We always know the rq used, let's just pass it around.
This seems to cut the size of scheduler core down a tiny bit:

Before:

  [linux]$ size kernel/sched/core.o.orig
     text    data     bss     dec     hex filename
    62760   16130    3876   82766   1434e kernel/sched/core.o.orig

After:

  [linux]$ size kernel/sched/core.o.patched
     text    data     bss     dec     hex filename
    62566   16130    3876   82572   1428c kernel/sched/core.o.patched

Probably speeds it up as well.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20130922142054.GA11499@redhat.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

This patch builds on patch 2 and periodically decays that max value to
do idle balancing per sched domain by approximately 1% per second. Also
decay the rq's max_idle_balance_cost value.
Signed-off-by: Jason Low <jason.low2@hp.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1379096813-3032-4-git-send-email-jason.low2@hp.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

In this patch, we keep track of the max cost we spend doing idle load balancing
for each sched domain. If the avg time the CPU remains idle is less then the
time we have already spent on idle balancing + the max cost of idle balancing
in the sched domain, then we don't continue to attempt the balance. We also
keep a per rq variable, max_idle_balance_cost, which keeps track of the max
time spent on newidle load balances throughout all its domains so that we can
determine the avg_idle's max value.

By using the max, we avoid overrunning the average. This further reduces the
chance we attempt balancing when the CPU is not idle for longer than the cost
to balance.
Signed-off-by: Jason Low <jason.low2@hp.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1379096813-3032-3-git-send-email-jason.low2@hp.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

When updating avg_idle, if the delta exceeds some max value, then avg_idle
gets set to the max, regardless of what the previous avg was. This can cause
avg_idle to often be overestimated.

This patch modifies the way we update avg_idle by always updating it with the
function call to update_avg() first. Then, if avg_idle exceeds the max, we set
it to the max.
Signed-off-by: Jason Low <jason.low2@hp.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1379096813-3032-2-git-send-email-jason.low2@hp.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

Currently new_dst_cpu is prevented from being reselected actually, not
dst_cpu. This can result in attempting to pull tasks to this_cpu twice.
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/281f59b6e596c718dd565ad267fc38f5b8e5c995.1379265590.git.vdavydov@parallels.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

Merge in the latest fixes before applying a dependent patch.
Signed-off-by: Ingo Molnar <mingo@kernel.org>

Patch a003a2 (sched: Consider runnable load average in move_tasks())
sets all top-level cfs_rqs' h_load to rq->avg.load_avg_contrib, which is
always 0. This mistype leads to all tasks having weight 0 when load
balancing in a cpu-cgroup enabled setup. There obviously should be sum
of weights of all runnable tasks there instead. Fix it.
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Reviewed-by: Paul Turner <pjt@google.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1379173186-11944-1-git-send-email-vdavydov@parallels.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

In busiest->group_imb case we can come to fix_small_imbalance() with
local->avg_load > busiest->avg_load. This can result in wrong imbalance
fix-up, because there is the following check there where all the
members are unsigned:

if (busiest->avg_load - local->avg_load + scaled_busy_load_per_task >=
    (scaled_busy_load_per_task * imbn)) {
	env->imbalance = busiest->load_per_task;
	return;
}

As a result we can end up constantly bouncing tasks from one cpu to
another if there are pinned tasks.

Fix it by substituting the subtraction with an equivalent addition in
the check.

[ The bug can be caught by running 2*N cpuhogs pinned to two logical cpus
  belonging to different cores on an HT-enabled machine with N logical
  cpus: just look at se.nr_migrations growth. ]
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/ef167822e5c5b2d96cf5b0e3e4f4bdff3f0414a2.1379252740.git.vdavydov@parallels.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

In busiest->group_imb case we can come to calculate_imbalance() with
local->avg_load >= busiest->avg_load >= sds->avg_load. This can result
in imbalance overflow, because it is calculated as follows

env->imbalance = min(
	max_pull * busiest->group_power,
	(sds->avg_load - local->avg_load) * local->group_power) / SCHED_POWER_SCALE;

As a result we can end up constantly bouncing tasks from one cpu to
another if there are pinned tasks.

Fix this by skipping the assignment and assuming imbalance=0 in case
local->avg_load > sds->avg_load.

