Skip to content
Permalink
Branch: pie
Commits on Sep 19, 2019
  1. msm: kgsl: remove duplicate A616 entries

    okitavera committed Sep 19, 2019
    am I drunk or something ? damn
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  2. defconfig: grus: enable minidump support

    okitavera committed Sep 19, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  3. input: goodix_ts: reinstantiate status under cfg process

    okitavera committed Sep 18, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  4. ARM: dts: sdm670: also align cluster to the cpu costs

    okitavera committed Sep 18, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 18, 2019
  1. Merge tag 'v4.9.193' of https://git.kernel.org/pub/scm/linux/kernel/g…

    okitavera committed Sep 18, 2019
    …it/stable/linux-stable
    
    This is the 4.9.193 stable release
    
    * tag 'v4.9.193' of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable:
      Linux 4.9.193
      vhost: make sure log_num < in_num
      clk: s2mps11: Add used attribute to s2mps11_dt_match
      scripts/decode_stacktrace: match basepath using shell prefix operator, not regex
      batman-adv: Only read OGM tvlv_len after buffer len check
      batman-adv: fix uninit-value in batadv_netlink_get_ifindex()
      vhost/test: fix build for vhost test
      ip6: fix skb leak in ip6frag_expire_frag_queue()
      xfrm: clean up xfrm protocol checks
      powerpc/tm: Fix FP/VMX unavailable exceptions inside a transaction
      drm/vmwgfx: Fix double free in vmw_recv_msg()
      sched/fair: Don't assign runtime for throttled cfs_rq
      ALSA: hda/realtek - Fix overridden device-specific initialization
      ALSA: hda - Fix potential endless loop at applying quirks
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 14, 2019
  1. defconfig: grus: enable and configure iasb

    okitavera committed Sep 13, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  2. cgroups: invoke iasb with 1sec boost when top-app changed

    okitavera committed Sep 13, 2019
    based on kerneltoast's commits :
    kernel: Boost whenever a zygote-forked process becomes a top app
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  3. kernel/sched: Introducing Input-Assisted SchedTune Boost

    okitavera committed Sep 13, 2019
    This is an event-based driver special for Dynamic SchedTune Boost.
    It can boost the schedtune by using the dynamic schedtune boost's API
    with specific duration, values, and conditions.
    
    By default, it will boost when touchscreen is :
    - Detecting a swipe action
    - Detecting a touch action with an "acceptable" duration
    
    It can be accessed from other modules with linux/iasb.h,
    and some options are accessible via sysfs at: /sys/module/iasb/
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  4. treewide: Revert cpu_input_boost and devfreq_boost driver

    okitavera committed Sep 12, 2019
    * Lord Sultan said those are deprecated
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 10, 2019
  1. Merge tag 'v4.9.192' of https://git.kernel.org/pub/scm/linux/kernel/g…

    okitavera committed Sep 10, 2019
    …it/stable/linux-stable
    
    This is the 4.9.192 stable release
    
    * tag 'v4.9.192' of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable:
      Linux 4.9.192
      mld: fix memory leak in mld_del_delrec()
      tcp: inherit timestamp on mtu probe
      net: stmmac: dwmac-rk: Don't fail if phy regulator is absent
      net: fix skb use after free in netpoll
      Revert "x86/apic: Include the LDR when clearing out APIC registers"
      spi: bcm2835aux: fix corruptions for longer spi transfers
      spi: bcm2835aux: remove dangerous uncontrolled read of fifo
      spi: bcm2835aux: unifying code between polling and interrupt driven code
      spi: bcm2835aux: ensure interrupts are enabled for shared handler
      libceph: allow ceph_buffer_put() to receive a NULL ceph_buffer
      KVM: arm/arm64: Only skip MMIO insn once
      ceph: fix buffer free while holding i_ceph_lock in fill_inode()
      ceph: fix buffer free while holding i_ceph_lock in __ceph_setxattr()
      IB/mlx4: Fix memory leaks
      Tools: hv: kvp: eliminate 'may be used uninitialized' warning
      ravb: Fix use-after-free ravb_tstamp_skb
      wimax/i2400m: fix a memory leak bug
      net: kalmia: fix memory leaks
      cx82310_eth: fix a memory leak bug
      net: myri10ge: fix memory leaks
      cxgb4: fix a memory leak bug
      gpio: Fix build error of function redefinition
      ibmveth: Convert multicast list size for little-endian system
      Bluetooth: btqca: Add a short delay before downloading the NVM
      net: tc35815: Explicitly check NET_IP_ALIGN is not zero in tc35815_rx
      net: tundra: tsi108: use spin_lock_irqsave instead of spin_lock_irq in IRQ context
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  2. Revert "writeback: hardcode dirty_expire_centisecs to 30s"

