These changes will merge back the enhancemnts done to improve the medium
side transfers. This was reverted temporarily till the fix identified.
Change-Id: Ib7ac92ecefe7ca3ff9f03716c51dc31b8322ee33
Signed-off-by: Mukesh Kumar Savaliya <msavaliy@codeaurora.org>
This reverts 'commit 9814eb7549 ("sd: Fix rw_max for devices
that report an optimal xfer size")'.
Max sectors limit for request queue(q->limits.max_sectors) is reduced
from 1024 to 16 sectors by above commit and causing performance impact
(reduced by upto 50%). Hence revert the change to fix the performance
issue observed on 8998.
Change-Id: I9abd2be6c90922ea9de818912f405e0d292a62b8
Signed-off-by: Sayali Lokhande <sayalil@codeaurora.org>
In split link panels, the four data lanes are split into
sublinks and the clock lane is shared among the sublink.
Add support for split link panels in mdss.
Change-Id: I40965c274a3591b0a00ca546052e7cb46967844d
Signed-off-by: Ashish Garg <ashigarg@codeaurora.org>
The binder allocator assumes that the thread that
called binder_open will never die for the lifetime of
that proc. That thread is normally the group_leader,
however it may not be. Use the group_leader instead
of current.
Bug: 35707103
Test: Created test case to open with temporary thread
Change-Id: Id693f74b3591f3524a8c6e9508e70f3e5a80c588
Signed-off-by: Todd Kjos <tkjos@google.com>
Signed-off-by: Martijn Coenen <maco@android.com>
Git-commit: 359795138d
Git-repo: https://android.googlesource.com/kernel/common/
Signed-off-by: Runmin Wang <runminw@codeaurora.org>
Adjust the PLL disable sequence as per the latest HW
programming guidelines to ensure there will not be any
stray clock glitches when PLL is turned OFF abruptly.
Change-Id: I6df35bbe18b0c42b43f38b9dd85c3502b2038928
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
When the supply to PLL digital domain is turned off,
it can result in certain PLL registers to get corrupted.
Make sure to re-program the PLL registers to the
power-on-reset value before starting to program the PLL again
to ensure that it locks reliably.
Change-Id: I63cac884cf11eae60b187f83654f5922a3342d66
Signed-off-by: Abhinav Kumar <abhinavk@codeaurora.org>
During clock gating (ufshcd_gate_work()), we first put the link hibern8 by
calling ufshcd_uic_hibern8_enter() and if ufshcd_uic_hibern8_enter()
returns success (0) then we gate all the clocks.
Now let’s zoom in to what ufshcd_uic_hibern8_enter() does internally:
It calls __ufshcd_uic_hibern8_enter() which on detecting the LINERESET,
initiates the full recovery and puts the link back to highest HS gear and
returns success (0) to ufshcd_uic_hibern8_enter() which is the issue as
link is still in active state due to recovery!
Now ufshcd_uic_hibern8_enter() returns success to ufshcd_gate_work() and
hence it goes ahead with gating the UFS clock while link is still in active
state hence I believe controller would raise UIC error interrupts. But when
we service the interrupt, clocks might have already been disabled!
This change fixes for this by returning failure from
__ufshcd_uic_hibern8_enter() if recovery succeeds as link is still not in
hibern8, upon receiving the error ufshcd_hibern8_enter() would initiate
retry to put the link state back into hibern8.
Change-Id: Ib550fb791fa4c582b8f2d317a7f5f7594acb0872
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
CPU hotplug operations take place in preemptible context. This leaves
the hotplugging thread at the mercy of overall system load and CPU
availability. If the hotplugging thread does not get an opportunity
to execute after it has already begun a hotplug operation, CPUs can
end up being stuck in a quasi online state. In the worst case a CPU
can be stuck in a state where the migration thread is parked while
another task is executing and changing affinity in a loop. This
combination can result in unbounded execution time for the running
task until the hotplugging thread gets the chance to run to complete
the hotplug operation.
Fix the said problem by ensuring that hotplug can only occur from
threads belonging to the RT sched class. This allows the hotplugging
thread priority on the CPU no matter what the system load or the
number of available CPUs are. If a SCHED_NORMAL task attempts to
hotplug a CPU, we temporarily elevate it's scheduling policy to RT.
Furthermore, we disallow hotplugging operations to begin if the
calling task belongs to the idle and deadline classes or those that
use the SCHED_BATCH policy.
Change-Id: Idbb1384626e6ddff46c0d2ce752eee68396c78af
Signed-off-by: Syed Rameez Mustafa <rameezmustafa@codeaurora.org>
Add support to configure base termination current which is the
upper boundary at which fuel gauge will signal an end of charge
to the charger during discharging.
Change-Id: Ied2f89fc8e132161d55a9bfbcca6738fef164b20
Signed-off-by: Subbaraman Narayanamurthy <subbaram@codeaurora.org>
When the battery temperature goes below -10 C, SOC may get stuck
to 100%. This is due to the SOC full loop adjusting down with
very high battery impedance. To mitigate this, SOC full loop has
to be frozen. Add support for it by adjusting ki_coeff_full_soc
dynamically based on battery temperature.