[ The bug can be caught by running 2*N cpuhogs pinned to two logical cpus
  belonging to different cores on an HT-enabled machine with N logical
  cpus: just look at se.nr_migrations growth. ]
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/8f596cc6bc0e5e655119dc892c9bfcad26e971f4.1379252740.git.vdavydov@parallels.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

Pull scheduler fixes from Ingo Molnar:
 "Misc fixes"

* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  sched: Fix comment for sched_info_depart
  sched/Documentation: Update sched-design-CFS.txt documentation
  sched/debug: Take PID namespace into account
  sched/fair: Fix small race where child->se.parent,cfs_rq might point to invalid ones

Pull perf fixes from Ingo Molnar:
 "Two small fixes"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf: Fix UAPI export of PERF_EVENT_IOC_ID
  perf/x86/intel: Fix Silvermont offcore masks

Without the following patch I have problems compiling code using
the new PERF_EVENT_IOC_ID ioctl().  It looks like u64 was used
instead of __u64
Signed-off-by: Vince Weaver <vincent.weaver@maine.edu>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Link: http://lkml.kernel.org/r/alpine.DEB.2.10.1309171450380.11444@vincent-weaver-1.um.maine.eduSigned-off-by: Ingo Molnar <mingo@kernel.org>

Pull KVM fixes from Gleb Natapov.

* 'fixes' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: VMX: set "blocked by NMI" flag if EPT violation happens during IRET from NMI
  kvm: free resources after canceling async_pf
  KVM: nEPT: reset PDPTR register cache on nested vmentry emulation
  KVM: mmu: allow page tables to be in read-only slots
  KVM: x86 emulator: emulate RETF imm

Pull HID updates from Jiri Kosina:
 "Fixes for CVE-2013-2897, CVE-2013-2895, CVE-2013-2897, CVE-2013-2894,
  CVE-2013-2893, CVE-2013-2891, CVE-2013-2890, CVE-2013-2889.

  All the bugs are triggerable only by specially crafted evil-on-purpose
  HW devices.  Fixes by Kees Cook and Benjamin Tissoires"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid:
  HID: lenovo-tpkbd: fix leak if tpkbd_probe_tp fails
  HID: multitouch: validate indexes details
  HID: logitech-dj: validate output report details
  HID: validate feature and input report details
  HID: lenovo-tpkbd: validate output report details
  HID: LG: validate HID output report details
  HID: steelseries: validate output report details
  HID: sony: validate HID output report details
  HID: zeroplus: validate output report details
  HID: provide a helper for validating hid reports

Starting from v3.10 (probably commit f91e2590: "tty: Signal
foreground group processes in hangup") disassociate_ctty() sends SIGCONT
if tty && on_exit.  This breaks LSB test-suite, in particular test8 in
_exit.c and test40 in sigcon5.c.

Put the "!on_exit" check back to restore the old behaviour.

Review by Peter Hurley:
 "Yes, this regression was introduced by me in that commit.  The effect
  of the regression is that ptys will receive a SIGCONT when, in similar
  circumstances, ttys would not.

  The fact that two test vectors accidentally tripped over this
  regression suggests that some other apps may as well.

  Thanks for catching this"

Cc: stable@vger.kernel.org # v3.10+
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Reported-by: Karel Srot <ksrot@redhat.com>
Reviewed-by: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Set "blocked by NMI" flag if EPT violation happens during IRET from NMI
otherwise NMI can be called recursively causing stack corruption.
Signed-off-by: Gleb Natapov <gleb@redhat.com>

Pull more tile architecture updates from Chris Metcalf:
 "This second batch of changes is just cleanup of various kinds from
  doing some tidying work in the sources.

  Some dead code is removed, comment typos fixed, whitespace and style
  issues cleaned up, and some header updates from our internal
  "upstream" architecture team"

* git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile:
  tile: remove stray blank space
  tile: <arch/> header updates from upstream
  tile: improve gxio iorpc autogenerated code style
  tile: double default VMALLOC space
  tile: remove stale arch/tile/kernel/futex_64.S
  tile: remove HUGE_VMAP dead code
  tile: use pmd_pfn() instead of casting via pte_t
  tile: fix typos in comment in arch/tile/kernel/unaligned.c

When we cancel 'async_pf_execute()', we should behave as if the work was
never scheduled in 'kvm_setup_async_pf()'.
Fixes a bug when we can't unload module because the vm wasn't destroyed.
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

After nested vmentry stale cache can be used to reload L2 PDPTR pointers
which will cause L2 guest to fail. Fix it by invalidating cache on nested
vmentry emulation.

https://bugzilla.kernel.org/show_bug.cgi?id=60830Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Page tables in a read-only memory slot will currently cause a triple
fault because the page walker uses gfn_to_hva and it fails on such a slot.

