Merge branch 'v4.4-16.09-android-tmp' into lsk-v4.4-16.09-android

* v4.4-16.09-android-tmp:
  unsafe_[get|put]_user: change interface to use a error target label
  usercopy: remove page-spanning test for now
  usercopy: fix overlap check for kernel text
  mm/slub: support left redzone
  Linux 4.4.21
  lib/mpi: mpi_write_sgl(): fix skipping of leading zero limbs
  regulator: anatop: allow regulator to be in bypass mode
  hwrng: exynos - Disable runtime PM on probe failure
  cpufreq: Fix GOV_LIMITS handling for the userspace governor
  metag: Fix atomic_*_return inline asm constraints
  scsi: fix upper bounds check of sense key in scsi_sense_key_string()
  ALSA: timer: fix NULL pointer dereference on memory allocation failure
  ALSA: timer: fix division by zero after SNDRV_TIMER_IOCTL_CONTINUE
  ALSA: timer: fix NULL pointer dereference in read()/ioctl() race
  ALSA: hda - Enable subwoofer on Dell Inspiron 7559
  ALSA: hda - Add headset mic quirk for Dell Inspiron 5468
  ALSA: rawmidi: Fix possible deadlock with virmidi registration
  ALSA: fireworks: accessing to user space outside spinlock
  ALSA: firewire-tascam: accessing to user space outside spinlock
  ALSA: usb-audio: Add sample rate inquiry quirk for B850V3 CP2114
  crypto: caam - fix IV loading for authenc (giv)decryption
  uprobes: Fix the memcg accounting
  x86/apic: Do not init irq remapping if ioapic is disabled
  vhost/scsi: fix reuse of &vq->iov[out] in response
  bcache: RESERVE_PRIO is too small by one when prio_buckets() is a power of two.
  ubifs: Fix assertion in layout_in_gaps()
  ovl: fix workdir creation
  ovl: listxattr: use strnlen()
  ovl: remove posix_acl_default from workdir
  ovl: don't copy up opaqueness
  wrappers for ->i_mutex access
  lustre: remove unused declaration
  timekeeping: Avoid taking lock in NMI path with CONFIG_DEBUG_TIMEKEEPING
  timekeeping: Cap array access in timekeeping_debug
  xfs: fix superblock inprogress check
  ASoC: atmel_ssc_dai: Don't unconditionally reset SSC on stream startup
  drm/msm: fix use of copy_from_user() while holding spinlock
  drm: Reject page_flip for !DRIVER_MODESET
  drm/radeon: fix radeon_move_blit on 32bit systems
  s390/sclp_ctl: fix potential information leak with /dev/sclp
  rds: fix an infoleak in rds_inc_info_copy
  powerpc/tm: Avoid SLB faults in treclaim/trecheckpoint when RI=0
  nvme: Call pci_disable_device on the error path.
  cgroup: reduce read locked section of cgroup_threadgroup_rwsem during fork
  block: make sure a big bio is split into at most 256 bvecs
  block: Fix race triggered by blk_set_queue_dying()
  ext4: avoid modifying checksum fields directly during checksum verification
  ext4: avoid deadlock when expanding inode size
  ext4: properly align shifted xattrs when expanding inodes
  ext4: fix xattr shifting when expanding inodes part 2
  ext4: fix xattr shifting when expanding inodes
  ext4: validate that metadata blocks do not overlap superblock
  net: Use ns_capable_noaudit() when determining net sysctl permissions
  kernel: Add noaudit variant of ns_capable()
  KEYS: Fix ASN.1 indefinite length object parsing
  drivers:hv: Lock access to hyperv_mmio resource tree
  cxlflash: Move to exponential back-off when cmd_room is not available
  netfilter: x_tables: check for size overflow
  drm/amdgpu/cz: enable/disable vce dpm even if vce pg is disabled
  cred: Reject inodes with invalid ids in set_create_file_as()
  fs: Check for invalid i_uid in may_follow_link()
  IB/IPoIB: Do not set skb truesize since using one linearskb
  udp: properly support MSG_PEEK with truncated buffers
  crypto: nx-842 - Mask XERS0 bit in return value
  cxlflash: Fix to avoid virtual LUN failover failure
  cxlflash: Fix to escalate LINK_RESET also on port 1
  tipc: fix nl compat regression for link statistics
  tipc: fix an infoleak in tipc_nl_compat_link_dump
  netfilter: x_tables: check for size overflow
  Bluetooth: Add support for Intel Bluetooth device 8265 [8087:0a2b]
  drm/i915: Check VBT for port presence in addition to the strap on VLV/CHV
  drm/i915: Only ignore eDP ports that are connected
  Input: xpad - move pending clear to the correct location
  net: thunderx: Fix link status reporting
  x86/hyperv: Avoid reporting bogus NMI status for Gen2 instances
  crypto: vmx - IV size failing on skcipher API
  tda10071: Fix dependency to REGMAP_I2C
  crypto: vmx - Fix ABI detection
  crypto: vmx - comply with ABIs that specify vrsave as reserved.
  HID: core: prevent out-of-bound readings
  lpfc: Fix DMA faults observed upon plugging loopback connector
  block: fix blk_rq_get_max_sectors for driver private requests
  irqchip/gicv3-its: numa: Enable workaround for Cavium thunderx erratum 23144
  clocksource: Allow unregistering the watchdog
  btrfs: Continue write in case of can_not_nocow
  blk-mq: End unstarted requests on dying queue
  cxlflash: Fix to resolve dead-lock during EEH recovery
  drm/radeon/mst: fix regression in lane/link handling.
  ecryptfs: fix handling of directory opening
  ALSA: hda: add AMD Polaris-10/11 AZ PCI IDs with proper driver caps
  drm: Balance error path for GEM handle allocation
  ntp: Fix ADJ_SETOFFSET being used w/ ADJ_NANO
  time: Verify time values in adjtimex ADJ_SETOFFSET to avoid overflow
  Input: xpad - correctly handle concurrent LED and FF requests
  net: thunderx: Fix receive packet stats
  net: thunderx: Fix for multiqset not configured upon interface toggle
  perf/x86/cqm: Fix CQM memory leak and notifier leak
  perf/x86/cqm: Fix CQM handling of grouping events into a cache_group
  s390/crypto: provide correct file mode at device register.
  proc: revert /proc/<pid>/maps [stack:TID] annotation
  intel_idle: Support for Intel Xeon Phi Processor x200 Product Family
  cxlflash: Fix to avoid unnecessary scan with internal LUNs
  Drivers: hv: vmbus: don't manipulate with clocksources on crash
  Drivers: hv: vmbus: avoid scheduling in interrupt context in vmbus_initiate_unload()
  Drivers: hv: vmbus: avoid infinite loop in init_vp_index()
  arcmsr: fixes not release allocated resource
  arcmsr: fixed getting wrong configuration data
  s390/pci_dma: fix DMA table corruption with > 4 TB main memory
  net/mlx5e: Don't modify CQ before it was created
  net/mlx5e: Don't try to modify CQ moderation if it is not supported
  mmc: sdhci: Do not BUG on invalid vdd
  UVC: Add support for R200 depth camera
  sched/numa: Fix use-after-free bug in the task_numa_compare
  ALSA: hda - add codec support for Kabylake display audio codec
  drm/i915: Fix hpd live status bits for g4x
  tipc: fix nullptr crash during subscription cancel
  arm64: Add workaround for Cavium erratum 27456
  net: thunderx: Fix for Qset error due to CQ full
  drm/radeon: fix dp link rate selection (v2)
  drm/amdgpu: fix dp link rate selection (v2)
  qla2xxx: Use ATIO type to send correct tmr response
  mmc: sdhci: 64-bit DMA actually has 4-byte alignment
  drm/atomic: Do not unset crtc when an encoder is stolen
  drm/i915/skl: Add missing SKL ids
  drm/i915/bxt: update list of PCIIDs
  hrtimer: Catch illegal clockids
  i40e/i40evf: Fix RSS rx-flow-hash configuration through ethtool
  mpt3sas: Fix for Asynchronous completion of timedout IO and task abort of timedout IO.
  mpt3sas: A correction in unmap_resources
  net: cavium: liquidio: fix check for in progress flag
  arm64: KVM: Configure TCR_EL2.PS at runtime
  irqchip/gic-v3: Make sure read from ICC_IAR1_EL1 is visible on redestributor
  pwm: lpc32xx: fix and simplify duty cycle and period calculations
  pwm: lpc32xx: correct number of PWM channels from 2 to 1
  pwm: fsl-ftm: Fix clock enable/disable when using PM
  megaraid_sas: Add an i/o barrier
  megaraid_sas: Fix SMAP issue
  megaraid_sas: Do not allow PCI access during OCR
  s390/cio: update measurement characteristics
  s390/cio: ensure consistent measurement state
  s390/cio: fix measurement characteristics memleak
  qeth: initialize net_device with carrier off
  lpfc: Fix external loopback failure.
  lpfc: Fix mbox reuse in PLOGI completion
  lpfc: Fix RDP Speed reporting.
  lpfc: Fix crash in fcp command completion path.
  lpfc: Fix driver crash when module parameter lpfc_fcp_io_channel set to 16
  lpfc: Fix RegLogin failed error seen on Lancer FC during port bounce
  lpfc: Fix the FLOGI discovery logic to comply with T11 standards
  lpfc: Fix FCF Infinite loop in lpfc_sli4_fcf_rr_next_index_get.
  cxl: Enable PCI device ID for future IBM CXL adapter
  cxl: fix build for GCC 4.6.x
  cxlflash: Enable device id for future IBM CXL adapter
  cxlflash: Resolve oops in wait_port_offline
  cxlflash: Fix to resolve cmd leak after host reset
  cxl: Fix DSI misses when the context owning task exits
  cxl: Fix possible idr warning when contexts are released
  Drivers: hv: vmbus: fix rescind-offer handling for device without a driver
  Drivers: hv: vmbus: serialize process_chn_event() and vmbus_close_internal()
  Drivers: hv: vss: run only on supported host versions
  drivers/hv: cleanup synic msrs if vmbus connect failed
  Drivers: hv: util: catch allocation errors
  tools: hv: report ENOSPC errors in hv_fcopy_daemon
  Drivers: hv: utils: run polling callback always in interrupt context
  Drivers: hv: util: Increase the timeout for util services
  lightnvm: fix missing grown bad block type
  lightnvm: fix locking and mempool in rrpc_lun_gc
  lightnvm: unlock rq and free ppa_list on submission fail
  lightnvm: add check after mempool allocation
  lightnvm: fix incorrect nr_free_blocks stat
  lightnvm: fix bio submission issue
  cxlflash: a couple off by one bugs
  fm10k: Cleanup exception handling for mailbox interrupt
  fm10k: Cleanup MSI-X interrupts in case of failure
  fm10k: reinitialize queuing scheme after calling init_hw
  fm10k: always check init_hw for errors
  fm10k: reset max_queues on init_hw_vf failure
  fm10k: Fix handling of NAPI budget when multiple queues are enabled per vector
  fm10k: Correct MTU for jumbo frames
  fm10k: do not assume VF always has 1 queue
  clk: xgene: Fix divider with non-zero shift value
  e1000e: fix division by zero on jumbo MTUs
  e1000: fix data race between tx_ring->next_to_clean
  ixgbe: Fix handling of NAPI budget when multiple queues are enabled per vector
  igb: fix NULL derefs due to skipped SR-IOV enabling
  igb: use the correct i210 register for EEMNGCTL
  igb: don't unmap NULL hw_addr
  i40e: Fix Rx hash reported to the stack by our driver
  i40e: clean whole mac filter list
  i40evf: check rings before freeing resources
  i40e: don't add zero MAC filter
  i40e: properly delete VF MAC filters
  i40e: Fix memory leaks, sideband filter programming
  i40e: fix: do not sleep in netdev_ops
  i40e/i40evf: Fix RS bit update in Tx path and disable force WB workaround
  i40evf: handle many MAC filters correctly
  i40e: Workaround fix for mss < 256 issue
