OSDN Git Service

sagit-ice-cold/kernel_xiaomi_msm8998.git
6 years agoMerge "msm: camera: Prevent use-after-free in v4l2_event_subscribe."
Linux Build Service Account [Sun, 26 Nov 2017 00:21:07 +0000 (16:21 -0800)]
Merge "msm: camera: Prevent use-after-free in v4l2_event_subscribe."

6 years agoMerge "Bluetooth: Fix for checking proper user-supplied buffers"
Linux Build Service Account [Sat, 25 Nov 2017 05:53:39 +0000 (21:53 -0800)]
Merge "Bluetooth: Fix for checking proper user-supplied buffers"

6 years agoMerge "Bluetooth - Fix for checking proper user-supplied buffers"
Linux Build Service Account [Sat, 25 Nov 2017 05:53:38 +0000 (21:53 -0800)]
Merge "Bluetooth - Fix for checking proper user-supplied buffers"

6 years agoMerge "msm: msm_bus: Add mutex lock for floor vote data"
Linux Build Service Account [Fri, 24 Nov 2017 20:07:41 +0000 (12:07 -0800)]
Merge "msm: msm_bus: Add mutex lock for floor vote data"

6 years agoBluetooth - Fix for checking proper user-supplied buffers
Balvinder Singh [Mon, 6 Nov 2017 06:20:46 +0000 (11:50 +0530)]
Bluetooth - Fix for checking proper user-supplied buffers

During patch download of devices, size validations & zero alloc
the buffers to ensure values passed are in range

CRs-fixed: 2084692
Change-Id: Ie1cd76fe68766d6d12d7262202e48c18ebe42274
Signed-off-by: Balvinder Singh <bpsingh@codeaurora.org>
6 years agoBluetooth: Fix for checking proper user-supplied buffers
Balvinder Singh [Mon, 6 Nov 2017 06:25:30 +0000 (11:55 +0530)]
Bluetooth: Fix for checking proper user-supplied buffers

During patch download procedure, size validation and zero
allocation of buffers are done to ensure values passed are
in permissible range

CRs-fixed: 2082452
Change-Id: I1c14e30f14af0279258efd0eadd8ab8ed6d8cf55
Signed-off-by: Balvinder Singh <bpsingh@codeaurora.org>
6 years agoMerge "msm: wlan: Update regulatory database"
Linux Build Service Account [Fri, 24 Nov 2017 07:38:13 +0000 (23:38 -0800)]
Merge "msm: wlan: Update regulatory database"

6 years agomsm: msm_bus: Add mutex lock for floor vote data
Odelu Kukatla [Thu, 23 Nov 2017 17:06:03 +0000 (22:36 +0530)]
msm: msm_bus: Add mutex lock for floor vote data

Floor vote data needs to be protected with mutex lock to
avoid double free of memory due to race condtion.

Change-Id: Ifaa01a14d273ccba6b9463aff3a41c0038b05f06
Signed-off-by: Odelu Kukatla <okukatla@codeaurora.org>
6 years agoMerge "msm: vidc: Fix memory leak issue"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:19 +0000 (14:01 -0800)]
Merge "msm: vidc: Fix memory leak issue"

6 years agoMerge "msm: vidc: Update extradata address mapping"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:17 +0000 (14:01 -0800)]
Merge "msm: vidc: Update extradata address mapping"

6 years agoMerge "msm: camera: isp: Enable epoch2 interrupt for pipeline 47"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:16 +0000 (14:01 -0800)]
Merge "msm: camera: isp: Enable epoch2 interrupt for pipeline 47"

6 years agoMerge "ath10k: Enable WoWLAN for the wcn3990 snoc wlan module"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:14 +0000 (14:01 -0800)]
Merge "ath10k: Enable WoWLAN for the wcn3990 snoc wlan module"

6 years agoMerge "scsi: ufs: Get TX and RX fsm state in case of abort"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:13 +0000 (14:01 -0800)]
Merge "scsi: ufs: Get TX and RX fsm state in case of abort"

6 years agoMerge "drm/msm: protect register dumps and debugfs with mutex"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:12 +0000 (14:01 -0800)]
Merge "drm/msm: protect register dumps and debugfs with mutex"

6 years agoMerge "USB: gsi: Ring Doorbell to GSI if check for suspend fails"
Linux Build Service Account [Thu, 23 Nov 2017 22:01:11 +0000 (14:01 -0800)]
Merge "USB: gsi: Ring Doorbell to GSI if check for suspend fails"

6 years agoMerge "ARM: dts: msm: update the new HDMI splash address on msm8996"
Linux Build Service Account [Thu, 23 Nov 2017 13:00:02 +0000 (05:00 -0800)]
Merge "ARM: dts: msm: update the new HDMI splash address on msm8996"

6 years agomsm: wlan: Update regulatory database
Rajeev Kumar Sirasanagandla [Thu, 23 Nov 2017 08:50:01 +0000 (14:20 +0530)]
msm: wlan: Update regulatory database

Remove 160 MHz band-width support for VN as per master sheet

CRs-Fixed: 2145374
Change-Id: Id0cda10bf3343bccda5eea08ae5369754bb40edb
Signed-off-by: Rajeev Kumar Sirasanagandla <rsirasan@codeaurora.org>
6 years agomsm: camera: isp: Enable epoch2 interrupt for pipeline 47
Lokesh Kumar Aakulu [Thu, 28 Sep 2017 10:58:19 +0000 (16:28 +0530)]
msm: camera: isp: Enable epoch2 interrupt for pipeline 47

