Use the no-log variants of the read{b,q,l}/write{b,q,l} APIs
to prevent flooding the MSM register trace buffer (RTB) logs
with memcpy/memset induced logging.
Change-Id: I16556e1d6f4abe00e6f33b8375a5a1839dfca34e
Signed-off-by: Trilok Soni <tsoni@codeaurora.org>
msm_write/read use write/read_relaxed function which will create
lots of RTB logging. Change the API to no_log version to remove
those RTB logs.
CRs-Fixed: 1062953
Change-Id: Id6524b7ae3e82e10ca651a2ca0de9223c18109da
Signed-off-by: Runmin Wang <runminw@codeaurora.org>
Add the VDD_APC CPR corner configuration corresponding to
speed-bin 1 parts whose performance cluster can operate
at a slower maximum clock rate compared to speed-bin 0 parts.
CRs-Fixed: 1057119
Change-Id: I01e2c9c8bafbb2be4c8d312a4212195c2a99f3ac
Signed-off-by: Osvaldo Banuelos <osvaldob@codeaurora.org>
Add support for a total of 16 fuse combos to cater to MSMCOBALT
parts blown with a speed-bin fuse equal to 1.
CRs-Fixed: 1057119
Change-Id: I778e33b9de9947dd33f4cb7142be8771d4afc7cb
Signed-off-by: Osvaldo Banuelos <osvaldob@codeaurora.org>
Update VDD_APC0 and VDD_APC1 CPR floor and ceiling voltages
of the highest fused corner on msmcobalt v2 parts to
adhere to latest hardware guidelines.
CRs-Fixed: 1057119
Change-Id: Icf06fe334558bfc4e4dedc9b1f18d51c99987966
Signed-off-by: Osvaldo Banuelos <osvaldob@codeaurora.org>
Add support for msmcobalt speed-bin 1 devices which can operate
with a performance cluster clock frequency of up to 2.208 GHz.
CRs-Fixed: 1057115
Change-Id: I2c733a1f0ee4baf978c3715aa3bd74a6b46ee6c2
Signed-off-by: Osvaldo Banuelos <osvaldob@codeaurora.org>
Glink core channel cleanup has a lot of code under spinlock with
preemption disabled this leads to deadlock scenerio.
Spinklock is used only for critical section, rest of the code is
be without spinlock.
CRs-Fixed: 1060407
Change-Id: I577dbff1cf2ee3711e1879aaa6dc48c72f98b98c
Signed-off-by: Dhoat Harpal <hdhoat@codeaurora.org>
Adding source and destination port number info
in the gso trace events to differentiate between
the flows.
CRs-Fixed: 982871
Change-Id: Idbae7f95dfd56293805b58e3c6626f5f6e07d08a
Acked-by: Ashwanth Goli <ashwanth@qti.qualcomm.com>
Signed-off-by: Ravinder Konka <rkonka@codeaurora.org>
[subashab@codeaurora.org: resolve trivial merge conflicts]
Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
With commit a3a3c169dd ("icnss: Prevent power collapse during
driver probe/remove"), NULL pointer dereference happens in probe
because of recent changes in the function. Fix the issue by using
right pointer.
CRs-Fixed: 1061279
Change-Id: I464579953fabe1bd42ced50f0bbfdfd3cda0a492
Signed-off-by: Prashanth Bhatta <bhattap@codeaurora.org>
Some HW monitors can do a better job of the sampling and the threshold
checking than the SW implementation in the governor. Update the governor's
API to add support for them.
Change-Id: Id4b5593a5ed3290684ba43ebebe2466ba0b730b6
Signed-off-by: Saravana Kannan <skannan@codeaurora.org>
Signed-off-by: David Keitel <dkeitel@codeaurora.org>
Add support for reading a speed-bin fuse for each cluster.
This allows for the selection of different OSM look-up tables
and thus different frequency configurations based upon device
fused values.
CRs-Fixed: 1057115
Change-Id: I9a864a2abb655e26fff5982b592b4f3c5dbfca24
Signed-off-by: Osvaldo Banuelos <osvaldob@codeaurora.org>
When Venus HW is not responding, before crashing the system,
print all the pending messages, current sessions and each
session info.
