OSDN Git Service

sagit-ice-cold/kernel_xiaomi_msm8998.git
7 years agoadreno_tz: Remove unnecessary devfreq NULL check in tz_handler()
Hareesh Gundu [Fri, 26 May 2017 10:20:25 +0000 (15:50 +0530)]
adreno_tz: Remove unnecessary devfreq NULL check in tz_handler()

Caller of tz_handler() is having the NULL check for devfreq,So
we don’t need unnecessary NULL pointer check.

CRs-Fixed: 2046922
Change-Id: I42eb70ab19b4a5af193521d77bca86447a1ecdb8
Signed-off-by: Hareesh Gundu <hareeshg@codeaurora.org>
7 years agoMerge "msm: kgsl: Return mementry only if pending_free is not set"
Linux Build Service Account [Thu, 25 May 2017 22:52:15 +0000 (15:52 -0700)]
Merge "msm: kgsl: Return mementry only if pending_free is not set"

7 years agoMerge "leds: leds-qpnp: Fix uninitialized variable use"
Linux Build Service Account [Thu, 25 May 2017 22:52:14 +0000 (15:52 -0700)]
Merge "leds: leds-qpnp: Fix uninitialized variable use"

7 years agoMerge "btfm: Don't program multi channel registers in wcn3990 for SCO"
Linux Build Service Account [Thu, 25 May 2017 22:52:07 +0000 (15:52 -0700)]
Merge "btfm: Don't program multi channel registers in wcn3990 for SCO"

7 years agoMerge "power: qpnp-fg-gen3: update default empty voltage to 2.85 V"
Linux Build Service Account [Thu, 25 May 2017 22:52:06 +0000 (15:52 -0700)]
Merge "power: qpnp-fg-gen3: update default empty voltage to 2.85 V"

7 years agoMerge "msm: kgsl: Don't delete uninitialized preemption timer"
Linux Build Service Account [Thu, 25 May 2017 22:52:05 +0000 (15:52 -0700)]
Merge "msm: kgsl: Don't delete uninitialized preemption timer"

7 years agoMerge "msm_11ad: fix device-tree property name"
Linux Build Service Account [Thu, 25 May 2017 22:52:05 +0000 (15:52 -0700)]
Merge "msm_11ad: fix device-tree property name"

7 years agoMerge "net: rmnet_data: Unset logical end points in bridge mode"
Linux Build Service Account [Thu, 25 May 2017 22:52:04 +0000 (15:52 -0700)]
Merge "net: rmnet_data: Unset logical end points in bridge mode"

7 years agoMerge "soc: qcom: smp2p: Fix simultaneous access of global loopback data"
Linux Build Service Account [Thu, 25 May 2017 22:52:03 +0000 (15:52 -0700)]
Merge "soc: qcom: smp2p: Fix simultaneous access of global loopback data"

7 years agoMerge "qcom: common_log: make dump registration as async"
Linux Build Service Account [Thu, 25 May 2017 22:52:02 +0000 (15:52 -0700)]
Merge "qcom: common_log: make dump registration as async"

7 years agoMerge "spmi-pmic-arb: add support to dispatch interrupt based on IRQ status"
Linux Build Service Account [Thu, 25 May 2017 22:52:01 +0000 (15:52 -0700)]
Merge "spmi-pmic-arb: add support to dispatch interrupt based on IRQ status"

7 years agoMerge "spcom: check error from ion_import_dma_buf()"
Linux Build Service Account [Thu, 25 May 2017 22:52:01 +0000 (15:52 -0700)]
Merge "spcom: check error from ion_import_dma_buf()"

7 years agoMerge "spcom: silent error messages until spss link is up"
Linux Build Service Account [Thu, 25 May 2017 22:52:00 +0000 (15:52 -0700)]
Merge "spcom: silent error messages until spss link is up"

7 years agoMerge "usb: pdphy: Widen voltage range setting for VDD regulator"
Linux Build Service Account [Thu, 25 May 2017 22:51:59 +0000 (15:51 -0700)]
Merge "usb: pdphy: Widen voltage range setting for VDD regulator"

7 years agoMerge "usb : dwc3: Initialize kernel stack variables properly"
Linux Build Service Account [Thu, 25 May 2017 22:51:58 +0000 (15:51 -0700)]
Merge "usb : dwc3: Initialize kernel stack variables properly"

7 years agoMerge "ASoC: msm: qdsp6v2: Add IEC61937 compressed HDMI pass-though"
Linux Build Service Account [Thu, 25 May 2017 22:51:57 +0000 (15:51 -0700)]
Merge "ASoC: msm: qdsp6v2: Add IEC61937 compressed HDMI pass-though"

7 years agoMerge "ASoC: msm: qdsp6v2: Extend pass through mode per stream"
Linux Build Service Account [Thu, 25 May 2017 22:51:56 +0000 (15:51 -0700)]
Merge "ASoC: msm: qdsp6v2: Extend pass through mode per stream"

7 years agoMerge "msm: ipa: Fix post init in workqueue context"
Linux Build Service Account [Thu, 25 May 2017 22:51:55 +0000 (15:51 -0700)]
Merge "msm: ipa: Fix post init in workqueue context"