Enable epoch2 interrupt and configure the line as asked
by input camif configuration from user library

Change-Id: Id4f623a907b385a17d63c35fc7746117b666d885
Signed-off-by: Lokesh Kumar Aakulu <lkumar@codeaurora.org>
6 years agoscsi: ufs: Get TX and RX fsm state in case of abort
Sayali Lokhande [Wed, 15 Nov 2017 10:24:33 +0000 (15:54 +0530)]
scsi: ufs: Get TX and RX fsm state in case of abort

If ufshcd_print_fsm_state() is called during UIC
commands failure, mutex lockup issue is observed
for uic_cmd_mutex. To fix this, read TX and RX
FSM states for UFS aborts cases only, for debugging
purpose.

Change-Id: I716251ec90ef406d9cff1f8a6aa6bc2f8f46926d
Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org>
6 years agoMerge "ath10k: configure the vdev listen interval before wow suspend"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:29 +0000 (19:51 -0800)]
Merge "ath10k: configure the vdev listen interval before wow suspend"

6 years agoMerge "msm: ipa: Fix to race condition proxy clock voting"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:28 +0000 (19:51 -0800)]
Merge "msm: ipa: Fix to race condition proxy clock voting"

6 years agoMerge "msm: ipa: UAPI security code changes"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:27 +0000 (19:51 -0800)]
Merge "msm: ipa: UAPI security code changes"

6 years agoMerge "drm/msm: implement HDMI teardown sequence for DRM driver"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:25 +0000 (19:51 -0800)]
Merge "drm/msm: implement HDMI teardown sequence for DRM driver"

6 years agoMerge "mhi: core: fix sequence for new transfer"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:24 +0000 (19:51 -0800)]
Merge "mhi: core: fix sequence for new transfer"

6 years agoMerge "msm: ipa3: Fix IPA aggregation force close call"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:23 +0000 (19:51 -0800)]
Merge "msm: ipa3: Fix IPA aggregation force close call"

6 years agoMerge "msm: sps: Update debug message format specifier"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:22 +0000 (19:51 -0800)]
Merge "msm: sps: Update debug message format specifier"

6 years agoMerge "Revert "scsi: ufshcd: Flush exception event before suspend""
Linux Build Service Account [Thu, 23 Nov 2017 03:51:21 +0000 (19:51 -0800)]
Merge "Revert "scsi: ufshcd: Flush exception event before suspend""

6 years agoMerge "scsi: ufs: Fix off-by-one bug in ufs debugfs driver"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:20 +0000 (19:51 -0800)]
Merge "scsi: ufs: Fix off-by-one bug in ufs debugfs driver"

6 years agoMerge "arm64: memory-hotplug: Add MEMORY_HOTPLUG, MEMORY_HOTREMOVE, MEMORY_PROBE"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:19 +0000 (19:51 -0800)]
Merge "arm64: memory-hotplug: Add MEMORY_HOTPLUG, MEMORY_HOTREMOVE, MEMORY_PROBE"

6 years agoMerge "msm: sde: disable support for v4l2 event subscription"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:17 +0000 (19:51 -0800)]
Merge "msm: sde: disable support for v4l2 event subscription"

6 years agoMerge "msm: mdss: hdmi: clear deep color bits in the HDMI controller"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:16 +0000 (19:51 -0800)]
Merge "msm: mdss: hdmi: clear deep color bits in the HDMI controller"

6 years agoMerge "smb-lib: allow hvdcp if pd is disabled"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:14 +0000 (19:51 -0800)]
Merge "smb-lib: allow hvdcp if pd is disabled"

6 years agoMerge "icnss: Send early modem crash notification to wlan driver"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:13 +0000 (19:51 -0800)]
Merge "icnss: Send early modem crash notification to wlan driver"

6 years agoMerge "usb: dwc3: Fix incorrect ep0 state on reset"
Linux Build Service Account [Thu, 23 Nov 2017 03:51:12 +0000 (19:51 -0800)]
Merge "usb: dwc3: Fix incorrect ep0 state on reset"

6 years agoath10k: Enable WoWLAN for the wcn3990 snoc wlan module
Sarada Prasanna Garnayak [Fri, 6 Oct 2017 10:59:41 +0000 (16:29 +0530)]
ath10k: Enable WoWLAN for the wcn3990 snoc wlan module

Register snoc bus layer suspend/resume PM ops and configure
the wakeup source for the device to adds support for WOWLAN
feature for the wcn3990 snoc wlan module.

CRs-Fixed: 2139961
Change-Id: Ic43414b965aee27564f8bbbe237c706f5262c2e4
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
6 years agonet: update rdev suspend method for the wowlan suspend/resume
Sarada Prasanna Garnayak [Wed, 22 Nov 2017 15:15:22 +0000 (20:45 +0530)]
net: update rdev suspend method for the wowlan suspend/resume

Pass wowlan configuration data as parameter in the rdev
suspend. The userspace wlan utility (iw, wpa_supplicant etc.)
configured the wow pattern and event through nl80211 cmd.

During the wlan system suspend the wlan device configured the
wakeup source and event as per the configuration data passed
by the wlan utility to wkae up the device from the sleep state
on wow pattern or event match.