  UPSTREAM: audit: fix a double fetch in audit_log_single_execve_arg()
  UPSTREAM: ARM: 8494/1: mm: Enable PXN when running non-LPAE kernel on LPAE processor
  FIXUP: sched/tune: update accouting before CPU capacity
  FIXUP: sched/tune: add fixes missing from a previous patch
  arm: Fix #if/#ifdef typo in topology.c
  arm: Fix build error "conflicting types for 'scale_cpu_capacity'"
  sched/walt: use do_div instead of division operator
  DEBUG: cpufreq: fix cpu_capacity tracing build for non-smp systems
  sched/walt: include missing header for arm_timer_read_counter()
  cpufreq: Kconfig: Fixup incorrect selection by CPU_FREQ_DEFAULT_GOV_SCHED
  sched/fair: Avoid redundant idle_cpu() call in update_sg_lb_stats()
  FIXUP: sched: scheduler-driven cpu frequency selection
  sched/rt: Add Kconfig option to enable panicking for RT throttling
  sched/rt: print RT tasks when RT throttling is activated
  UPSTREAM: sched: Fix a race between __kthread_bind() and sched_setaffinity()
  sched/fair: Favor higher cpus only for boosted tasks
  vmstat: make vmstat_updater deferrable again and shut down on idle
  sched/fair: call OPP update when going idle after migration
  sched/cpufreq_sched: fix thermal capping events
  sched/fair: Picking cpus with low OPPs for tasks that prefer idle CPUs
  FIXUP: sched/tune: do initialization as a postcore_initicall
  DEBUG: sched: add tracepoint for RD overutilized
  sched/tune: Introducing a new schedtune attribute prefer_idle
  sched: use util instead of capacity to select busy cpu
  arch_timer: add error handling when the MPM global timer is cleared
  FIXUP: sched: Fix double-release of spinlock in move_queued_task
  FIXUP: sched/fair: Fix hang during suspend in sched_group_energy
  FIXUP: sched: fix SchedFreq integration for both PELT and WALT
  sched: EAS: Avoid causing spikes to max-freq unnecessarily
  FIXUP: sched: fix set_cfs_cpu_capacity when WALT is in use
  sched/walt: Accounting for number of irqs pending on each core
  sched: Introduce Window Assisted Load Tracking (WALT)
  sched/tune: fix PB and PC cuts indexes definition
  sched/fair: optimize idle cpu selection for boosted tasks
  FIXUP: sched/tune: fix accounting for runnable tasks
  sched/tune: use a single initialisation function
  sched/{fair,tune}: simplify fair.c code
  FIXUP: sched/tune: fix payoff calculation for boost region
  sched/tune: Add support for negative boost values
  FIX: sched/tune: move schedtune_nornalize_energy into fair.c
  FIX: sched/tune: update usage of boosted task utilisation on CPU selection
  sched/fair: add tunable to set initial task load
  sched/fair: add tunable to force selection at cpu granularity
  sched: EAS: take cstate into account when selecting idle core
  sched/cpufreq_sched: Consolidated update
  FIXUP: sched: fix build for non-SMP target
  DEBUG: sched/tune: add tracepoint on P-E space filtering
  DEBUG: sched/tune: add tracepoint for energy_diff() values
  DEBUG: sched/tune: add tracepoint for task boost signal
  arm: topology: Define TC2 energy and provide it to the scheduler
  CHROMIUM: sched: update the average of nr_running
  DEBUG: schedtune: add tracepoint for schedtune_tasks_update() values
  DEBUG: schedtune: add tracepoint for CPU boost signal
  DEBUG: schedtune: add tracepoint for SchedTune configuration update
  DEBUG: sched: add energy procfs interface
  DEBUG: sched,cpufreq: add cpu_capacity change tracepoint
  DEBUG: sched: add tracepoint for CPU load/util signals
  DEBUG: sched: add tracepoint for task load/util signals
  DEBUG: sched: add tracepoint for cpu/freq scale invariance
  sched/fair: filter energy_diff() based on energy_payoff value
  sched/tune: add support to compute normalized energy
  sched/fair: keep track of energy/capacity variations
  sched/fair: add boosted task utilization
  sched/{fair,tune}: track RUNNABLE tasks impact on per CPU boost value
  sched/tune: compute and keep track of per CPU boost value
  sched/tune: add initial support for CGroups based boosting
  sched/fair: add boosted CPU usage
  sched/fair: add function to convert boost value into "margin"
  sched/tune: add sysctl interface to define a boost value
  sched/tune: add detailed documentation
  fixup! sched/fair: jump to max OPP when crossing UP threshold
  fixup! sched: scheduler-driven cpu frequency selection
  sched: rt scheduler sets capacity requirement
  sched: deadline: use deadline bandwidth in scale_rt_capacity
  sched: remove call of sched_avg_update from sched_rt_avg_update
  sched/cpufreq_sched: add trace events
  sched/fair: jump to max OPP when crossing UP threshold
  sched/fair: cpufreq_sched triggers for load balancing
  sched/{core,fair}: trigger OPP change request on fork()
  sched/fair: add triggers for OPP change requests
  sched: scheduler-driven cpu frequency selection
  cpufreq: introduce cpufreq_driver_is_slow
  sched: Consider misfit tasks when load-balancing
  sched: Add group_misfit_task load-balance type
  sched: Add per-cpu max capacity to sched_group_capacity
  sched: Do eas idle balance regardless of the rq avg idle value
  arm64: Enable max freq invariant scheduler load-tracking and capacity support
  arm: Enable max freq invariant scheduler load-tracking and capacity support
  sched: Update max cpu capacity in case of max frequency constraints
  cpufreq: Max freq invariant scheduler load-tracking and cpu capacity support
  arm64, topology: Updates to use DT bindings for EAS costing data
  sched: Support for extracting EAS energy costs from DT
  Documentation: DT bindings for energy model cost data required by EAS
  sched: Disable energy-unfriendly nohz kicks
  sched: Consider a not over-utilized energy-aware system as balanced
  sched: Energy-aware wake-up task placement
  sched: Determine the current sched_group idle-state
  sched, cpuidle: Track cpuidle state index in the scheduler
  sched: Add over-utilization/tipping point indicator
  sched: Estimate energy impact of scheduling decisions
  sched: Extend sched_group_energy to test load-balancing decisions
  sched: Calculate energy consumption of sched_group
  sched: Highest energy aware balancing sched_domain level pointer
  sched: Relocated cpu_util() and change return type
  sched: Compute cpu capacity available at current frequency
  arm64: Cpu invariant scheduler load-tracking and capacity support
  arm: Cpu invariant scheduler load-tracking and capacity support
  sched: Introduce SD_SHARE_CAP_STATES sched_domain flag
  sched: Initialize energy data structures
  sched: Introduce energy data structures
  sched: Make energy awareness a sched feature
  sched: Documentation for scheduler energy cost model
  sched: Prevent unnecessary active balance of single task in sched group
  sched: Enable idle balance to pull single task towards cpu with higher capacity
  sched: Consider spare cpu capacity at task wake-up
  sched: Add cpu capacity awareness to wakeup balancing
  sched: Store system-wide maximum cpu capacity in root domain
  arm: Update arch_scale_cpu_capacity() to reflect change to define
  arm64: Enable frequency invariant scheduler load-tracking support
  arm: Enable frequency invariant scheduler load-tracking support
  cpufreq: Frequency invariant scheduler load-tracking support
  sched/fair: Fix new task's load avg removed from source CPU in wake_up_new_task()
  FROMLIST: pstore: drop pmsg bounce buffer
  UPSTREAM: usercopy: remove page-spanning test for now
  UPSTREAM: usercopy: force check_object_size() inline
  BACKPORT: usercopy: fold builtin_const check into inline function
  UPSTREAM: x86/uaccess: force copy_*_user() to be inlined
  UPSTREAM: HID: core: prevent out-of-bound readings
  Android: Fix build breakages.
  UPSTREAM: tty: Prevent ldisc drivers from re-using stale tty fields
  UPSTREAM: netfilter: nfnetlink: correctly validate length of batch messages
  cpuset: Make cpusets restore on hotplug
  UPSTREAM: mm/slub: support left redzone
  UPSTREAM: Make the hardened user-copy code depend on having a hardened allocator
  Android: MMC/UFS IO Latency Histograms.
  UPSTREAM: usercopy: fix overlap check for kernel text
  UPSTREAM: usercopy: avoid potentially undefined behavior in pointer math
  UPSTREAM: unsafe_[get|put]_user: change interface to use a error target label
  BACKPORT: arm64: mm: fix location of _etext
  BACKPORT: ARM: 8583/1: mm: fix location of _etext
  BACKPORT: Don't show empty tag stats for unprivileged uids
  UPSTREAM: tcp: fix use after free in tcp_xmit_retransmit_queue()
  ANDROID: base-cfg: drop SECCOMP_FILTER config
  UPSTREAM: [media] xc2028: unlock on error in xc2028_set_config()
  UPSTREAM: [media] xc2028: avoid use after free
  ANDROID: base-cfg: enable SECCOMP config
  ANDROID: rcu_sync: Export rcu_sync_lockdep_assert
  RFC: FROMLIST: cgroup: reduce read locked section of cgroup_threadgroup_rwsem during fork
  RFC: FROMLIST: cgroup: avoid synchronize_sched() in __cgroup_procs_write()
  RFC: FROMLIST: locking/percpu-rwsem: Optimize readers and reduce global impact
  net: ipv6: Fix ping to link-local addresses.
  ipv6: fix endianness error in icmpv6_err
  ANDROID: dm: android-verity: Allow android-verity to be compiled as an independent module
  backporting: a brief introduce of backported feautures on 4.4
  Linux 4.4.20
  sysfs: correctly handle read offset on PREALLOC attrs
  hwmon: (iio_hwmon) fix memory leak in name attribute
  ALSA: line6: Fix POD sysfs attributes segfault
  ALSA: line6: Give up on the lock while URBs are released.
  ALSA: line6: Remove double line6_pcm_release() after failed acquire.
  ACPI / SRAT: fix SRAT parsing order with both LAPIC and X2APIC present
  ACPI / sysfs: fix error code in get_status()
  ACPI / drivers: replace acpi_probe_lock spinlock with mutex
  ACPI / drivers: fix typo in ACPI_DECLARE_PROBE_ENTRY macro
  staging: comedi: ni_mio_common: fix wrong insn_write handler
  staging: comedi: ni_mio_common: fix AO inttrig backwards compatibility
  staging: comedi: comedi_test: fix timer race conditions
  staging: comedi: daqboard2000: bug fix board type matching code
  USB: serial: option: add WeTelecom 0x6802 and 0x6803 products
  USB: serial: option: add WeTelecom WM-D200
  USB: serial: mos7840: fix non-atomic allocation in write path
  USB: serial: mos7720: fix non-atomic allocation in write path
  USB: fix typo in wMaxPacketSize validation
  usb: chipidea: udc: don't touch DP when controller is in host mode
  USB: avoid left shift by -1
  dmaengine: usb-dmac: check CHCR.DE bit in usb_dmac_isr_channel()
  crypto: qat - fix aes-xts key sizes
  crypto: nx - off by one bug in nx_of_update_msc()
  Input: i8042 - set up shared ps2_cmd_mutex for AUX ports
  Input: i8042 - break load dependency between atkbd/psmouse and i8042
  Input: tegra-kbc - fix inverted reset logic
  btrfs: properly track when rescan worker is running