Change-Id: I6218568712b025bc4427770bf86d142d50f40c7b
Signed-off-by: Subbaraman Narayanamurthy <subbaram@codeaurora.org>
Add support to handle system error and shutdown notification
from MHI host.
CRs-Fixed: 2022936
Change-Id: Id36097dffd7571490d7d53d2e496bfe024702a42
Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
The bit mask macro GEN_FIELD(24,8) causes overflow as it tries to shift by
32 bits. Using kernel's builtin GENMASK macro to avoid such overflows.
Change-Id: I614a68eb009eb8b223a4da45349124a70c0a8134
Signed-off-by: Mahesh Sivasubramanian <msivasub@codeaurora.org>
The source_vm* variable is used to hold enum values of type
vmid{}, which has negative values (VMID_INVAL = -1). This variable
is declared as u32 in few functions. Change it to type int.
Change-Id: I4cc546e79a20a843d6bd096ab696913ab40dde6d
Signed-off-by: Sudarshan Rajagopalan <sudaraja@codeaurora.org>
Two notify requests are being queued for one available modem
response. So for the first GET_ENCAPSULATED_RESPONSE we provide
the host with the available response. Now for the next
GET_ENCAPSULATED_COMMAND we notify the host that the response is
available even before the modem is ready with a response because
of the extra notify request queued on the interrupt endpoint.
This causes a STALL for the next GET_ENCAPSULATED_RESPONSE request.
This is caused because we are queueing a notify request from the
completion handler of the interrupt endpoint request when the
response queue is not empty.
Fix this by queuing a notify request when a new response is
available only after the current resposne is send to the Host.
Change-Id: If84bc315f2be910503328cc6b0e21be342c6eb37
Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org>
Signed-off-by: Mayank Rana <mrana@codeaurora.org>
Add support to handle system error and shutdown
notifications from mhi host.
CRs-Fixed: 1097560
Change-Id: Ied6c907586aa4dc2ed3b1a7c19305877144b3b21
Signed-off-by: Sujeev Dias <sdias@codeaurora.org>
Driver register doesn't have to be synchronous with probe call
back. Keep the driver registered till unregister gets called and
never reset ops during probe failure.
CRs-fixed: 2029329
Change-Id: I61331c7f33b29b0bc4833a8e4c52ee94f17660e7
Signed-off-by: Anurag Chouhan <achouhan@codeaurora.org>
Soundwire control data is NULL if there is no soundwire slave
device enumerated.
Add null check to avoid panic.
Change-Id: Ief60d69c36c2a9831825f38da2c9a3f6dd5f13dc
CRs-Fixed: 2020293
Signed-off-by: Meng Wang <mwang@codeaurora.org>
Glink_pkt_read calls glink_queue_rx_intent without mutex lock.
This result in race condition where glink_queue_rx_intent generates
same intent id for two intents.
Mutex lock is added to avoid race condition in glink_queue_rx_intent.
CRs-Fixed: 2026579
Change-Id: Iedb9a8874193479154ebb932fa8ff80703402215
Signed-off-by: Dhoat Harpal <hdhoat@codeaurora.org>
Check for legacy PM4 commands instead of adreno version to calculate
ringbuffer space for PM4 commands that write to memory.
Change-Id: I5d1d4cfbc70bc73ddee9ee752de24aae154a04dc
Signed-off-by: Lynus Vaz <lvaz@codeaurora.org>
The new "fst_link_loss" sysfs can be used to enable/disable the fst
link loss behavior for specific connection.
Change-Id: I7afcf86a575ca3add581c8823a3d5567483abcb0
Signed-off-by: Hamad Kadmany <hkadmany@codeaurora.org>
During recovery, there are cases where in WLAN driver running on
APPS can access WLAN copy engine register causing exceptions
because of WLAN hardware in reset state.
Add voltage regulator and clk in icnss device node to vote for WLAN
hardware resources by the icnss platform driver during WLAN recovery.
CRs-Fixed: 2029173
Change-Id: I00bbad40ee152c8dac187ff1f541e34b254589cb
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
During recovery, there are cases where in WLAN driver running on
APPS can access WLAN copy engine register causing exceptions
because of WLAN hardware in reset state.
Add voltage regulator and clk in icnss device node to vote for WLAN
hardware resources by the icnss platform driver during WLAN recovery.
CRs-Fixed: 2029102
Change-Id: Iba785acc391fb56ea068199598ddc14d20bb3481
Signed-off-by: Sarada Prasanna Garnayak <sgarna@codeaurora.org>
In case wil6210 resume fails, wil6210 suspend function will try
to access the suspended device in the next kernel suspend.
To prevent that, add wil_status_suspended flag to indicate if the
device is already suspended and clear it only if the resume succeeds.
Change-Id: I548e5b516cfa62389ce5f6cafcf1568206f7c2a9
Signed-off-by: Maya Erez <merez@codeaurora.org>