CRs-Fixed: 2147919
Change-Id: Iaca21724e8dd5a9c18e104c44ee6fe73752594e6
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
6 years agoath10k: configure the vdev listen interval before wow suspend
Sarada Prasanna Garnayak [Sat, 11 Nov 2017 08:37:04 +0000 (14:07 +0530)]
ath10k: configure the vdev listen interval before wow suspend

During the WLAN WOWLAN mode the default listen interval needs
to be configured for each vdev to receive and process the WLAN
WOW pattern match found by the WLAN firmware.

Improper listen interval configuration for the vdev leads to WOW
pattern match missing by the WLAN module and a result WLAN module
failed to wakeup the target as per the WOW pattern configured
during WOWLAN suspend method.

CRs-Fixed: 2141933
Change-Id: I4b4e03c134ecea328bbda55748b1dd633697167a
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
6 years agomsm: ipa: Fix to race condition proxy clock voting
Mohammed Javid [Mon, 13 Nov 2017 18:13:27 +0000 (23:43 +0530)]
msm: ipa: Fix to race condition proxy clock voting

Added code changes to fix the race condition during
proxy clock voating in IPA.

Change-Id: I1fee17836bf58d25f461643ea1e60bcd06182372
Acked-by: Ashok Vuyyuru <avuyyuru@qti.qualcomm.com>
Signed-off-by: Mohammed Javid <mjavid@codeaurora.org>
6 years agomsm: ipa3: Fix IPA aggregation force close call
Mohammed Javid [Mon, 6 Nov 2017 12:19:56 +0000 (17:49 +0530)]
msm: ipa3: Fix IPA aggregation force close call

In IPA driver uninitialized structure causing some
garbage values result into IPA data stall.
Added code changes to reset zero uninitialized structure.

Change-Id: I3b98041f8a96ea1d571eeaa60fc84457f6309d45
Acked-by: Ashok Vuyyuru <avuyyuru@qti.qualcomm.com>
Signed-off-by: Mohammed Javid <mjavid@codeaurora.org>
6 years agoRevert "scsi: ufshcd: Flush exception event before suspend"
Subhash Jadavani [Thu, 16 Nov 2017 22:32:28 +0000 (14:32 -0800)]
Revert "scsi: ufshcd: Flush exception event before suspend"

This reverts commit de406d12ef760a7536a486e19016e8e38251f1bc.

In case of a background operations exception, there could be a deadlock:
-000|__switch_to()
-001|context_switch(inline)
-001|__schedule()
-002|schedule()
-003|schedule_timeout()
-004|do_wait_for_common(inline)
-004|__wait_for_common(inline)
-004|wait_for_common()
-005|wait_for_completion()
-006|destroy_work_on_stack(inline)
-006|flush_work()
-007|ufshcd_suspend()
-008|ufshcd_runtime_suspend()
-009|ufshcd_pltfrm_runtime_suspend()
-010|pm_generic_runtime_suspend()
-011|__rpm_callback(inline)
-011|rpm_callback()
-012|rpm_suspend()
-013|rpm_idle()
-014|__pm_runtime_idle()
-015|ufshcd_exception_event_handler()
-016|__read_once_size(inline)
-016|static_key_count(inline)
-016|static_key_false(inline)
-016|trace_workqueue_execute_end(inline)
-016|process_one_work()
-017|__read_once_size(inline)
-017|list_empty(inline)
-017|worker_thread()
-018|kthread()
-019|ret_from_fork(asm)

Exception handler tried to suspend UFS synchronously and suspend tries to
flush the currently running exception handler work which results into
deadlock. Revert this change until we find proper fix for this.

Change-Id: I2b8efdf583d08a276b987914c1037619eb8b25f7
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
6 years agoARM: dts: msm: update the new HDMI splash address on msm8996
Rahul Sharma [Tue, 21 Nov 2017 15:36:18 +0000 (21:06 +0530)]
ARM: dts: msm: update the new HDMI splash address on msm8996

HDMI continuous splash memory addrerss has been changed in bootloader.
This change updates the splash memory address to support the smooth
LK to kernel handoff.

Change-Id: If9c8f20b8215a4ea835b0e070feb2b51364772e6
Signed-off-by: Rahul Sharma <sharah@codeaurora.org>
6 years agomhi: core: fix sequence for new transfer
jiad [Tue, 21 Nov 2017 04:35:01 +0000 (12:35 +0800)]
mhi: core: fix sequence for new transfer

sequence for both rddm and fw download are initialized to be 0.
They are incremented by 1 in bhi_probe() and upon real transfer
takes place.

Issue is in SSR case, sequence is incrementing in a way like 1,
3, 5 and etc as observed. The increment in bhi_probe() seems
unnecessay as there's no real transfer.

Fix is to initialize sequence to 1 and increment by 1 when real
transfer takes place in bhi_bhie_transfer().

Change-Id: I210840cd1f7fecc300ecb69da8a50fbb1d200562
Signed-off-by: Jia Ding <jiad@codeaurora.org>
CRs-Fixed: 2147433

6 years agodrm/msm: protect register dumps and debugfs with mutex
Lloyd Atkinson [Tue, 26 Sep 2017 17:13:31 +0000 (13:13 -0400)]
drm/msm: protect register dumps and debugfs with mutex

Add a mutex to serialize sde debug register dumps for
back-to-back dumps, as well as any debugfs access.