  btrfs: waiting on qgroup rescan should not always be interruptible
  fs/seq_file: fix out-of-bounds read
  gpio: Fix OF build problem on UM
  usb: renesas_usbhs: gadget: fix return value check in usbhs_mod_gadget_probe()
  megaraid_sas: Fix probing cards without io port
  mpt3sas: Fix resume on WarpDrive flash cards
  cdc-acm: fix wrong pipe type on rx interrupt xfers
  i2c: cros-ec-tunnel: Fix usage of cros_ec_cmd_xfer()
  mfd: cros_ec: Add cros_ec_cmd_xfer_status() helper
  aacraid: Check size values after double-fetch from user
  ARC: Elide redundant setup of DMA callbacks
  ARC: Call trace_hardirqs_on() before enabling irqs
  ARC: use correct offset in pt_regs for saving/restoring user mode r25
  ARC: build: Better way to detect ISA compatible toolchain
  drm/i915: fix aliasing_ppgtt leak
  drm/amdgpu: record error code when ring test failed
  drm/amd/amdgpu: sdma resume fail during S4 on CI
  drm/amdgpu: skip TV/CV in display parsing
  drm/amdgpu: avoid a possible array overflow
  drm/amdgpu: fix amdgpu_move_blit on 32bit systems
  drm/amdgpu: Change GART offset to 64-bit
  iio: fix sched WARNING "do not call blocking ops when !TASK_RUNNING"
  sched/nohz: Fix affine unpinned timers mess
  sched/cputime: Fix NO_HZ_FULL getrusage() monotonicity regression
  of: fix reference counting in of_graph_get_endpoint_by_regs
  arm64: dts: rockchip: add reset saradc node for rk3368 SoCs
  mac80211: fix purging multicast PS buffer queue
  s390/dasd: fix hanging device after clear subchannel
  EDAC: Increment correct counter in edac_inc_ue_error()
  pinctrl/amd: Remove the default de-bounce time
  iommu/arm-smmu: Don't BUG() if we find aborting STEs with disable_bypass
  iommu/arm-smmu: Fix CMDQ error handling
  iommu/dma: Don't put uninitialised IOVA domains
  xhci: Make sure xhci handles USB_SPEED_SUPER_PLUS devices.
  USB: serial: ftdi_sio: add PIDs for Ivium Technologies devices
  USB: serial: ftdi_sio: add device ID for WICED USB UART dev board
  USB: serial: option: add support for Telit LE920A4
  USB: serial: option: add D-Link DWM-156/A3
  USB: serial: fix memleak in driver-registration error path
  xhci: don't dereference a xhci member after removing xhci
  usb: xhci: Fix panic if disconnect
  xhci: always handle "Command Ring Stopped" events
  usb/gadget: fix gadgetfs aio support.
  usb: gadget: fsl_qe_udc: off by one in setup_received_handle()
  USB: validate wMaxPacketValue entries in endpoint descriptors
  usb: renesas_usbhs: Use dmac only if the pipe type is bulk
  usb: renesas_usbhs: clear the BRDYSTS in usbhsg_ep_enable()
  USB: hub: change the locking in hub_activate
  USB: hub: fix up early-exit pathway in hub_activate
  usb: hub: Fix unbalanced reference count/memory leak/deadlocks
  usb: define USB_SPEED_SUPER_PLUS speed for SuperSpeedPlus USB3.1 devices
  usb: dwc3: gadget: increment request->actual once
  usb: dwc3: pci: add Intel Kabylake PCI ID
  usb: misc: usbtest: add fix for driver hang
  usb: ehci: change order of register cleanup during shutdown
  crypto: caam - defer aead_set_sh_desc in case of zero authsize
  crypto: caam - fix echainiv(authenc) encrypt shared descriptor
  crypto: caam - fix non-hmac hashes
  genirq/msi: Make sure PCI MSIs are activated early
  genirq/msi: Remove unused MSI_FLAG_IDENTITY_MAP
  um: Don't discard .text.exit section
  ACPI / CPPC: Prevent cpc_desc_ptr points to the invalid data
  ACPI: CPPC: Return error if _CPC is invalid on a CPU
  mmc: sdhci-acpi: Reduce Baytrail eMMC/SD/SDIO hangs
  PCI: Limit config space size for Netronome NFP4000
  PCI: Add Netronome NFP4000 PF device ID
  PCI: Limit config space size for Netronome NFP6000 family
  PCI: Add Netronome vendor and device IDs
  PCI: Support PCIe devices with short cfg_size
  NVMe: Don't unmap controller registers on reset
  ALSA: hda - Manage power well properly for resume
  libnvdimm, nd_blk: mask off reserved status bits
  perf intel-pt: Fix occasional decoding errors when tracing system-wide
  vfio/pci: Fix NULL pointer oops in error interrupt setup handling
  virtio: fix memory leak in virtqueue_add()
  parisc: Fix order of EREFUSED define in errno.h
  arm64: Define AT_VECTOR_SIZE_ARCH for ARCH_DLINFO
  ALSA: usb-audio: Add quirk for ELP HD USB Camera
  ALSA: usb-audio: Add a sample rate quirk for Creative Live! Cam Socialize HD (VF0610)
  powerpc/eeh: eeh_pci_enable(): fix checking of post-request state
  SUNRPC: allow for upcalls for same uid but different gss service
  SUNRPC: Handle EADDRNOTAVAIL on connection failures
  tools/testing/nvdimm: fix SIGTERM vs hotplug crash
  uprobes/x86: Fix RIP-relative handling of EVEX-encoded instructions
  x86/mm: Disable preemption during CR3 read+write
  hugetlb: fix nr_pmds accounting with shared page tables
  mm: SLUB hardened usercopy support
  mm: SLAB hardened usercopy support
  s390/uaccess: Enable hardened usercopy
  sparc/uaccess: Enable hardened usercopy
  powerpc/uaccess: Enable hardened usercopy
  ia64/uaccess: Enable hardened usercopy
  arm64/uaccess: Enable hardened usercopy
  ARM: uaccess: Enable hardened usercopy
  x86/uaccess: Enable hardened usercopy
  x86: remove more uaccess_32.h complexity
  x86: remove pointless uaccess_32.h complexity
  x86: fix SMAP in 32-bit environments
  Use the new batched user accesses in generic user string handling
  Add 'unsafe' user access functions for batched accesses
  x86: reorganize SMAP handling in user space accesses
  mm: Hardened usercopy
  mm: Implement stack frame object validation
  mm: Add is_migrate_cma_page
  Linux 4.4.19
  Documentation/module-signing.txt: Note need for version info if reusing a key
  module: Invalidate signatures on force-loaded modules
  dm flakey: error READ bios during the down_interval
  rtc: s3c: Add s3c_rtc_{enable/disable}_clk in s3c_rtc_setfreq()
  lpfc: fix oops in lpfc_sli4_scmd_to_wqidx_distr() from lpfc_send_taskmgmt()
  ACPI / EC: Work around method reentrancy limit in ACPICA for _Qxx
  x86/platform/intel_mid_pci: Rework IRQ0 workaround
  PCI: Mark Atheros AR9485 and QCA9882 to avoid bus reset
  MIPS: hpet: Increase HPET_MIN_PROG_DELTA and decrease HPET_MIN_CYCLES
  MIPS: Don't register r4k sched clock when CPUFREQ enabled
  MIPS: mm: Fix definition of R6 cache instruction
  SUNRPC: Don't allocate a full sockaddr_storage for tracing
  Input: elan_i2c - properly wake up touchpad on ASUS laptops
  target: Fix ordered task CHECK_CONDITION early exception handling
  target: Fix max_unmap_lba_count calc overflow
  target: Fix race between iscsi-target connection shutdown + ABORT_TASK
  target: Fix missing complete during ABORT_TASK + CMD_T_FABRIC_STOP
  target: Fix ordered task target_setup_cmd_from_cdb exception hang
  iscsi-target: Fix panic when adding second TCP connection to iSCSI session
  ubi: Fix race condition between ubi device creation and udev
  ubi: Fix early logging
  ubi: Make volume resize power cut aware
  of: fix memory leak related to safe_name()
  IB/mlx4: Fix memory leak if QP creation failed
  IB/mlx4: Fix error flow when sending mads under SRIOV
  IB/mlx4: Fix the SQ size of an RC QP
  IB/IWPM: Fix a potential skb leak
  IB/IPoIB: Don't update neigh validity for unresolved entries
  IB/SA: Use correct free function
  IB/mlx5: Return PORT_ERR in Active to Initializing tranisition
  IB/mlx5: Fix post send fence logic
  IB/mlx5: Fix entries check in mlx5_ib_resize_cq
  IB/mlx5: Fix returned values of query QP
  IB/mlx5: Fix entries checks in mlx5_ib_create_cq
  IB/mlx5: Fix MODIFY_QP command input structure
  ALSA: hda - Fix headset mic detection problem for two dell machines
  ALSA: hda: add AMD Bonaire AZ PCI ID with proper driver caps
  ALSA: hda/realtek - Can't adjust speaker's volume on a Dell AIO
  ALSA: hda: Fix krealloc() with __GFP_ZERO usage
  mm/hugetlb: avoid soft lockup in set_max_huge_pages()
  mtd: nand: fix bug writing 1 byte less than page size
  block: fix bdi vs gendisk lifetime mismatch
  block: add missing group association in bio-cloning functions
  metag: Fix __cmpxchg_u32 asm constraint for CMP
  ftrace/recordmcount: Work around for addition of metag magic but not relocations
  balloon: check the number of available pages in leak balloon
  drm/i915/dp: Revert "drm/i915/dp: fall back to 18 bpp when sink capability is unknown"
  drm/i915: Never fully mask the the EI up rps interrupt on SNB/IVB
  drm/edid: Add 6 bpc quirk for display AEO model 0.
  drm: Restore double clflush on the last partial cacheline
  drm/nouveau/fbcon: fix font width not divisible by 8
  drm/nouveau/gr/nv3x: fix instobj write offsets in gr setup
  drm/nouveau: check for supported chipset before booting fbdev off the hw
  drm/radeon: support backlight control for UNIPHY3
  drm/radeon: fix firmware info version checks
  drm/radeon: Poll for both connect/disconnect on analog connectors
  drm/radeon: add a delay after ATPX dGPU power off
  drm/amdgpu/gmc7: add missing mullins case
  drm/amdgpu: fix firmware info version checks
  drm/amdgpu: Disable RPM helpers while reprobing connectors on resume
  drm/amdgpu: support backlight control for UNIPHY3
  drm/amdgpu: Poll for both connect/disconnect on analog connectors
  drm/amdgpu: add a delay after ATPX dGPU power off
  w1:omap_hdq: fix regression
  netlabel: add address family checks to netlbl_{sock,req}_delattr()
  ARM: dts: sunxi: Add a startup delay for fixed regulator enabled phys
  audit: fix a double fetch in audit_log_single_execve_arg()
  iommu/amd: Update Alias-DTE in update_device_table()
  iommu/amd: Init unity mappings only for dma_ops domains
  iommu/amd: Handle IOMMU_DOMAIN_DMA in ops->domain_free call-back
  iommu/vt-d: Return error code in domain_context_mapping_one()
  iommu/exynos: Suppress unbinding to prevent system failure
  drm/i915: Don't complain about lack of ACPI video bios
  nfsd: don't return an unhashed lock stateid after taking mutex
  nfsd: Fix race between FREE_STATEID and LOCK
  nfs: don't create zero-length requests
  MIPS: KVM: Propagate kseg0/mapped tlb fault errors
  MIPS: KVM: Fix gfn range check in kseg0 tlb faults
  MIPS: KVM: Add missing gfn range check
  MIPS: KVM: Fix mapped fault broken commpage handling
  random: add interrupt callback to VMBus IRQ handler
  random: print a warning for the first ten uninitialized random users
  random: initialize the non-blocking pool via add_hwgenerator_randomness()
  CIFS: Fix a possible invalid memory access in smb2_query_symlink()
  cifs: fix crash due to race in hmac(md5) handling
  cifs: Check for existing directory when opening file with O_CREAT
  fs/cifs: make share unaccessible at root level mountable
  jbd2: make journal y2038 safe
  ARC: mm: don't loose PTE_SPECIAL in pte_modify()
  remoteproc: Fix potential race condition in rproc_add
  ovl: disallow overlayfs as upperdir
  HID: uhid: fix timeout when probe races with IO
