bcmdhd wifi: update to 1.201.59.4

rfkill-wlan, solve rockchip_wifi_mac_addr error

HDMI: add mutex lock to edid modelist operation.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

esp8089 wifi: solve warning in 64bit, new method

rk: iommu: support iommu region guard

Add an guard area after each iommu region, it is used to detect cross-border access

Signed-off-by: CMY <cmy@rock-chips.com>
Signed-off-by: XXM <xxm@rock-chips.com>

rk fb: act_height should be 2 pix align for interlace output

Signed-off-by: hjc <hjc@rock-chips.com>

rm64: rockchip: rockchip_defconfig disable EARLY_PRINTK

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm64: rockchip: check tf version later

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

vcodec: add soft reset after reading dec registers

Sometimes decoder got a timeout error and can not recover from error status.
We need to add soft reset after reading dec registers.

Signed-off-by: Herman Chen <herman.chen@rock-chips.com>

vcodec: fix enable race condition when power on/off

When multi-instances power on at same time the condition enabled maybe
set mutli-times. Use atomic function to replace orginal
read/compare/change operation to avoid multi-instances.
Also add power on/off count to debugfs for checking

Signed-off-by: Herman Chen <herman.chen@rock-chips.com>

add func: check tf ver

rk3368 lcdc: update hdmi overscan config path

Signed-off-by: hjc <hjc@rock-chips.com>

rk fb: fix in interlace mode switch between ntsc and pal lead to pagefault

Signed-off-by: hjc <hjc@rock-chips.com>

rk3368 lcdc: add wait frame start for uboot logo switch kernel logo and enable iommu

Signed-off-by: hjc <hjc@rock-chips.com>

rk fb: fix uboot logo switch to kernel logo splash screen

Signed-off-by: hjc <hjc@rock-chips.com>

dvfs: rockchip: add cpu thermal management test nodes

Signed-off-by: Xiao Feng <xf@rock-chips.com>

Makefile: revert to 899286fb6a7a20f5b5c5d0a9a22e4fc39b2d9e70

display-sys: fix compile warning

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

display-sys: function mode_string check input parameter.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

modify smc call func id to Standard sip id for efuse and pm-rk3368

rk3368 lcdc: fix iommu crash when resume

Signed-off-by: hjc <hjc@rock-chips.com>

arm64: rockchip: rk3368: refactor efuse code

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm64: rockchip: rk3368: dts: add rockchip,rk3368-efuse-256

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

fix rga mmu buf cause system crash when use ion mmu buf

Signed-off-by: Shengqin.Zhang <zsq@rock-chips.com>

rk3368: iommu: fix region free position when iovmm_activate failed

Signed-off-by: Simon <xxm@rock-chips.com>

Revert "mmc: rk_sdmmc: manually zero desc after allocated on ARM64 platform."

This reverts commit e68d16d33646fabf36e782b300404f112eb94d1d.

mmc: dw_mmc: init desc in dw_mci_idmac_init

Set 0 to des1 in 32bit case.
Otherwise the random value of des1 will be used in
dw_mci_translate_sglist: IDMAC_SET_BUFFER1_SIZE(desc, length)

Signed-off-by: Fei Wang <w.f@huawei.com>
Signed-off-by: Zhangfei Gao <zhangfei.gao@linaro.org>
Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>

dts: arm64: rk3368: fix wrong property for esd

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
cc: Sam Guo <sam.guo@rock-chips.com>
cc: Xiao Yao <xiaoyao@rock-chips.com>

CVBS: Add gm7122 driver.

      If you use it, should add board information to dts file, such as i2c address, sleep pin.
      like that:
gm7122_tve@44 {
compatible = "gm7122_tve";
reg = <0x44>;
rockchip,source = <0>; //0: LCDC0; 1: LCDC1
rockchip,prop = <PRMRY>;//<EXTEND>
gpio-reset = <&gpio0 GPIO_A1 GPIO_ACTIVE_HIGH>;
gpio-sleep = <&gpio0 GPIO_C6 GPIO_ACTIVE_HIGH>;
status = "okay";
};

Signed-off-by: ShenZhengyi <szy@rock-chips.com>
Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

HDMI: No need to set audio when hdmi is not connected.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

arm64: kernel: add system revision support

Bug 1248538

Change-Id: I272a264ba2904dfc7a2c6d74ab4fbfdd301b8985
Signed-off-by: Peng Du <pdu@nvidia.com>
Reviewed-on: http://git-master/r/260756
Reviewed-by: Alexander Van Brunt <avanbrunt@nvidia.com>

HDMI: register display type according lcdc property.

