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB 0*8192kB 0*16384kB = 0kB
	0 total pagecache pages
	0 pages in swap cache
	Swap cache stats: add 0, delete 0, find 0/0
	Free swap  = 0kB
	Total swap = 0kB
	819200 pages RAM
	0 pages HighMem/MovableOnly
	817481 pages reserved
	0 pages cma reserved
	0 pages hwpoisoned

the reason is that the managed memory is too low (only 110MB) while the
rest of the the 50GB is still waiting for the deferred intialization to
be done.  update_defer_init estimates the initial memoty to initialize
to 2GB at least but it doesn't consider any memory allocated in that
range.  In this particular case we've had

	Reserving 4096MB of memory at 128MB for crashkernel (System RAM: 51200MB)

so the low 2GB is mostly depleted.

Fix this by considering memblock allocations in the initial static
initialization estimation.  Move the max_initialise to
reset_deferred_meminit and implement a simple memblock_reserved_memory
helper which iterates all reserved blocks and sums the size of all that
start below the given address.  The cumulative size is than added on top
of the initial estimation.  This is still not ideal because
reset_deferred_meminit doesn't consider holes and so reservation might
be above the initial estimation whihch we ignore but let's make the
logic simpler until we really need to handle more complicated cases.

Fixes: 3a80a7fa ("mm: meminit: initialise a subset of struct pages if CONFIG_DEFERRED_STRUCT_PAGE_INIT is set")
Link: http://lkml.kernel.org/r/20170531104010.GI27783@dhcp22.suse.czSigned-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Tested-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: <stable@vger.kernel.org>	[4.2+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

KVM uses get_user_pages() to resolve its stage2 faults.  KVM sets the
FOLL_HWPOISON flag causing faultin_page() to return -EHWPOISON when it
finds a VM_FAULT_HWPOISON.  KVM handles these hwpoison pages as a
special case.  (check_user_page_hwpoison())

When huge pages are involved, this doesn't work so well.
get_user_pages() calls follow_hugetlb_page(), which stops early if it
receives VM_FAULT_HWPOISON from hugetlb_fault(), eventually returning
-EFAULT to the caller.  The step to map this to -EHWPOISON based on the
FOLL_ flags is missing.  The hwpoison special case is skipped, and
-EFAULT is returned to user-space, causing Qemu or kvmtool to exit.

Instead, move this VM_FAULT_ to errno mapping code into a header file
and use it from faultin_page() and follow_hugetlb_page().

With this, KVM works as expected.

This isn't a problem for arm64 today as we haven't enabled
MEMORY_FAILURE, but I can't see any reason this doesn't happen on x86
too, so I think this should be a fix.  This doesn't apply earlier than
stable's v4.11.1 due to all sorts of cleanup.

[james.morse@arm.com: add vm_fault_to_errno() call to faultin_page()]
suggested.
  Link: http://lkml.kernel.org/r/20170525171035.16359-1-james.morse@arm.com
[akpm@linux-foundation.org: coding-style fixes]
Link: http://lkml.kernel.org/r/20170524160900.28786-1-james.morse@arm.comSigned-off-by: James Morse <james.morse@arm.com>
Acked-by: Punit Agrawal <punit.agrawal@arm.com>
Acked-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Cc: <stable@vger.kernel.org>	[4.11.1+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Kefeng reported that when running the follow test, the mlock count in
meminfo will increase permanently:

 [1] testcase
 linux:~ # cat test_mlockal
 grep Mlocked /proc/meminfo
  for j in `seq 0 10`
  do
 	for i in `seq 4 15`
 	do
 		./p_mlockall >> log &
 	done
 	sleep 0.2
 done
 # wait some time to let mlock counter decrease and 5s may not enough
 sleep 5
 grep Mlocked /proc/meminfo

 linux:~ # cat p_mlockall.c
 #include <sys/mman.h>
 #include <stdlib.h>
 #include <stdio.h>

 #define SPACE_LEN	4096

 int main(int argc, char ** argv)
 {
	 	int ret;
	 	void *adr = malloc(SPACE_LEN);
	 	if (!adr)
	 		return -1;

	 	ret = mlockall(MCL_CURRENT | MCL_FUTURE);
	 	printf("mlcokall ret = %d\n", ret);

	 	ret = munlockall();
	 	printf("munlcokall ret = %d\n", ret);

	 	free(adr);
	 	return 0;
	 }

In __munlock_pagevec() we should decrement NR_MLOCK for each page where
we clear the PageMlocked flag.  Commit 1ebb7cc6 ("mm: munlock: batch
NR_MLOCK zone state updates") has introduced a bug where we don't
decrement NR_MLOCK for pages where we clear the flag, but fail to
isolate them from the lru list (e.g.  when the pages are on some other
cpu's percpu pagevec).  Since PageMlocked stays cleared, the NR_MLOCK
accounting gets permanently disrupted by this.