OVMF uses such a page table; however, real hardware seems to be fine with
that as long as the accessed/dirty bits are set.  Save whether the slot
is readonly, and later check it when updating the accessed and dirty bits.
Reviewed-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Reviewed-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Opcode CA

This gets used by a DOS based NetWare guest.
Signed-off-by: Bruce Rogers <brogers@suse.com>
Reviewed-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Pull timer code update from Thomas Gleixner:
 - armada SoC clocksource overhaul with a trivial merge conflict
 - Minor improvements to various SoC clocksource drivers

* 'timers/core' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  clocksource: armada-370-xp: Add detailed clock requirements in devicetree binding
  clocksource: armada-370-xp: Get reference fixed-clock by name
  clocksource: armada-370-xp: Replace WARN_ON with BUG_ON
  clocksource: armada-370-xp: Fix device-tree binding
  clocksource: armada-370-xp: Introduce new compatibles
  clocksource: armada-370-xp: Use CLOCKSOURCE_OF_DECLARE
  clocksource: armada-370-xp: Simplify TIMER_CTRL register access
  clocksource: armada-370-xp: Use BIT()
  ARM: timer-sp: Set dynamic irq affinity
  ARM: nomadik: add dynamic irq flag to the timer
  clocksource: sh_cmt: 32-bit control register support
  clocksource: em_sti: Convert to devm_* managed helpers

The compat sys_llseek() definition addition added a bogus space
on an otherwise-blank line.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>

The hardware architecture descriptor headers have been updated, in
particular to reflect some larger MMIO fields on the mPIPE shims for
controlling the network hardware, from the recent Gx72 release.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>

Fix some whitespace style issues in some auto-generated files.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>

With per-cpu data as well as loaded kernel modules coming from
the vmalloc arena, we get close to the line all the time and
occasionally need more than we had, so just double it up by default.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>

Pull CIFS fixes from Steve French:
 "Two minor cifs fixes and a minor documentation cleanup for cifs.txt"

* 'for-next' of git://git.samba.org/sfrench/cifs-2.6:
  cifs: update cifs.txt and remove some outdated infos
  cifs: Avoid calling unlock_page() twice in cifs_readpage() when using fscache
  cifs: Do not take a reference to the page in cifs_readpage_worker()

Pull UBI fixes from Artem Bityutskiy:
 "Just a single fastmap fix plus a regression fix"

* tag 'upstream-3.12-rc1' of git://git.infradead.org/linux-ubi:
  UBI: Fix invalidate_fastmap()
  UBI: Fix PEB leak in wear_leveling_worker()

Pull ubifs fix from Artem Bityutskiy:
 "Just one patch which fixes the power-cut recovery testing mode.

  I'll start using a single UBI/UBIFS tree instead of 2 trees from now
  on.  So in the future you'll get 1 small pull request instead of 2
  tiny ones"

* tag 'upstream-3.12-rc1' of git://git.infradead.org/linux-ubifs:
  UBIFS: remove invalid warn msg with tst_recovery enabled

sched_info_depart seems to be only called from
sched_info_switch(), so only on involuntary task switch.

Fix the comment to match.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Link: http://lkml.kernel.org/r/20130916083036.GA1113@redhat.comSigned-off-by: Ingo Molnar <mingo@kernel.org>

Pull MIPS fixes from Ralf Baechle:
 "These are four patches for three construction sites:

   - Fix register decoding for the combination of multi-core processors
     and multi-threading.

   - Two more fixes that are part of the ongoing DECstation resurrection
     work.  One of these touches a DECstation-only network driver.

   - Finally Markos' trivial build fix for the AP/SP support.

  (With this applied now all MIPS defconfigs are building again)"

* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
  MIPS: kernel: vpe: Make vpe_attrs an array of pointers.
  MIPS: Fix SMP core calculations when using MT support.
  MIPS: DECstation I/O ASIC DMA interrupt handling fix
  MIPS: DECstation HRT initialization rearrangement

Pull x86 platform updates from Matthew Garrett:
 "Nothing amazing here, almost entirely cleanups and minor bugfixes and
  one bit of hardware enablement in the amilo-rfkill driver"

* 'for_linus' of git://cavan.codon.org.uk/platform-drivers-x86:
  platform/x86: panasonic-laptop: reuse module_acpi_driver
  samsung-laptop: fix config build error
  platform: x86: remove unnecessary platform_set_drvdata()
  amilo-rfkill: Enable using amilo-rfkill with the FSC Amilo L1310.
  wmi: parse_wdg() should return kernel error codes
  hp_wmi: Fix unregister order in hp_wmi_rfkill_setup()
  platform: replace strict_strto*() with kstrto*()
  x86: irst: use module_acpi_driver to simplify the code
  x86: smartconnect: use module_acpi_driver to simplify the code
  platform samsung-q10: use ACPI instead of direct EC calls
  thinkpad_acpi: add the ability setting TPACPI_LED_NONE by quirk
  thinkpad_acpi: return -NODEV while operating uninitialized LEDs

Pull misc SCSI driver updates from James Bottomley:
 "This patch set is a set of driver updates (megaraid_sas, fnic, lpfc,
  ufs, hpsa) we also have a couple of bug fixes (sd out of bounds and
  ibmvfc error handling) and the first round of esas2r checker fixes and
  finally the much anticipated big endian additions for megaraid_sas"

* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (47 commits)
  [SCSI] fnic: fnic Driver Tuneables Exposed through CLI
  [SCSI] fnic: Kernel panic while running sh/nosh with max lun cfg
  [SCSI] fnic: Hitting BUG_ON(io_req->abts_done) in fnic_rport_exch_reset
  [SCSI] fnic: Remove QUEUE_FULL handling code
  [SCSI] fnic: On system with >1.1TB RAM, VIC fails multipath after boot up
  [SCSI] fnic: FC stat param seconds_since_last_reset not getting updated
  [SCSI] sd: Fix potential out-of-bounds access
  [SCSI] lpfc 8.3.42: Update lpfc version to driver version 8.3.42
  [SCSI] lpfc 8.3.42: Fixed issue of task management commands having a fixed timeout
  [SCSI] lpfc 8.3.42: Fixed inconsistent spin lock usage.
  [SCSI] lpfc 8.3.42: Fix driver's abort loop functionality to skip IOs already getting aborted
  [SCSI] lpfc 8.3.42: Fixed failure to allocate SCSI buffer on PPC64 platform for SLI4 devices
  [SCSI] lpfc 8.3.42: Fix WARN_ON when driver unloads
  [SCSI] lpfc 8.3.42: Avoided making pci bar ioremap call during dual-chute WQ/RQ pci bar selection
  [SCSI] lpfc 8.3.42: Fixed driver iocbq structure's iocb_flag field running out of space
  [SCSI] lpfc 8.3.42: Fix crash on driver load due to cpu affinity logic
  [SCSI] lpfc 8.3.42: Fixed logging format of setting driver sysfs attributes hard to interpret
  [SCSI] lpfc 8.3.42: Fixed back to back RSCNs discovery failure.
  [SCSI] lpfc 8.3.42: Fixed race condition between BSG I/O dispatch and timeout handling
  [SCSI] lpfc 8.3.42: Fixed function mode field defined too small for not recognizing dual-chute mode
  ...

Pull SLAB update from Pekka Enberg:
 "Nothing terribly exciting here apart from Christoph's kmalloc
  unification patches that brings sl[aou]b implementations closer to
  each other"

* 'slab/next' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/linux:
  slab: Use correct GFP_DMA constant
  slub: remove verify_mem_not_deleted()
  mm/sl[aou]b: Move kmallocXXX functions to common code
  mm, slab_common: add 'unlikely' to size check of kmalloc_slab()
  mm/slub.c: beautify code for removing redundancy 'break' statement.
  slub: Remove unnecessary page NULL check
  slub: don't use cpu partial pages on UP
  mm/slub: beautify code for 80 column limitation and tab alignment
  mm/slub: remove 'per_cpu' which is useless variable

Pull input update from Dmitry Torokhov:
 "The only change is David Hermann's new EVIOCREVOKE evdev ioctl that
  allows safely passing file descriptors to input devices to session
  processes and later being able to stop delivery of events through
  these fds so that inactive sessions will no longer receive user input
  that does not belong to them"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
  Input: evdev - add EVIOCREVOKE ioctl

Sedat points out that I transposed some letters in "LRU" and wrote "RLU"
instead in one of the new comments explaining the flow.  Let's just fix
it.
Reported-by: Sedat Dilek <sedat.dilek@jpberlin.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Matt found that commit 27a7c642 ("partitions/efi: account for pmbr
size in lba") caused his GPT formatted eMMC device not to boot.  The
reason is that this commit enforced Linux to always check the lesser of
the whole disk or 2Tib for the pMBR size in LBA.  While most disk
partitioning tools out there create a pMBR with these characteristics,
Microsoft does not, as it always sets the entry to the maximum 32-bit
limitation - even though a drive may be smaller than that[1].

Loosen this check and only verify that the size is either the whole disk
or 0xFFFFFFFF.  No tool in its right mind would set it to any value
other than these.

[1] http://thestarman.pcministry.com/asm/mbr/GPT.htm#GPTPTReported-and-tested-by: Matt Porter <matt.porter@linaro.org>
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull writeback fix from Wu Fengguang:
 "A trivial writeback fix"

* tag 'writeback-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/wfg/linux:
  writeback: Do not sort b_io list only because of block device inode

The LRU list changes interacted badly with our nr_dentry_unused
accounting, and even worse with the new DCACHE_LRU_LIST bit logic.

This introduces helper functions to make sure everything follows the
proper dcache d_lru list rules: the dentry cache is complicated by the
fact that some of the hotpaths don't even want to look at the LRU list
at all, and the fact that we use the same list entry in the dentry for
both the LRU list and for our temporary shrinking lists when removing
things from the LRU.

The helper functions temporarily have some extra sanity checking for the
flag bits that have to match the current LRU state of the dentry.  We'll
remove that before the final 3.12 release, but considering how easy it
is to get wrong, this first cleanup version has some very particular
sanity checking.
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>