OSDN Git Service

sagit-ice-cold/kernel_xiaomi_msm8998.git
7 years agoMerge "Revert "sd: Fix rw_max for devices that report an optimal xfer size""
Linux Build Service Account [Thu, 6 Apr 2017 21:33:10 +0000 (14:33 -0700)]
Merge "Revert "sd: Fix rw_max for devices that report an optimal xfer size""

7 years agoMerge "spi: spi_qsd: Disable packing for unaliged non DMA mode transfers"
Linux Build Service Account [Thu, 6 Apr 2017 21:33:09 +0000 (14:33 -0700)]
Merge "spi: spi_qsd: Disable packing for unaliged non DMA mode transfers"

7 years agoMerge "cpu-hotplug: Always use real time scheduling when hotplugging a CPU"
Linux Build Service Account [Thu, 6 Apr 2017 21:33:08 +0000 (14:33 -0700)]
Merge "cpu-hotplug: Always use real time scheduling when hotplugging a CPU"

7 years agoMerge "binder: use group leader instead of open thread"
Linux Build Service Account [Thu, 6 Apr 2017 21:33:07 +0000 (14:33 -0700)]
Merge "binder: use group leader instead of open thread"

7 years agoMerge "pps-gpio: Timestamp alignment with boottime"
Linux Build Service Account [Thu, 6 Apr 2017 21:33:06 +0000 (14:33 -0700)]
Merge "pps-gpio: Timestamp alignment with boottime"

7 years agoMerge "sched/debug: Make the "Preemption disabled at ..." message more useful"
Linux Build Service Account [Thu, 6 Apr 2017 12:39:40 +0000 (05:39 -0700)]
Merge "sched/debug: Make the "Preemption disabled at ..." message more useful"

7 years agoMerge "leds: qpnp-flash-v2: Fix pinctrl configuration"
Linux Build Service Account [Thu, 6 Apr 2017 12:39:39 +0000 (05:39 -0700)]
Merge "leds: qpnp-flash-v2: Fix pinctrl configuration"

7 years agoMerge "msm: vidc: Re-calculate buffer requirement"
Linux Build Service Account [Thu, 6 Apr 2017 12:39:38 +0000 (05:39 -0700)]
Merge "msm: vidc: Re-calculate buffer requirement"

7 years agoMerge "msm: vidc: Prioritize debugfs timeout control over dtsi entry"
Linux Build Service Account [Thu, 6 Apr 2017 12:39:37 +0000 (05:39 -0700)]
Merge "msm: vidc: Prioritize debugfs timeout control over dtsi entry"

7 years agoMerge "ARM: dts: msm: add support for sharp wuxga split link panel"
Linux Build Service Account [Thu, 6 Apr 2017 12:39:36 +0000 (05:39 -0700)]
Merge "ARM: dts: msm: add support for sharp wuxga split link panel"

7 years agoMerge "msmcortex: Enable PPP related defconfigs"
Linux Build Service Account [Thu, 6 Apr 2017 12:39:35 +0000 (05:39 -0700)]
Merge "msmcortex: Enable PPP related defconfigs"

7 years agospi: spi_qsd: Disable packing for unaliged non DMA mode transfers
Mukesh Kumar Savaliya [Fri, 31 Mar 2017 07:09:11 +0000 (12:39 +0530)]
spi: spi_qsd: Disable packing for unaliged non DMA mode transfers

SPI core has a characterstic that it always sends data in FIFO word size
when we compress the data and fill the FIFO. This helps to improve the
overall througput but sometimes for the data transfers which is not
aligned to FIFO Word [i.e 4] it pads extra bytes and sends out which can
mislead spi slave device.

Do not enable packing if it's a FIFO mode transfer and data size is not
aligned to FIFO word size.

Change-Id: I28f4e3e38db4b882f229c00aa54aabdc72d2c139
Signed-off-by: Mukesh Kumar Savaliya <msavaliy@codeaurora.org>
7 years agopps-gpio: Timestamp alignment with boottime
Naresh Munagala [Wed, 18 May 2016 11:41:09 +0000 (17:11 +0530)]
pps-gpio: Timestamp alignment with boottime

GPS driver which makes use of PPS timestamp expects time
with respect to boot time. Existing pps_get_ts function
doesn't return time from the boot so replaced this function
with get_monotonic_boottime function which returns elapsed time
from the boot.

Change-Id: I8f656a1c4ef43f6314d4ab295a177125bcc8195e
CRs-Fixed: 1012438
Signed-off-by: Naresh Munagala <nareshm@codeaurora.org>
7 years agospi: spi_qsd: Merge back throughput improvements done for small transfers
Mukesh Kumar Savaliya [Tue, 4 Apr 2017 05:47:38 +0000 (11:17 +0530)]
spi: spi_qsd: Merge back throughput improvements done for small transfers

These changes will merge back the enhancemnts done to improve the medium
side transfers. This was reverted temporarily till the fix identified.