Change-Id: I7f6186575df021ce48e18ea980f9c2108b756283
Signed-off-by: Lloyd Atkinson <latkinso@codeaurora.org>
6 years agomsm: sde: disable support for v4l2 event subscription
Alan Kwong [Thu, 12 Oct 2017 15:49:55 +0000 (11:49 -0400)]
msm: sde: disable support for v4l2 event subscription

V4l2 event subscription is not used by rotator, and
default implementation on 4.9 is causing failure during
file release. Disable v4l2 event subscription for rotator
to avoid invalid file release.

Change-Id: I486f2eba7112852a1c818efeeb86033f17139b33
Signed-off-by: Alan Kwong <akwong@codeaurora.org>
6 years agoMerge "msm: ba: avoid system crash in msm_ba_open()"
Linux Build Service Account [Tue, 21 Nov 2017 14:44:58 +0000 (06:44 -0800)]
Merge "msm: ba: avoid system crash in msm_ba_open()"

6 years agoMerge "Revert "ais: query field info and pass to user space""
Linux Build Service Account [Tue, 21 Nov 2017 14:44:57 +0000 (06:44 -0800)]
Merge "Revert "ais: query field info and pass to user space""

6 years agoMerge "Prevent potential double frees in sg driver"
Linux Build Service Account [Tue, 21 Nov 2017 14:44:56 +0000 (06:44 -0800)]
Merge "Prevent potential double frees in sg driver"

6 years agoMerge "audio: ion: switch to msm ion on gvm."
Linux Build Service Account [Tue, 21 Nov 2017 14:44:55 +0000 (06:44 -0800)]
Merge "audio: ion: switch to msm ion on gvm."

6 years agoMerge "msm: camera: Add support to reconfig regupdate fail buffer"
Linux Build Service Account [Tue, 21 Nov 2017 14:44:53 +0000 (06:44 -0800)]
Merge "msm: camera: Add support to reconfig regupdate fail buffer"

6 years agoarm64: memory-hotplug: Add MEMORY_HOTPLUG, MEMORY_HOTREMOVE, MEMORY_PROBE
Scott Branden [Fri, 28 Apr 2017 08:39:53 +0000 (14:09 +0530)]
arm64: memory-hotplug: Add MEMORY_HOTPLUG, MEMORY_HOTREMOVE, MEMORY_PROBE

Add memory-hotplug support for ARM64 platform.

This requires addition of
ARCH_ENABLE_MEMORY_HOTPLUG and ARCH_ENABLE_MEMORY_HOTREMOVE config options.

MEMORY_PROBE config option is added to support
/sys/devices/system/memory/probe functionality.

In addition architecture specific arch_add_memory and
arch_remove memory management functions are added.

Change-Id: I1fc4e8c93cfe1b9722b2fc8d811390e6cdfa7995
Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Patch-mainline: linux-kernel @ 11 Apr 2017, 18:24
Signed-off-by: Srivatsa Vaddagiri <vatsa@codeaurora.org>
6 years agodrm/msm: implement HDMI teardown sequence for DRM driver
Abhinav Kumar [Sat, 18 Nov 2017 03:53:39 +0000 (19:53 -0800)]
drm/msm: implement HDMI teardown sequence for DRM driver

Currently the teardown sequence in the DRM HDMI driver is
incorrect and it places the HDMI controller in an incorrect
state during cable unplug OR device teardown.
Fix the teardown sequence and follow the hardware programming
sequence correctly to avoid artifacts.

Change-Id: Ifb9ec303fa710409087a74c03435c83823d02763
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
6 years agodrm/msm: disable HDCP before turning off timing engine
Abhinav Kumar [Tue, 21 Nov 2017 05:49:04 +0000 (21:49 -0800)]
drm/msm: disable HDCP before turning off timing engine

As per the hardware programming sequence, HDCP authentication
process must start after turning ON the timing engine and the
authentication process should be stopped before turning OFF
the timing engine.

Implement this sequence during the HDMI turn OFF sequence.

Also, make sure to wait for one frame before disabling the DDC
while turning OFF the HDCP. This is required to make sure we
receive the HDCP encryption disabled interrupt before disabling
the HW.

Change-Id: I3174f51fe1b6ab578db2d38caaf761444e6ebd8f
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
6 years agoscsi: ufs: Fix off-by-one bug in ufs debugfs driver
Sayali Lokhande [Fri, 17 Nov 2017 04:57:18 +0000 (10:27 +0530)]
scsi: ufs: Fix off-by-one bug in ufs debugfs driver

When getting string from userspace by simple_write_to_buffer
in ufs_qcom_dbg_testbus_cfg_write() function, null byte may
be written out of bounds of configuration buffer if return
value is same as size of buffer, causing off-by-one bug.
This change passes correct available size of configuration
buffer to simple_write_to_buffer function.

Change-Id: If6085e806aa34622a9a57b02f563869c51011d2f
Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org>
6 years agoRevert "ais: query field info and pass to user space"
Andy Sun [Mon, 20 Nov 2017 06:10:53 +0000 (14:10 +0800)]
Revert "ais: query field info and pass to user space"

This reverts commit 3375a292eaf2230a2ad3b90dfc16c9c3ad1f9ce7.
Need to change the solution and re-implement this feature.

