The charger therm on RRADC only reads valid values if a charger is
present. Add a check accordingly.
Change-Id: I74c6c15e1e72d4d2ef3ffd698f130b7d0fce6df4
Signed-off-by: Ashay Jaiswal <ashayj@codeaurora.org>
Add the proper clock handle to USB dt node for noc_aggr_clk for
SDM660.
Change-Id: Id9ec52123ce2e45b466cc6cd74696f52154efdae
Signed-off-by: Vijayavardhan Vennapusa <vvreddy@codeaurora.org>
Increase the size of host to target CE4 ring
in order to hold more data packets in the ring and
avoid the unnecessary drops at hif layer.
CRs-Fixed: 2017758
Change-Id: Ib344c7035a79b0730a8b39d2441857b363e18efd
Signed-off-by: Govind Singh <govinds@codeaurora.org>
TCP TX throttling logic is changed a little after
3.19-rc1 kernel, the TCP sending limit will be smaller,
which will throttle the TCP packets to the host driver.
The TCP UP LINK throughput will drop heavily. In order to
fix this issue, need to orphan the socket buffer asap, which
will call skb's destructor to notify the TCP stack that the
SKB buffer is unowned. And then the TCP stack will pump more
packets to host driver.
The TX packets might be dropped for UDP case in the iperf
testing. So need to be protected by follow control
Change-Id: I1c77313fbdb579e708d4a5a6bcd063c7c34a6d39
Signed-off-by: Govind Singh <govinds@codeaurora.org>
On certain PMICs, PMIC peripheral drivers like haptics need to
read MISC peripheral register. Add support for clients to read
it by using qpnp_misc_read_reg().
Change-Id: Id5dfd9e440a8861b56572dab50839d1583535882
Signed-off-by: Subbaraman Narayanamurthy <subbaram@codeaurora.org>
Update the config option to "CONFIG_QCOM_TLB_HANDLER" so el1_sync
will call into EL2 TLB fault handler. Default TLB handler in EL1
is changed into a dummy handler.
Change-Id: Id4008d03002928bd24affeba0ef4250664ea3a48
Signed-off-by: Runmin Wang <runminw@codeaurora.org>
Fix possible buffer overflow by reading 'resp->total_domains' from the
qmi response message since 'resp->total_domains' indicate total number
of matching domains found by servreg.
'resp->domain_list_len' indicates the domains that could be sent in one
response which should not be greater than 'resp->total_domains'.
CRs-Fixed: 2009016
Change-Id: I614561c5f9bc996689129bc098baaffc9b59c377
Signed-off-by: Puja Gupta <pujag@codeaurora.org>
Die temperature samples 10 times to take average to reduce error,
which extends time in reading and causes delay during boot up, where
the data is not really insterested.
To improve it, only take one sample for the first few user temp
readings.
Change-Id: Ifc10de8a624f1d50d303673ff553d4801337481b
Signed-off-by: Harry Yang <harryy@codeaurora.org>
Fix ref counting of the secure buffers. This change releases
the ref count on the secure buffers during the free operation.
This allows the buffers to be ref counted until hw has done
fetching from the buffers.
CRs-Fixed: 2012323
Change-Id: Ic24c39e75d502ed73453d0c6d78fd76448737697
Signed-off-by: Abhijit Kulkarni <kabhijit@codeaurora.org>
Enable RCU_STALL_WATCHDOG_BITE flag to induce the
watchdog bite for every rcu stall instead of just
printing warning message. This option helps for
postmortem analysis on rcu stalls by collecting
complete ram dumps and cpu context.
Change-Id: I6badcd034bf75a10f211eb506701a68efccb43c3
Signed-off-by: Prasad Sodagudi <psodagud@codeaurora.org>
In case of high load on the system the old timeout (10ms)
is not sufficient for GSI to stop a channel.
This change is to increase GSI stop command to 20ms.
Change-Id: I6f86e6bf9a0254daee937e05927c8152402b2ffe
CRs-Fixed: 2015928
Acked-by: Ady Abraham <adya@qti.qualcommm.com>
Signed-off-by: Skylar Chang <chiaweic@codeaurora.org>
Every RCU stall need to be debugged, So collect the ram
dumps on every RCU stall to debug further by inducing
non secure watchdog bite whenever rcu stall detected.
Change-Id: I6c1cfddc92f06b48c3f22fe9970b70f2ec670bf6
Signed-off-by: Prasad Sodagudi <psodagud@codeaurora.org>
For every __pm_stay_awake call, there must be a
__pm_relax call to ensure there is no stale
wakelock is held. with the current logic
__pm_stay_awake is called whenever wakelock ref
cnt is non-zero and released only when it is 0.
This leads to wakelock being held even when
it is not required. Make a change to acquire
wakelock only when it is not held previously.
Change-Id: I8c61c184506c705fd5866de8a606f642c5de3d5b
Acked-by: Chaitanya Pratapa <cpratapa@qti.qualcomm.com>
Signed-off-by: Utkarsh Saxena <usaxena@codeaurora.org>