OSDN Git Service

sagit-ice-cold/kernel_xiaomi_msm8998.git
7 years agoMerge "tty: serial: msm: fix potential race b/w startup and irq handling"
Linux Build Service Account [Wed, 3 May 2017 02:30:29 +0000 (19:30 -0700)]
Merge "tty: serial: msm: fix potential race b/w startup and irq handling"

7 years agoMerge "msm: ipa3: fix a race condition on clock vote from suspend"
Linux Build Service Account [Wed, 3 May 2017 02:30:28 +0000 (19:30 -0700)]
Merge "msm: ipa3: fix a race condition on clock vote from suspend"

7 years agoMerge "fs: ecryptfs: Cache to be cleared after file write"
Linux Build Service Account [Wed, 3 May 2017 02:30:27 +0000 (19:30 -0700)]
Merge "fs: ecryptfs: Cache to be cleared after file write"

7 years agoMerge "msm: ipa: Add mutext protection when updating notify req cache"
Linux Build Service Account [Wed, 3 May 2017 02:30:26 +0000 (19:30 -0700)]
Merge "msm: ipa: Add mutext protection when updating notify req cache"

7 years agoMerge "msm: camera: Fix NULL pointer dereference in msm_vb2.c."
Linux Build Service Account [Wed, 3 May 2017 02:30:19 +0000 (19:30 -0700)]
Merge "msm: camera: Fix NULL pointer dereference in msm_vb2.c."

7 years agoMerge "ASoC: msm: qdsp6v2: Clear pass-through mode on routing close"
Linux Build Service Account [Tue, 2 May 2017 16:07:43 +0000 (09:07 -0700)]
Merge "ASoC: msm: qdsp6v2: Clear pass-through mode on routing close"

7 years agoMerge "ASoC: msm-lsm-client: use kzalloc instead of kmalloc"
Linux Build Service Account [Tue, 2 May 2017 16:07:42 +0000 (09:07 -0700)]
Merge "ASoC: msm-lsm-client: use kzalloc instead of kmalloc"

7 years agoMerge "icnss: remove pre-alloc memory leak check in platform driver"
Linux Build Service Account [Tue, 2 May 2017 16:07:41 +0000 (09:07 -0700)]
Merge "icnss: remove pre-alloc memory leak check in platform driver"

7 years agoMerge "Merge remote-tracking branch 'remotes/quic/dev/msm-4.4-8996au' into msm-4.4"
Linux Build Service Account [Tue, 2 May 2017 16:07:40 +0000 (09:07 -0700)]
Merge "Merge remote-tracking branch 'remotes/quic/dev/msm-4.4-8996au' into msm-4.4"

7 years agoMerge "input: synaptics: fix for buggy code poined by SIL tool"
Linux Build Service Account [Tue, 2 May 2017 16:07:38 +0000 (09:07 -0700)]
Merge "input: synaptics: fix for buggy code poined by SIL tool"

7 years agoMerge "usb: gadget: MIDI: Add support for SuperSpeed enumeration"
Linux Build Service Account [Tue, 2 May 2017 16:07:37 +0000 (09:07 -0700)]
Merge "usb: gadget: MIDI: Add support for SuperSpeed enumeration"

7 years agoMerge "diag: Open glink channels in workqueue context"
Linux Build Service Account [Tue, 2 May 2017 16:07:36 +0000 (09:07 -0700)]
Merge "diag: Open glink channels in workqueue context"

7 years agoMerge "clk: Add support to dump state of all clocks into ftrace"
Linux Build Service Account [Tue, 2 May 2017 16:07:35 +0000 (09:07 -0700)]
Merge "clk: Add support to dump state of all clocks into ftrace"

7 years agoMerge "SDM660: ADSPRPC: Align the size of IOVA to 128MB"
Linux Build Service Account [Tue, 2 May 2017 16:07:35 +0000 (09:07 -0700)]
Merge "SDM660: ADSPRPC: Align the size of IOVA to 128MB"

7 years agoMerge "esoc: mdm-4x: Separate out callback function for 9x45"
Linux Build Service Account [Tue, 2 May 2017 16:07:32 +0000 (09:07 -0700)]
Merge "esoc: mdm-4x: Separate out callback function for 9x45"

7 years agoMerge "ARM: dts: msm: Add support mdm9x45 external modem"
Linux Build Service Account [Tue, 2 May 2017 16:07:32 +0000 (09:07 -0700)]
Merge "ARM: dts: msm: Add support mdm9x45 external modem"

7 years agoMerge "esoc: Add provision to handle shutdown request in userspace"
Linux Build Service Account [Tue, 2 May 2017 16:07:31 +0000 (09:07 -0700)]
Merge "esoc: Add provision to handle shutdown request in userspace"

7 years agoMerge "soc: qcom: ssr: Add option to relax on ssr failures"
Linux Build Service Account [Tue, 2 May 2017 16:07:30 +0000 (09:07 -0700)]
Merge "soc: qcom: ssr: Add option to relax on ssr failures"

7 years agoMerge "msm/drm: Move msm_drm_config configuration into the GPUs"
Linux Build Service Account [Tue, 2 May 2017 16:07:18 +0000 (09:07 -0700)]
Merge "msm/drm: Move msm_drm_config configuration into the GPUs"

7 years agotty: serial: msm: fix potential race b/w startup and irq handling
Neeraj Upadhyay [Wed, 26 Apr 2017 05:07:15 +0000 (10:37 +0530)]
tty: serial: msm: fix potential race b/w startup and irq handling