Change-Id: Ib7ac92ecefe7ca3ff9f03716c51dc31b8322ee33
Signed-off-by: Mukesh Kumar Savaliya <msavaliy@codeaurora.org>
7 years agoRevert "sd: Fix rw_max for devices that report an optimal xfer size"
Sayali Lokhande [Thu, 6 Apr 2017 04:26:52 +0000 (09:56 +0530)]
Revert "sd: Fix rw_max for devices that report an optimal xfer size"

This reverts 'commit 9814eb75495b ("sd: Fix rw_max for devices
that report an optimal xfer size")'.
Max sectors limit for request queue(q->limits.max_sectors) is reduced
from 1024 to 16 sectors by above commit and causing performance impact
(reduced by upto 50%). Hence revert the change to fix the performance
issue observed on 8998.

Change-Id: I9abd2be6c90922ea9de818912f405e0d292a62b8
Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org>
7 years agoMerge "clk: mdss: adjust PLL disable sequence to avoid glitch"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:33 +0000 (20:21 -0700)]
Merge "clk: mdss: adjust PLL disable sequence to avoid glitch"

7 years agoMerge "scsi: ufs: fix error handing during hibern8 enter"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:33 +0000 (20:21 -0700)]
Merge "scsi: ufs: fix error handing during hibern8 enter"

7 years agoMerge "msm: ipa: continue probe on ipc log failure"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:32 +0000 (20:21 -0700)]
Merge "msm: ipa: continue probe on ipc log failure"

7 years agoMerge "msm: mhi_rmnet: add support for shutdown and system error notification"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:31 +0000 (20:21 -0700)]
Merge "msm: mhi_rmnet: add support for shutdown and system error notification"

7 years agoMerge "leds: qpnp-flash-v2: Fix some issues"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:30 +0000 (20:21 -0700)]
Merge "leds: qpnp-flash-v2: Fix some issues"

7 years agoMerge "usb: gadget: gsi: Fix incorrect repsonse available notifications"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:29 +0000 (20:21 -0700)]
Merge "usb: gadget: gsi: Fix incorrect repsonse available notifications"

7 years agoMerge "icnss: Make driver register asynchronous to probe"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:28 +0000 (20:21 -0700)]
Merge "icnss: Make driver register asynchronous to probe"

7 years agoMerge "ASoC: wcd934x: add null check for soundwire control data"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:27 +0000 (20:21 -0700)]
Merge "ASoC: wcd934x: add null check for soundwire control data"

7 years agoMerge "smb138x: support usbin-usbin with external rsense"
Linux Build Service Account [Thu, 6 Apr 2017 03:21:27 +0000 (20:21 -0700)]
Merge "smb138x: support usbin-usbin with external rsense"

7 years agobinder: use group leader instead of open thread
Martijn Coenen [Tue, 7 Mar 2017 14:51:18 +0000 (15:51 +0100)]
binder: use group leader instead of open thread

The binder allocator assumes that the thread that
called binder_open will never die for the lifetime of
that proc. That thread is normally the group_leader,
however it may not be. Use the group_leader instead
of current.

Bug: 35707103
Test: Created test case to open with temporary thread

Change-Id: Id693f74b3591f3524a8c6e9508e70f3e5a80c588
Signed-off-by: Todd Kjos <tkjos@google.com>
Signed-off-by: Martijn Coenen <maco@android.com>
Git-commit: 359795138dc5440e09c58025e28ec1b38d648c09
Git-repo: https://android.googlesource.com/kernel/common/
Signed-off-by: Runmin Wang <runminw@codeaurora.org>
7 years agoclk: mdss: adjust PLL disable sequence to avoid glitch
Abhinav Kumar [Fri, 24 Feb 2017 19:56:21 +0000 (11:56 -0800)]
clk: mdss: adjust PLL disable sequence to avoid glitch

Adjust the PLL disable sequence as per the latest HW
programming guidelines to ensure there will not be any
stray clock glitches when PLL is turned OFF abruptly.

Change-Id: I6df35bbe18b0c42b43f38b9dd85c3502b2038928
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
7 years agoclk: mdss: re-program PLL registers to power-on-reset value
Abhinav Kumar [Mon, 20 Feb 2017 05:01:15 +0000 (21:01 -0800)]
clk: mdss: re-program PLL registers to power-on-reset value

When the supply to PLL digital domain is turned off,
it can result in certain PLL registers to get corrupted.

Make sure to re-program the PLL registers to the
power-on-reset value before starting to program the PLL again
to ensure that it locks reliably.

