• mm: Don't warn on page allocation failures for OOM-killed processes • Readahead: Optimize divide/multiply by power of 2 using L/R shift - Fix • Readahead: Optimize divide/multiply by power of 2 using L/R shift • mm: Stop kswapd early when nothing's waiting for it to free pages • mm: vmpressure: Don't cache the window size • mm: vmpressure: Interpret zero scanned pages as 100% pressure • mm: vmpressure: Don't exclude any allocation types • mm: vmpressure: Make local functions and global variables static • mm: skip swap readahead when process is exiting • mm: release the spinlock on zap_pte_range • mm/vmalloc.c: move 'area->pages' after if statement • mm: mmap: increase sockets maximum memory size pgoff for 32bits • mm, memcg: use consistent gfp flags during readahead • mm/slab_common: Align all caches' objects to hardware cachelines • Revert "mm: export add_swap_extent()" • mm: allow __GFP_MOVABLE to use CMA memory • base: remove default blocking of qcom_rx_wakelock and update • md: disable dm-verity • ASoC: msm-cpe: Fix msm_cpe_lsm_ioctl_compat error handling • scsi: ufs: disable auto hibern8 • msm: sps: Fix uninitialized result usage when an invalid IRQ is found • media: v4l2-ioctl: Use a larger on-stack video copy buffer • ALSA: control_compat: Don't dynamically allocate single-use structs • ALSA: control: Don't dynamically allocate a single struct instance • drivers: power: bcl: queue work on system_power_efficient_wq • mm: Increase ratelimit pages value • sched/core: nr_migrate = 128 increases number of tasks to iterate in a single balance run. • sched/fair: Make scheduling period more efficient • soc/qcom: move icnss_initialize to async probe • mm: vmstat: use power efficient workingqueues • platform: ipa: use power efficient workingqueues • workqueue: queue everything on system's power efficient wq • power: qpnp-fg-gen3: queue work on system_power_efficient_wq • drivers: mdss: queue work on system_power_efficient_wq • thermal: msm-tsens: Reduce workqueue priority • thermal: qpnp-adc-tm: Remove workqueue high-priority flag • BACKPORT: thermal: core: using power_efficient_wq for thermal worker • block: queue work on power efficient wq • sched: Enable Dynamic Schedtune Boost By Default • SIO: update for linux 3.18 • lowmemorykiller: Partial revert settings • block: fiops: fix build after merge f67e041a2899b60e3fa786736e9d0f28a2869668 • block: bfq: fix build after merge f67e041a2899b60e3fa786736e9d0f28a2869668 • defconfig: Enable I/O Scheduler BFQ,SIO,FIOPS & Set BFQ as default • block: modify fiops • FIOPS: update for linux 3.18 • Add SIO and FIOPS i/o schedulers • block: do not merge requests without consulting with io scheduler • block: bfq: Update backing_dev_info to be a pointer • block, bfq: update to latest bfq-v8-v4.4 state • block, bfq: add Early Queue Merge (EQM) to BFQ-v7r11 for 4.4.0 • block: introduce the BFQ-v7r11 I/O sched for 4.4.0 • block: cgroups, kconfig, build bits for BFQ-v7r11-4.4.0 • block: disable iostats & set permission read-only • cpufreq: create permanent "interactive" tunables for any real cpu governor used • sched: add non-EAS, HMP specific scheduler dummy tunables • devfreq_boost: fix error: expected '}' • devfreq_boost: Add DEVFREQ_GPUBW and Update Description • kernel: Add DEVFREQ_GPUBW For Kprofiles • defconfig: Decrease Freq CPUBW and Add GPUBW Boost • kernel: Fix identifier cpu_input_boost • defconfig: Enable CPU_INPUT_BOOST and Calculate DEVFREQ_BOOST_CPUBW • cpu_input_boost: Expose compile-time settings as module parameters • cpu_input_boost: Mark boost kthread as performance critical • cpu_input_boost: Introduce driver for event-based CPU boosting • defconfig: Added KPROFILES • kernel: Update For Kprofiles • Adapt Kernel Profiles • Add 'drivers/misc/kprofiles/' from commit '52df4eb2b8f1202ff74d3b5257c0d06946a165aa' • msm_cci: fix inconsistent mutex_lock • qdsp6v2: fix inconsistent mutex_lock • wcd9335: fix inconsistent mutex_lock • msm: qdsp6v2: fix inconsistent spin_lock • slimbus: fix inconsistent mutex_lock • msm: mdss: fix inconsistent mutex_lock • msm: mdss: fix inconsistent mutex_lock • msm: mdss: fix inconsistent mutex_lock • power: smb-lib: use USB options in low ICL range • qcom: msm-core: uninterruptible wait - you can kiss my arse goodbye • thermal-core: Fix invalid sensor data being accepted • ARM: dts: {msm8998,sdm660}: increase linux,cma from 32M to 128M • ARM: dts: sdm660-regulator: Add cooling devices for VDD_MX and VDD_CX • mm: Always overcommit memory by default • mm/slub.c: branch optimization in free slowpath • staging: sync: Add a function to check if a fence has been signaled • fix derp • msm: kgsl: Actually free the sharedmem when fence is expired • drivers: gpu: msm a5xx: Remove overwriting A5XX_PC_DBG_ECO_CNTL register • drivers: gpu: drm: msm_smmu: Specify sync probe for msmdrm_smmu • msm: kgsl: Reduce latency while processing ioctls • msm: kgsl: Don't try to wait for fences that have been signaled • adreno/kgsl: remove debug fs support completely • msm: kgsl: Don't allocate memory dynamically for drawobj sync structs • Revert "msm: kgsl: Add a sync fence timeout callback" • BACKPORT: msm: kgsl: Keep dispatcher halted during SUSPEND state • Revert "msm: kgsl: Disallow L2PC during wake up from SLUMBER" • msm: kgsl: Dispatch commands using a master kthread • Revert "msm: kgsl: Defer issue commands to worker thread" • msm: kgsl: Relax adreno spin idle tight loop • msm: kgsl: Avoid dynamically allocating small command buffers • msm: kgsl: Don't allocate memory dynamically for temp command buffers • ARM: dts: sdm660-gpu: enable unmap fast support • adreno: enable fake gpu after 1st boot v2 • adreno: new interface to fake gpu model after boot • kgsl: add an api to query adreno kgsl resume/suspend states • msm: kgsl: Add support to make use of iommu_unmap_fast • iommu: Introduce Interface for IOMMU TLB Flushing • mdss,adreno: move to async probe • msm: kgsl: Use interruptible waits • dts: qcom: delete identifier #include simple-thermal.dtsi • qpnp-fg-gen3: 1000mA minimum charging <= 90% • msm: kgsl: Relax CPU latency requirements to save power • gpu: msm: Fix derp • gpu: msm: Compile with -funroll-loops optimization • msm: kgsl: Report correct GPU frequency for clock_mhz • gpu: msm: Disable adreno driver logging • ARM64: dts/qcom: Disable CoreSight DT entries for sdm660 • pm660l dtsi: use analog dimming • dts: fix pa_therm0 temperature reading • drivers: gpu: remove unused POPP code POPP is now removed, remove this useless code to fix error while compiling • msm: kgsl: Remove POPP • mm: Boost when memory pressure becomes high • dtc: Silence warnings • dtc: update warning settings for new bus and node/property name checks • scripts/Makefile.lib: speed up build process • scripts/dtc: Update to upstream version 1.4.4-Android-build • Revert "adreno: disable snapshot, coresight and trace" • adreno: disable snapshot, coresight and trace • drivers: gpu: Increase adreno bus data limit • gpu: msm: adreno_dispatch: increase maximum ringbuffer inflight • msm: kgsl: Increase worker thread priority • msm: kgsl: Wake GPU upon receiving an ioctl rather than upon touch input • msm: kgsl: Mark IRQ and worker thread as performance critical • msm: mdss: Don't cache the address of a stack variable in timings init • msm: mdss: Mark IRQ and important kthreads as performance critical • msm: mdss: Don't add event timer for unused autorefresh feature • msm-thermal: omit unset nodes • thermal: qpnp-adc-tm: Fix aborting suspend • msm_thermal: ignore sensor -19 • devfreq: msm_adreno_tz: Decrease busy time ceiling • kernel: Boost DDR bus for a short amount of time when zygote forks • msm: mdss: Boost DDR bus when committing a new frame • kthread: Make queuing_blocked() available in include/linux/kthread.h • cpu-boost: Boost priority of notifier_block • cpu-boost: Adapt kthread on msm-4.4 • cpu-boost: Rework scheduling setup • cpu-boost: Hardcode as top-app only for DSB • cpu-boost: Update functions for newer Dynamic Schedtune Boost changes • cpu-boost: Reset Dynamic SchedTune Boost only if it is currently active • cpu-boost: Implement Dynamic SchedTune Boost v3