  EDAC: Correct channel count limit
  Bluetooth: Fix l2cap_sock_setsockopt() with optname BT_RCVMTU
  spi: pxa2xx: Clear all RFT bits in reset_sccr1() on Intel Quark
  i2c: efm32: fix a failure path in efm32_i2c_probe()
  s5p-mfc: Add release callback for memory region devs
  s5p-mfc: Set device name for reserved memory region devs
  hp-wmi: Fix wifi cannot be hard-unblocked
  dm: set DMF_SUSPENDED* _before_ clearing DMF_NOFLUSH_SUSPENDING
  sur40: fix occasional oopses on device close
  sur40: lower poll interval to fix occasional FPS drops to ~56 FPS
  Fix RC5 decoding with Fintek CIR chipset
  vb2: core: Skip planes array verification if pb is NULL
  videobuf2-v4l2: Verify planes array in buffer dequeueing
  media: dvb_ringbuffer: Add memory barriers
  media: usbtv: prevent access to free'd resources
  mfd: qcom_rpm: Parametrize also ack selector size
  mfd: qcom_rpm: Fix offset error for msm8660
  intel_pstate: Fix MSR_CONFIG_TDP_x addressing in core_get_max_pstate()
  s390/cio: allow to reset channel measurement block
  KVM: nVMX: Fix memory corruption when using VMCS shadowing
  KVM: VMX: handle PML full VMEXIT that occurs during event delivery
  KVM: MTRR: fix kvm_mtrr_check_gfn_range_consistency page fault
  KVM: PPC: Book3S HV: Save/restore TM state in H_CEDE
  KVM: PPC: Book3S HV: Pull out TM state save/restore into separate procedures
  arm64: mm: avoid fdt_check_header() before the FDT is fully mapped
  arm64: dts: rockchip: fixes the gic400 2nd region size for rk3368
  pinctrl: cherryview: prevent concurrent access to GPIO controllers
  Bluetooth: hci_intel: Fix null gpio desc pointer dereference
  gpio: intel-mid: Remove potentially harmful code
  gpio: pca953x: Fix NBANK calculation for PCA9536
  tty/serial: atmel: fix RS485 half duplex with DMA
  serial: samsung: Fix ERR pointer dereference on deferred probe
  tty: serial: msm: Don't read off end of tx fifo
  arm64: Fix incorrect per-cpu usage for boot CPU
  arm64: debug: unmask PSTATE.D earlier
  arm64: kernel: Save and restore UAO and addr_limit on exception entry
  USB: usbfs: fix potential infoleak in devio
  usb: renesas_usbhs: fix NULL pointer dereference in xfer_work()
  USB: serial: option: add support for Telit LE910 PID 0x1206
  usb: dwc3: fix for the isoc transfer EP_BUSY flag
  usb: quirks: Add no-lpm quirk for Elan
  usb: renesas_usbhs: protect the CFIFOSEL setting in usbhsg_ep_enable()
  usb: f_fs: off by one bug in _ffs_func_bind()
  usb: gadget: avoid exposing kernel stack
  UPSTREAM: usb: gadget: configfs: add mutex lock before unregister gadget
  ANDROID: dm-verity: adopt changes made to dm callbacks
  UPSTREAM: ecryptfs: fix handling of directory opening
  ANDROID: net: core: fix UID-based routing
  ANDROID: net: fib: remove duplicate assignment
  FROMLIST: proc: Fix timerslack_ns CAP_SYS_NICE check when adjusting self
  ANDROID: dm verity fec: pack the fec_header structure
  ANDROID: dm: android-verity: Verify header before fetching table
  ANDROID: dm: allow adb disable-verity only in userdebug
  ANDROID: dm: mount as linear target if eng build
  ANDROID: dm: use default verity public key
  ANDROID: dm: fix signature verification flag
  ANDROID: dm: use name_to_dev_t
  ANDROID: dm: rename dm-linear methods for dm-android-verity
  ANDROID: dm: Minor cleanup
  ANDROID: dm: Mounting root as linear device when verity disabled
  ANDROID: dm-android-verity: Rebase on top of 4.1
  ANDROID: dm: Add android verity target
  ANDROID: dm: fix dm_substitute_devices()
  ANDROID: dm: Rebase on top of 4.1
  CHROMIUM: dm: boot time specification of dm=
  Implement memory_state_time, used by qcom,cpubw
  Revert "panic: Add board ID to panic output"
  usb: gadget: f_accessory: remove duplicate endpoint alloc
  BACKPORT: brcmfmac: defer DPC processing during probe
  FROMLIST: proc: Add LSM hook checks to /proc/<tid>/timerslack_ns
  FROMLIST: proc: Relax /proc/<tid>/timerslack_ns capability requirements
  UPSTREAM: ppp: defer netns reference release for ppp channel
  cpuset: Add allow_attach hook for cpusets on android.
  UPSTREAM: KEYS: Fix ASN.1 indefinite length object parsing
  ANDROID: sdcardfs: fix itnull.cocci warnings
  android-recommended.cfg: enable fstack-protector-strong
  Linux 4.4.18
  mm: memcontrol: fix memcg id ref counter on swap charge move
  mm: memcontrol: fix swap counter leak on swapout from offline cgroup
  mm: memcontrol: fix cgroup creation failure after many small jobs
  ext4: fix reference counting bug on block allocation error
  ext4: short-cut orphan cleanup on error
  ext4: validate s_reserved_gdt_blocks on mount
  ext4: don't call ext4_should_journal_data() on the journal inode
  ext4: fix deadlock during page writeback
  ext4: check for extents that wrap around
  crypto: scatterwalk - Fix test in scatterwalk_done
  crypto: gcm - Filter out async ghash if necessary
  fs/dcache.c: avoid soft-lockup in dput()
  fuse: fix wrong assignment of ->flags in fuse_send_init()
  fuse: fuse_flush must check mapping->flags for errors
  fuse: fsync() did not return IO errors
  sysv, ipc: fix security-layer leaking
  block: fix use-after-free in seq file
  x86/syscalls/64: Add compat_sys_keyctl for 32-bit userspace
  drm/i915: Pretend cursor is always on for ILK-style WM calculations (v2)
  x86/mm/pat: Fix BUG_ON() in mmap_mem() on QEMU/i386
  x86/pat: Document the PAT initialization sequence
  x86/xen, pat: Remove PAT table init code from Xen
  x86/mtrr: Fix PAT init handling when MTRR is disabled
  x86/mtrr: Fix Xorg crashes in Qemu sessions
  x86/mm/pat: Replace cpu_has_pat with boot_cpu_has()
  x86/mm/pat: Add pat_disable() interface
  x86/mm/pat: Add support of non-default PAT MSR setting
  devpts: clean up interface to pty drivers
  random: strengthen input validation for RNDADDTOENTCNT
  apparmor: fix ref count leak when profile sha1 hash is read
  Revert "s390/kdump: Clear subchannel ID to signal non-CCW/SCSI IPL"
  KEYS: 64-bit MIPS needs to use compat_sys_keyctl for 32-bit userspace
  arm: oabi compat: add missing access checks
  cdc_ncm: do not call usbnet_link_change from cdc_ncm_bind
  i2c: i801: Allow ACPI SystemIO OpRegion to conflict with PCI BAR
  x86/mm/32: Enable full randomization on i386 and X86_32
  HID: sony: do not bail out when the sixaxis refuses the output report
  PNP: Add Broadwell to Intel MCH size workaround
  PNP: Add Haswell-ULT to Intel MCH size workaround
  scsi: ignore errors from scsi_dh_add_device()
  ipath: Restrict use of the write() interface
  tcp: consider recv buf for the initial window scale
  qed: Fix setting/clearing bit in completion bitmap
  net/irda: fix NULL pointer dereference on memory allocation failure
  net: bgmac: Fix infinite loop in bgmac_dma_tx_add()
  bonding: set carrier off for devices created through netlink
  ipv4: reject RTNH_F_DEAD and RTNH_F_LINKDOWN from user space
  tcp: enable per-socket rate limiting of all 'challenge acks'
  tcp: make challenge acks less predictable
  arm64: relocatable: suppress R_AARCH64_ABS64 relocations in vmlinux
  arm64: vmlinux.lds: make __rela_offset and __dynsym_offset ABSOLUTE
  Linux 4.4.17
  vfs: fix deadlock in file_remove_privs() on overlayfs
  intel_th: Fix a deadlock in modprobing
  intel_th: pci: Add Kaby Lake PCH-H support
  net: mvneta: set real interrupt per packet for tx_done
  libceph: apply new_state before new_up_client on incrementals
  libata: LITE-ON CX1-JB256-HP needs lower max_sectors
  i2c: mux: reg: wrong condition checked for of_address_to_resource return value
  posix_cpu_timer: Exit early when process has been reaped
  media: fix airspy usb probe error path
  ipr: Clear interrupt on croc/crocodile when running with LSI
  SCSI: fix new bug in scsi_dev_info_list string matching
  RDS: fix rds_tcp_init() error path
  can: fix oops caused by wrong rtnl dellink usage
  can: fix handling of unmodifiable configuration options fix
  can: c_can: Update D_CAN TX and RX functions to 32 bit - fix Altera Cyclone access
  can: at91_can: RX queue could get stuck at high bus load
  perf/x86: fix PEBS issues on Intel Atom/Core2
  ovl: handle ATTR_KILL*
  sched/fair: Fix effective_load() to consistently use smoothed load
  mmc: block: fix packed command header endianness
  block: fix use-after-free in sys_ioprio_get()
  qeth: delete napi struct when removing a qeth device
  platform/chrome: cros_ec_dev - double fetch bug in ioctl
  clk: rockchip: initialize flags of clk_init_data in mmc-phase clock
  spi: sun4i: fix FIFO limit
  spi: sunxi: fix transfer timeout
  namespace: update event counter when umounting a deleted dentry
  9p: use file_dentry()
  ext4: verify extent header depth
  ecryptfs: don't allow mmap when the lower fs doesn't support it
  Revert "ecryptfs: forbid opening files without mmap handler"
  locks: use file_inode()
  power_supply: power_supply_read_temp only if use_cnt > 0
  cgroup: set css->id to -1 during init
  pinctrl: imx: Do not treat a PIN without MUX register as an error
  pinctrl: single: Fix missing flush of posted write for a wakeirq
  pvclock: Add CPU barriers to get correct version value
  Input: tsc200x - report proper input_dev name
  Input: xpad - validate USB endpoint count during probe
  Input: wacom_w8001 - w8001_MAX_LENGTH should be 13
  Input: xpad - fix oops when attaching an unknown Xbox One gamepad
  Input: elantech - add more IC body types to the list
  Input: vmmouse - remove port reservation
  ALSA: timer: Fix leak in events via snd_timer_user_tinterrupt
  ALSA: timer: Fix leak in events via snd_timer_user_ccallback
  ALSA: timer: Fix leak in SNDRV_TIMER_IOCTL_PARAMS
  xenbus: don't bail early from xenbus_dev_request_and_reply()
  xenbus: don't BUG() on user mode induced condition
  xen/pciback: Fix conf_space read/write overlap check.
  ARC: unwind: ensure that .debug_frame is generated (vs. .eh_frame)
  arc: unwind: warn only once if DW2_UNWIND is disabled
  kernel/sysrq, watchdog, sched/core: Reset watchdog on all CPUs while processing sysrq-w
  pps: do not crash when failed to register
  vmlinux.lds: account for destructor sections
  mm, meminit: ensure node is online before checking whether pages are uninitialised
  mm, meminit: always return a valid node from early_pfn_to_nid
  mm, compaction: prevent VM_BUG_ON when terminating freeing scanner
  fs/nilfs2: fix potential underflow in call to crc32_le
  mm, compaction: abort free scanner if split fails
  mm, sl[au]b: add __GFP_ATOMIC to the GFP reclaim mask
  dmaengine: at_xdmac: double FIFO flush needed to compute residue
  dmaengine: at_xdmac: fix residue corruption
  dmaengine: at_xdmac: align descriptors on 64 bits
  x86/quirks: Add early quirk to reset Apple AirPort card
  x86/quirks: Reintroduce scanning of secondary buses
  x86/quirks: Apply nvidia_bugs quirk only on root bus
  USB: OHCI: Don't mark EDs as ED_OPER if scheduling fails