Change-Id: Iabd38bd5c0c062590215c3f22e9383853efa889c
Signed-off-by: Andy Sun <bins@codeaurora.org>
6 years agoMerge "msm: camera: isp: don't set stripe_rd reload bit during hw reset"
Linux Build Service Account [Mon, 20 Nov 2017 23:38:49 +0000 (15:38 -0800)]
Merge "msm: camera: isp: don't set stripe_rd reload bit during hw reset"

6 years agoMerge "USB: core: Add USB_DEVICE_ERROR uevent for enumeration timeout"
Linux Build Service Account [Mon, 20 Nov 2017 23:38:48 +0000 (15:38 -0800)]
Merge "USB: core: Add USB_DEVICE_ERROR uevent for enumeration timeout"

6 years agoicnss: Send early modem crash notification to wlan driver
Yuanyuan Liu [Fri, 10 Nov 2017 20:12:17 +0000 (12:12 -0800)]
icnss: Send early modem crash notification to wlan driver

When SSR happending during WLAN interface up, a deadlock will happen
in many places where it waits for FW response. When wlan module starts,
it will hold a rtnl_lock and wait for FW response. If SSR happened at
this point, it will send ssr notification to all the registed modules
one by one. It will send notification to IPA first, which will also
tried to get a hold of rtnl_lock. Since ssr notification is synchronized,
IPA won't be able to return because of failing to acquire the lock and
wlan driver won't release the rtnl_lock until it timeout and crash.
Only graceful to avoid this deadlock is to get early modem
crash notification from platform via a path which does not acquire
RTNL lock and this way once driver has received early crash
notification it can reject interface up.

Modem notification comes to icnss before IPA, hence send
early crash notification in modem SSR handler so that wlan driver
can release RTNL lock as soon as possible.

CRs-Fixed: 2141970
Change-Id: If77c54a2a39ce5cf0e129eb5c6599cd3fd722886
Signed-off-by: Yuanyuan Liu <yuanliu@codeaurora.org>
6 years agosmb-lib: allow hvdcp if pd is disabled
Abhijeet Dharmapurikar [Fri, 25 Aug 2017 21:23:09 +0000 (14:23 -0700)]
smb-lib: allow hvdcp if pd is disabled

Currently when pd is disabled, we do not allow hvdcp. To fix that
once hvdcp timeout happens and if pd is disabled, force a pd_active
= 0. This will cause us to run legacy workaround and rerun apsd if
the charger was detected as hvdcp.

Importantly, once legacy bit workaround is run, wait 400mS for
typeC to debounce. Currently we wait for 100mS only and that
causes a removal detection since the CC lines are not settled within
100mS.

Also it may take up to 650mS for the vbus to drop once CC lines are
floated during the workaround. The current wait is just 500mS,
bump it up to a 1 second.

Change-Id: I17a7341e96d6efccde6270bc69d79c3f215b83e4
Signed-off-by: Abhijeet Dharmapurikar <adharmap@codeaurora.org>
6 years agomsm: ipa: UAPI security code changes
Mohammed Javid [Mon, 7 Aug 2017 17:45:48 +0000 (23:15 +0530)]
msm: ipa: UAPI security code changes

Added code changes for UAPI security code
userspace inputs validation

Change-Id: I5f8b7c9cd27a9d8ca721a98f1cbf05c14588dd85
Acked-by: Ashok Vuyyuru <avuyyuru@qti.qualcomm.com>
Acked-by: Abhishek Choubey <abchoube@qti.qualcomm.com>
Signed-off-by: Mohammed Javid <mjavid@codeaurora.org>
6 years agousb: dwc3: Fix incorrect ep0 state on reset
Sriharsha Allenki [Thu, 16 Nov 2017 11:30:09 +0000 (17:00 +0530)]
usb: dwc3: Fix incorrect ep0 state on reset

According to the databook ep0 should be in setup
phase during reset. If host issues reset between
control transfers, ep0 will be  in an invalid state.
Fix this my issuing stall and restart on ep0 if it
is not in setup phase.

CRs-Fixed: 2136658
Change-Id: I6dc20c2735a6ce772533ccb5b63ba5d1b01f89d7
Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org>
6 years agomsm: ba: avoid system crash in msm_ba_open()
Jingtao Chen [Fri, 17 Nov 2017 09:14:41 +0000 (17:14 +0800)]
msm: ba: avoid system crash in msm_ba_open()

If it failed to get ba device, return NULL in msm_ba_open() to
avoid system crash on some specific board without adv7481 chip.

Change-Id: I5a3af414a683d852027b84b9ba0d64e2ee450e61
Signed-off-by: Jingtao Chen <jingtaoc@codeaurora.org>
6 years agomsm: mdss: hdmi: clear deep color bits in the HDMI controller
Yuan Zhao [Tue, 24 Oct 2017 08:21:42 +0000 (16:21 +0800)]
msm: mdss: hdmi: clear deep color bits in the HDMI controller

Clear the deep color bits in the HDMI controller in case
the mode does not need OR support deep color.