7 years agoMerge "spcom: increment channel reference count after successful glink_open()"
Linux Build Service Account [Thu, 25 May 2017 22:51:54 +0000 (15:51 -0700)]
Merge "spcom: increment channel reference count after successful glink_open()"

7 years agoMerge "USB: f_qc_rndis: Prevent use-after-free for _rndis_qc"
Linux Build Service Account [Thu, 25 May 2017 22:51:53 +0000 (15:51 -0700)]
Merge "USB: f_qc_rndis: Prevent use-after-free for _rndis_qc"

7 years agoMerge "ARM: dts: msm: Add iris support on sdm630"
Linux Build Service Account [Thu, 25 May 2017 22:51:53 +0000 (15:51 -0700)]
Merge "ARM: dts: msm: Add iris support on sdm630"

7 years agoMerge "mmc: block: Disable clock scaling during shutdown"
Linux Build Service Account [Thu, 25 May 2017 22:51:51 +0000 (15:51 -0700)]
Merge "mmc: block: Disable clock scaling during shutdown"

7 years agoMerge "drm/sde: only set split registers for DSI interfaces"
Linux Build Service Account [Thu, 25 May 2017 22:51:50 +0000 (15:51 -0700)]
Merge "drm/sde: only set split registers for DSI interfaces"

7 years agoMerge "SDM660: ADSPRPC: Fix for NULL pointer dereference"
Linux Build Service Account [Thu, 25 May 2017 22:51:49 +0000 (15:51 -0700)]
Merge "SDM660: ADSPRPC: Fix for NULL pointer dereference"

7 years agoMerge "drm/msm: wait fence complete before flip"
Linux Build Service Account [Thu, 25 May 2017 22:51:48 +0000 (15:51 -0700)]
Merge "drm/msm: wait fence complete before flip"

7 years agoMerge "soc: qcom: Fix failure prints in rpm stats and log"
Linux Build Service Account [Thu, 25 May 2017 22:51:48 +0000 (15:51 -0700)]
Merge "soc: qcom: Fix failure prints in rpm stats and log"

7 years agoMerge "policy_engine: Don't allow MODE change if no active connection"
Linux Build Service Account [Thu, 25 May 2017 22:51:47 +0000 (15:51 -0700)]
Merge "policy_engine: Don't allow MODE change if no active connection"

7 years agoMerge "power: qpnp-smb2: remove workaround for USBIN supplied VCONN"
Linux Build Service Account [Thu, 25 May 2017 22:51:46 +0000 (15:51 -0700)]
Merge "power: qpnp-smb2: remove workaround for USBIN supplied VCONN"

7 years agoMerge "msm: ais: Security fixes for ispif and cpp"
Linux Build Service Account [Thu, 25 May 2017 22:51:44 +0000 (15:51 -0700)]
Merge "msm: ais: Security fixes for ispif and cpp"

7 years agoMerge "ARM: dts: msm: Add USB2 device node for SDM660"
Linux Build Service Account [Thu, 25 May 2017 22:51:42 +0000 (15:51 -0700)]
Merge "ARM: dts: msm: Add USB2 device node for SDM660"

7 years agoMerge "coresight-remote-etm: Adds missing lock to avoid race condition"
Linux Build Service Account [Thu, 25 May 2017 22:51:40 +0000 (15:51 -0700)]
Merge "coresight-remote-etm: Adds missing lock to avoid race condition"

7 years agoMerge "Merge remote-tracking branch 'origin/tmp-285c137' into msm-4.4"
Linux Build Service Account [Thu, 25 May 2017 22:51:39 +0000 (15:51 -0700)]
Merge "Merge remote-tracking branch 'origin/tmp-285c137' into msm-4.4"

7 years agoMerge "USB: composite: Advertise USB version as 0x200 if max speed is HS"
Linux Build Service Account [Thu, 25 May 2017 22:51:38 +0000 (15:51 -0700)]
Merge "USB: composite: Advertise USB version as 0x200 if max speed is HS"

7 years agoMerge "ARM: dts: msm: Update ref-clk node for sdm660"
Linux Build Service Account [Thu, 25 May 2017 22:51:37 +0000 (15:51 -0700)]
Merge "ARM: dts: msm: Update ref-clk node for sdm660"

7 years agoMerge "usb: phy-msm-qusb-v2: Add support for reference counting of regulators"
Linux Build Service Account [Thu, 25 May 2017 22:51:36 +0000 (15:51 -0700)]
Merge "usb: phy-msm-qusb-v2: Add support for reference counting of regulators"

7 years agoMerge "ASoC: msm-cpe: Fix range checking in function fw_name_store"
Linux Build Service Account [Thu, 25 May 2017 22:51:34 +0000 (15:51 -0700)]
Merge "ASoC: msm-cpe: Fix range checking in function fw_name_store"

7 years agoMerge "input: synaptics_dsx: fix mutex init"
Linux Build Service Account [Thu, 25 May 2017 22:51:34 +0000 (15:51 -0700)]
Merge "input: synaptics_dsx: fix mutex init"