There is a potential race for tx/rx dma requests between
msm_startup() and irq handling, for cases where dmas are not
available. This results in irq path trying to do dma mapping,
resulting in data abort. For example, consider below scenario
where rx handler reads the intermediate value of dma->chan,
set in msm_request_rx_dma(), and tries to do dma mapping,
which results in data abort.

uart_port_startup()
  msm_startup()
   request_irq()
   ...
   msm_request_rx_dma()
    ...
    dma->chan = dma_request_slave_channel_reason(dev, "rx");
    <UART RX IRQ>
     msm_uart_irq()
      msm_handle_rx_dm()
       msm_start_rx_dma()
        dma->desc = dma_map_single()
         <data abort>

Change-Id: Icf5d48f2718c3c6a855ffd3d10988a93f8281d78
Signed-off-by: Neeraj Upadhyay <neeraju@codeaurora.org>
7 years agofs: ecryptfs: Cache to be cleared after file write
Neeraj Soni [Tue, 2 May 2017 10:02:48 +0000 (15:32 +0530)]
fs: ecryptfs: Cache to be cleared after file write

This ensures that caches are maintained and no stale
data is present.

Change-Id: Ie8e92197a5a4179d422a523d3d6dad48bdf0b5d2
Signed-off-by: Neeraj Soni <neersoni@codeaurora.org>
7 years agoinput: synaptics: fix for buggy code poined by SIL tool
Andrew Chant [Fri, 31 Mar 2017 10:03:48 +0000 (15:33 +0530)]
input: synaptics: fix for buggy code poined by SIL tool

Place file offset validity checks under mutex for
synaptics_dsx_rmi_dev.c touch driver.

Git-repo: https://android.googlesource.com/kernel/msm
Git-commit: e1fb1600fc222337989e3084d68df929882deae5
Change-Id: I2c32babbccb483547204cb2843973abf97e988a5
Signed-off-by: Andrew Chant <achant@google.com>
[srkupp@codeaurora.org: This change is a fix for buggy
code pointed by sil after merging the above commit.}
Signed-off-by: Srinivasa Rao Kuppala <srkupp@codeaurora.org>
Signed-off-by: Shantanu Jain <shjain@codeaurora.org>
7 years agoMerge "drm/msm: Add PLL_DELTA property to HDMI connector"
Linux Build Service Account [Tue, 2 May 2017 06:57:09 +0000 (23:57 -0700)]
Merge "drm/msm: Add PLL_DELTA property to HDMI connector"

7 years agoMerge "ARM: dts: msm: Set 1.848V as the minimum voltage for LDO24 on msm8998"
Linux Build Service Account [Tue, 2 May 2017 06:57:08 +0000 (23:57 -0700)]
Merge "ARM: dts: msm: Set 1.848V as the minimum voltage for LDO24 on msm8998"

7 years agoMerge "power_supply: Add REAL_TYPE power_supply_property"
Linux Build Service Account [Tue, 2 May 2017 06:57:05 +0000 (23:57 -0700)]
Merge "power_supply: Add REAL_TYPE power_supply_property"

7 years agoMerge "qpnp-smb2: make qnovo vote for current and voltage"
Linux Build Service Account [Tue, 2 May 2017 06:57:04 +0000 (23:57 -0700)]
Merge "qpnp-smb2: make qnovo vote for current and voltage"

7 years agoMerge "qcom: qpnp-smb2: fix cleanup path"
Linux Build Service Account [Tue, 2 May 2017 06:57:04 +0000 (23:57 -0700)]
Merge "qcom: qpnp-smb2: fix cleanup path"

7 years agoMerge "qcom: battery: Fix using stale votable pointers"
Linux Build Service Account [Tue, 2 May 2017 06:57:03 +0000 (23:57 -0700)]
Merge "qcom: battery: Fix using stale votable pointers"

7 years agoMerge "ARM: dts: msm: Add iommu coherent test device for msm8998"
Linux Build Service Account [Tue, 2 May 2017 06:57:02 +0000 (23:57 -0700)]
Merge "ARM: dts: msm: Add iommu coherent test device for msm8998"

7 years agoMerge "msm: ipa3: fix channel stop retry logic"
Linux Build Service Account [Tue, 2 May 2017 06:57:02 +0000 (23:57 -0700)]
Merge "msm: ipa3: fix channel stop retry logic"

7 years agoMerge "ARM: boot: Silence 'zimage-dtb is ready' message"
Linux Build Service Account [Tue, 2 May 2017 06:57:01 +0000 (23:57 -0700)]
Merge "ARM: boot: Silence 'zimage-dtb is ready' message"

7 years agoMerge "esoc: mdm-4x: Add support for mdm9x45 and apq8096"
Linux Build Service Account [Tue, 2 May 2017 06:56:58 +0000 (23:56 -0700)]
Merge "esoc: mdm-4x: Add support for mdm9x45 and apq8096"

7 years agoMerge "defconfig: disable slub debug on sdm660"
Linux Build Service Account [Tue, 2 May 2017 06:56:57 +0000 (23:56 -0700)]
Merge "defconfig: disable slub debug on sdm660"

7 years agoMerge "mmc: cmdq_hci: Avoid releasing clock twice during DCMD error"
Linux Build Service Account [Tue, 2 May 2017 06:56:56 +0000 (23:56 -0700)]
Merge "mmc: cmdq_hci: Avoid releasing clock twice during DCMD error"