Change-Id: I77a5cd74bc90c249d8902437a7b87a07259a2c35
Signed-off-by: Yuan Zhao <yzhao@codeaurora.org>
6 years agoMerge "cnss2: fix PCIe D3hot when suspending link"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:23 +0000 (07:45 -0800)]
Merge "cnss2: fix PCIe D3hot when suspending link"

6 years agoMerge "FROMLIST: binder: fix proc->files use-after-free"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:22 +0000 (07:45 -0800)]
Merge "FROMLIST: binder: fix proc->files use-after-free"

6 years agoMerge "ais: query field info and pass to user space"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:21 +0000 (07:45 -0800)]
Merge "ais: query field info and pass to user space"

6 years agoMerge "msm: kgsl: Update the dispatcher timer properly"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:20 +0000 (07:45 -0800)]
Merge "msm: kgsl: Update the dispatcher timer properly"

6 years agoMerge "clk: qcom: mdss: fix the divider programming for DisplayPort PLL"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:19 +0000 (07:45 -0800)]
Merge "clk: qcom: mdss: fix the divider programming for DisplayPort PLL"

6 years agoMerge "msm: camera: Version changes for new csid on msm8953"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:18 +0000 (07:45 -0800)]
Merge "msm: camera: Version changes for new csid on msm8953"

6 years agoMerge "msm: camera: Changing return condition after NULL check"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:17 +0000 (07:45 -0800)]
Merge "msm: camera: Changing return condition after NULL check"

6 years agoMerge "msm: vidc: Fix instance formats initialization issue"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:16 +0000 (07:45 -0800)]
Merge "msm: vidc: Fix instance formats initialization issue"

6 years agoMerge "msm: mdss: dp: fix programming of DisplayPort SDP registers"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:15 +0000 (07:45 -0800)]
Merge "msm: mdss: dp: fix programming of DisplayPort SDP registers"

6 years agoMerge "msm: mdss: adjust mdss_mdp_get_plane_sizes parameters init order"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:13 +0000 (07:45 -0800)]
Merge "msm: mdss: adjust mdss_mdp_get_plane_sizes parameters init order"

6 years agoMerge "msm: kgsl: Use the bind lock when dumping entries"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:12 +0000 (07:45 -0800)]
Merge "msm: kgsl: Use the bind lock when dumping entries"

6 years agoMerge "msm: ais: sensor: actuator: avoid accessing out of bound memory"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:11 +0000 (07:45 -0800)]
Merge "msm: ais: sensor: actuator: avoid accessing out of bound memory"

6 years agoMerge "drivers: qdsp6v2: Add mutex unlock to properly release lock"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:10 +0000 (07:45 -0800)]
Merge "drivers: qdsp6v2: Add mutex unlock to properly release lock"

6 years agoMerge "security: pfe: Return proper error code"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:09 +0000 (07:45 -0800)]
Merge "security: pfe: Return proper error code"

6 years agoMerge "ARM: dts: msm: Fix tavil codec reset issue on sdm660"
Linux Build Service Account [Sun, 19 Nov 2017 15:45:08 +0000 (07:45 -0800)]
Merge "ARM: dts: msm: Fix tavil codec reset issue on sdm660"

6 years agomsm: sps: Update debug message format specifier
Jishnu Prakash [Tue, 14 Nov 2017 13:29:34 +0000 (18:59 +0530)]
msm: sps: Update debug message format specifier

Restrict printing of kernel virtual addresses in SPS driver.
In debug code, %p is used to print virtual addresses of
kernel objects, which can be exploited by attackers. It is
replaced with %pK, which hides these values if kptr_restrict
is set (default on Android).

Change-Id: I57585fa655abc01b2e8d694c8f31b7617bbf4ec7
Signed-off-by: Jishnu Prakash <jprakash@codeaurora.org>
6 years agoMerge "ath10k: check the pipe credit update enabled flag before set"
Linux Build Service Account [Sat, 18 Nov 2017 08:35:27 +0000 (00:35 -0800)]
Merge "ath10k: check the pipe credit update enabled flag before set"

6 years agoMerge "power: qpnp-fg-gen3: update battery profile based on Rconn configuration"
Linux Build Service Account [Sat, 18 Nov 2017 08:35:25 +0000 (00:35 -0800)]
Merge "power: qpnp-fg-gen3: update battery profile based on Rconn configuration"

6 years agoMerge "DRM: SDE: Fix the spelling in DTS parser"
Linux Build Service Account [Sat, 18 Nov 2017 00:00:55 +0000 (16:00 -0800)]
Merge "DRM: SDE: Fix the spelling in DTS parser"

6 years agoMerge "ARM: dts: msm: Add support for MSM8996 Mizar platform"
Linux Build Service Account [Sat, 18 Nov 2017 00:00:54 +0000 (16:00 -0800)]
Merge "ARM: dts: msm: Add support for MSM8996 Mizar platform"

6 years agoath10k: check the pipe credit update enabled flag before set
Sarada Prasanna Garnayak [Sat, 11 Nov 2017 07:24:50 +0000 (12:54 +0530)]
ath10k: check the pipe credit update enabled flag before set

Check the credit update flag has enabled for the pipe
before setting the flag for credit report update request
on the wcn3990 WLAN module.

CRs-Fixed: 2141924
Change-Id: Ic628fa49934c110b7abccefb45a2a354dc462791
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
6 years agoFROMLIST: binder: fix proc->files use-after-free
Todd Kjos [Fri, 10 Nov 2017 23:30:27 +0000 (15:30 -0800)]
FROMLIST: binder: fix proc->files use-after-free

(from https://patchwork.kernel.org/patch/10058587/)

proc->files cleanup is initiated by binder_vma_close. Therefore
a reference on the binder_proc is not enough to prevent the
files_struct from being released while the binder_proc still has
a reference. This can lead to an attempt to dereference the
stale pointer obtained from proc->files prior to proc->files
cleanup. This has been seen once in task_get_unused_fd_flags()
when __alloc_fd() is called with a stale "files".