Change-Id: I63cac884cf11eae60b187f83654f5922a3342d66
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
7 years agoscsi: ufs: fix error handing during hibern8 enter
Subhash Jadavani [Fri, 24 Mar 2017 21:44:01 +0000 (14:44 -0700)]
scsi: ufs: fix error handing during hibern8 enter

During clock gating (ufshcd_gate_work()), we first put the link hibern8 by
calling ufshcd_uic_hibern8_enter() and if ufshcd_uic_hibern8_enter()
returns success (0) then we gate all the clocks.
Now let’s zoom in to what ufshcd_uic_hibern8_enter() does internally:
It calls __ufshcd_uic_hibern8_enter() which on detecting the LINERESET,
initiates the full recovery and puts the link back to highest HS gear and
returns success (0) to ufshcd_uic_hibern8_enter() which is the issue as
link is still in active state due to recovery!
Now ufshcd_uic_hibern8_enter() returns success to ufshcd_gate_work() and
hence it goes ahead with gating the UFS clock while link is still in active
state hence I believe controller would raise UIC error interrupts. But when
we service the interrupt, clocks might have already been disabled!

This change fixes for this by returning failure from
__ufshcd_uic_hibern8_enter() if recovery succeeds as link is still not in
hibern8, upon receiving the error ufshcd_hibern8_enter() would initiate
retry to put the link state back into hibern8.

Change-Id: Ib550fb791fa4c582b8f2d317a7f5f7594acb0872
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
7 years agocpu-hotplug: Always use real time scheduling when hotplugging a CPU
Syed Rameez Mustafa [Fri, 16 Dec 2016 19:59:05 +0000 (11:59 -0800)]
cpu-hotplug: Always use real time scheduling when hotplugging a CPU

CPU hotplug operations take place in preemptible context. This leaves
the hotplugging thread at the mercy of overall system load and CPU
availability. If the hotplugging thread does not get an opportunity
to execute after it has already begun a hotplug operation, CPUs can
end up being stuck in a quasi online state. In the worst case a CPU
can be stuck in a state where the migration thread is parked while
another task is executing and changing affinity in a loop. This
combination can result in unbounded execution time for the running
task until the hotplugging thread gets the chance to run to complete
the hotplug operation.

Fix the said problem by ensuring that hotplug can only occur from
threads belonging to the RT sched class. This allows the hotplugging
thread priority on the CPU no matter what the system load or the
number of available CPUs are. If a SCHED_NORMAL task attempts to
hotplug a CPU, we temporarily elevate it's scheduling policy to RT.
Furthermore, we disallow hotplugging operations to begin if the
calling task belongs to the idle and deadline classes or those that
use the SCHED_BATCH policy.

Change-Id: Idbb1384626e6ddff46c0d2ce752eee68396c78af
Signed-off-by: Syed Rameez Mustafa <rameezmustafa@codeaurora.org>
7 years agomsm: mhi_rmnet: add support for shutdown and system error notification
Sujeev Dias [Fri, 3 Feb 2017 19:13:58 +0000 (11:13 -0800)]
msm: mhi_rmnet: add support for shutdown and system error notification

Add support to handle system error and shutdown notification
from MHI host.

CRs-Fixed: 2022936
Change-Id: Id36097dffd7571490d7d53d2e496bfe024702a42
Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
7 years agoMerge "msm: mdss: dp: fix notifications for link training tests"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:10 +0000 (12:17 -0700)]
Merge "msm: mdss: dp: fix notifications for link training tests"

7 years agoMerge "ARM: dts: msm: Update MDSS max mixer/pipe width for sdm630"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:09 +0000 (12:17 -0700)]
Merge "ARM: dts: msm: Update MDSS max mixer/pipe width for sdm630"

7 years agoMerge "crypto: msm: Use appropriate logging function"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:08 +0000 (12:17 -0700)]
Merge "crypto: msm: Use appropriate logging function"

7 years agoMerge "qcom: smb1351: update CHARGING_ENABLE reporting logic"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:05 +0000 (12:17 -0700)]
Merge "qcom: smb1351: update CHARGING_ENABLE reporting logic"

7 years agoMerge "power_supply: add HW_CURRENT_MAX power_supply property"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:03 +0000 (12:17 -0700)]
Merge "power_supply: add HW_CURRENT_MAX power_supply property"

7 years agoMerge "ARM: dts: msm: add regulators and clk in icnss device node for sdm630"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:02 +0000 (12:17 -0700)]
Merge "ARM: dts: msm: add regulators and clk in icnss device node for sdm630"

7 years agoMerge "sdm660: ADSPRPC: Null check file session ctx during mmap create"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:01 +0000 (12:17 -0700)]
Merge "sdm660: ADSPRPC: Null check file session ctx during mmap create"

7 years agoMerge "ARM: dts: msm: add regulators and clk in icnss device node for sdm660"
Linux Build Service Account [Wed, 5 Apr 2017 19:17:00 +0000 (12:17 -0700)]
Merge "ARM: dts: msm: add regulators and clk in icnss device node for sdm660"

7 years agoMerge "ARM: dts: msm: update itech and st1031ga battery profile parameters"
Linux Build Service Account [Wed, 5 Apr 2017 19:16:59 +0000 (12:16 -0700)]
Merge "ARM: dts: msm: update itech and st1031ga battery profile parameters"

7 years agoMerge "ARM: dts: msm: enable SDR104 workaround for MSM8998"
Linux Build Service Account [Wed, 5 Apr 2017 19:16:57 +0000 (12:16 -0700)]
Merge "ARM: dts: msm: enable SDR104 workaround for MSM8998"