7 years agoMerge "cfg80211: Add macros to indicate backport support for FILS"
Linux Build Service Account [Tue, 2 May 2017 06:56:54 +0000 (23:56 -0700)]
Merge "cfg80211: Add macros to indicate backport support for FILS"

7 years agoMerge "ASoC: msm: q6dspv2: fix APR deregistration logic in ASM during ADSP SSR"
Linux Build Service Account [Tue, 2 May 2017 06:56:52 +0000 (23:56 -0700)]
Merge "ASoC: msm: q6dspv2: fix APR deregistration logic in ASM during ADSP SSR"

7 years agoMerge "defconfig: msm: add dcc config for perf on sdm660"
Linux Build Service Account [Tue, 2 May 2017 06:56:51 +0000 (23:56 -0700)]
Merge "defconfig: msm: add dcc config for perf on sdm660"

7 years agoMerge "msm: pcie: switch GPIO to sleep state before asserting PERST"
Linux Build Service Account [Tue, 2 May 2017 06:56:49 +0000 (23:56 -0700)]
Merge "msm: pcie: switch GPIO to sleep state before asserting PERST"

7 years agoMerge "msm: sde: Avoid use of uninitialized variable"
Linux Build Service Account [Tue, 2 May 2017 06:56:48 +0000 (23:56 -0700)]
Merge "msm: sde: Avoid use of uninitialized variable"

7 years agousb: gadget: MIDI: Add support for SuperSpeed enumeration
Ajay Agarwal [Wed, 26 Apr 2017 05:24:12 +0000 (10:54 +0530)]
usb: gadget: MIDI: Add support for SuperSpeed enumeration

Currently MIDI function supports only upto HighSpeed
enumeration. Add descriptors for SuperSpeed mode
and bind them to enable SuperSpeed enumeration of
USB MIDI function.

Change-Id: I0451dabf91e88503ab588dadbfbe6a2b76e2351b
Signed-off-by: Ajay Agarwal <ajaya@codeaurora.org>
7 years agoSDM660: ADSPRPC: Align the size of IOVA to 128MB
Tharun Kumar Merugu [Thu, 27 Apr 2017 12:13:27 +0000 (17:43 +0530)]
SDM660: ADSPRPC: Align the size of IOVA to 128MB

Align the size of the IOVA that is being passed
in the arm_iommu_create_mapping() to 128MB.

Change-Id: Ia554c2157d6c46b2f3848f993a7e61ff7f029547
Acked-by: Chenna Kesava Raju <chennak@qti.qualcomm.com>
Signed-off-by: Tharun Kumar Merugu <mtharu@codeaurora.org>
7 years agoicnss: remove pre-alloc memory leak check in platform driver
Hardik Kantilal Patel [Thu, 27 Apr 2017 06:25:45 +0000 (11:55 +0530)]
icnss: remove pre-alloc memory leak check in platform driver

The WLAN host driver is allocating the memory from pre-alloc pool
during insmod/wlan start up before WLAN driver register and
release the pre-alloc memory after driver unregister/remove.
The Pre-alloc memory leak check and reset in Icnss platform driver
on probe failure and after remove will leads to invalid memory
leak stat and dangling pointer for wlan host driver allocated memory
from the pre-alloc memory pool.

To fix the above issue remove the pre-allaoc memory leak
check and pre-alloc memory pool reset from the icnss platform
driver and export symbol for the pre-alloc memory leak check
and pre-alloc memory pool reset.

CRs-Fixed: 2039483
Change-Id: Id9f01c9d2b5184fbb58935eaf11fd21b50b47908
Signed-off-by: Hardik Kantilal Patel <hkpatel@codeaurora.org>
7 years agoesoc: mdm-4x: Separate out callback function for 9x45
Arun KS [Thu, 20 Apr 2017 13:11:33 +0000 (18:41 +0530)]
esoc: mdm-4x: Separate out callback function for 9x45

Reset and power off timing is different. Hence need a
separate callback function.

Change-Id: I9714a4449b1fbeab84017da4be17f5ca2cb6cab6
Signed-off-by: Arun KS <arunks@codeaurora.org>
7 years agoARM: dts: msm: Add support mdm9x45 external modem
Arun KS [Fri, 24 Mar 2017 06:18:36 +0000 (11:48 +0530)]
ARM: dts: msm: Add support mdm9x45 external modem

This change defines upto 3 instances of external mdm9x45 modems in
device-tree which would need to be controlled by esoc (external esoc)
driver. The device-tree nodes allows the configuration of the external
modems (like the GPIO pins used to communicate status of modem) to be
specified.

Change-Id: I7e609f7549a02cd3322db76b00dc30137ed68953
Signed-off-by: Arun KS <arunks@codeaurora.org>
7 years agoesoc: Add provision to handle shutdown request in userspace
Arun KS [Thu, 2 Mar 2017 12:42:54 +0000 (18:12 +0530)]
esoc: Add provision to handle shutdown request in userspace

In certain scenarios, modem shutdown requests are handled in
userspace. Enhance request engine of esoc driver to send
shutdown requests to userspace.

Also, during a shutdown, avoid setting status to 0, if line is
not a power source. There can be multiple mdms monitoring status
line. This can otherwise be misinterpreted as an unexpected reset
by other mdms.