7 years agoMerge "soc: qcom: msm_perf: Initialize Array elements"
Linux Build Service Account [Thu, 25 May 2017 22:51:33 +0000 (15:51 -0700)]
Merge "soc: qcom: msm_perf: Initialize Array elements"

7 years agoMerge "ARM: dts: msm: Add flash handle to camera1 device on SDM660"
Linux Build Service Account [Thu, 25 May 2017 22:51:32 +0000 (15:51 -0700)]
Merge "ARM: dts: msm: Add flash handle to camera1 device on SDM660"

7 years agoMerge "drm/msm: deal with arbitrary # of cmd buffers"
Linux Build Service Account [Thu, 25 May 2017 22:51:31 +0000 (15:51 -0700)]
Merge "drm/msm: deal with arbitrary # of cmd buffers"

7 years agoMerge "msm: camera: isp : Fix race condition in close sequence"
Linux Build Service Account [Thu, 25 May 2017 22:51:30 +0000 (15:51 -0700)]
Merge "msm: camera: isp : Fix race condition in close sequence"

7 years agoMerge "ARM: dts: msm: Add entry for video throttle clock on sdm660 and sdm630"
Linux Build Service Account [Thu, 25 May 2017 22:51:29 +0000 (15:51 -0700)]
Merge "ARM: dts: msm: Add entry for video throttle clock on sdm660 and sdm630"

7 years agoMerge "firmware_class: Avoid pending list corruption"
Linux Build Service Account [Thu, 25 May 2017 22:51:28 +0000 (15:51 -0700)]
Merge "firmware_class: Avoid pending list corruption"

7 years agoMerge "ASoC: msm: qdsp6v2: Fix voice call and record SSR test"
Linux Build Service Account [Thu, 25 May 2017 22:51:27 +0000 (15:51 -0700)]
Merge "ASoC: msm: qdsp6v2: Fix voice call and record SSR test"

7 years agoMerge "usb: policy_engine: Close pd phy after exceeding SINK hard reset count"
Linux Build Service Account [Thu, 25 May 2017 22:51:26 +0000 (15:51 -0700)]
Merge "usb: policy_engine: Close pd phy after exceeding SINK hard reset count"

7 years agomsm: kgsl: Return mementry only if pending_free is not set
Deepak Kumar [Mon, 8 May 2017 09:59:03 +0000 (15:29 +0530)]
msm: kgsl: Return mementry only if pending_free is not set

Return mementry from kgsl_sharedmem_find only if pending_free
is not set for that mementry. This is necessary to avoid use
of a mementry after it is already marked for free.

Change-Id: I23111e9c82a88ccbda2ab259074c38d91f9ff5cb
Signed-off-by: Deepak Kumar <dkumar@codeaurora.org>
7 years agoleds: leds-qpnp: Fix uninitialized variable use
Ankit Sharma [Mon, 22 May 2017 11:20:52 +0000 (16:50 +0530)]
leds: leds-qpnp: Fix uninitialized variable use

Fix use of uninitialized variable.

CRs-Fixed: 2050786
Change-Id: I3e0ac770df89347b878c9fe6eeaf5e4d8d53537c
Signed-off-by: Ankit Sharma <ansharma@codeaurora.org>
7 years agomsm: kgsl: Don't delete uninitialized preemption timer
Wenbin Wang [Thu, 30 Mar 2017 06:10:19 +0000 (14:10 +0800)]
msm: kgsl: Don't delete uninitialized preemption timer

If we delete uninitialized timer on CONFIG_DEBUG_OBJECTS
disabled build del_timer_sync() will block for ever. For
all A3xx targets preemption timer is not initialized, but
dispatcher fault handler is trying to delete the
uninitialized preemption timer. Fix this issue by adding
a preemption check before we delete it.

CRs-Fixed: 2023690
Change-Id: I2c51a0b2286b82bf2eb5ee68d923dd9585f07f00
Signed-off-by: Wenbin Wang <wwenbin@codeaurora.org>
Signed-off-by: Abhilash Kumar <krabhi@codeaurora.org>
7 years agopower: qpnp-fg-gen3: update default empty voltage to 2.85 V
Subbaraman Narayanamurthy [Fri, 19 May 2017 20:50:44 +0000 (13:50 -0700)]
power: qpnp-fg-gen3: update default empty voltage to 2.85 V

Based on the hardware recommendation, update the default empty
voltage from 2.8 V to 2.85 V. This is to prevent undervolting
in low temperatures (e.g. -20 C) when battery voltage drops below
2.8 V but the battery SOC may not reach 0% as the charger is
configured to issue GP_FAULT2 when battery voltage goes below
2.8 V. This helps in shutting down graciously.

Change-Id: Ifc68b825541c09d985d890ee7200ff8e0af2ba0f
Signed-off-by: Subbaraman Narayanamurthy <subbaram@codeaurora.org>
7 years agomsm_11ad: fix device-tree property name
Hamad Kadmany [Mon, 22 May 2017 10:33:32 +0000 (13:33 +0300)]
msm_11ad: fix device-tree property name

Change keep_radio_on_during_sleep DT property to follow
the naming convention.