Conflicts:
	arch/arm/kernel/topology.c
	arch/arm64/include/asm/arch_gicv3.h
	arch/arm64/kernel/topology.c
	block/bio.c
	drivers/cpufreq/Kconfig
	drivers/md/Makefile
	drivers/media/dvb-core/dvb_ringbuffer.c
	drivers/media/tuners/tuner-xc2028.c
	drivers/misc/Kconfig
	drivers/misc/Makefile
	drivers/mmc/core/host.c
	drivers/scsi/ufs/ufshcd.c
	drivers/scsi/ufs/ufshcd.h
	drivers/usb/dwc3/gadget.c
	drivers/usb/gadget/configfs.c
	fs/ecryptfs/file.c
	include/linux/mmc/core.h
	include/linux/mmc/host.h
	include/linux/mmzone.h
	include/linux/sched.h
	include/linux/sched/sysctl.h
	include/trace/events/power.h
	include/trace/events/sched.h
	init/Kconfig
	kernel/cpuset.c
	kernel/exit.c
	kernel/sched/Makefile
	kernel/sched/core.c
	kernel/sched/cputime.c
	kernel/sched/fair.c
	kernel/sched/features.h
	kernel/sched/rt.c
	kernel/sched/sched.h
	kernel/sched/stop_task.c
	kernel/sched/tune.c
	lib/Kconfig.debug
	mm/Makefile
	mm/vmstat.c