Change-Id: I9c20a86e76f892cc61dbfb814202b26e5cce3e96
Signed-off-by: Arun KS <arunks@codeaurora.org>
Signed-off-by: Srivatsa Vaddagiri <vatsa@codeaurora.org>
7 years agoesoc: Update SSR driver with crash status
Arun KS [Mon, 6 Mar 2017 07:55:56 +0000 (13:25 +0530)]
esoc: Update SSR driver with crash status

During an unexpected reset or error fatal, update the
crash status to SSR. This is important for the drivers
listening at SSR related kernel notifier calls, where
crash status is also passed as a data payload.

Change-Id: Ide0634d0139a84b5988fa87e709877f3028029ef
Signed-off-by: Arun KS <arunks@codeaurora.org>
7 years agosoc: qcom: ssr: Add option to relax on ssr failures
Arun KS [Mon, 6 Mar 2017 07:51:45 +0000 (13:21 +0530)]
soc: qcom: ssr: Add option to relax on ssr failures

SSR failures are considered fatal and results in system panic.
In certain scenarios system can continue to work even with a
failed subsystem.

Add an option in subsystem descriptor to relax on ssr failures.

Change-Id: I86dcaa615d6443937077880d9a91070d9c22ea1f
Signed-off-by: Arun KS <arunks@codeaurora.org>
[satyap@codeaurora.org: trivial merge conflict resolution]
Signed-off-by: Satya Durga Srinivasu Prabhala <satyap@codeaurora.org>
7 years agoclk: Add support to dump state of all clocks into ftrace
Amit Nischal [Mon, 6 Mar 2017 11:10:58 +0000 (16:40 +0530)]
clk: Add support to dump state of all clocks into ftrace

Add clk_state event to record the state of all the clocks
into ftrace. The clock event could be triggered by using
the "trace_clocks" debugfs entry and it would dump the
current state of all clocks in ftrace logs.

Change-Id: I28b6574fe1d96472833a93e7b251dbba6c6eae49
Signed-off-by: Amit Nischal <anischal@codeaurora.org>
7 years agopower_supply: Add REAL_TYPE power_supply_property
Fenglin Wu [Tue, 25 Apr 2017 12:51:06 +0000 (20:51 +0800)]
power_supply: Add REAL_TYPE power_supply_property

Add REAL_TYPE power_supply property to record the real time charger
type.

Change-Id: I5fb2e3e3e782bcac0f8dd6071a830bcf370ebbd4
Signed-off-by: Fenglin Wu <fenglinw@codeaurora.org>
7 years agoqpnp-smb2: make qnovo vote for current and voltage
Abhijeet Dharmapurikar [Fri, 28 Apr 2017 02:20:31 +0000 (19:20 -0700)]
qpnp-smb2: make qnovo vote for current and voltage

Currently the code simply enforces Qnovo's current and voltage bypassing
other voters. This is not desired. Make Qnovo vote via the FCC and FV
votables. The only other vote it should skip is the votes coming from
battery profile.

Change-Id: I5c794ea209a8ea2a61d834e2f619d7ccfd02fed9
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqcom: qpnp-smb2: fix cleanup path
Ashay Jaiswal [Thu, 27 Apr 2017 11:46:26 +0000 (17:16 +0530)]
qcom: qpnp-smb2: fix cleanup path

Fix the cleanup path of probe failure to make sure
all the resources get released in proper order.

Change-Id: Ie482c9856569ea708a8fa186049ab778a8e5be12
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqcom: battery: Fix using stale votable pointers
Harry Yang [Thu, 6 Apr 2017 07:02:30 +0000 (00:02 -0700)]
qcom: battery: Fix using stale votable pointers

The code flow between battery.c and smblib.c could end up with
stale references in smblib. This is when if pl_init fails for
some reason after creating the votables, while smblib obtains
references to them, those references become invalid.

Fix this by calling pl_init early in smb2 driver's probe such that if
it fails smb2 driver exits early.

Also change the name of pl_(de)init() functions to more appropriate
name - qcom_batt_(de)init().

Change-Id: I58f79d26e6cc8524e792a23185ff6fc8cfdffa75
Signed-off-by: Harry Yang <harryy@codeaurora.org>
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoMerge remote-tracking branch 'remotes/quic/dev/msm-4.4-8996au' into msm-4.4
Zhiqiang Tu [Thu, 27 Apr 2017 02:51:03 +0000 (10:51 +0800)]
Merge remote-tracking branch 'remotes/quic/dev/msm-4.4-8996au' into msm-4.4

Conflicts:
arch/arm/boot/dts/qcom/msm8996-auto-cdp.dtsi
drivers/gpu/drm/msm/Makefile

Change-Id: Ief80c28ff1422fd71a0c3d2041531e2ab078ee7a
Signed-off-by: Zhiqiang Tu <ztu@codeaurora.org>
7 years agoARM: dts: msm: Add iommu coherent test device for msm8998
Sudarshan Rajagopalan [Thu, 27 Apr 2017 20:38:21 +0000 (13:38 -0700)]
ARM: dts: msm: Add iommu coherent test device for msm8998

Added a new iommu test device which is dma-coherent. This
test device will be used for validating buffer coherency.

Change-Id: Iea1c57fd2cd5d71de2f26932d21cae8102fb918e
Signed-off-by: Sudarshan Rajagopalan <sudaraja@codeaurora.org>
7 years agomsm: pcie: switch GPIO to sleep state before asserting PERST
Tony Truong [Sat, 29 Apr 2017 01:17:04 +0000 (18:17 -0700)]
msm: pcie: switch GPIO to sleep state before asserting PERST

Before asserting PERST, PCIe bus driver should switch GPIO to
sleep state. This will prevent host from missing any events after
the link is off.