If property of lcdc connected by HDMI is EXTERN,
we register HDMI as a aux display device.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

display-sys: delete unused code.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

mmc: rk_sdmmc: fix compile warning

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>

mmc: core: add wrapper for timeout value

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
cc: Xiao yao <xiaoyao@rock-chips.com>

mmc: rk_sdmmc: do more timeout recovery operation

  We stop slaver and rebank our setting for better
compatibility.

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Reviewed-by: Roger Hu <roger.hu@rock-chips.com>
Tested-by: Allen Lee <allen.lee@rock-chips.com>
cc: Xiao yao <xiaoyao@rock-chips.com>

mmc: core: skyrocket timeout for cmd erase_group_xxx

  It shocked me that a buggy emmc need quite a long period
of time for setting its erase_group_xxx. Increase timeout
value does not harm to normal case which just under constrain.

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Tested-by: Yongzhong Yang <yongzhong.yang@rock-chips.com>
cc: Xiao yao <xiaoyao@rock-chips.com>

mmc: rk_sdmmc: disable controller int before enter deepsleep

   Generally when machine enter deepsleep, with no cirrcumstance,
host can interact with slavery devices. So disbale our int generater
ourself. By occasionly, we find reset controller will trigger an
unexpected SEB int. It exists a race condition for reset and IO timing
before switched to gpio and confuse state machine.

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Tested-and-Reviewed-by: Xiao yao <xiaoyao@rock-chips.com>

mmc: rk_sdmmc: add platform hook for special pm ops during deepsleep

  We register a platform hook and restore all of our regs if platform does
need to cutoff controller's power-supply during deepsleep in period of time.
Make sure your have added "controller-power-down" property AND comment out
"keep-power-in-suspend" in related dts file.

Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Xiaoyao <xiaoyao@rock-chips.com>
Reviewed-by: Chenjh <chenjh@rock-chips.com>
Tested-by: Chenjh <chenjh@rock-chips.com>

dtsi: ddr_freq: add vop bandwidth freq table

Signed-off-by: Xiao Feng <xf@rock-chips.com>

fb: add sysfs node lcdcid.

Now we have many fb devices, each vop layer is mapped to a fb device.
If there is 2 vop, we can not distinguish the vop between these fb devices.
Add sysfs node lcdcid to get the lcdc id mapped by fb device.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

rfkill-wlan: skip can't find rockchip,grf property error in none 3368 platform

rk3368: pmic: rk818:adjust dcdc voltage step

Signed-off-by: zhangqing <zhangqing@rock-chips.com>

ASoC: es8316: use startup/shutdown to manager power consumption.

use startup/shutdown to reduce power consumption, every time codec
standby, we shutdown codec. so, there is no need extra operation in
suspend.

Signed-off-by: Sugar Zhang <sugar.zhang@rock-chips.com>

arm: rockchip: vcodec_service: allow no rockchip,grf in dts

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

clk: rockchip: allow no rockchip,grf in dts on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

ASoC: dmaengine: use dma_request_slave_channel to request dma chan.

if i2s module probe before dma module, dma chan resources are not ready.
use dma_request_slave_channel to reinitialize dma chan if not ready.

Signed-off-by: Sugar Zhang <sugar.zhang@rock-chips.com>

HDMI: support show edid status.

Use /sys/class/display/HDMI/debug node to
show sink EDID status and raw status.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

display-sys: debug node support to show debug information.

Add a callback function getdebug() to show debug information.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

HDMI: just list common format when edid is not valid.

If EDID is not valid, list following common format:
1920x1080p-60
1920x1080p-50
1280x720p-60
1280x720p-50
720px576p-50
720px480p-60

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

esp8089 wifi: solve warning in 64bit

video: rockchip: rga2: fix compilation error on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

video: rockchip: fix compilation error on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

usb: rockchip: fix compilation error on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

spi: rockchip: fix compilation error on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

pwm: rockchip: fix compilation error on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm: rockchip: select MFD_SYSCON

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

usb: dwc_otg: build driver depends on arm or arm64

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm: rockchip: cpu.h: only support on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm: rockchip: efuse: fix compilation error on arm

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm: rockchip: use read_cpuid_part instead of read_cpuid_part_number

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

rk: gcc-wrapper.py ignore sysctl_net_core.c:24

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

rkwifi: dhd_config: fix compilation warning

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

rk1000: fix compilation warning

Merge branch develop-3.10-next

ddrfreq: add vop bandwidth freq table in dts, in order to support more boards

Signed-off-by: Xiao Feng <xf@rock-chips.com>

ddrfreq: fix new_freq=0 when the current load of ddr is between high_load and low_load

Signed-off-by: Xiao Feng <xf@rock-chips.com>

VPU, set the power-on flag after power domain already on.