7 years agoMerge "msm: kgsl: Use legacy PM4 check instead of adreno version"
Linux Build Service Account [Wed, 5 Apr 2017 19:16:56 +0000 (12:16 -0700)]
Merge "msm: kgsl: Use legacy PM4 check instead of adreno version"

7 years agousb: gadget: gsi: Fix incorrect repsonse available notifications
Sriharsha Allenki [Mon, 3 Apr 2017 09:14:45 +0000 (14:44 +0530)]
usb: gadget: gsi: Fix incorrect repsonse available notifications

Two notify requests are being queued for one available modem
response. So for the first GET_ENCAPSULATED_RESPONSE we provide
the host with the available response. Now for the next
GET_ENCAPSULATED_COMMAND we notify the host that the response is
available even before the modem is ready with a response because
of the extra notify request queued on the interrupt endpoint.
This causes a STALL for the next GET_ENCAPSULATED_RESPONSE request.
This is caused because we are queueing a notify request from the
completion handler of the interrupt endpoint request when the
response queue is not empty.
Fix this by queuing a notify request when a new response is
available only after the current resposne is send to the Host.

Change-Id: If84bc315f2be910503328cc6b0e21be342c6eb37
Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org>
Signed-off-by: Mayank Rana <mrana@codeaurora.org>
7 years agomhi: mhi_uci: add support for shutdown and system error notification
Sujeev Dias [Tue, 31 Jan 2017 19:59:23 +0000 (11:59 -0800)]
mhi: mhi_uci: add support for shutdown and system error notification

Add support to handle system error and shutdown
notifications from mhi host.

CRs-Fixed: 1097560
Change-Id: Ied6c907586aa4dc2ed3b1a7c19305877144b3b21
Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
7 years agoicnss: Make driver register asynchronous to probe
Anurag Chouhan [Fri, 24 Mar 2017 10:37:45 +0000 (16:07 +0530)]
icnss: Make driver register asynchronous to probe

Driver register doesn't have to be synchronous with probe call
back. Keep the driver registered till unregister gets called and
never reset ops during probe failure.

CRs-fixed: 2029329
Change-Id: I61331c7f33b29b0bc4833a8e4c52ee94f17660e7
Signed-off-by: Anurag Chouhan <achouhan@codeaurora.org>
7 years agoASoC: wcd934x: add null check for soundwire control data
Meng Wang [Tue, 21 Mar 2017 03:16:24 +0000 (11:16 +0800)]
ASoC: wcd934x: add null check for soundwire control data

Soundwire control data is NULL if there is no soundwire slave
device enumerated.
Add null check to avoid panic.

Change-Id: Ief60d69c36c2a9831825f38da2c9a3f6dd5f13dc
CRs-Fixed: 2020293
Signed-off-by: Meng Wang <mwang@codeaurora.org>
7 years agoMerge "msm: ADSPRPC: Unmap remote heap buffer only for ADSP SSR"
Linux Build Service Account [Wed, 5 Apr 2017 11:05:40 +0000 (04:05 -0700)]
Merge "msm: ADSPRPC: Unmap remote heap buffer only for ADSP SSR"

7 years agoMerge "power: smb-lib: update USBIN_V_VOTER usage"
Linux Build Service Account [Wed, 5 Apr 2017 11:05:39 +0000 (04:05 -0700)]
Merge "power: smb-lib: update USBIN_V_VOTER usage"

7 years agoMerge "msm: pcie: support PCIe MSI QGIC with stage 1 SMMU enabled"
Linux Build Service Account [Wed, 5 Apr 2017 11:05:38 +0000 (04:05 -0700)]
Merge "msm: pcie: support PCIe MSI QGIC with stage 1 SMMU enabled"

7 years agoMerge "Revert "defconfig: msm: Enable RCU_STALL_WATCHDOG_BITE flag""
Linux Build Service Account [Wed, 5 Apr 2017 11:05:37 +0000 (04:05 -0700)]
Merge "Revert "defconfig: msm: Enable RCU_STALL_WATCHDOG_BITE flag""

7 years agoMerge "Input: stm: optimize the laser sensor driver"
Linux Build Service Account [Wed, 5 Apr 2017 11:05:36 +0000 (04:05 -0700)]
Merge "Input: stm: optimize the laser sensor driver"

7 years agoMerge "ARM: dts: msm: Update camera clock sources for sdm630"
Linux Build Service Account [Wed, 5 Apr 2017 11:05:35 +0000 (04:05 -0700)]
Merge "ARM: dts: msm: Update camera clock sources for sdm630"

7 years agoMerge "ARM: dts: msm: Add flash and OIS node to camera sensor mtp for msm8998"
Linux Build Service Account [Wed, 5 Apr 2017 11:05:34 +0000 (04:05 -0700)]
Merge "ARM: dts: msm: Add flash and OIS node to camera sensor mtp for msm8998"

7 years agomsm: kgsl: Use legacy PM4 check instead of adreno version
Lynus Vaz [Mon, 27 Mar 2017 12:50:00 +0000 (18:20 +0530)]
msm: kgsl: Use legacy PM4 check instead of adreno version

Check for legacy PM4 commands instead of adreno version to calculate
ringbuffer space for PM4 commands that write to memory.