Change-Id: Ie57d339da02dde6cbb3c41d2be52ee0602227193
Signed-off-by: Tony Truong <truong@codeaurora.org>
7 years agosmb138x: use chg src bit to disable parallel charger
Abhijeet Dharmapurikar [Tue, 25 Apr 2017 01:33:55 +0000 (18:33 -0700)]
smb138x: use chg src bit to disable parallel charger

smb1355 doesn't have any facility to suspend input. One has to
disable charging using the chg_en command bit.

So, when parallel charging needs to be disabled, configure
the charger to ignore the chg_en input pin and instead use the
command register. The command register is always programmed to
disable charging.

When parallel charging needs to be enabled, configure it to
follow the chg_en input pin.

This will work for both smb1355 and smb1381 parallel chargers.

Change-Id: I50294f40927641c2e0a7c7e4e7d263592086d3a9
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqpnp-smb2: don't stomp over D+/D- during hard reset
Abhijeet Dharmapurikar [Mon, 24 Apr 2017 21:06:54 +0000 (14:06 -0700)]
qpnp-smb2: don't stomp over D+/D- during hard reset

Currently when we are in hard reset, the driver requests to remove
the float of D+/D- when VBUS drops and once VBUS is back, it requests
to float D+/D- again. This behaviour ends up stomping over D+/D- lines
leading to noncompliance with PD spec.

Fix it by ensuring we only run the CC2 workaround while VBUS dips
in hard reset.

Change-Id: I2bb9364b890b3a227dd3ea3e0a34a415b600e21a
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agoqpnp-smb2: report main charger's share in its current_max
Abhijeet Dharmapurikar [Mon, 24 Apr 2017 20:33:07 +0000 (13:33 -0700)]
qpnp-smb2: report main charger's share in its current_max

Currently main power supply reports the value of the input current limit
set on the usb path. This is incorrect, it should report its own
share. When parallel is disabled, it will end up reporting a value
close to current limit, rounded by 25mA steps.

Note that it should report INT_MAX when there is no limit set on
the ICL - the case for hvdcp chargers.

CRs-Fixed: 2037898
Change-Id: I9859f3d759644a08c6f762e929c471a41b1cdf8f
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
7 years agomsm: ipa3: fix channel stop retry logic
Skylar Chang [Mon, 24 Apr 2017 17:05:05 +0000 (10:05 -0700)]
msm: ipa3: fix channel stop retry logic

Stopping a PROD channel might result in a timeout because the
channel is not empty, or IPA is busy. Whether or not to retry
a stop operation is determined by channel owner.
This change removes the common retry logic of retrying to stop
a PROD channel and adds a retry only for the scenario it is
actually needed.

Change-Id: I7ac1e81f7f99de2b0c3162aa5aaea2102a450838
CRs-Fixed: 2037955
Acked-by: Ady Abraham <adya@qti.qualcomm.com>
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
7 years agomsm: ipa3: fix a race condition on clock vote from suspend
Skylar Chang [Fri, 24 Feb 2017 19:22:03 +0000 (11:22 -0800)]
msm: ipa3: fix a race condition on clock vote from suspend

There is a potential race condition where suspend interrupt
handling might be running from two different context in parallel.
Add a mutex to ensure mutual exclusion when voting for IPA clocks.

CRs-Fixed: 2037199
Change-Id: I7f14866696b6c9f5e239f3de2c4bad6991097ecb
Acked-by: Ady Abraham <adya@qti.qualcomm.com>
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
7 years agodefconfig: disable slub debug on sdm660
Vinayak Menon [Mon, 1 May 2017 09:18:32 +0000 (14:48 +0530)]
defconfig: disable slub debug on sdm660

Disable default enablement of slub debug on sdm660
debug defconfig to prevent skb allocation failure
due to increase in order of allocation.

Change-Id: If63822a5abf3b297a34b95e5e5f973556631bd68
Signed-off-by: Vinayak Menon <vinmenon@codeaurora.org>
7 years agodefconfig: enable uid stats on sdm660
Vinayak Menon [Mon, 1 May 2017 09:15:35 +0000 (14:45 +0530)]
defconfig: enable uid stats on sdm660

'commit 6a61b529b4a9 ("ANDROID: uid_sys_stats: rename
uid_cputime.c to uid_sys_stats.c")' renamed CONFIG_UID_CPUTIME
to CONFIG_UID_SYS_STATS and enabled the new config on relevant
defconfigs except sdm660. Enable it and sync the config to
remove the reference to CONFIG_UID_CPUTIME. Task IO accounting
is enabled as CONFIG_UID_SYS_STATS depends on it.

Change-Id: Ib7163cf090c4fe86030469a257c49be1f6558984
Signed-off-by: Vinayak Menon <vinmenon@codeaurora.org>
7 years agodiag: Open glink channels in workqueue context
Mohit Aggarwal [Mon, 1 May 2017 10:01:32 +0000 (15:31 +0530)]
diag: Open glink channels in workqueue context

Currently, on loading WDSP, diag is trying to close socket
transport and opening glink channels. Due to this, deadlock
is happening and wdsp channels are not getting opened properly
which leads to WDSP log stall. This patch fixes the issue by
moving opening of channels in workqueue context.