Change-Id: I6450cfdfe38a4cf5b1b1ccb40b5506a34aeb076a
Signed-off-by: Hamad Kadmany <hkadmany@codeaurora.org>
Signed-off-by: Maya Erez <merez@codeaurora.org>
7 years agousb : dwc3: Initialize kernel stack variables properly
Sai Krishna Juturi [Wed, 17 May 2017 05:36:35 +0000 (11:06 +0530)]
usb : dwc3: Initialize kernel stack variables properly

If kernel stack variables are not initialized properly,
there  is a chance of kernel information disclosure.
So, initialize kernel stack variables with null characters.

CRs-fixed: 2042592
Change-Id: I213c0e5c7f67824c2cecace276ff2f8f81599d51
Signed-off-by: Sai Krishna Juturi <jsaikrishna@codeaurora.org>
7 years agomsm: ais: Security fixes for ispif and cpp
Terence Ho [Fri, 19 May 2017 19:05:14 +0000 (15:05 -0400)]
msm: ais: Security fixes for ispif and cpp

Check the user parameters in ispif to prevent buffer overflow.
Fix out-of-bounds frame or command buffer access in cpp.

CRs-fixed: 1097390 2002207
Change-Id: Ifdaf223d1b1c418596c2fb43d970dad728abd4a3
Signed-off-by: Terence Ho <terenceh@codeaurora.org>
Signed-off-by: Andy Sun <bins@codeaurora.org>
7 years agobtfm: Don't program multi channel registers in wcn3990 for SCO
Satish Kodishala [Thu, 18 May 2017 14:41:20 +0000 (20:11 +0530)]
btfm: Don't program multi channel registers in wcn3990 for SCO

For SCO use cases, we should not program multi chanel
registers. Programming this leading to silence randomly
in SCO Rx/Tx.

CRs-Fixed: 2048423
Change-Id: I5e41591fb81b5bd3cf9358819c0a62a5cfe60e13
Signed-off-by: Satish Kodishala <skodisha@codeaurora.org>
7 years agospcom: check error from ion_import_dma_buf()
Amir Samuelov [Mon, 15 May 2017 07:38:00 +0000 (10:38 +0300)]
spcom: check error from ion_import_dma_buf()

Check ion_import_dma_buf() returned handle is not NULL or error code,
by using IS_ERR_OR_NULL() macro.
Check channel name is null terminated on create channel command.
Check minimal command size from user space.
Fix spelling.

Change-Id: I81631c1844982952498eaf2bff84ff90578720b2
Signed-off-by: Amir Samuelov <amirs@codeaurora.org>
7 years agospcom: silent error messages until spss link is up
Amir Samuelov [Sun, 21 May 2017 11:00:09 +0000 (14:00 +0300)]
spcom: silent error messages until spss link is up

After spss reset (a.k.a SSR) the spdaemon tries to open spcom char dev.
Avoid flooding the kernel log with error messages until the spss link
is up.

Change-Id: I7abd37f50fa4c172e07c8551c287b6ccf0511a44
Signed-off-by: Amir Samuelov <amirs@codeaurora.org>
7 years agospcom: increment channel reference count after successful glink_open()
Amir Samuelov [Thu, 18 May 2017 13:10:01 +0000 (16:10 +0300)]
spcom: increment channel reference count after successful glink_open()

spcom_open() should increment the ref count only after successful glink
open the channel.
Otherwise, the following attempts to open the channel will fail.
glink_open() failure was noticed after SP reset (SSR), when attempting
to open the channel before the link between SP and HLOS is up.

Change-Id: I8be28ce810f54cc3fa95d14c7244921acf7edaa7
Signed-off-by: Amir Samuelov <amirs@codeaurora.org>
7 years agoMerge "qpnp-smb2: fix FV and FCC reporting"
Linux Build Service Account [Sat, 20 May 2017 08:37:19 +0000 (01:37 -0700)]
Merge "qpnp-smb2: fix FV and FCC reporting"

7 years agoMerge "msm: vidc: Cancel userptr validation during secure decode"
Linux Build Service Account [Sat, 20 May 2017 08:37:18 +0000 (01:37 -0700)]
Merge "msm: vidc: Cancel userptr validation during secure decode"

7 years agoMerge "ASoC: msm: qdspv2: Rename HDMI drift mixer controls"
Linux Build Service Account [Sat, 20 May 2017 08:37:17 +0000 (01:37 -0700)]
Merge "ASoC: msm: qdspv2: Rename HDMI drift mixer controls"

7 years agoMerge "ion: fix logical-not comparison warning"
Linux Build Service Account [Sat, 20 May 2017 08:37:17 +0000 (01:37 -0700)]
Merge "ion: fix logical-not comparison warning"

7 years agoMerge "coresight-tmc: Adds lock to avoid race condition"
Linux Build Service Account [Sat, 20 May 2017 08:37:16 +0000 (01:37 -0700)]
Merge "coresight-tmc: Adds lock to avoid race condition"

7 years agoMerge "msm: wlan: Update regulatory database"
Linux Build Service Account [Sat, 20 May 2017 08:37:15 +0000 (01:37 -0700)]
Merge "msm: wlan: Update regulatory database"