Change-Id: I5d1d4cfbc70bc73ddee9ee752de24aae154a04dc
Signed-off-by: Lynus Vaz <lvaz@codeaurora.org>
7 years agoARM: dts: msm: add regulators and clk in icnss device node for sdm630
Sarada Prasanna Garnayak [Thu, 16 Mar 2017 14:22:47 +0000 (19:52 +0530)]
ARM: dts: msm: add regulators and clk in icnss device node for sdm630

During recovery, there are cases where in WLAN driver running on
APPS can access WLAN copy engine register causing exceptions
because of WLAN hardware in reset state.

Add voltage regulator and clk in icnss device node to vote for WLAN
hardware resources by the icnss platform driver during WLAN recovery.

CRs-Fixed: 2029173
Change-Id: I00bbad40ee152c8dac187ff1f541e34b254589cb
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
7 years agoARM: dts: msm: add regulators and clk in icnss device node for sdm660
Sarada Prasanna Garnayak [Thu, 16 Mar 2017 13:35:27 +0000 (19:05 +0530)]
ARM: dts: msm: add regulators and clk in icnss device node for sdm660

During recovery, there are cases where in WLAN driver running on
APPS can access WLAN copy engine register causing exceptions
because of WLAN hardware in reset state.

Add voltage regulator and clk in icnss device node to vote for WLAN
hardware resources by the icnss platform driver during WLAN recovery.

CRs-Fixed: 2029102
Change-Id: Iba785acc391fb56ea068199598ddc14d20bb3481
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
7 years agocrypto: msm: Use appropriate logging function
Neeraj Soni [Wed, 5 Apr 2017 06:27:57 +0000 (11:57 +0530)]
crypto: msm: Use appropriate logging function

Dynamic debug can be used to get the logs

Change-Id: I6510da111019022149190fd753a5da1cc816fffa
Signed-off-by: Neeraj Soni <neersoni@codeaurora.org>
7 years agomsm: ADSPRPC: Unmap remote heap buffer only for ADSP SSR
Tharun Kumar Merugu [Thu, 30 Mar 2017 09:03:56 +0000 (14:33 +0530)]
msm: ADSPRPC: Unmap remote heap buffer only for ADSP SSR

Unmap remote heap buffer only for ADSP SSR

Change-Id: I9abff20cfdcc4588a3a377e73d38d3ea91316d25
Acked-by: Himateja Reddy <hmreddy@qti.qualcomm.com>
Signed-off-by: Tharun Kumar Merugu <mtharu@codeaurora.org>
7 years agoqcom: smb1351: update CHARGING_ENABLE reporting logic
Ashay Jaiswal [Tue, 4 Apr 2017 11:27:57 +0000 (16:57 +0530)]
qcom: smb1351: update CHARGING_ENABLE reporting logic

In case of parallel charging mode "parallel_charger_suspended"
variable tracks the charging enable state, update CHARGING_ENABLE
reporting to use the same.

CRs-Fixed: 2014572
Change-Id: Ib7246f407ad103343b7587b9de3ac938fb63767d
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
7 years agoqcom: smb2: disable parallel in restricted/thermal conditions
Ashay Jaiswal [Tue, 28 Mar 2017 11:49:47 +0000 (17:19 +0530)]
qcom: smb2: disable parallel in restricted/thermal conditions

Add voting to disable parallel charger in case of restricted
charging and when system temperature level changes from normal.
Parallel charger gets re-enabled once thermal condition/FCC
restrictions are removed.

CRs-Fixed: 2014572
Change-Id: Ic26b3d93f2f3d582a2eb3c7b9ea0d27bbad24a50
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
7 years agoqcom: battery: update parallel algorithm for ICL change
Abhijeet Dharmapurikar [Thu, 23 Mar 2017 21:07:07 +0000 (14:07 -0700)]
qcom: battery: update parallel algorithm for ICL change

Update the parallel algorithm for ICL change request.
Following steps are performed for every ICL change:
- disable parallel charger using ICL_CHANGE_VOTER
- update the new ICL and re-run AICL.
- re-enable parallel charger by removing vote of
  ICL_CHANGE_VOTER.

While at it, update the 'smblib_get_prop_usb_online'
function to use voter library's locked API to get
the usb_icl_vote, this fixes the lockup that happens
when charging is enabled/disabled from usb_icl_votable's
callback function.

CRs-Fixed: 2014572
Change-Id: I7deb6a50d67471ab1aa5e1db6fff880574b4bafb
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqcom: battery: remove ICL_REDUCTION support
Abhijeet Dharmapurikar [Thu, 23 Mar 2017 21:04:05 +0000 (14:04 -0700)]
qcom: battery: remove ICL_REDUCTION support

Make battery library to directly update the split current
of the main charger via power_supply framework's set_property
API using CURRENT_MAX property. ICL_REDUCTION property
is no longer required.