CRs-Fixed: 2030858
Change-Id: Ib6153b04cdda80275a94b16842032a19b2a5bb75
Signed-off-by: Mohit Aggarwal <maggarwa@codeaurora.org>
7 years agommc: cmdq_hci: Avoid releasing clock twice during DCMD error
Vijay Viswanath [Thu, 27 Apr 2017 09:39:45 +0000 (15:09 +0530)]
mmc: cmdq_hci: Avoid releasing clock twice during DCMD error

When erase DCMD commands fail, a completion notification may get raised
by host controller for the task in addition to error interrupt. This
can result in the done function of mmc request getting called twice
and release the clock twice and mess up the clock request counter. The
done function should be called only once for DCMD commands except flush
commands.

Change-Id: Ie3a96e48ca98872cf686049eb5caa060f143f9cb
Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
7 years agommc: queue: Remove BUG_ON() in cmdq_thread
Vijay Viswanath [Fri, 17 Mar 2017 13:26:18 +0000 (18:56 +0530)]
mmc: queue: Remove BUG_ON() in cmdq_thread

If any discard requests fails, the mmc driver should not cause kernel
panic. The error handling and informing block layer about error will be
done by completion context.

Change-Id: I0c88ad851ad12c392a557bcb94d3b21dfc62d1f8
Signed-off-by: Vijay Viswanath <vviswana@codeaurora.org>
7 years agocfg80211: Add macros to indicate backport support for FILS
Vidyullatha Kanchanapally [Mon, 10 Apr 2017 10:06:44 +0000 (15:36 +0530)]
cfg80211: Add macros to indicate backport support for FILS

This change adds two macros, one to indicate backport support for the
new connect done event which takes connection response parameters as a
structure instead of function arguments and the other macro is to
indicate backport support for FILS shared key authentication offload to
driver.

Change-Id: I6fb562741bb2abacbe1c7c7ffbf4b2afb25e555b
CRs-Fixed: 2028536
Signed-off-by: Vidyullatha Kanchanapally <vidyullatha@codeaurora.org>
7 years agocfg80211: Add support for FILS shared key authentication offload
Vidyullatha Kanchanapally [Thu, 30 Mar 2017 21:22:34 +0000 (00:22 +0300)]
cfg80211: Add support for FILS shared key authentication offload

Enhance nl80211 and cfg80211 connect request and response APIs to
support FILS shared key authentication offload. The new nl80211
attributes can be used to provide additional information to the driver
to establish a FILS connection. Also enhance the set/del PMKSA to allow
support for adding and deleting PMKSA based on FILS cache identifier.

Add a new feature flag that drivers can use to advertize support for
FILS shared key authentication and association in station mode when
using their own SME.

Change-Id: I75d47301b767e691d6240fd6c20c28b9bfaaa571
Signed-off-by: Vidyullatha Kanchanapally <vkanchan@qti.qualcomm.com>
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Git-commit: a3caf7440dedd2399f90f27ff11ac390bf03e6c4
Git-repo: git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next.git
CRs-Fixed: 2028536
[vidyullatha@codeaurora.org: backport to kernel 4.4 - This commit includes
the changes from following commits in include/uapi/linux/nl80211.h to
compile for msm-4.4 -
4a4b8169501b18c3450ac735a7e277b24886a651 :
cfg80211: Accept multiple RSSI thresholds for CQM]
Signed-off-by: Vidyullatha Kanchanapally <vidyullatha@codeaurora.org>
7 years agowireless: define cipher/AKM suites using a macro
Johannes Berg [Thu, 26 Jan 2017 16:15:44 +0000 (17:15 +0100)]
wireless: define cipher/AKM suites using a macro

The spec writes cipher/AKM suites as something like 00-0F-AC:9,
but the part after the colon isn't hex, it's decimal, so that
we've already had a few mistakes (in other code, or unmerged
patches) to e.g. write 0x000FAC10 instead of 0x000FAC0A.

Use a macro to avoid that problem.

Reviewed-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Git-commit: 228c8c6b1f4376788e9d5ab00d50b10228eb40d3
Git-repo: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
CRs-Fixed: 2028536
Change-Id: I1507de1ee316182d429186ff33a932675013ba7d
Signed-off-by: Vidyullatha Kanchanapally <vidyullatha@codeaurora.org>
7 years agocfg80211: Use a structure to pass connect response params
Vidyullatha Kanchanapally [Thu, 30 Mar 2017 21:22:33 +0000 (00:22 +0300)]
cfg80211: Use a structure to pass connect response params

Currently the connect event from driver takes all the connection
response parameters as arguments. With support for new features these
response parameters can grow. Use a structure to pass these parameters
rather than passing them as function arguments.

Signed-off-by: Vidyullatha Kanchanapally <vkanchan@qti.qualcomm.com>
Signed-off-by: Jouni Malinen <jouni@qca.qualcomm.com>
[add to documentation]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Git-commit: 5349a0f7bfbdd7d81b8418c707dcd1439c714647
Git-repo: git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next.git
CRs-Fixed: 2028536
Change-Id: I340a96d052647f79248ef8aa2e0af2b6ba979b2d
Signed-off-by: Vidyullatha Kanchanapally <vidyullatha@codeaurora.org>
7 years agoMerge "msm: kgsl: Perform cache operation with kernel address"
Linux Build Service Account [Mon, 1 May 2017 07:11:59 +0000 (00:11 -0700)]
Merge "msm: kgsl: Perform cache operation with kernel address"

7 years agoMerge "mmc: core: Update the logic of controlling clk scaling through sysfs"
Linux Build Service Account [Mon, 1 May 2017 07:11:58 +0000 (00:11 -0700)]
Merge "mmc: core: Update the logic of controlling clk scaling through sysfs"

7 years agodefconfig: msm: add dcc config for perf on sdm660
Saranya Chidura [Mon, 3 Apr 2017 08:59:22 +0000 (14:29 +0530)]
defconfig: msm: add dcc config for perf on sdm660

Add the DCC_CONFIG option in perf defconfig for SDM660
target.