Change-Id: I243a43231ca56a6362076fa6301827e1b0493be5
Signed-off-by: Runmin Wang <runminw@codeaurora.org>

1 files changed, 734 insertions, 22 deletions

diff --git a/kernel/sched/tune.c b/kernel/sched/tune.c
index ee2af8e0b5ce..3c2e21c5a5a0 100644
--- a/kernel/sched/tune.c
+++ b/kernel/sched/tune.c
@@ -1,13 +1,108 @@
 #include <linux/cgroup.h>
 #include <linux/err.h>
+#include <linux/kernel.h>
 #include <linux/percpu.h>
 #include <linux/printk.h>
+#include <linux/rcupdate.h>
 #include <linux/slab.h>
 
+#include <trace/events/sched.h>
+
 #include "sched.h"
+#include "tune.h"
+
+#ifdef CONFIG_CGROUP_SCHEDTUNE
+static bool schedtune_initialized = false;
+#endif
 
 unsigned int sysctl_sched_cfs_boost __read_mostly;
 
+extern struct target_nrg schedtune_target_nrg;
+
+/* Performance Boost region (B) threshold params */
+static int perf_boost_idx;
+
+/* Performance Constraint region (C) threshold params */
+static int perf_constrain_idx;
+
+/**
+ * Performance-Energy (P-E) Space thresholds constants
+ */
+struct threshold_params {
+	int nrg_gain;
+	int cap_gain;
+};
+
+/*
+ * System specific P-E space thresholds constants
+ */
+static struct threshold_params
+threshold_gains[] = {
+	{ 0, 5 }, /*   < 10% */
+	{ 1, 5 }, /*   < 20% */
+	{ 2, 5 }, /*   < 30% */
+	{ 3, 5 }, /*   < 40% */
+	{ 4, 5 }, /*   < 50% */
+	{ 5, 4 }, /*   < 60% */
+	{ 5, 3 }, /*   < 70% */
+	{ 5, 2 }, /*   < 80% */
+	{ 5, 1 }, /*   < 90% */
+	{ 5, 0 }  /* <= 100% */
+};
+
+static int
+__schedtune_accept_deltas(int nrg_delta, int cap_delta,
+			  int perf_boost_idx, int perf_constrain_idx)
+{
+	int payoff = -INT_MAX;
+	int gain_idx = -1;
+
+	/* Performance Boost (B) region */
+	if (nrg_delta >= 0 && cap_delta > 0)
+		gain_idx = perf_boost_idx;
+	/* Performance Constraint (C) region */
+	else if (nrg_delta < 0 && cap_delta <= 0)
+		gain_idx = perf_constrain_idx;
+
+	/* Default: reject schedule candidate */
+	if (gain_idx == -1)
+		return payoff;
+
+	/*
+	 * Evaluate "Performance Boost" vs "Energy Increase"
+	 *
+	 * - Performance Boost (B) region
+	 *
+	 *   Condition: nrg_delta > 0 && cap_delta > 0
+	 *   Payoff criteria:
+	 *     cap_gain / nrg_gain  < cap_delta / nrg_delta =
+	 *     cap_gain * nrg_delta < cap_delta * nrg_gain
+	 *   Note that since both nrg_gain and nrg_delta are positive, the
+	 *   inequality does not change. Thus:
+	 *
+	 *     payoff = (cap_delta * nrg_gain) - (cap_gain * nrg_delta)
+	 *
+	 * - Performance Constraint (C) region
+	 *
+	 *   Condition: nrg_delta < 0 && cap_delta < 0
+	 *   payoff criteria:
+	 *     cap_gain / nrg_gain  > cap_delta / nrg_delta =
+	 *     cap_gain * nrg_delta < cap_delta * nrg_gain
+	 *   Note that since nrg_gain > 0 while nrg_delta < 0, the
+	 *   inequality change. Thus:
+	 *
+	 *     payoff = (cap_delta * nrg_gain) - (cap_gain * nrg_delta)
+	 *
+	 * This means that, in case of same positive defined {cap,nrg}_gain
+	 * for both the B and C regions, we can use the same payoff formula
+	 * where a positive value represents the accept condition.
+	 */
+	payoff  = cap_delta * threshold_gains[gain_idx].nrg_gain;
+	payoff -= nrg_delta * threshold_gains[gain_idx].cap_gain;
+
+	return payoff;
+}
+
 #ifdef CONFIG_CGROUP_SCHEDTUNE
 
 /*
@@ -52,6 +147,15 @@ struct schedtune {
 	bool colocate_update_disabled;
 #endif
 
+	/* Performance Boost (B) region threshold params */
+	int perf_boost_idx;
+
+	/* Performance Constraint (C) region threshold params */
+	int perf_constrain_idx;
+
+	/* Hint to bias scheduling of tasks on that SchedTune CGroup
+	 * towards idle CPUs */
+	int prefer_idle;
 };
 
 static inline struct schedtune *css_st(struct cgroup_subsys_state *css)
@@ -88,8 +192,42 @@ root_schedtune = {
 	.colocate = false,
 	.colocate_update_disabled = false,
 #endif
+	.perf_boost_idx = 0,
+	.perf_constrain_idx = 0,
+	.prefer_idle = 0,
 };
 
+int
+schedtune_accept_deltas(int nrg_delta, int cap_delta,
+			struct task_struct *task)
+{
+	struct schedtune *ct;
+	int perf_boost_idx;
+	int perf_constrain_idx;
+
+	/* Optimal (O) region */
+	if (nrg_delta < 0 && cap_delta > 0) {
+		trace_sched_tune_filter(nrg_delta, cap_delta, 0, 0, 1, 0);
+		return INT_MAX;
+	}
+
+	/* Suboptimal (S) region */
+	if (nrg_delta > 0 && cap_delta < 0) {
+		trace_sched_tune_filter(nrg_delta, cap_delta, 0, 0, -1, 5);
+		return -INT_MAX;
+	}
+
+	/* Get task specific perf Boost/Constraints indexes */
+	rcu_read_lock();
+	ct = task_schedtune(task);
+	perf_boost_idx = ct->perf_boost_idx;
+	perf_constrain_idx = ct->perf_constrain_idx;
+	rcu_read_unlock();
+
+	return __schedtune_accept_deltas(nrg_delta, cap_delta,
+			perf_boost_idx, perf_constrain_idx);
+}
+
 /*
  * Maximum number of boost groups to support
  * When per-task boosting is used we still allow only limited number of
@@ -119,13 +257,16 @@ static struct schedtune *allocated_group[BOOSTGROUPS_COUNT] = {
  */
 struct boost_groups {
 	/* Maximum boost value for all RUNNABLE tasks on a CPU */
-	unsigned boost_max;
+	bool idle;
+	int boost_max;
 	struct {
 		/* The boost for tasks on that boost group */
-		unsigned boost;
+		int boost;
 		/* Count of RUNNABLE tasks on that boost group */
 		unsigned tasks;
 	} group[BOOSTGROUPS_COUNT];
+	/* CPU's boost group locking */
+	raw_spinlock_t lock;
 };
 
 /* Boost groups affecting each CPU in the system */
@@ -246,7 +387,342 @@ static inline void init_sched_boost(struct schedtune *st) { }
 