    okitavera committed Sep 10, 2019
    This reverts commit 99bc58d.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  3. cpu_input_boost: check touch interval before boosting

    okitavera committed Sep 10, 2019
    record touch down and continuous touch,
    and then decide whether it will boost or not depends on
    the minimal touch threshold.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  4. cpu_input_boost: move stune boosting into *_boost_bit

    okitavera committed Sep 10, 2019
    set/clear_boost_bit() is always be called anytime when
    boost/unboosting scene is happening. Use it instead.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  5. kernel/cgroups: optimize boosting mechanism

    okitavera committed Sep 9, 2019
    reduce boost time to 500ms, and also simplify task checking
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  6. kernel/fork: increase the boost time for zygote-forking

    okitavera committed Sep 9, 2019
    instead of short-boosting, we just keep it little longer.
    since opening an app is quite heavy and takes some times.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  7. kernel/fork: nuke devfreq boost on zygote-forks scene

    okitavera committed Sep 9, 2019
    boosting the bandwidth is unnecessary, drop it.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 9, 2019
  1. sched/tune: drop hardcoded TA stune boost

    okitavera committed Sep 9, 2019
    This reverts commit e15c52d and 02ec1dc.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 8, 2019
  1. defconfig: grus: enable 32-bit vDSO

    okitavera committed Sep 8, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 7, 2019
  1. drivers: thermal: fix invalid buffer size on boost_store

    okitavera committed Sep 7, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Commits on Sep 6, 2019
  1. Merge tag 'v4.9.191' of https://git.kernel.org/pub/scm/linux/kernel/g…