7 years agoMerge "msm_11ad: clear affinity hint upon low bus request"
Linux Build Service Account [Sat, 20 May 2017 08:37:13 +0000 (01:37 -0700)]
Merge "msm_11ad: clear affinity hint upon low bus request"

7 years agoMerge "ARM: dts: msm: Add support for multiple neutrino on msm8996"
Linux Build Service Account [Sat, 20 May 2017 08:37:13 +0000 (01:37 -0700)]
Merge "ARM: dts: msm: Add support for multiple neutrino on msm8996"

7 years agoMerge "drm/msm: Don't allow zero sized buffer objects"
Linux Build Service Account [Sat, 20 May 2017 08:37:12 +0000 (01:37 -0700)]
Merge "drm/msm: Don't allow zero sized buffer objects"

7 years agoMerge "msm: camera: sensor: Validate sensor related name length"
Linux Build Service Account [Sat, 20 May 2017 08:37:11 +0000 (01:37 -0700)]
Merge "msm: camera: sensor: Validate sensor related name length"

7 years agoMerge "msm: camera: sensor: Validate sensor name during probe"
Linux Build Service Account [Sat, 20 May 2017 08:37:10 +0000 (01:37 -0700)]
Merge "msm: camera: sensor: Validate sensor name during probe"

7 years agoqpnp-smb2: fix FV and FCC reporting
Abhijeet Dharmapurikar [Mon, 15 May 2017 20:38:53 +0000 (13:38 -0700)]
qpnp-smb2: fix FV and FCC reporting

Currently the read of battery's voltage_max property (FV) and battery's
current_max property (FCC) returns the value of DEFAULT_VOTER voter.

But such a voter doesn't exist on battery's voltage_max property. Fix it
by returning the value voted by BATT_PROFILE_VOTER.

Change-Id: I2dac98f2ce9a423d5045788dbe120591f10c5a9e
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agosmb-lib: do not disable parallel based only on main's settled current
Abhijeet Dharmapurikar [Thu, 18 May 2017 22:37:56 +0000 (15:37 -0700)]
smb-lib: do not disable parallel based only on main's settled current

Currently main charger ends up voting to disable parallel(USBIN_I_VOTER)
if the settled current is less than 1.4Amps. It could be that the
parallel path is drawing some current and the draw is higher than
1.4Amps.

Move this disabling to battery, where the parallel charger's input current
is taken in to consideration.

Also while voting to disable, make sure that the main's input is limited.
If there is more room for ICL to increase that means more power could be
drawn from the charger and so we must keep the parallel charging enabled.
If the ICL ceiling itself if 1.4Amps then disable parallel charging,
because even if ICL has room to increase it wont cross 1.4Amps.

The main's input limited state is set while AICL steps up - this
ensures that we keep parallel disabled while AICL is running and settled
is less than 1.4Amps.

When the charger is removed, the main's input limited status is unset,
however we want to keep the parallel charger disabled. In such situations
the main's settled current is 0mA - use this to vote for disabling
parallel charging.

For any ICL change vote request:
Disable parallel using USBIN_I_VOTER if requested current is less than
1.4Amps and to un-vote (i.e. when ICL vote is higher than 1.4Amps) rely
on event from AICL done or status_change delayed work.

Note that USBIN_I_VOTER is effective only for non-QC chargers. For QC
chargers USBIN_V_VOTER continues to keep charging enabled since parallel
need not be disabled is a lower current is requested with a QC charger.

Change-Id: I61d7cd161a87de2904470b2d870058a0fe199116
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agonet: rmnet_data: Unset logical end points in bridge mode
Ashwanth Goli [Fri, 24 Feb 2017 18:03:03 +0000 (11:03 -0700)]
net: rmnet_data: Unset logical end points in bridge mode

We clean up the logical end points only for the un-registering device
in bridge mode. However, the other physical end point's local end
point still points to the the un-registered device.

This may lead up to a crash if one of the physical devices in bridge
mode is un-registered. Fix this by unsetting the local endpoint.

It is still possible that packets in a different context across cores
might try to access this data. This usually manifests as packets
requesting a very large headroom. Handle this by dropping these stale
skb's.

CRs-Fixed: 1098513
Change-Id: I1ba4d877a6ed3eca66946fe056938f0927bcd9a5
Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: Ashwanth Goli <ashwanth@codeaurora.org>
7 years agommc: block: Disable clock scaling during shutdown
Vijay Viswanath [Tue, 16 May 2017 05:04:21 +0000 (10:34 +0530)]
mmc: block: Disable clock scaling during shutdown

During shutdown of sdhc drivers, power off notification (PON) is sent
to eMMC device. But it doesn't prevent the clock scaling framework
from kicking in and sending commands to card after PON is sent. So
exit clock scaling framework of sdhc before PON is sent.