The fix is to always use get_files_struct() to obtain struct_files
so that the refcount on the files_struct is used to prevent
a premature free. proc->files is removed since we get it every
time.

Bug: 69164715
Change-Id: I6431027d3d569e76913935c21885201505627982
Signed-off-by: Todd Kjos <tkjos@google.com>
Git-commit: 1d6b43334b45144beb0dc8236f9eb3b6cf3b2995
Git-repo: https://android.googlesource.com/kernel/common
Signed-off-by: Kyle Yan <kyan@codeaurora.org>
6 years agoMerge "msm: camera: Enable (2+1) lane csiphy combo mode"
Linux Build Service Account [Fri, 17 Nov 2017 03:11:13 +0000 (19:11 -0800)]
Merge "msm: camera: Enable (2+1) lane csiphy combo mode"

6 years agocnss2: fix PCIe D3hot when suspending link
jiad [Thu, 16 Nov 2017 08:54:09 +0000 (16:54 +0800)]
cnss2: fix PCIe D3hot when suspending link

When doing WLAN SSR tests on Rome PCIe platform, system hang
is observed. After debugging, setting D3hot to PCIe link when
suspending leads to the hang.

Setting D3hot is a new change, which is added for Napier platform.
But this code path shares with legacy Rome platform. In previous
implementations for Rome PCIe platforms, setting D3hot to link
is not needed.

Fix is to set D3hot for Napier specific platforms.

Change-Id: Ia6bcb1b898204c0a8fd63355dd1f0808021182c4
CRs-Fixed: 2145090
Signed-off-by: Jia Ding <jiad@codeaurora.org>
6 years agoDRM: SDE: Fix the spelling in DTS parser
Camus Wong [Mon, 13 Nov 2017 21:20:02 +0000 (16:20 -0500)]
DRM: SDE: Fix the spelling in DTS parser

Fix one of the mode properties' spelling error.

Change-Id: I6cb0f31aa62b825773b97099fcac3da8a2b94af6
Signed-off-by: Camus Wong <camusw@codeaurora.org>
6 years agoclk: qcom: mdss: fix the divider programming for DisplayPort PLL
Padmanabhan Komanduru [Wed, 15 Nov 2017 13:21:05 +0000 (18:51 +0530)]
clk: qcom: mdss: fix the divider programming for DisplayPort PLL

Fix the divider programming of DisplayPort PLL with the correct
value. Without this, display doesn't up fine with certain
resolutions on some sinks when link rate is 5.4 GHz.

Change-Id: I7c5a452a9df757240a1c6c3d371bd46a16f98efd
Signed-off-by: Padmanabhan Komanduru <pkomandu@codeaurora.org>
6 years agomsm: camera: Version changes for new csid on msm8953
Shankar Ravi [Thu, 17 Dec 2015 09:22:12 +0000 (14:52 +0530)]
msm: camera: Version changes for new csid on msm8953

msm8953 uses a new version of csid. Adding new
configuration file and making necessary driver changes.

Change-Id: I5c0ce7f43a4ccd55bd18461bc910e4dcb0c23bb0
Signed-off-by: Shankar Ravi <rshankar@codeaurora.org>
6 years agomsm: camera: Changing return condition after NULL check
Samyukta Mogily [Wed, 12 Oct 2016 12:43:50 +0000 (18:13 +0530)]
msm: camera: Changing return condition after NULL check

If return is inside the if condition, rc value is returned
only for that case. Hence shifting the return rc statement
outside the if loop, to return the value of rc for all cases.

Change-Id: I2f96cd0d301f580c0e7019746bade4e125661a36
CRs-Fixed: 1076948
Signed-off-by: Samyukta Mogily <smogily@codeaurora.org>
6 years agoUSB: gsi: Ring Doorbell to GSI if check for suspend fails
Vijayavardhan Vennapusa [Wed, 15 Nov 2017 05:22:11 +0000 (10:52 +0530)]
USB: gsi: Ring Doorbell to GSI if check for suspend fails

Currently GSI driver is clearing Doorbell to GSI in gsi_suspend(),
when it receives function suspend and notifying suspend to GSI in work
queue context. In ipa_suspend_work_handler(), driver is first checking
whether GSI wrapper is ready to suspend before doing xdci_suspend(). This
check fails if function resume received before ipa_suspend_work_handler()
gets chance to run. Due to this, state machine still stays at CONNECTED
state. When function resume is received, state machine is not handling
resume event and results in not ringing Doorbell to GSI. This eventually
causes data stall. Hence fix the issue by ringing Doorbell to GSI if in
case check for suspend fails in ipa_suspend_work_handler.

Change-Id: Ifbc03b7edde6e1be23c326118e89f805b148f9ff
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
6 years agoSDE: Fix the minimum value in display property
Camus Wong [Sat, 11 Nov 2017 05:41:10 +0000 (00:41 -0500)]
SDE: Fix the minimum value in display property

The minium value of u32 is zero instead of -1.  -1 will become the
largest value in u32.