    okitavera committed Sep 6, 2019
    …it/stable/linux-stable
    
    This is the 4.9.191 stable release
    
    * tag 'v4.9.191' of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable:
      Linux 4.9.191
      mm/zsmalloc.c: fix build when CONFIG_COMPACTION=n
      x86/ptrace: fix up botched merge of spectrev1 fix
      i2c: piix4: Fix port selection for AMD Family 16h Model 30h
      KVM: arm/arm64: vgic-v2: Handle SGI bits in GICD_I{S,C}PENDR0 as WI
      KVM: arm/arm64: vgic: Fix potential deadlock when ap_list is long
      mac80211: fix possible sta leak
      Revert "cfg80211: fix processing world regdomain when non modular"
      VMCI: Release resource if the work is already queued
      stm class: Fix a double free of stm_source_device
      mmc: core: Fix init of SD cards reporting an invalid VDD range
      mmc: sdhci-of-at91: add quirk for broken HS200
      uprobes/x86: Fix detection of 32-bit user mode
      ptrace,x86: Make user_64bit_mode() available to 32-bit builds
      USB: storage: ums-realtek: Whitelist auto-delink support
      USB: storage: ums-realtek: Update module parameter description for auto_delink_en
      usb: host: xhci: rcar: Fix typo in compatible string matching
      usb: host: ohci: fix a race condition between shutdown and irq
      usb: chipidea: udc: don't do hardware access if gadget has stopped
      USB: cdc-wdm: fix race between write and disconnect due to flag abuse
      usb-storage: Add new JMS567 revision to unusual_devs
      mm/zsmalloc.c: fix race condition in zs_destroy_pool
      x86/apic: Include the LDR when clearing out APIC registers
      x86/apic: Do not initialize LDR and DFR for bigsmp
      KVM: x86: Don't update RIP or do single-step on faulting emulation
      ALSA: seq: Fix potential concurrent access to the deleted pool
      ALSA: line6: Fix memory leak at line6_init_pcm() error path
      tcp: make sure EPOLLOUT wont be missed
      ALSA: usb-audio: Fix an OOB bug in parse_audio_mixer_unit
      ALSA: usb-audio: Fix a stack buffer overflow bug in check_input_term
      tcp: fix tcp_rtx_queue_tail in case of empty retransmit queue
      scsi: ufs: Fix RX_TERMINATION_FORCE_ENABLE define value
      watchdog: bcm2835_wdt: Fix module autoload
      tools: hv: fix KVP and VSS daemons exit code
      usb: host: fotg2: restart hcd after port reset
      i2c: emev2: avoid race when unregistering slave client
      xen/blkback: fix memory leaks
      usb: gadget: composite: Clear "suspended" on reset/disconnect
      iommu/dma: Handle SG length overflow better
      dmaengine: ste_dma40: fix unneeded variable warning
      x86/CPU/AMD: Clear RDRAND CPUID bit on AMD family 15h/16h
      Revert "perf test 6: Fix missing kvm module load for s390"
      xfs: fix missing ILOCK unlock when xfs_setattr_nonsize fails due to EDQUOT
      mm/zsmalloc.c: migration can leave pages in ZS_EMPTY indefinitely
      mm, page_owner: handle THP splits correctly
      genirq: Properly pair kobject_del() with kobject_add()
      dm table: fix invalid memory accesses with too high sector number
      dm space map metadata: fix missing store of apply_bops() return value
      dm btree: fix order of block initialization in btree_split_beneath
      x86/boot: Fix boot regression caused by bootparam sanitizing
      x86/boot: Save fields explicitly, zero out everything else
      x86/apic: Handle missing global clockevent gracefully
      x86/retpoline: Don't clobber RFLAGS during CALL_NOSPEC on i386
      userfaultfd_release: always remove uffd flags and clear vm_userfaultfd_ctx
      gpiolib: never report open-drain/source lines as 'input' to user-space
      Revert "dm bufio: fix deadlock with loop device"
      HID: wacom: Correct distance scale for 2nd-gen Intuos devices
      HID: wacom: correct misreported EKR ring values
      selftests: kvm: Adding config fragments
      perf pmu-events: Fix missing "cpu_clk_unhalted.core" event
      drm/vmwgfx: fix memory leak when too many retries have occurred
      x86/lib/cpu: Address missing prototypes warning
      libata: add SG safety checks in SFF pio transfers
      net: hisilicon: Fix dma_map_single failed on arm64
      net: hisilicon: fix hip04-xmit never return TX_BUSY
      net: hisilicon: make hip04_tx_reclaim non-reentrant
      net: cxgb3_main: Fix a resource leak in a error path in 'init_one()'
      HID: input: fix a4tech horizontal wheel custom usage
      NFSv4: Fix a potential sleep while atomic in nfs4_do_reclaim()
      can: peak_usb: force the string buffer NULL-terminated
      can: sja1000: force the string buffer NULL-terminated
      perf bench numa: Fix cpu0 binding
      isdn: hfcsusb: Fix mISDN driver crash caused by transfer buffer on the stack
      isdn: mISDN: hfcsusb: Fix possible null-pointer dereferences in start_isoc_chain()
      net: usb: qmi_wwan: Add the BroadMobi BM818 card
      ASoC: ti: davinci-mcasp: Correct slot_width posed constraint
      st_nci_hci_connectivity_event_received: null check the allocation
      st21nfca_connectivity_event_received: null check the allocation
      can: dev: call netif_carrier_off() in register_candev()
      bonding: Force slave speed check after link state recovery for 802.3ad
      ASoC: dapm: Fix handling of custom_stop_condition on DAPM graph walks
      netfilter: ebtables: fix a memory leak bug in compat
      MIPS: kernel: only use i8253 clocksource with periodic clockevent
      HID: Add 044f:b320 ThrustMaster, Inc. 2 in 1 DT
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
    