Change-Id: Ibfe097732042458965db529f235a7fd28cfabaab
Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
7 years agommc: core: Avoid returning error when clok scaling devfreq is removed
Vijay Viswanath [Fri, 19 May 2017 10:56:33 +0000 (16:26 +0530)]
mmc: core: Avoid returning error when clok scaling devfreq is removed

When resume clock scaling is not possible because there is no devfreq
framework, it is not an error. So avoid printing an error and return
success. This is necessary in certain scenarios like during shutdown
where we unregister the clock scaling devfreq.

Change-Id: Ib37eabdcdbbb58a336ee87220e5758bb559625af
Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
7 years agoARM: dts: msm: Add USB2 device node for SDM660
Vijayavardhan Vennapusa [Tue, 18 Apr 2017 06:31:19 +0000 (12:01 +0530)]
ARM: dts: msm: Add USB2 device node for SDM660

Add USB2 device node for SDM660 for using secondary USB2 port
in host only mode.

Change-Id: I8ff06411902bcafbb5b3f429634760d4c44a7b20
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
7 years agoARM: dts: msm: Add iris support on sdm630
Tanvi Aggarwal [Fri, 5 May 2017 12:33:46 +0000 (18:03 +0530)]
ARM: dts: msm: Add iris support on sdm630

Add IRIS camera support on sdm630 mtp platform.
This is required for secure camera feature.

Change-Id: Ic18eb896e7e2099debcf755dca4278c25295f8b9
CRs-Fixed: 2042077
Signed-off-by: Tanvi Aggarwal <tanvia@codeaurora.org>
7 years agosoc: qcom: Fix failure prints in rpm stats and log
Maulik Shah [Thu, 18 May 2017 04:25:59 +0000 (09:55 +0530)]
soc: qcom: Fix failure prints in rpm stats and log

Correct failure prints in rpm stats and rpm log drivers.

Change-Id: I4a2d4a8cf4df4c9c1d750a65beb7454041fc06c2
Signed-off-by: Maulik Shah <mkshah@codeaurora.org>
7 years agoASoC: msm: qdsp6v2: Add IEC61937 compressed HDMI pass-though
Ben Romberger [Fri, 7 Apr 2017 00:23:15 +0000 (17:23 -0700)]
ASoC: msm: qdsp6v2: Add IEC61937 compressed HDMI pass-though

Add support for IEC61937 which is an audio packaging
standard used for compressed HDMI pass-through. Any
audio format can be sent through HDMI pass-through
formatted as IEC61937.

Change-Id: I09821add306e2aeae6530cecbac189471a55e589
Signed-off-by: Ben Romberger <bromberg@codeaurora.org>
7 years agomsm: vidc: Cancel userptr validation during secure decode
Prabhakar Reddy Krishnappa [Fri, 19 May 2017 00:43:17 +0000 (17:43 -0700)]
msm: vidc: Cancel userptr validation during secure decode

Secure decode doesn't need userptr and uses ion fd instead. We were
conducting userptr validation regardless of the decode mode (secure/
unsecure). This forced user to populate userptr with a dummy value
during QBUF in secure mode decode on both output and capture ports
to avoid a userptr validation fail.

CRs-Fixed: 2049213
Change-Id: I0060efb52792201a2634072f648a537ebb02d17c
Signed-off-by: Prabhakar Reddy Krishnappa <prkrishn@codeaurora.org>
7 years agoARM: dts: msm: Remove TCSR_QUSB2_PHY_CLAMP_DIG_N_1P8 for SDM660
Vijayavardhan Vennapusa [Thu, 4 May 2017 07:40:10 +0000 (13:10 +0530)]
ARM: dts: msm: Remove TCSR_QUSB2_PHY_CLAMP_DIG_N_1P8 for SDM660

Remove TCSR_QUSB2_PHY_CLAMP_DIG_N_1P8 resource entry for SDM660
which is not required to set now during disconnect as writing to this
register causes USB3.0 port and USB2.0 ports working independently.

Change-Id: Ifb91c4e493ca1ecd31ef649c9e85acd809b9eec6
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
7 years agousb-phy-qusb: powerdown PHY during disconnect to avoid leakage
Vijayavardhan Vennapusa [Thu, 4 May 2017 07:30:54 +0000 (13:00 +0530)]
usb-phy-qusb: powerdown PHY during disconnect to avoid leakage

Currently PHY driver is using TCSR_QUSB2PHY_CLAMP_DIG_N_1P8 register
to avoid leakage issue seen during disconnect. But this could fail
USB ports working independently if this register is used as that
could clamp both inputs going to two QUSB2 PHY on platforms having
USB2 port. To fix this, it is recommended to powerdown QUSB2 PHY
by setting powerdown bit in  QUSB2PHY_PORT_POWER_DOWN register
during disconnect and removing writing 0x0 to TCSR register. Removed
updating TCSR register from dts file as separate change.

Change-Id: I6b0915faa346e5d88310106eeddb3c7a1bfb1d08
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
7 years agoUSB:dwc3-msm: Don't schedule work if pm_qos_latency is zero
Vijayavardhan Vennapusa [Wed, 26 Apr 2017 05:40:15 +0000 (11:10 +0530)]
USB:dwc3-msm: Don't schedule work if pm_qos_latency is zero

Currently driver is scheduling pm_qos_work again even though
pm_qos_latency is not passed from dts file. Add a check for
pm_qos_latency and don't schedule work if pm_qos_latency value
is zero. Also remove use of static variable for last_irq_count and
add the variable for the same in dwc3 structure for the case where
multi DWC3 usages are used.