Change-Id: I9f927176bcf4505cecc2c8b23a9a8c9896d45b3a
Signed-off-by: Saranya Chidura <schidura@codeaurora.org>
7 years agoesoc: mdm-4x: Add support for mdm9x45 and apq8096
Arun KS [Mon, 16 Jan 2017 12:30:37 +0000 (18:00 +0530)]
esoc: mdm-4x: Add support for mdm9x45 and apq8096

Add mdm_ops for mdm9x45 and apq8064.

Change-Id: Iea167175b9bd35a515d15a72897947a889093c03
Signed-off-by: Arun KS <arunks@codeaurora.org>
Signed-off-by: Srivatsa Vaddagiri <vatsa@codeaurora.org>
7 years agoesoc: Add support for autoboot
Arun KS [Mon, 16 Jan 2017 12:17:03 +0000 (17:47 +0530)]
esoc: Add support for autoboot

Some of the external SoC are flash based and can boot independently.
Extend esoc driver to support such auto boot esocs.

This patch also adds support for primary esoc. Primary esoc are
esoc that control secondary esoc such as modems. Primary esoc have
control over reset/poweroff of secondary esoc. Secondary esoc don't
have control over reset/poweroff of primary esoc. In general modems
are considered as secondary esoc while apps processor is considered
as primary esoc.

Change-Id: Id02417fcd122ac108cf75d3381ee7955f0f8f783
Signed-off-by: Arun KS <arunks@codeaurora.org>
Signed-off-by: Srivatsa Vaddagiri <vatsa@codeaurora.org>
7 years agoesoc: Add err_fatal signal status to clink_ops
Arun KS [Mon, 16 Jan 2017 09:57:48 +0000 (15:27 +0530)]
esoc: Add err_fatal signal status to clink_ops

Auto_boot esoc devices can boot and crash before
esoc driver comes up. But there is no way for the
user space code to know that it has crashed by looking
at status line alone. Hence, create a new ioctl entry
to export status of err_fatal line to user space.

Change-Id: Ie7d6115c749d4c63f06aefca29ba457d38eccc7f
Signed-off-by: Arun KS <arunks@codeaurora.org>
7 years agoesoc: Fix integration with SSR driver
Arun KS [Mon, 16 Jan 2017 09:09:52 +0000 (14:39 +0530)]
esoc: Fix integration with SSR driver

SSR driver expects subsystem descriptor device type to be
struct platform_device, whereas esoc driver is passing a
struct device instead. This results in a NULL pointer crash
during platform_get_irq in subsystem_restart.c. Fix esoc
driver to pass platform_device.

Change-Id: I93d02623cb7ac14ea3171c3792a4c52c73a74dea
Signed-off-by: Arun KS <arunks@codeaurora.org>
Signed-off-by: Srivatsa Vaddagiri <vatsa@codeaurora.org>
7 years agoMerge "ARM: dts: msm: move panel bindings to different msm8998 platforms"
Linux Build Service Account [Sat, 29 Apr 2017 14:00:57 +0000 (07:00 -0700)]
Merge "ARM: dts: msm: move panel bindings to different msm8998 platforms"

7 years agoMerge "ARM: dts: msm: Remove I/O coherence for the GPU"
Linux Build Service Account [Sat, 29 Apr 2017 14:00:56 +0000 (07:00 -0700)]
Merge "ARM: dts: msm: Remove I/O coherence for the GPU"

7 years agoMerge "ARM: dts: msm: Change panel settings for sdm660 QRD"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:52 +0000 (22:10 -0700)]
Merge "ARM: dts: msm: Change panel settings for sdm660 QRD"

7 years agoMerge "ARM: dts: msm: control VCI register switch for rm67195"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:50 +0000 (22:10 -0700)]
Merge "ARM: dts: msm: control VCI register switch for rm67195"

7 years agoMerge "drm/msm: Add explicit sync operations"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:49 +0000 (22:10 -0700)]
Merge "drm/msm: Add explicit sync operations"

7 years agoMerge "ASoC: qdsp6v2: Remove Eagle code"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:48 +0000 (22:10 -0700)]
Merge "ASoC: qdsp6v2: Remove Eagle code"

7 years agoMerge "Merge branch 'android-4.4@b834e92' into branch 'msm-4.4'"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:46 +0000 (22:10 -0700)]
Merge "Merge branch 'android-4.4@b834e92' into branch 'msm-4.4'"

7 years agoMerge "power: smb-lib: Fix charger type reporting"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:45 +0000 (22:10 -0700)]
Merge "power: smb-lib: Fix charger type reporting"

7 years agoMerge "ath10k: Move logging prior to ath10k core destroy"
Linux Build Service Account [Sat, 29 Apr 2017 05:10:44 +0000 (22:10 -0700)]
Merge "ath10k: Move logging prior to ath10k core destroy"

7 years agoMerge "ARM: dts: msm: configure supply property for DSI on msm8998."
Linux Build Service Account [Sat, 29 Apr 2017 05:10:43 +0000 (22:10 -0700)]
Merge "ARM: dts: msm: configure supply property for DSI on msm8998."