Note that we are not doing initial ICL vote for USBIN USBIN,
instead when time comes to split aicl, we will read the input
current limit and if nothing is set, check what the hw determined
as the max limit and use that.

For a PD charger case where the apsd result will be unknown,
the current limit is expected to be set by the time aicl settled is
called. The 30 second delaying of parallel enabling, should give
PD engine ample time to set the current limits.

While at it clean up the way settled current change is tracked. Create
a variable that specifically indicates the total settled current seen
last time settled current was adjusted. This makes compare to an updated
settled current value easy.

CRs-Fixed: 2014572
Change-Id: I040b65e6380f2c12350ea44bf32e6981ff126f18
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agopower_supply: add HW_CURRENT_MAX power_supply property
Ashay Jaiswal [Thu, 23 Mar 2017 05:04:18 +0000 (10:34 +0530)]
power_supply: add HW_CURRENT_MAX power_supply property

Add support for HW_CURRENT_MAX property which returns maximum
input current that can be set for an adapter. This property will
be specifically used by parallel charger to cast an initial ICL
vote.

CRs-Fixed: 2014572
Change-Id: I49e80e6b25f40791d9035bc1567fdf881914b1e8
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqpnp: smb-lib: move usb_icl_votable to battery library
Ashay Jaiswal [Wed, 22 Mar 2017 17:48:51 +0000 (23:18 +0530)]
qpnp: smb-lib: move usb_icl_votable to battery library

Move usb_icl_votable creation and it's callback in the
battery.c file, in preparation for an upcoming change where
usb_icl_votable callback will be able to
disable parallel charger when a low input current
request is set.

The new callback uses CURRENT_MAX property to be set on main
psy to set the ICL limit of primary charger.

The limit setting code in main charger path where the votable
call resides currently relies a lot on client being present or not.
Since the votable is being moved away, it won't have information
about the client. So use the current parameter of INT_MAX to
indicate no clients present.

Also it could be that main psy is not instantiated by the time first
icl vote comes in. Ensure that we rerun_election on usb_icl_votable
the instant main psy is seen.

CRs-Fixed: 2014572
Change-Id: Ie449af086ed9218b40ea83158b69e8f8e73edda3
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqcom: smb2: disable h/w autonomous parallel control
Ashay Jaiswal [Tue, 21 Mar 2017 17:16:06 +0000 (22:46 +0530)]
qcom: smb2: disable h/w autonomous parallel control

The charger driver ensures that parallel charging remains
disabled as long as AICL is below threshole (1400mA) using
it's software voting mechanism(USBIN_I_VOTER), thus disable
hardware's autonomous control of input current based parallel
charging.

CRs-Fixed: 2014572
Change-Id: Ie991e0c1d16d63b6def6fb3379019114d20d60f7
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqcom: smb-lib: add delay for parallel enable
Ashay Jaiswal [Tue, 21 Mar 2017 06:54:16 +0000 (12:24 +0530)]
qcom: smb-lib: add delay for parallel enable

Currently, the driver enables parallel charger as soon
as a favorable charger is seen. This causes many, unnecessary
splits of AICL results as AICL runs for the first time. Besides
the detection logic could further update charger type and rerun
AICL.

Its best to simply delay enabling parallel charger for 30 seconds.
This ensures that type detection and its AICL run to completion and
parallel charging starts with a stable AICL result.

To realize this
- implement a workqueue which will be scheduled/cancelled 30 seconds
  later when vbus plugin interrupt happens.
- The workqueue signifies that 30 seconds have elapsed and parallel
  charging should be enabled by PL_DELAY_VOTER.
- remove PL_DELAY_HDVP_VOTER since PL_DELAY_VOTER will
  ensure parallel gets enabled after 30 seconds of insertion.
- unvote from PL_DELAY_VOTER only when the typeC removal event happens.
  A PR_SWAP during a sink session will keep the PL_DELAY_VOTER's vote
  enabled, we expect CHG_STATE_VOTER to disable parallel since device will
  be supplying VBUS and not charging the battery. This also means that
  when another PR_SWAP happens and the device starts operating in sink mode
  again, PL_DELAY_VOTER's vote will remain enabled and parallel charging
  would begin when other conditions become favorable. IOW PL_DELAY_VOTER
  tracks 30 second after physical insertion and remains unaffected by
  PR_SWAPs.

CRs-Fixed: 2014572
Change-Id: Ibe10768e61c6d2661bc8946f7f6b9be06f0c28e6
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agopower: smb-lib: update USBIN_V_VOTER usage
Ashay Jaiswal [Tue, 21 Mar 2017 05:45:51 +0000 (11:15 +0530)]
power: smb-lib: update USBIN_V_VOTER usage

The current code enables parallel charging if a high voltage
charger is seen (QC or PD) or a high current charger is seen.