Previous register operation before power on.
Software reset vpu inside vdpu interrupt handle.

Signed-off-by: Alpha Lin <alpha.lin@rock-chips.com>

ASoC: es8316: reduce standby power consumption around ~uA.

Signed-off-by: Sugar Zhang <sugar.zhang@rock-chips.com>

rk fb:extern rockchip_get_screen_type for mipi dsi

Signed-off-by: hjc <hjc@rock-chips.com>

rk fb: fix hdmi/cvbs display error when uboot display logo

Signed-off-by: hjc <hjc@rock-chips.com>

HDMI: Not filter the input 3dmode with EDID information.

If input 3dmode is not list in EDID, we print a
warning log and send it to sink, no longer return
error.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

HDMI: fix edid 3d information offset parse error.

In EDID 3D descriptor, 3d information is after
4K format. But we did not increase the offset of
3d content if 4K format is zero, which make 3d
information error.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

HDMI: Not filter the input 3dmode with EDID information.

If input 3dmode is not list in EDID, we print a
warning log and send it to sink, no longer return
error.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

HDMI: fix edid 3d information offset parse error.

In EDID 3D descriptor, 3d information is after
4K format. But we did not increase the offset of
3d content if 4K format is zero, which make 3d
information error.

Signed-off-by: Zheng Yang <zhengyang@rock-chips.com>

rk_fb: compatible for mid hdmi switch screen

Signed-off-by: Mark Yao <mark.yao@rock-chips.com>

rockchip: video: fix logo display when switch screen

Signed-off-by: Mark Yao <mark.yao@rock-chips.com>

input: keyboard: rk_keys: add wake lock

New version Android will never call EPOLLWAKEUP or EVIOCSSUSPENDBLOCK.
So we need add wake lock in driver.

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

input: keyboard: rk_keys: use jiffies rather than msecs_to_jiffies

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

input: keyboard: rk_keys: fix CodingStyle issues only

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

rk_fb: compatible for mid hdmi switch screen

Signed-off-by: Mark Yao <mark.yao@rock-chips.com>

rk3368: cpufreq: fix pm_notifier return NOTIFY_BAD

the big cluster mask is random when down cpu1~cpu7, so we get the error policy cpu in pm_notifier.

Signed-off-by: Xiao Feng <xf@rock-chips.com>

arm64: rockchip: cpuquiet: be quiet when cpu up/down

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

cpufreq: rockchip_big_little: minor cleanups

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

rockchip: hmp: adjust hmp_up/down_threshold

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

cpuquiet: add rockchip cpuquiet bl_balanced governor driver

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

cpuquiet: add rockchip cpuquiet driver

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

cpuquiet: Update averaging of nr_runnables

Doing a Exponential moving average per nr_running++/-- does not
guarantee a fixed sample rate which induces errors if there are lots of
threads being enqueued/dequeued from the rq (Linpack mt). Instead of
keeping track of the avg, the scheduler now keeps track of the integral
of nr_running and allows the readers to perform filtering on top.

Implemented a proper exponential moving average for the runnables
governor and a straight 100ms average for the balanced governor. Tweaked
the thresholds for the runnables governor to minimize latency. Also,
decreased sample_rate for the runnables governor to the absolute minimum
of 10msecs.

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

cpuquiet: PM QoS: Add max/min online cpus as PM QoS parameter

Signed-off-by: Huang, Tao <huangtao@rock-chips.com>

arm64: add cpu power configs to arm64/Kconfig

Add cpuquiet to proper location

Bug 1453967

Change-Id: I03b35ed12acdf5c6fc62c3bddc8b5a2a018a1303
Signed-off-by: Krishna Sitaraman <ksitaraman@nvidia.com>
Reviewed-on: http://git-master/r/396582
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Adeel Raza <araza@nvidia.com>
Reviewed-by: Alexander Van Brunt <avanbrunt@nvidia.com>

cpuquiet: Added cpuquiet to driver Makefile

Change-Id: I8a3fd51df81de8d4375b9ddabbaf4fed5277b17d
Signed-off-by: Sai Charan Gurrappadi <sgurrappadi@nvidia.com>
Reviewed-on: http://git-master/r/123849
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Peter Boonstoppel <pboonstoppel@nvidia.com>
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>
Rebase-Id: R90846374c2567592a285664e65941d6534017c8a

drivers: cpuquiet: Fix runnable gov device_busy

On device_busy, the runnable governor stops its sample timer and cancels
any queued runnable_work. However, it is possible for the cpuquiet
platform driver to call device_busy on cpuquiet_quiesence/wake_cpu calls
which can be invoked from the queued runnable_worker thread. This will
lead to the runnable_worker thread waiting infinitely on it self
finishing in cancel_work_sync.