 #endif /* CONFIG_SCHED_HMP */
 
+static void
+schedtune_cpu_update(int cpu)
+{
+	struct boost_groups *bg;
+	int boost_max;
+	int idx;
+
+	bg = &per_cpu(cpu_boost_groups, cpu);
+
+	/* The root boost group is always active */
+	boost_max = bg->group[0].boost;
+	for (idx = 1; idx < BOOSTGROUPS_COUNT; ++idx) {
+		/*
+		 * A boost group affects a CPU only if it has
+		 * RUNNABLE tasks on that CPU
+		 */
+		if (bg->group[idx].tasks == 0)
+			continue;
+
+		boost_max = max(boost_max, bg->group[idx].boost);
+	}
+	/* Ensures boost_max is non-negative when all cgroup boost values
+	 * are neagtive. Avoids under-accounting of cpu capacity which may cause
+	 * task stacking and frequency spikes.*/
+	boost_max = max(boost_max, 0);
+	bg->boost_max = boost_max;
+}
+
+static int
+schedtune_boostgroup_update(int idx, int boost)
+{
+	struct boost_groups *bg;
+	int cur_boost_max;
+	int old_boost;
+	int cpu;
+
+	/* Update per CPU boost groups */
+	for_each_possible_cpu(cpu) {
+		bg = &per_cpu(cpu_boost_groups, cpu);
+
+		/*
+		 * Keep track of current boost values to compute the per CPU
+		 * maximum only when it has been affected by the new value of
+		 * the updated boost group
+		 */
+		cur_boost_max = bg->boost_max;
+		old_boost = bg->group[idx].boost;
+
+		/* Update the boost value of this boost group */
+		bg->group[idx].boost = boost;
+
+		/* Check if this update increase current max */
+		if (boost > cur_boost_max && bg->group[idx].tasks) {
+			bg->boost_max = boost;
+			trace_sched_tune_boostgroup_update(cpu, 1, bg->boost_max);
+			continue;
+		}
+
+		/* Check if this update has decreased current max */
+		if (cur_boost_max == old_boost && old_boost > boost) {
+			schedtune_cpu_update(cpu);
+			trace_sched_tune_boostgroup_update(cpu, -1, bg->boost_max);
+			continue;
+		}
+
+		trace_sched_tune_boostgroup_update(cpu, 0, bg->boost_max);
+	}
+
+	return 0;
+}
+
+#define ENQUEUE_TASK  1
+#define DEQUEUE_TASK -1
+
+static inline void
+schedtune_tasks_update(struct task_struct *p, int cpu, int idx, int task_count)
+{
+	struct boost_groups *bg = &per_cpu(cpu_boost_groups, cpu);
+	int tasks = bg->group[idx].tasks + task_count;
+
+	/* Update boosted tasks count while avoiding to make it negative */
+	bg->group[idx].tasks = max(0, tasks);
+
+	trace_sched_tune_tasks_update(p, cpu, tasks, idx,
+			bg->group[idx].boost, bg->boost_max);
+
+	/* Boost group activation or deactivation on that RQ */
+	if (tasks == 1 || tasks == 0)
+		schedtune_cpu_update(cpu);
+}
+
+/*
+ * NOTE: This function must be called while holding the lock on the CPU RQ
+ */
+void schedtune_enqueue_task(struct task_struct *p, int cpu)
+{
+	struct boost_groups *bg = &per_cpu(cpu_boost_groups, cpu);
+	unsigned long irq_flags;
+	struct schedtune *st;
+	int idx;
+
+	if (!unlikely(schedtune_initialized))
+		return;
+
+	/*
+	 * When a task is marked PF_EXITING by do_exit() it's going to be
+	 * dequeued and enqueued multiple times in the exit path.
+	 * Thus we avoid any further update, since we do not want to change
+	 * CPU boosting while the task is exiting.
+	 */
+	if (p->flags & PF_EXITING)
+		return;
+
+	/*
+	 * Boost group accouting is protected by a per-cpu lock and requires
+	 * interrupt to be disabled to avoid race conditions for example on
+	 * do_exit()::cgroup_exit() and task migration.
+	 */
+	raw_spin_lock_irqsave(&bg->lock, irq_flags);
+	rcu_read_lock();
+
+	st = task_schedtune(p);
+	idx = st->idx;
+
+	schedtune_tasks_update(p, cpu, idx, ENQUEUE_TASK);
+
+	rcu_read_unlock();
+	raw_spin_unlock_irqrestore(&bg->lock, irq_flags);
+}
+
+int schedtune_allow_attach(struct cgroup_taskset *tset)
+{
+	/* We always allows tasks to be moved between existing CGroups */
+	return 0;
+}
+
+int schedtune_can_attach(struct cgroup_taskset *tset)
+{
+	struct task_struct *task;
+	struct cgroup_subsys_state *css;
+	struct boost_groups *bg;
+	unsigned long irq_flags;
+	unsigned int cpu;
+	struct rq *rq;
+	int src_bg; /* Source boost group index */
+	int dst_bg; /* Destination boost group index */
+	int tasks;
+
+	if (!unlikely(schedtune_initialized))
+		return 0;
+
+
+	cgroup_taskset_for_each(task, css, tset) {
+
+		/*
+		 * Lock the CPU's RQ the task is enqueued to avoid race
+		 * conditions with migration code while the task is being
+		 * accounted
+		 */
+		rq = lock_rq_of(task, &irq_flags);
+
+		if (!task->on_rq) {
+			unlock_rq_of(rq, task, &irq_flags);
+			continue;
+		}
+
+		/*
+		 * Boost group accouting is protected by a per-cpu lock and requires
+		 * interrupt to be disabled to avoid race conditions on...
+		 */
+		cpu = cpu_of(rq);
+		bg = &per_cpu(cpu_boost_groups, cpu);
+		raw_spin_lock(&bg->lock);
+
+		dst_bg = css_st(css)->idx;
+		src_bg = task_schedtune(task)->idx;
+
+		/*
+		 * Current task is not changing boostgroup, which can
+		 * happen when the new hierarchy is in use.
+		 */
+		if (unlikely(dst_bg == src_bg)) {
+			raw_spin_unlock(&bg->lock);
+			unlock_rq_of(rq, task, &irq_flags);
+			continue;
+		}
+
+		/*
+		 * This is the case of a RUNNABLE task which is switching its
+		 * current boost group.
+		 */
+
+		/* Move task from src to dst boost group */
+		tasks = bg->group[src_bg].tasks - 1;
+		bg->group[src_bg].tasks = max(0, tasks);
+		bg->group[dst_bg].tasks += 1;
+
+		raw_spin_unlock(&bg->lock);
+		unlock_rq_of(rq, task, &irq_flags);
+
+		/* Update CPU boost group */
+		if (bg->group[src_bg].tasks == 0 || bg->group[dst_bg].tasks == 1)
+			schedtune_cpu_update(task_cpu(task));
+
+	}
+
+	return 0;
+}
+
+void schedtune_cancel_attach(struct cgroup_taskset *tset)
+{
+	/* This can happen only if SchedTune controller is mounted with
+	 * other hierarchies ane one of them fails. Since usually SchedTune is
+	 * mouted on its own hierarcy, for the time being we do not implement
+	 * a proper rollback mechanism */
+	WARN(1, "SchedTune cancel attach not implemented");
+}
+
+/*
+ * NOTE: This function must be called while holding the lock on the CPU RQ
+ */
+void schedtune_dequeue_task(struct task_struct *p, int cpu)
+{
+	struct boost_groups *bg = &per_cpu(cpu_boost_groups, cpu);
+	unsigned long irq_flags;
+	struct schedtune *st;
+	int idx;
+
+	if (!unlikely(schedtune_initialized))
+		return;
+
+	/*
+	 * When a task is marked PF_EXITING by do_exit() it's going to be
+	 * dequeued and enqueued multiple times in the exit path.
+	 * Thus we avoid any further update, since we do not want to change
+	 * CPU boosting while the task is exiting.
+	 * The last dequeue is already enforce by the do_exit() code path
+	 * via schedtune_exit_task().
+	 */
+	if (p->flags & PF_EXITING)
+		return;
+
+	/*
+	 * Boost group accouting is protected by a per-cpu lock and requires
+	 * interrupt to be disabled to avoid race conditions on...
+	 */
+	raw_spin_lock_irqsave(&bg->lock, irq_flags);
+	rcu_read_lock();
+
+	st = task_schedtune(p);
+	idx = st->idx;
+
+	schedtune_tasks_update(p, cpu, idx, DEQUEUE_TASK);
+
+	rcu_read_unlock();
+	raw_spin_unlock_irqrestore(&bg->lock, irq_flags);
+}
+
+void schedtune_exit_task(struct task_struct *tsk)
+{
+	struct schedtune *st;
+	unsigned long irq_flags;
+	unsigned int cpu;
+	struct rq *rq;
+	int idx;
+
+	if (!unlikely(schedtune_initialized))
+		return;
+
+	rq = lock_rq_of(tsk, &irq_flags);
+	rcu_read_lock();
+
+	cpu = cpu_of(rq);
+	st = task_schedtune(tsk);
+	idx = st->idx;
+	schedtune_tasks_update(tsk, cpu, idx, DEQUEUE_TASK);
+
+	rcu_read_unlock();
+	unlock_rq_of(rq, tsk, &irq_flags);
+}
+
+int schedtune_cpu_boost(int cpu)
+{
+	struct boost_groups *bg;
+
+	bg = &per_cpu(cpu_boost_groups, cpu);
+	return bg->boost_max;
+}
+
+int schedtune_task_boost(struct task_struct *p)
+{
+	struct schedtune *st;
+	int task_boost;
+
+	/* Get task boost value */
+	rcu_read_lock();
+	st = task_schedtune(p);
+	task_boost = st->boost;
+	rcu_read_unlock();
+
+	return task_boost;
+}
+
+int schedtune_prefer_idle(struct task_struct *p)
+{
+	struct schedtune *st;
+	int prefer_idle;
+
+	/* Get prefer_idle value */
+	rcu_read_lock();
+	st = task_schedtune(p);
+	prefer_idle = st->prefer_idle;
+	rcu_read_unlock();
+
+	return prefer_idle;
+}
+
 static u64
+prefer_idle_read(struct cgroup_subsys_state *css, struct cftype *cft)
+{
+	struct schedtune *st = css_st(css);
+
+	return st->prefer_idle;
+}
+
+static int
+prefer_idle_write(struct cgroup_subsys_state *css, struct cftype *cft,
+	    u64 prefer_idle)
+{
+	struct schedtune *st = css_st(css);
+	st->prefer_idle = prefer_idle;
+
+	return 0;
+}
+
+static s64
 boost_read(struct cgroup_subsys_state *css, struct cftype *cft)
 {
 	struct schedtune *st = css_st(css);
@@ -256,16 +732,37 @@ boost_read(struct cgroup_subsys_state *css, struct cftype *cft)
 
 static int
 boost_write(struct cgroup_subsys_state *css, struct cftype *cft,
-	    u64 boost)
+	    s64 boost)
 {
 	struct schedtune *st = css_st(css);
+	unsigned threshold_idx;
+	int boost_pct;
 
-	if (boost < 0 || boost > 100)
+	if (boost < -100 || boost > 100)
 		return -EINVAL;
+	boost_pct = boost;
+
+	/*
+	 * Update threshold params for Performance Boost (B)
+	 * and Performance Constraint (C) regions.
+	 * The current implementatio uses the same cuts for both
+	 * B and C regions.
+	 */
+	threshold_idx = clamp(boost_pct, 0, 99) / 10;
+	st->perf_boost_idx = threshold_idx;
+	st->perf_constrain_idx = threshold_idx;
 
 	st->boost = boost;
-	if (css == &root_schedtune.css)
+	if (css == &root_schedtune.css) {
 		sysctl_sched_cfs_boost = boost;
+		perf_boost_idx  = threshold_idx;
+		perf_constrain_idx  = threshold_idx;
+	}
+
+	/* Update CPU boost */
+	schedtune_boostgroup_update(st->idx, st->boost);
+
+	trace_sched_tune_config(st->boost);
 
 	return 0;
 }
@@ -289,8 +786,13 @@ static void schedtune_attach(struct cgroup_taskset *tset)
 static struct cftype files[] = {
 	{
 		.name = "boost",
-		.read_u64 = boost_read,
-		.write_u64 = boost_write,
+		.read_s64 = boost_read,
+		.write_s64 = boost_write,
+	},
+	{
+		.name = "prefer_idle",
+		.read_u64 = prefer_idle_read,
+		.write_u64 = prefer_idle_write,
 	},
 #ifdef CONFIG_SCHED_HMP
 	{
@@ -315,26 +817,19 @@ static struct cftype files[] = {
 static int
 schedtune_boostgroup_init(struct schedtune *st)
 {
-	/* Keep track of allocated boost groups */
-	allocated_group[st->idx] = st;
-
-	return 0;
-}
-
-static int
-schedtune_init(void)
-{
 	struct boost_groups *bg;
 	int cpu;
 