Change-Id: I55e1e3a7d48fbea0a421802aae176ac57a48869f
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
7 years agocoresight-tmc: Adds lock to avoid race condition
Mukesh Ojha [Fri, 5 May 2017 08:44:56 +0000 (14:14 +0530)]
coresight-tmc: Adds lock to avoid race condition

There can be a case of race in the call to tmc_etr_bam_disable(). The call
from usb_notifier() to tmc_etr_bam_disable() is within a lock
'drvdata->mem_lock' but the call from tmc_disable is not under any lock
which can lead to race.

This patch adds the missing lock.

CRs-Fixed: 2047196
Change-Id: I3986b498ae8e2b2ddfe8f5b3315fdefab197d11c
Signed-off-by: Mukesh Ojha <mojha@codeaurora.org>
7 years agoSDM660: ADSPRPC: Fix for NULL pointer dereference
Tharun Kumar Merugu [Tue, 16 May 2017 18:10:05 +0000 (23:40 +0530)]
SDM660: ADSPRPC: Fix for NULL pointer dereference

Fixing NULL pointer dereference in fastrpc_device_ioctl

Change-Id: I240e6eae69f9ab4831352a8bf160d5f0e8a98b1b
Acked-by: Chenna Kesava Raju <chennak@qti.qualcomm.com>
Signed-off-by: Tharun Kumar Merugu <mtharu@codeaurora.org>
7 years agomsm: ipa: Fix post init in workqueue context
Utkarsh Saxena [Wed, 3 May 2017 08:34:30 +0000 (14:04 +0530)]
msm: ipa: Fix post init in workqueue context

ipa post init is accessed from user space
context from ops structure.

While user space program triggered post init
there could be a chance, where the
user space context may get killed at any time,
due to which the memory allocation in post init
gets failed and in turn which clean up the
IPA driver.

Because the memory allocation of the
sigkill pending process will fail.

Make changes in ops structure write func-pointer to
queue a work instead of directly invoking post init.

So that ipa post init will execute in workqueue context.

Change-Id: Ia11b3a868d960111422c357bcc9924ef7b220e67
Acked-by: Mohammed Javid <mjavid@qti.qualcomm.com>
Signed-off-by: Utkarsh Saxena <usaxena@codeaurora.org>
7 years agopolicy_engine: Don't allow MODE change if no active connection
Vijayavardhan Vennapusa [Tue, 16 May 2017 09:25:51 +0000 (14:55 +0530)]
policy_engine: Don't allow MODE change if no active connection

Currently driver is allowing MODE change even if there is no active
cable connection and is waiting for to complete.  For this MODE
change to complete,  driver first sets power role to none to do
type C disconnect and then switch to new mode.  Here disconnect won't
come as there is no active cable connection. Fix this issue by returning
error in case there is no active connection.

Change-Id: Ie423bd8b61ab7b2624f98c84e8d8e75b6c853b00
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
7 years agoion: fix logical-not comparison warning
Sudarshan Rajagopalan [Mon, 1 May 2017 22:52:29 +0000 (15:52 -0700)]
ion: fix logical-not comparison warning

Fixes compiler warning when a logical-not on a variable
is applied while comparing it with value zero.

Change-Id: I92aa16bd0d57a0d59ecd26eef1ac92220332998d
Signed-off-by: Sudarshan Rajagopalan <sudaraja@codeaurora.org>
7 years agomsm_11ad: clear affinity hint upon low bus request
Dedy Lansky [Tue, 16 May 2017 12:18:02 +0000 (15:18 +0300)]
msm_11ad: clear affinity hint upon low bus request

When driver calls ops_bus_request with high bandwidth requirement,
affinity of wil6210 interrupt is set to first big core. However this
setting is not cleared when driver lowers bandwidth requirement.
This causes WARN_ON_ONCE(desc->affinity_hint) upon wil6210 rmmod.

Fix this by clearing affinity hint upon low bandwidth request.

Change-Id: I87b6b7ec9b369b84a9d3724d92a821a1302d1f83
Signed-off-by: Dedy Lansky <dlansky@codeaurora.org>
7 years agoASoC: msm: qdspv2: Rename HDMI drift mixer controls
Manish Dewangan [Wed, 17 May 2017 01:30:05 +0000 (07:00 +0530)]
ASoC: msm: qdspv2: Rename HDMI drift mixer controls

Update HDMI drift mixer controls name to be aligned with
drift mixer controls for other devices(speaker, BT).

CRs-Fixed: 2036899
Change-Id: If7aa29ea9511b65de71ff12143a1c34d977de2b8
Signed-off-by: Manish Dewangan <manish@codeaurora.org>
7 years agoARM: dts: msm: Update ref-clk node for sdm660
Sayali Lokhande [Wed, 17 May 2017 06:26:25 +0000 (11:56 +0530)]
ARM: dts: msm: Update ref-clk node for sdm660

As a part of 'commit 64db44d65ef5 ("ARM: dts: msm:
move ref-clk from phy to ufs node in sdm660")',
ref-clk node is moved from phy to ufs and ref-clk
related property should be prefixed with "qcom,".
This change updates the same for sdm660 mtp so as
to turn-off the ref-clk supply when link is off.