     Conflicts:
    	drivers/usb/gadget/composite.c
    	fs/userfaultfd.c
    	sound/usb/mixer.c
  2. blu_schedutil: drop remaining unused bits

    okitavera committed Sep 4, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  3. defconfig: grus: cleanup boosting values

    okitavera committed Sep 3, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  4. devfreq_boost: drop boosting on wakeup too

    okitavera committed Sep 3, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  5. cpu_input_boost: drop boosting on wakeup

    okitavera committed Sep 3, 2019
    it's unnecessary for me since it already boosted
    by several input handlers that already registered before
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  6. ARM64: dts: sdm670: re-apply cost-data from the legacy kernel

    okitavera committed Aug 29, 2019
    It contains aligned busy and idle-cost-data to match LPM level number and big-little ratio from Pixel 3A.
    It can reduce slight battery cost with no performance regression (tested on UIBench).
    
    cherry-picked from bonito pie b4s4:
    2b8df89945bbe1f6a523e957f25004a47149a3da
    145d99f13674091e8882bed53bc4e06dda065459
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  7. defconfig: grus: revert back to cfq for better life

    okitavera committed Aug 29, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  8. defconfig: grus: enable exposure adjustment

    okitavera committed Aug 23, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  9. defconfig: grus: regenerate defconfig

    okitavera committed Aug 22, 2019
    - don't show timing info on printk
    - keep devfreq boosting a little-bit longer
    - follow google recommended kernel timing
    - utilize small cluster more than big ones
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  10. fakeconfig: remove unnecessary audit keys

    okitavera committed Aug 21, 2019
    CONFIG_AUDIT only is enough to fool the vts
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  11. defconfig: grus: disable per-process reclaim

    okitavera committed Aug 20, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  12. defconfig: grus: cpu-boosting adjustment

    okitavera committed Aug 20, 2019
    Add dynamic stune boost default value.
    Keep boost on wake litte longer, needed for in-display fingerprint scanning.
    Increase default big cluster boost freq, since boosting is mostly happens on
    the top-app scenario were the big-cluster is being used.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  13. cpu_input_boost: adjustable Dynamic SchedTune value via .config

    okitavera committed Aug 19, 2019
    accessible via CONFIG_DYNAMIC_STUNE_BOOST_VALUE
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  14. defconfig: disable auditing support

    okitavera committed Aug 19, 2019
    99% users don't need for auditing infrastructure.
    And because SELinux is still works without it, so better to nuke it
    to decrease the possibility of kernel attack surface and also
    reduce a possibility of fast path regression.
    
    Of course with fakeconfig too,
    since this action is definitely prohibited by Android VTS
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  15. kernel: HACK: Introduce fake dotconfig

    okitavera committed Aug 19, 2019
    This patch does creating a fake /proc/config.gz.
    It will allow us to specify which keys that we want to "tell"
    the system that we have but not actually present in the kernel binary.
    
    It does the job by merging $(objtree)/.config and $(srctree)/.fakeconfig,
    and then use the merged config (in this case, $(objtree)/.fakeconfig)
    as the content of config_data.gz.
    
    This hacks is disabled by default.
    Use KBUILD_FAKECONFIG=1 in order to activate it.
    
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
  16. ARM: dts: ea8076: disable delay for backlight update

    okitavera committed Aug 19, 2019
    Signed-off-by: Nanda Oktavera <codeharuka.yusa@gmail.com>
Older
You can’t perform that action at this time.