This means that parallel charger could remain enabled when PD
transitions its voltage even though it has reduced the current
to 500mA. It is desirable to disable parallel charging during
such voltage transition.

So prepare for this change by first ensuring that only a QC
charger votes for USBIN_V_VOTER. The PD charger, even while
operating at higher voltages will need to depend on USBIN_I_VOTER,
just like a simple DCP charger, to enable parallel charger.

CRs-Fixed: 2014572
Change-Id: I03dfcff44b30b06918ebf176043873fb141f485f
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
7 years agoMerge "defconfig: msm: Enable CONFIG_UID_CPUTIME flag"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:32 +0000 (20:07 -0700)]
Merge "defconfig: msm: Enable CONFIG_UID_CPUTIME flag"

7 years agoMerge "mmc: core: prefer SDR50 mode over DDR50 for SD card"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:31 +0000 (20:07 -0700)]
Merge "mmc: core: prefer SDR50 mode over DDR50 for SD card"

7 years agoMerge "ASoC: msm: qdsp6v2: Add support for 7 channel mapping"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:30 +0000 (20:07 -0700)]
Merge "ASoC: msm: qdsp6v2: Add support for 7 channel mapping"

7 years agoMerge "mmc: core: Increase the runtime PM reference count in try_claim_host"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:29 +0000 (20:07 -0700)]
Merge "mmc: core: Increase the runtime PM reference count in try_claim_host"

7 years agoMerge "msm: kgsl: Remove incorrect check to cap AB vote to IB vote"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:29 +0000 (20:07 -0700)]
Merge "msm: kgsl: Remove incorrect check to cap AB vote to IB vote"

7 years agoMerge "msm: gsi: continue probe on ipc log failure"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:28 +0000 (20:07 -0700)]
Merge "msm: gsi: continue probe on ipc log failure"

7 years agoMerge "msm: kgsl: Disable memory retention for core clock before entering SLUMBER"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:27 +0000 (20:07 -0700)]
Merge "msm: kgsl: Disable memory retention for core clock before entering SLUMBER"

7 years agoMerge "mmc: core: Remove unused variable cached_ext_csd"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:26 +0000 (20:07 -0700)]
Merge "mmc: core: Remove unused variable cached_ext_csd"

7 years agoMerge "ARM: dts: msm: Add dt support for platforms with audio over USBC"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:25 +0000 (20:07 -0700)]
Merge "ARM: dts: msm: Add dt support for platforms with audio over USBC"

7 years agoMerge "block: Make del_gendisk() safer for disks without queues"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:24 +0000 (20:07 -0700)]
Merge "block: Make del_gendisk() safer for disks without queues"

7 years agoMerge "scsi: ufs: scale up the gear in 2 steps"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:23 +0000 (20:07 -0700)]
Merge "scsi: ufs: scale up the gear in 2 steps"

7 years agoMerge "drivers/base: cpu: Add node for core control isolation"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:22 +0000 (20:07 -0700)]
Merge "drivers/base: cpu: Add node for core control isolation"

7 years agoMerge "ASoC: msm: qdsp6v2: add route for TERT_TDM_RX_4 playback"
Linux Build Service Account [Wed, 5 Apr 2017 03:07:21 +0000 (20:07 -0700)]
Merge "ASoC: msm: qdsp6v2: add route for TERT_TDM_RX_4 playback"

7 years agomsm: pcie: support PCIe MSI QGIC with stage 1 SMMU enabled
Tony Truong [Fri, 24 Mar 2017 01:00:34 +0000 (18:00 -0700)]
msm: pcie: support PCIe MSI QGIC with stage 1 SMMU enabled

When Stage 1 SMMU is enabled, the QGIC doorbell address needs to
be mapped or else there will be a translation fault when an endpoint
tries to trigger an interrupt via MSI. PCIe host driver will map
this address on behalf of the client.

Change-Id: I7fdbe62daeb5dbecc459e4d9bc7832785f5b9fb7
Signed-off-by: Tony Truong <truong@codeaurora.org>
7 years agomsm: pcie: update teardown sequence for PCIe MSI interrupt
Tony Truong [Wed, 29 Mar 2017 19:16:51 +0000 (12:16 -0700)]
msm: pcie: update teardown sequence for PCIe MSI interrupt

Update the teardown sequence for PCIe MSI to support
multiple endpoints.