CRs-Fixed: 1059604
Change-Id: I8d39eaaa425e5e6fa97704c9bc15414f9a0f61ce
Signed-off-by: Praneeth Paladugu <ppaladug@codeaurora.org>
When USB RNDIS network device is bridged with other network device
on windows host machine, it has been observed that many MSG PACKET
FILTER packet with zero and non-zero value are send from host to
device. USB GSI state machine handles those and results into case
where TRB buffer's address is updated wrongly causing bus error
with IPA GSI. Fix this issue as below:
- Move MSG PACKET FILTER handling as part of ipa_work_handler()
- Add blocking USB3 ringing doorbell on receiving MSG PACKET FILTER
with non-zero value.
- Before processing HOST_NRDY event, see already posted event. If it
is HOST_READY event, then ignore HOST_NRDY and HOST_READY to avoid
performing operations related to HOST_NRDY and HOST_READY again.
CRs-Fixed: 1059577
Change-Id: I3ac8f740d7b84212da1d799b563a6e8ef1c9b493
Signed-off-by: Mayank Rana <mrana@codeaurora.org>
On MSMCOBALT, the GPU GX GDSC enable status register must be
polled to see if the GPU (and top level CPR sensors) are powered
before attempting to perform a CPR aging measurement.
Specify the address of this register as well the bitmask and
expected enable state masked register value in the VDD_GFX CPR
device node.
Change-Id: I55d5fb0c799dfec73830e8e97dcff79cd045b29c
Signed-off-by: David Collins <collinsd@codeaurora.org>
Some platforms have a special register that must be checked in
order to verify that it is possible to perform a CPR aging
measurement. For example, on MSMCOBALT, the GPU head switch
power state must be checked to see if a graphics CPR aging
measurement can take place.
Add support for this kind of custom aging allowed register
verification via three device tree properties: the address of
the register, the mask to use on the register, and the expected
masked value to read if aging measurements are allowed.
Change-Id: I7bcb88002bdc0609e835bc9ce29c84d93684e445
Signed-off-by: David Collins <collinsd@codeaurora.org>
Change the CPR sensor used for graphics aging measurements on
MSMCOBALT from 17 to 29. Sensor 29 is in the top level of the
GPU so it is guaranteed to be usable for aging measurements
whenever the GPU is powered regardless of the GPU operating
state.
Also bypass CPR sensors 0 - 23 during graphics CPR aging
measurements so that the measurements will succeed even if the
hardware power state of these CPR sensors changes during a
given measurement.
Change-Id: I3c0dd344e0192a375f31fefea7ef6559fab0652d
CRs-Fixed: 1061372
Signed-off-by: David Collins <collinsd@codeaurora.org>
Correct the condition checked in the CPR aging error cleanup path
so that a return code of 0 (success) is not returned when the
aging measurement fails. Also, correct the order of the
arguments in the aging failure error message.
Change-Id: Ica925da95b5ab35f63bb0355470475733080a88c
CRs-Fixed: 1061372
Signed-off-by: David Collins <collinsd@codeaurora.org>
During power up operation check whether power up
failure is due to SSR or noise over slimbus.
If power up failure is due to SSR prevent retry
operation.
This change also do allocate DMA memory for slimbus
bulk trasactions. Memory should be DMA for slimbus
transactions as using non-DMA memory will end up
in creating bounce buffers and affect the performance.
Change-Id: Iaca520d92c4ee2d73f8e751699779e6cfc9ddb87
Signed-off-by: Dilip Kota <dkota@codeaurora.org>
Not all memebers of config_32 are set before they are used which
might lead to invalid values being passed and used. To fix this issue
initialize all member variables of struct config_32 to 0 before
assigning specific values individually.
CRs-Fixed: 1058826
Change-Id: Ifea3a6e8bf45481c65a4455ee64318304798fee2
Signed-off-by: Laxminath Kasam <lkasam@codeaurora.org>
1. Select COMMON_CLK_MSM for msmfalcon when building
for 32-bit.
2. Select RATIONAL when COMMON_CLK_MSM is selected as
it's using an api from rational library.
Change-Id: I5b8fa962718a5ae44dfd18a13285715580ee0dbc
Signed-off-by: Srinivas Ramana <sramana@codeaurora.org>