7 years agoMerge "soc: qcom: Register for indication cb after obtaining state of remote pd"
Linux Build Service Account [Fri, 28 Apr 2017 18:12:25 +0000 (11:12 -0700)]
Merge "soc: qcom: Register for indication cb after obtaining state of remote pd"

7 years agoMerge "icnss: Add support to handle SSR within SSR"
Linux Build Service Account [Fri, 28 Apr 2017 18:12:24 +0000 (11:12 -0700)]
Merge "icnss: Add support to handle SSR within SSR"

7 years agoMerge "qcom: PMRESR register read returns zero for un supported group events"
Linux Build Service Account [Fri, 28 Apr 2017 18:12:23 +0000 (11:12 -0700)]
Merge "qcom: PMRESR register read returns zero for un supported group events"

7 years agoMerge "arm64: Change cpu_resume() to enable mmu early then access sleep_sp by va"
Linux Build Service Account [Fri, 28 Apr 2017 18:12:20 +0000 (11:12 -0700)]
Merge "arm64: Change cpu_resume() to enable mmu early then access sleep_sp by va"

7 years agoMerge "arm64: kasan: clear stale stack poison"
Linux Build Service Account [Fri, 28 Apr 2017 18:12:20 +0000 (11:12 -0700)]
Merge "arm64: kasan: clear stale stack poison"

7 years agoMerge "Revert "arm64: Change cpu_resume() to enable mmu early ...""
Linux Build Service Account [Fri, 28 Apr 2017 18:12:19 +0000 (11:12 -0700)]
Merge "Revert "arm64: Change cpu_resume() to enable mmu early ...""

7 years agoMerge "msm: ADSPRPC: Initialize the list for global gfa maps"
Linux Build Service Account [Fri, 28 Apr 2017 18:12:18 +0000 (11:12 -0700)]
Merge "msm: ADSPRPC: Initialize the list for global gfa maps"

7 years agoARM: dts: msm: move panel bindings to different msm8998 platforms
Chandan Uddaraju [Fri, 14 Apr 2017 16:46:38 +0000 (09:46 -0700)]
ARM: dts: msm: move panel bindings to different msm8998 platforms

Include the panel dtsi files as part of all the different
msm8998 platform specific device tree files. This will
separate panel properties from SOC specific MDSS binding.

Change-Id: I423a53b4601447d0c7be2bdc041b36495f99da3b
Signed-off-by: Chandan Uddaraju <chandanu@codeaurora.org>
7 years agosoc: qcom: Register for indication cb after obtaining state of remote pd
Avaneesh Kumar Dwivedi [Fri, 21 Apr 2017 15:45:03 +0000 (21:15 +0530)]
soc: qcom: Register for indication cb after obtaining state of remote pd

This change moves the registration of indication call back after inquiring
the state of remote PD, this is logical flow since in any case just after
registration we are inquiring the state and doing client notification.

With existing arrangement of code, sometime there is occurring a race
condition between inquiring the remote pd state and indication call back.

Change-Id: I2d4d5e0dc7afde9dfb89747b878c26862532bec4
Signed-off-by: Avaneesh Kumar Dwivedi <akdwived@codeaurora.org>
7 years agosoc: qcom: Clear the memory before freeing it up
Avaneesh Kumar Dwivedi [Mon, 24 Apr 2017 16:21:42 +0000 (21:51 +0530)]
soc: qcom: Clear the memory before freeing it up

It is a case of write after free, this is causing page allocation
failure due to corruption. This is due to freeing up of segments
allocated for venus subsystem, when venus fw loading fail midway.

Change-Id: I0019a05b1d1336dcf361264607597430e5f1625a
Signed-off-by: Avaneesh Kumar Dwivedi <akdwived@codeaurora.org>
7 years agoath10k: Move logging prior to ath10k core destroy
Govind Singh [Fri, 28 Apr 2017 11:20:41 +0000 (16:50 +0530)]
ath10k: Move logging prior to ath10k core destroy

ath10k core destroy frees the debug references and
this is leading to crash when ATH10K_DBG_SNOC mask is
defined.

Fix this by moving logs to prior ath10k core destroy.

Change-Id: If4fd96fdfd9faaf19480b6d523c501747f56d40e
Signed-off-by: Govind Singh <govinds@codeaurora.org>
7 years agoath10k: wait for FW ready during initialization
Rakesh Pillai [Wed, 26 Apr 2017 13:31:06 +0000 (19:01 +0530)]
ath10k: wait for FW ready during initialization

Since the qmi service in snoc driver gets registered late,
it misses the first time FW ready is sent. This causes the
wait on FW ready to fail and eventually the driver loading fails.

Proceed with the driver initialization only once the FW ready
indication arrives. Handle error in case the wait for these
events timeout.

Change-Id: Ib20ddb3a2f8b5b48936cc97b38f637f31e4e0100
Signed-off-by: Rakesh Pillai <pillair@codeaurora.org>
7 years agoMerge "input: misc: hbtp-input: Support for FB_BLANK_NORMAL event"
Linux Build Service Account [Fri, 28 Apr 2017 08:45:05 +0000 (01:45 -0700)]
Merge "input: misc: hbtp-input: Support for FB_BLANK_NORMAL event"

7 years agoMerge "msm: camera: isp: Rate limit tasklet overflow logs"
Linux Build Service Account [Fri, 28 Apr 2017 08:45:04 +0000 (01:45 -0700)]
Merge "msm: camera: isp: Rate limit tasklet overflow logs"