Fix it by counting the number of page whose PageMlock flag is cleared.

Fixes: 1ebb7cc6 (" mm: munlock: batch NR_MLOCK zone state updates")
Link: http://lkml.kernel.org/r/1495678405-54569-1-git-send-email-xieyisheng1@huawei.comSigned-off-by: Yisheng Xie <xieyisheng1@huawei.com>
Reported-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Tested-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Joern Engel <joern@logfs.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michel Lespinasse <walken@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Xishi Qiu <qiuxishi@huawei.com>
Cc: zhongjiang <zhongjiang@huawei.com>
Cc: Hanjun Guo <guohanjun@huawei.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

On failing to migrate a page, soft_offline_huge_page() performs the
necessary update to the hugepage ref-count.

But when !hugepage_migration_supported() , unmap_and_move_hugepage()
also decrements the page ref-count for the hugepage.  The combined
behaviour leaves the ref-count in an inconsistent state.

This leads to soft lockups when running the overcommitted hugepage test
from mce-tests suite.

  Soft offlining pfn 0x83ed600 at process virtual address 0x400000000000
  soft offline: 0x83ed600: migration failed 1, type 1fffc00000008008 (uptodate|head)
  INFO: rcu_preempt detected stalls on CPUs/tasks:
   Tasks blocked on level-0 rcu_node (CPUs 0-7): P2715
    (detected by 7, t=5254 jiffies, g=963, c=962, q=321)
    thugetlb_overco R  running task        0  2715   2685 0x00000008
    Call trace:
      dump_backtrace+0x0/0x268
      show_stack+0x24/0x30
      sched_show_task+0x134/0x180
      rcu_print_detail_task_stall_rnp+0x54/0x7c
      rcu_check_callbacks+0xa74/0xb08
      update_process_times+0x34/0x60
      tick_sched_handle.isra.7+0x38/0x70
      tick_sched_timer+0x4c/0x98
      __hrtimer_run_queues+0xc0/0x300
      hrtimer_interrupt+0xac/0x228
      arch_timer_handler_phys+0x3c/0x50
      handle_percpu_devid_irq+0x8c/0x290
      generic_handle_irq+0x34/0x50
      __handle_domain_irq+0x68/0xc0
      gic_handle_irq+0x5c/0xb0

Address this by changing the putback_active_hugepage() in
soft_offline_huge_page() to putback_movable_pages().

This only triggers on systems that enable memory failure handling
(ARCH_SUPPORTS_MEMORY_FAILURE) but not hugepage migration
(!ARCH_ENABLE_HUGEPAGE_MIGRATION).

I imagine this wasn't triggered as there aren't many systems running
this configuration.

[akpm@linux-foundation.org: remove dead comment, per Naoya]
Link: http://lkml.kernel.org/r/20170525135146.32011-1-punit.agrawal@arm.comReported-by: Manoj Iyer <manoj.iyer@canonical.com>
Tested-by: Manoj Iyer <manoj.iyer@canonical.com>
Suggested-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Signed-off-by: Punit Agrawal <punit.agrawal@arm.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: <stable@vger.kernel.org>	[3.14+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

We currently have two related PMD vs PTE races in the DAX code.  These
can both be easily triggered by having two threads reading and writing
simultaneously to the same private mapping, with the key being that
private mapping reads can be handled with PMDs but private mapping
writes are always handled with PTEs so that we can COW.

Here is the first race:

  CPU 0					CPU 1

  (private mapping write)
  __handle_mm_fault()
    create_huge_pmd() - FALLBACK
    handle_pte_fault()
      passes check for pmd_devmap()

					(private mapping read)
					__handle_mm_fault()
					  create_huge_pmd()
					    dax_iomap_pmd_fault() inserts PMD

      dax_iomap_pte_fault() does a PTE fault, but we already have a DAX PMD
      			  installed in our page tables at this spot.

Here's the second race:

  CPU 0					CPU 1

  (private mapping read)
  __handle_mm_fault()
    passes check for pmd_none()
    create_huge_pmd()
      dax_iomap_pmd_fault() inserts PMD

  (private mapping write)
  __handle_mm_fault()
    create_huge_pmd() - FALLBACK
					(private mapping read)
					__handle_mm_fault()
					  passes check for pmd_none()
					  create_huge_pmd()

    handle_pte_fault()
      dax_iomap_pte_fault() inserts PTE
					    dax_iomap_pmd_fault() inserts PMD,
					       but we already have a PTE at
					       this spot.