Change-Id: Iee0fcaaaa34108c35d836d6136f8c1a0964f0fa4
Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org>
7 years agoinput: synaptics_dsx: fix mutex init
Mohan Pallaka [Mon, 8 May 2017 18:51:41 +0000 (11:51 -0700)]
input: synaptics_dsx: fix mutex init

Initialize mutex before using in irq handler.

CRs-fixed: 2047720
Change-Id: I5ebfbc22f2a64dde239fcbb13eb26be50543b484
Signed-off-by: Mohan Pallaka <mpallaka@codeaurora.org>
7 years agoMerge "msm: ADSPRPC: Copy process name buffer to kernel space"
Linux Build Service Account [Tue, 16 May 2017 23:08:16 +0000 (16:08 -0700)]
Merge "msm: ADSPRPC: Copy process name buffer to kernel space"

7 years agoMerge "usb: gadget: f_cdev: Fix NULL pointer dereference in cser_free_inst"
Linux Build Service Account [Tue, 16 May 2017 23:08:15 +0000 (16:08 -0700)]
Merge "usb: gadget: f_cdev: Fix NULL pointer dereference in cser_free_inst"

7 years agoMerge "msm: mink: Avoid handling output params if TZ op returns an err"
Linux Build Service Account [Tue, 16 May 2017 23:08:14 +0000 (16:08 -0700)]
Merge "msm: mink: Avoid handling output params if TZ op returns an err"

7 years agoMerge "defconfig: msm: Remove CONFIG_TIMER_STATS"
Linux Build Service Account [Tue, 16 May 2017 23:08:13 +0000 (16:08 -0700)]
Merge "defconfig: msm: Remove CONFIG_TIMER_STATS"

7 years agoMerge "usb: phy: qusb: Support specifying vdda33 levels from device tree"
Linux Build Service Account [Tue, 16 May 2017 23:08:12 +0000 (16:08 -0700)]
Merge "usb: phy: qusb: Support specifying vdda33 levels from device tree"

7 years agoMerge "ARM: dts: msm: Enable support for guest virtual platform"
Linux Build Service Account [Tue, 16 May 2017 23:08:11 +0000 (16:08 -0700)]
Merge "ARM: dts: msm: Enable support for guest virtual platform"

7 years agoMerge "ARM: dts: msm: always enable mdss_mdp node for msm8998 based MOB3 platform"
Linux Build Service Account [Tue, 16 May 2017 23:08:08 +0000 (16:08 -0700)]
Merge "ARM: dts: msm: always enable mdss_mdp node for msm8998 based MOB3 platform"

7 years agoMerge "msm: msm_bus: Add mutex to protect node_list"
Linux Build Service Account [Tue, 16 May 2017 23:08:07 +0000 (16:08 -0700)]
Merge "msm: msm_bus: Add mutex to protect node_list"

7 years agoMerge "msm_11ad: add support to PCIe D3hot in system suspend"
Linux Build Service Account [Tue, 16 May 2017 23:08:06 +0000 (16:08 -0700)]
Merge "msm_11ad: add support to PCIe D3hot in system suspend"

7 years agousb: phy-msm-qusb-v2: Add support for reference counting of regulators
Hemant Kumar [Fri, 28 Apr 2017 02:39:58 +0000 (19:39 -0700)]
usb: phy-msm-qusb-v2: Add support for reference counting of regulators

Upon cable disconnect PMI calls qusb_phy_update_dpdm() to turn off phy
regulators before notifying vbus off to usb. As a result phy power down
sequence is executed with phy regulators turned off. This may result into
improper phy line state.  Fix this issue by adding a reference counter to
keep track of all regulator enable and disable requests and only disable
regulator when ref count becomes zero. Also add a mutex in
qusb_phy_enable_power() API to prevent any race condition in enabling and
disabling phy regulators between the callers i.e. PMI and phy driver.

Change-Id: I620f2b8cbf4f9271db81d5a517f1ee2a13c57f27
Signed-off-by: Hemant Kumar <hemantk@codeaurora.org>
7 years agoUSB: phy: qusb: Turn on vdd along with 1p8/3p3 LDOs when PMI requests
Hemant Kumar [Fri, 28 Apr 2017 00:51:11 +0000 (17:51 -0700)]
USB: phy: qusb: Turn on vdd along with 1p8/3p3 LDOs when PMI requests

QUSB PHY requires VDD, 1p8 and 3p1 regulators to remove any unwanted
pull downs on DP/DM lines. These pull downs may result in incorrect charger
detection by PMI. Avoid incorrect charger detection by turning on VDD,
1p8 and 3p1 whenever PMI requests DP/DM to be floating.

Change-Id: I27e68d5ebf2f49fb6571ad777318483f9c123407
Signed-off-by: Hemant Kumar <hemantk@codeaurora.org>