Fix this by removing the cancel_work_sync call as it isn't necessary
because the worker doesn't requeue it self. Stopping the timer is
sufficient.

Bug 200002511

Change-Id: Ic69061b12d4dbb96b259ed9b62a0475271bd36d0
Signed-off-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
(cherry picked from commit b9a2a2cfa91d14935fdef626517d738399af4e4d)
Reviewed-on: http://git-master/r/415315
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>
Reviewed-by: Kerwin Wan <kerwinw@nvidia.com>

cpuquiet: Add Sysfs node for nr_run_thresholds

Added sysfs node that exposes the nr_runnable threshold at which cores
are quiesenced/woken up. To write into this node:

echo 1core# 2core# ... n_core# > nr_run_thresholds

Bug 1427140

Change-Id: I7cabd63692cf92f7fe8d1064437b1d73002d4ba5
Signed-off-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
Reviewed-on: http://git-master/r/369017
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>

cpuquiet: Support X cpus for runnables governor

Runnables govenror will now work properly for any number of cpus as
defined by NR_CPUS. Switched the ordering of nr_run_threshold init to
ensure that array is initialized correctly to UINT_MAX for NR_CPUS-1 cpu.

Bug 1427140

Change-Id: Ic612406ceaf80dd7e0b6693906df339f8b6512aa
Signed-off-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
Reviewed-on: http://git-master/r/369016
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Peter Boonstoppel <pboonstoppel@nvidia.com>
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>

cpuquiet: Clean up sysfs memory properly

Runnable sysfs node wasn't being free'd in stop via a kobject_put call
because the sysfs node wasn't allocated with kobject_create. Now do an
explicity kfree on stop after the kobject_put.

Change-Id: I5ba554312757380dfb2e814f75832b28d4c7edf8
Signed-off-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
Reviewed-on: http://git-master/r/346645
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Peter Boonstoppel <pboonstoppel@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>

cpuquiet: Fix locking in driver_unregister

cpuquiet_switch_governors needs to be protected with the cpuquiet_lock.

Change-Id: I948d1190a270b9e8b732eaad80a303560fe9f282
Signed-off-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
Reviewed-on: http://git-master/r/345013
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Peter Boonstoppel <pboonstoppel@nvidia.com>
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>

cpuquiet: move back cpuquiet_lock

Moved cpuquiet_lock to the back of cpuquiet_switch_governor().
cpuquiet_switch_governor() has to be protected by cpuquiet_lock.
In the runnable cpuquiet_governor case, it can occur a kernel panic or BUG_ON
because schedule_work() and INIT_WORK() can be performed at the same time.

Bug 1410816

Change-Id: I6d3d69bd8896e5425a762cbc4a58343a319edf29
Signed-off-by: Jubeom Kim <jubeomk@nvidia.com>
(cherry picked from commit 8502d0b7c057f96b7cd0495b49e37fc968d4bf8a)
Reviewed-on: http://git-master/r/337335
Reviewed-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
Tested-by: Sai Gurrappadi <sgurrappadi@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Diwakar Tundlam <dtundlam@nvidia.com>

cpuquiet: balanced: WQ_RESCUER has been removed

Signed-off-by: Dan Willemsen <dwillemsen@nvidia.com>

ARM: tegra: cpuquiet: augment no_lp to force both LP and G

- tegra_cpuquiet 'no_lp' node now accepts '-1' to force LP cluster on
- when a cluster is forced the switches take place immediately without
  any delay

Bug 1247636

Change-Id: I1ec07acc3cf51419a4770542c6a1c90b92e0bf66
Signed-off-by: Simo Melenius <smelenius@nvidia.com>
Reviewed-on: http://git-master/r/208772
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
(cherry picked from commit ab16bbb6f581da538448d2a64e03dd272b8fb985)
Reviewed-on: http://git-master/r/210358

ARM: tegra: cpuquiet: make userspace governor actions synchronous

Userspace expects changes to happen synchronously. Implement this by waiting
with a (configureable) timeout for the action to happen.

Bug 1220065

Change-Id: I81301719707e4baf2b3aea62c38fc771ffe1205d
Signed-off-by: Peter De Schrijver <pdeschrijver@nvidia.com>
Reviewed-on: http://git-master/r/200013
Reviewed-by: Simone Willett <swillett@nvidia.com>
Tested-by: Simone Willett <swillett@nvidia.com>