The core of the issue is that while there is isolation between faults to
the same range in the DAX fault handlers via our DAX entry locking,
there is no isolation between faults in the code in mm/memory.c.  This
means for instance that this code in __handle_mm_fault() can run:

	if (pmd_none(*vmf.pmd) && transparent_hugepage_enabled(vma)) {
		ret = create_huge_pmd(&vmf);

But by the time we actually get to run the fault handler called by
create_huge_pmd(), the PMD is no longer pmd_none() because a racing PTE
fault has installed a normal PMD here as a parent.  This is the cause of
the 2nd race.  The first race is similar - there is the following check
in handle_pte_fault():

	} else {
		/* See comment in pte_alloc_one_map() */
		if (pmd_devmap(*vmf->pmd) || pmd_trans_unstable(vmf->pmd))
			return 0;

So if a pmd_devmap() PMD (a DAX PMD) has been installed at vmf->pmd, we
will bail and retry the fault.  This is correct, but there is nothing
preventing the PMD from being installed after this check but before we
actually get to the DAX PTE fault handlers.

In my testing these races result in the following types of errors:

  BUG: Bad rss-counter state mm:ffff8800a817d280 idx:1 val:1
  BUG: non-zero nr_ptes on freeing mm: 15

Fix this issue by having the DAX fault handlers verify that it is safe
to continue their fault after they have taken an entry lock to block
other racing faults.

[ross.zwisler@linux.intel.com: improve fix for colliding PMD & PTE entries]
  Link: http://lkml.kernel.org/r/20170526195932.32178-1-ross.zwisler@linux.intel.com
Link: http://lkml.kernel.org/r/20170522215749.23516-2-ross.zwisler@linux.intel.comSigned-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Reported-by: Pawel Lebioda <pawel.lebioda@intel.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Matthew Wilcox <mawilcox@microsoft.com>
Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Pawel Lebioda <pawel.lebioda@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Xiong Zhou <xzhou@redhat.com>
Cc: Eryu Guan <eguan@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

When the pmd_devmap() checks were added by 5c7fb56e ("mm, dax:
dax-pmd vs thp-pmd vs hugetlbfs-pmd") to add better support for DAX huge
pages, they were all added to the end of if() statements after existing
pmd_trans_huge() checks.  So, things like:

  -       if (pmd_trans_huge(*pmd))
  +       if (pmd_trans_huge(*pmd) || pmd_devmap(*pmd))

When further checks were added after pmd_trans_unstable() checks by
commit 7267ec00 ("mm: postpone page table allocation until we have
page to map") they were also added at the end of the conditional:

  +       if (pmd_trans_unstable(fe->pmd) || pmd_devmap(*fe->pmd))

This ordering is fine for pmd_trans_huge(), but doesn't work for
pmd_trans_unstable().  This is because DAX huge pages trip the bad_pmd()
check inside of pmd_none_or_trans_huge_or_clear_bad() (called by
pmd_trans_unstable()), which prints out a warning and returns 1.  So, we
do end up doing the right thing, but only after spamming dmesg with
suspicious looking messages:

  mm/pgtable-generic.c:39: bad pmd ffff8808daa49b88(84000001006000a5)

Reorder these checks in a helper so that pmd_devmap() is checked first,
avoiding the error messages, and add a comment explaining why the
ordering is important.

Fixes: commit 7267ec00 ("mm: postpone page table allocation until we have page to map")
Link: http://lkml.kernel.org/r/20170522215749.23516-1-ross.zwisler@linux.intel.comSigned-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Pawel Lebioda <pawel.lebioda@intel.com>
Cc: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Matthew Wilcox <mawilcox@microsoft.com>
Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Xiong Zhou <xzhou@redhat.com>
Cc: Eryu Guan <eguan@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Roman Gushchin has reported that the OOM killer can trivially selects
next OOM victim when a thread doing memory allocation from page fault
path was selected as first OOM victim.