Change-Id: I1fbdb840bf3677e30d4d27a50503c5cc70ece272
Signed-off-by: Tony Truong <truong@codeaurora.org>
7 years agoMerge "pinctrl: lpi: Avoid initial SSR notifications at bootup"
Linux Build Service Account [Tue, 4 Apr 2017 19:10:28 +0000 (12:10 -0700)]
Merge "pinctrl: lpi: Avoid initial SSR notifications at bootup"

7 years agoMerge "qcom: smb-lib: disable VCONN operations for micro USB mode"
Linux Build Service Account [Tue, 4 Apr 2017 19:10:27 +0000 (12:10 -0700)]
Merge "qcom: smb-lib: disable VCONN operations for micro USB mode"

7 years agoMerge "wil6210: remove HALP voting in debugfs ioblob"
Linux Build Service Account [Tue, 4 Apr 2017 19:10:26 +0000 (12:10 -0700)]
Merge "wil6210: remove HALP voting in debugfs ioblob"

7 years agoMerge "ASoC: msm: support 5 and 7 channel for TDM hw param"
Linux Build Service Account [Tue, 4 Apr 2017 19:10:25 +0000 (12:10 -0700)]
Merge "ASoC: msm: support 5 and 7 channel for TDM hw param"

7 years agoMerge "ASoC: msm: Fix mute on speaker in VoWLAN calls"
Linux Build Service Account [Tue, 4 Apr 2017 19:10:24 +0000 (12:10 -0700)]
Merge "ASoC: msm: Fix mute on speaker in VoWLAN calls"

7 years agoASoC: msm: qdsp6v2: Add support for 7 channel mapping
Rohit Kumar [Tue, 14 Feb 2017 06:45:45 +0000 (12:15 +0530)]
ASoC: msm: qdsp6v2: Add support for 7 channel mapping

Add support for 7 channel mapping in asm to support
7 channel playback.

CRs-Fixed: 1114041
Change-Id: I0daf306d4495d3966406d05fc6328b2956a2d11f
Signed-off-by: Rohit Kumar <rohitkr@codeaurora.org>
7 years agommc: core: Increase the runtime PM reference count in try_claim_host
Veerabhadrarao Badiganti [Tue, 4 Apr 2017 07:08:24 +0000 (12:38 +0530)]
mmc: core: Increase the runtime PM reference count in try_claim_host

Runtime PM reference count is being increased in mmc_claim_host() and
is decreased in mmc_release_host(). This reference count is kept
during the complete cycle of a claim -> release host.

Same need to be done even in mmc_try_claim_host() as well. Increase
the runtime PM reference count by invoking pm_runtime_get_sync() from
mmc_try_claim_host() upon first successful claim.

Without this change the runtime PM reference count goes for a toss
since count is not getting incremented in mmc_try_claim_host() but is
getting decremented in mmc_release_host().

Change-Id: I77836875b4700a4bf3dbde2bf1abdf2ad36c4cac
Signed-off-by: Veerabhadrarao Badiganti <vbadigan@codeaurora.org>
7 years agoMerge "ASoC: mbhc: Modify special headset detection"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:17 +0000 (03:08 -0700)]
Merge "ASoC: mbhc: Modify special headset detection"

7 years agoMerge "ASoC: msm-lsm-client: cleanup ioctl functions"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:17 +0000 (03:08 -0700)]
Merge "ASoC: msm-lsm-client: cleanup ioctl functions"

7 years agoMerge "ASoC: wcd934x: Fix sequence for efuse sensing"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:16 +0000 (03:08 -0700)]
Merge "ASoC: wcd934x: Fix sequence for efuse sensing"

7 years agoMerge "drivers: soc: apr: remove GLINK_TX_REQ_INTENT flag in apr write"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:15 +0000 (03:08 -0700)]
Merge "drivers: soc: apr: remove GLINK_TX_REQ_INTENT flag in apr write"

7 years agoMerge "ARM: dts: msm: add support for secondary TDM on msm8996 adp/cdp"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:14 +0000 (03:08 -0700)]
Merge "ARM: dts: msm: add support for secondary TDM on msm8996 adp/cdp"

7 years agoMerge "ARM: dts: msm: add support for secondary MI2S on msm8996 adp/cdp"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:13 +0000 (03:08 -0700)]
Merge "ARM: dts: msm: add support for secondary MI2S on msm8996 adp/cdp"

7 years agoMerge "msm: mdss: Validating data values before passing on as operand"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:12 +0000 (03:08 -0700)]
Merge "msm: mdss: Validating data values before passing on as operand"

7 years agoMerge "msm: thermal-dev: Protect IOCTL from race condition"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:10 +0000 (03:08 -0700)]
Merge "msm: thermal-dev: Protect IOCTL from race condition"

7 years agoMerge "msm: ipa3: fix Q6 statuses exception path"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:09 +0000 (03:08 -0700)]
Merge "msm: ipa3: fix Q6 statuses exception path"

7 years agoMerge "mhi: mhi_uci: redesign mhi_uci driver"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:08 +0000 (03:08 -0700)]
Merge "mhi: mhi_uci: redesign mhi_uci driver"

7 years agoMerge "mhi: core: add support for subsystem restart and shutdown"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:08 +0000 (03:08 -0700)]
Merge "mhi: core: add support for subsystem restart and shutdown"

7 years agoMerge "usb: dwc3: Use high priority worker as bottom half handler"
Linux Build Service Account [Tue, 4 Apr 2017 10:08:04 +0000 (03:08 -0700)]
Merge "usb: dwc3: Use high priority worker as bottom half handler"