+	/* Keep track of allocated boost groups */
+	allocated_group[st->idx] = st;
+
 	/* Initialize the per CPU boost groups */
 	for_each_possible_cpu(cpu) {
 		bg = &per_cpu(cpu_boost_groups, cpu);
-		memset(bg, 0, sizeof(struct boost_groups));
+		bg->group[st->idx].boost = 0;
+		bg->group[st->idx].tasks = 0;
 	}
 
-	pr_info("  schedtune configured to support %d boost groups\n",
-		BOOSTGROUPS_COUNT);
 	return 0;
 }
 
@@ -344,10 +839,8 @@ schedtune_css_alloc(struct cgroup_subsys_state *parent_css)
 	struct schedtune *st;
 	int idx;
 
-	if (!parent_css) {
-		schedtune_init();
+	if (!parent_css)
 		return &root_schedtune.css;
-	}
 
 	/* Allow only single level hierachies */
 	if (parent_css != &root_schedtune.css) {
@@ -386,6 +879,9 @@ out:
 static void
 schedtune_boostgroup_release(struct schedtune *st)
 {
+	/* Reset this boost group */
+	schedtune_boostgroup_update(st->idx, 0);
+
 	/* Keep track of allocated boost groups */
 	allocated_group[st->idx] = NULL;
 }
@@ -402,12 +898,55 @@ schedtune_css_free(struct cgroup_subsys_state *css)
 struct cgroup_subsys schedtune_cgrp_subsys = {
 	.css_alloc	= schedtune_css_alloc,
 	.css_free	= schedtune_css_free,
+	.allow_attach   = schedtune_allow_attach,
+	.can_attach     = schedtune_can_attach,
+	.cancel_attach  = schedtune_cancel_attach,
 	.legacy_cftypes	= files,
 	.early_init	= 1,
 	.allow_attach	= subsys_cgroup_allow_attach,
 	.attach		= schedtune_attach,
 };
 
+static inline void
+schedtune_init_cgroups(void)
+{
+	struct boost_groups *bg;
+	int cpu;
+
+	/* Initialize the per CPU boost groups */
+	for_each_possible_cpu(cpu) {
+		bg = &per_cpu(cpu_boost_groups, cpu);
+		memset(bg, 0, sizeof(struct boost_groups));
+	}
+
+	pr_info("schedtune: configured to support %d boost groups\n",
+		BOOSTGROUPS_COUNT);
+
+	schedtune_initialized = true;
+}
+
+#else /* CONFIG_CGROUP_SCHEDTUNE */
+
+int
+schedtune_accept_deltas(int nrg_delta, int cap_delta,
+			struct task_struct *task)
+{
+	/* Optimal (O) region */
+	if (nrg_delta < 0 && cap_delta > 0) {
+		trace_sched_tune_filter(nrg_delta, cap_delta, 0, 0, 1, 0);
+		return INT_MAX;
+	}
+
+	/* Suboptimal (S) region */
+	if (nrg_delta > 0 && cap_delta < 0) {
+		trace_sched_tune_filter(nrg_delta, cap_delta, 0, 0, -1, 5);
+		return -INT_MAX;
+	}
+
+	return __schedtune_accept_deltas(nrg_delta, cap_delta,
+			perf_boost_idx, perf_constrain_idx);
+}
+
 #endif /* CONFIG_CGROUP_SCHEDTUNE */
 
 int
@@ -416,10 +955,183 @@ sysctl_sched_cfs_boost_handler(struct ctl_table *table, int write,
 			       loff_t *ppos)
 {
 	int ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
+	unsigned threshold_idx;
+	int boost_pct;
 
 	if (ret || !write)
 		return ret;
 
+	if (sysctl_sched_cfs_boost < -100 || sysctl_sched_cfs_boost > 100)
+		return -EINVAL;
+	boost_pct = sysctl_sched_cfs_boost;
+
+	/*
+	 * Update threshold params for Performance Boost (B)
+	 * and Performance Constraint (C) regions.
+	 * The current implementatio uses the same cuts for both
+	 * B and C regions.
+	 */
+	threshold_idx = clamp(boost_pct, 0, 99) / 10;
+	perf_boost_idx = threshold_idx;
+	perf_constrain_idx = threshold_idx;
+
 	return 0;
 }
 
+#ifdef CONFIG_SCHED_DEBUG
+static void
+schedtune_test_nrg(unsigned long delta_pwr)
+{
+	unsigned long test_delta_pwr;
+	unsigned long test_norm_pwr;
+	int idx;
+
+	/*
+	 * Check normalization constants using some constant system
+	 * energy values
+	 */
+	pr_info("schedtune: verify normalization constants...\n");
+	for (idx = 0; idx < 6; ++idx) {
+		test_delta_pwr = delta_pwr >> idx;
+
+		/* Normalize on max energy for target platform */
+		test_norm_pwr = reciprocal_divide(
+					test_delta_pwr << SCHED_LOAD_SHIFT,
+					schedtune_target_nrg.rdiv);
+
+		pr_info("schedtune: max_pwr/2^%d: %4lu => norm_pwr: %5lu\n",
+			idx, test_delta_pwr, test_norm_pwr);
+	}
+}
+#else
+#define schedtune_test_nrg(delta_pwr)
+#endif
+
+/*
+ * Compute the min/max power consumption of a cluster and all its CPUs
+ */
+static void
+schedtune_add_cluster_nrg(
+		struct sched_domain *sd,
+		struct sched_group *sg,
+		struct target_nrg *ste)
+{
+	struct sched_domain *sd2;
+	struct sched_group *sg2;
+
+	struct cpumask *cluster_cpus;
+	char str[32];
+
+	unsigned long min_pwr;
+	unsigned long max_pwr;
+	int cpu;
+
+	/* Get Cluster energy using EM data for the first CPU */
+	cluster_cpus = sched_group_cpus(sg);
+	snprintf(str, 32, "CLUSTER[%*pbl]",
+		 cpumask_pr_args(cluster_cpus));
+
+	min_pwr = sg->sge->idle_states[sg->sge->nr_idle_states - 1].power;
+	max_pwr = sg->sge->cap_states[sg->sge->nr_cap_states - 1].power;
+	pr_info("schedtune: %-17s min_pwr: %5lu max_pwr: %5lu\n",
+		str, min_pwr, max_pwr);
+
+	/*
+	 * Keep track of this cluster's energy in the computation of the
+	 * overall system energy
+	 */
+	ste->min_power += min_pwr;
+	ste->max_power += max_pwr;
+
+	/* Get CPU energy using EM data for each CPU in the group */
+	for_each_cpu(cpu, cluster_cpus) {
+		/* Get a SD view for the specific CPU */
+		for_each_domain(cpu, sd2) {
+			/* Get the CPU group */
+			sg2 = sd2->groups;
+			min_pwr = sg2->sge->idle_states[sg2->sge->nr_idle_states - 1].power;
+			max_pwr = sg2->sge->cap_states[sg2->sge->nr_cap_states - 1].power;
+
+			ste->min_power += min_pwr;
+			ste->max_power += max_pwr;
+
+			snprintf(str, 32, "CPU[%d]", cpu);
+			pr_info("schedtune: %-17s min_pwr: %5lu max_pwr: %5lu\n",
+				str, min_pwr, max_pwr);
+
+			/*
+			 * Assume we have EM data only at the CPU and
+			 * the upper CLUSTER level
+			 */
+			BUG_ON(!cpumask_equal(
+				sched_group_cpus(sg),
+				sched_group_cpus(sd2->parent->groups)
+				));
+			break;
+		}
+	}
+}
+
+/*
+ * Initialize the constants required to compute normalized energy.
+ * The values of these constants depends on the EM data for the specific
+ * target system and topology.
+ * Thus, this function is expected to be called by the code
+ * that bind the EM to the topology information.
+ */
+static int
+schedtune_init(void)
+{
+	struct target_nrg *ste = &schedtune_target_nrg;
+	unsigned long delta_pwr = 0;
+	struct sched_domain *sd;
+	struct sched_group *sg;
+
+	pr_info("schedtune: init normalization constants...\n");
+	ste->max_power = 0;
+	ste->min_power = 0;
+
+	rcu_read_lock();
+
+	/*
+	 * When EAS is in use, we always have a pointer to the highest SD
+	 * which provides EM data.
+	 */
+	sd = rcu_dereference(per_cpu(sd_ea, cpumask_first(cpu_online_mask)));
+	if (!sd) {
+		if (energy_aware())
+			pr_warn("schedtune: no energy model data\n");
+		goto nodata;
+	}
+
+	sg = sd->groups;
+	do {
+		schedtune_add_cluster_nrg(sd, sg, ste);
+	} while (sg = sg->next, sg != sd->groups);
+
+	rcu_read_unlock();
+
+	pr_info("schedtune: %-17s min_pwr: %5lu max_pwr: %5lu\n",
+		"SYSTEM", ste->min_power, ste->max_power);
+
+	/* Compute normalization constants */
+	delta_pwr = ste->max_power - ste->min_power;
+	ste->rdiv = reciprocal_value(delta_pwr);
+	pr_info("schedtune: using normalization constants mul: %u sh1: %u sh2: %u\n",
+		ste->rdiv.m, ste->rdiv.sh1, ste->rdiv.sh2);
+
+	schedtune_test_nrg(delta_pwr);
+
+#ifdef CONFIG_CGROUP_SCHEDTUNE
+	schedtune_init_cgroups();
+#else
+	pr_info("schedtune: configured to support global boosting only\n");
+#endif
+
+	return 0;
+
+nodata:
+	rcu_read_unlock();
+	return -EINVAL;
+}
+postcore_initcall(schedtune_init);