    allocate invoked oom-killer: gfp_mask=0x14280ca(GFP_HIGHUSER_MOVABLE|__GFP_ZERO), nodemask=(null),  order=0, oom_score_adj=0
    allocate cpuset=/ mems_allowed=0
    CPU: 1 PID: 492 Comm: allocate Not tainted 4.12.0-rc1-mm1+ #181
    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014
    Call Trace:
     oom_kill_process+0x219/0x3e0
     out_of_memory+0x11d/0x480
     __alloc_pages_slowpath+0xc84/0xd40
     __alloc_pages_nodemask+0x245/0x260
     alloc_pages_vma+0xa2/0x270
     __handle_mm_fault+0xca9/0x10c0
     handle_mm_fault+0xf3/0x210
     __do_page_fault+0x240/0x4e0
     trace_do_page_fault+0x37/0xe0
     do_async_page_fault+0x19/0x70
     async_page_fault+0x28/0x30
    ...
    Out of memory: Kill process 492 (allocate) score 899 or sacrifice child
    Killed process 492 (allocate) total-vm:2052368kB, anon-rss:1894576kB, file-rss:4kB, shmem-rss:0kB
    allocate: page allocation failure: order:0, mode:0x14280ca(GFP_HIGHUSER_MOVABLE|__GFP_ZERO), nodemask=(null)
    allocate cpuset=/ mems_allowed=0
    CPU: 1 PID: 492 Comm: allocate Not tainted 4.12.0-rc1-mm1+ #181
    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014
    Call Trace:
     __alloc_pages_slowpath+0xd32/0xd40
     __alloc_pages_nodemask+0x245/0x260
     alloc_pages_vma+0xa2/0x270
     __handle_mm_fault+0xca9/0x10c0
     handle_mm_fault+0xf3/0x210
     __do_page_fault+0x240/0x4e0
     trace_do_page_fault+0x37/0xe0
     do_async_page_fault+0x19/0x70
     async_page_fault+0x28/0x30
    ...
    oom_reaper: reaped process 492 (allocate), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
    ...
    allocate invoked oom-killer: gfp_mask=0x0(), nodemask=(null),  order=0, oom_score_adj=0
    allocate cpuset=/ mems_allowed=0
    CPU: 1 PID: 492 Comm: allocate Not tainted 4.12.0-rc1-mm1+ #181
    Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014
    Call Trace:
     oom_kill_process+0x219/0x3e0
     out_of_memory+0x11d/0x480
     pagefault_out_of_memory+0x68/0x80
     mm_fault_error+0x8f/0x190
     ? handle_mm_fault+0xf3/0x210
     __do_page_fault+0x4b2/0x4e0
     trace_do_page_fault+0x37/0xe0
     do_async_page_fault+0x19/0x70
     async_page_fault+0x28/0x30
    ...
    Out of memory: Kill process 233 (firewalld) score 10 or sacrifice child
    Killed process 233 (firewalld) total-vm:246076kB, anon-rss:20956kB, file-rss:0kB, shmem-rss:0kB

There is a race window that the OOM reaper completes reclaiming the
first victim's memory while nothing but mutex_trylock() prevents the
first victim from calling out_of_memory() from pagefault_out_of_memory()
after memory allocation for page fault path failed due to being selected
as an OOM victim.

This is a side effect of commit 9a67f648 ("mm: consolidate
GFP_NOFAIL checks in the allocator slowpath") because that commit
silently changed the behavior from

    /* Avoid allocations with no watermarks from looping endlessly */

to

    /*
     * Give up allocations without trying memory reserves if selected
     * as an OOM victim
     */

in __alloc_pages_slowpath() by moving the location to check TIF_MEMDIE
flag.  I have noticed this change but I didn't post a patch because I
thought it is an acceptable change other than noise by warn_alloc()
because !__GFP_NOFAIL allocations are allowed to fail.  But we
overlooked that failing memory allocation from page fault path makes
difference due to the race window explained above.

While it might be possible to add a check to pagefault_out_of_memory()
that prevents the first victim from calling out_of_memory() or remove
out_of_memory() from pagefault_out_of_memory(), changing
pagefault_out_of_memory() does not suppress noise by warn_alloc() when
allocating thread was selected as an OOM victim.  There is little point
with printing similar backtraces and memory information from both
out_of_memory() and warn_alloc().

Instead, if we guarantee that current thread can try allocations with no
watermarks once when current thread looping inside
__alloc_pages_slowpath() was selected as an OOM victim, we can follow "who
can use memory reserves" rules and suppress noise by warn_alloc() and
prevent memory allocations from page fault path from calling
pagefault_out_of_memory().

If we take the comment literally, this patch would do

  -    if (test_thread_flag(TIF_MEMDIE))
  -        goto nopage;
  +    if (alloc_flags == ALLOC_NO_WATERMARKS || (gfp_mask & __GFP_NOMEMALLOC))
  +        goto nopage;

because gfp_pfmemalloc_allowed() returns false if __GFP_NOMEMALLOC is
given.  But if I recall correctly (I couldn't find the message), the
condition is meant to apply to only OOM victims despite the comment.
Therefore, this patch preserves TIF_MEMDIE check.

Fixes: 9a67f648 ("mm: consolidate GFP_NOFAIL checks in the allocator slowpath")
Link: http://lkml.kernel.org/r/201705192112.IAF69238.OQOHSJLFOFFMtV@I-love.SAKURA.ne.jpSigned-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reported-by: Roman Gushchin <guro@fb.com>
Tested-by: Roman Gushchin <guro@fb.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: <stable@vger.kernel.org>	[4.11]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>