Change-Id: I20fcab7d5912d6da7c4afe1ec7a86333767b0bf1
Signed-off-by: Camus Wong <camusw@codeaurora.org>
6 years agomsm: mdss: adjust mdss_mdp_get_plane_sizes parameters init order
Benjamin Chan [Tue, 14 Nov 2017 05:27:17 +0000 (00:27 -0500)]
msm: mdss: adjust mdss_mdp_get_plane_sizes parameters init order

Parameter mdss_mdp_plane_sizes must be cleared to 0 before returning
under an error condition, otherwise caller function will use the
uninitialized mdss_mdp_plane_sizes values and caused incorrect
operation.

Change-Id: I856b17ce9e917cc450040463ec34b7309d34b9b5
Signed-off-by: Benjamin Chan <bkchan@codeaurora.org>
6 years agoARM: dts: msm: Add support for MSM8996 Mizar platform
Bharathraj Nagaraju [Wed, 2 Aug 2017 17:02:01 +0000 (10:02 -0700)]
ARM: dts: msm: Add support for MSM8996 Mizar platform

Add device tree files needed to support MSM8996 based
mizar platform.

Change-Id: Ife6c6659b981b37c6b9d9cbb0a0c106488f1e07f
Signed-off-by: Bharathraj Nagaraju <snbraj@codeaurora.org>
6 years agoUSB: core: Add USB_DEVICE_ERROR uevent for enumeration timeout
Chandana Kishori Chiluveru [Tue, 17 May 2016 06:13:54 +0000 (11:43 +0530)]
USB: core: Add USB_DEVICE_ERROR uevent for enumeration timeout

Some userspace modules expect USB driver to send event to userspace, when
usb device enumeration failure with device descriptor read errors.
Add USB_DEVICE_ERROR uevent in the USB driver to explicitly send the ERROR
uevent for such case.

Change-Id: I7128869ff0700ab90d4c949de24cd03c5c90e22e
Signed-off-by: Chandana Kishori Chiluveru <cchiluve@codeaurora.org>
6 years agoais: query field info and pass to user space
Andy Sun [Mon, 25 Sep 2017 02:08:42 +0000 (10:08 +0800)]
ais: query field info and pass to user space

Current path to query field info: SOF ISR -> tasklet -> query thread
with high priority -> ba driver -> adv7481.

Pass field type to user space if they are valid after verification.

Change-Id: Id9926236389200446092dc9abb688ee1f83ab0c3
Signed-off-by: Andy Sun <bins@codeaurora.org>
6 years agomsm: camera: Add support to reconfig regupdate fail buffer
Lokesh Kumar Aakulu [Mon, 9 Oct 2017 09:05:59 +0000 (14:35 +0530)]
msm: camera: Add support to reconfig regupdate fail buffer

When reg update isn't issued for given settings current
buffer is marked as drop buffer. To avoid such drop bu-
ffers, reconfigure same buffer to next SOF and propaga-
te that error to MCT and IFACE modules and make sure
request frame is delayed by one SOF Frame.

Change-Id: Iaec5eef3f26f941cc54825d40815c047ccebb3bc
Signed-off-by: Lokesh Kumar Aakulu <lkumar@codeaurora.org>
6 years agomsm: camera: isp: don't set stripe_rd reload bit during hw reset
Gaoxiang Chen [Tue, 31 Oct 2017 02:57:06 +0000 (10:57 +0800)]
msm: camera: isp: don't set stripe_rd reload bit during hw reset

During reset isp HW, don't set stripe_rd reload bit, otherwise
it might cause iommu page fault, which actually is a fasle alarm.

CRs-Fixed: 2135311
Change-Id: I9be0400640edbf687ecfe08edcfdc1c7f7de25b6
Signed-off-by: Gaoxiang Chen <gaochen@codeaurora.org>
6 years agomsm: kgsl: Use the bind lock when dumping entries
Lynus Vaz [Mon, 30 Oct 2017 12:20:43 +0000 (17:50 +0530)]
msm: kgsl: Use the bind lock when dumping entries

Take the bind lock when we dump information about sparse memory
allocations through the debugfs node.

Change-Id: I955d64d06a259a2bbe4d85a33d68887ee01aad07
Signed-off-by: Lynus Vaz <lvaz@codeaurora.org>
6 years agopower: qpnp-fg-gen3: update battery profile based on Rconn configuration
Subbaraman Narayanamurthy [Fri, 10 Nov 2017 23:37:29 +0000 (15:37 -0800)]
power: qpnp-fg-gen3: update battery profile based on Rconn configuration

Currently, Rconn is configured once in fg_hw_init() if it is not
configured already. However, if the driver loads the profile
later, it will overwrite the Rconn value configured during
fg_hw_init. To fix this problem, configure Rconn after handling
battery profile loading.

Usually the Rconn configuration is setup differently from what
the profile has for it. We check if the profile in SRAM is bit
to bit matching with the profile in device tree. We may come
across a mismatch for Rconn configuration and conclude that the
profile isn't matching. Fix this by initializing the Rconn
configuration in the profile read from device tree to the value
read from SRAM.

CRs-Fixed: 2142441
Change-Id: I2b7ac8cd6efe811527c29bc5cd0fa43b77da7b15
Signed-off-by: Subbaraman Narayanamurthy <subbaram@codeaurora.org>