This change allows kernel modules to make calls to register external
display interfaces and deregister them.
Change-Id: Ia7236bd458565bbbdec4a8355c40a1c5a1bcab86
Signed-off-by: Samantha Tran <samtran@codeaurora.org>
mhi_reg_write_enqueue API stores reg, val and valid which can go
out of order. In case valid is set to true before val is set,
offload worker running on another core ends up writing stale value
to register. Another possibility is valid being set to true but not
visible to other cores. When offload worker gets a chance to run,
this results into skipping register write. Fix these issues by
adding smp_wmb() between stores of val and valid and another after
valid is set.
Change-Id: I3b930e7fad4252d34386de525491f94997b34f36
Signed-off-by: Hemant Kumar <hemantk@codeaurora.org>
Add new GPU chip id and required code to
Support A619 GPU.
Change-Id: I9b4e8b35e536486c63661b050a2242538d873c2d
Signed-off-by: Hareesh Gundu <hareeshg@codeaurora.org>
CVP SSR crashed when derefencing the pointer to a
shared memory buffer descriptor.
Change-Id: I3f7b690d90496efdf0451b5500578fee636f81fe
Signed-off-by: George Shen <sqiao@codeaurora.org>
Currently, the async job member of the ioctl structure is not
being initialized properly while making an unmap call to the
remote subsystem because of which an indeterminate value is used
to incorrectly mark this RPC call as an asynchronous job.
Initialize the async job field to NULL before making the RPC call
to unmap memory on DSP.
Change-Id: I611e9419793da32f73b9997c7dbe50e3d1e7e81b
Acked-by: Thyagarajan Venkatanarayanan <venkatan@qti.qualcomm.com>
Signed-off-by: Himateja Reddy <hmreddy@codeaurora.org>
Adding numerator and denominator to sensor data
for increasing scalability of driver.
Change-Id: Iba90d890b70feae41a273fd51d1a76ab61c5211b
Signed-off-by: Rohit Bandi <rohitbandi@codeaurora.org>
Add support for CDSP DCVS and CDSP hardware cooling devices to QMI
cooling device driver. These cooling devices can be used to mitigate
CDSP subsystem at high temperature via qmi interface.
Change-Id: Ia9906b07ebca4bd6158b68e6e100405ebaa30b60
Signed-off-by: Manaf Meethalavalappu Pallikunhi <manafm@codeaurora.org>
Do not hard code resource address in PDC initialization
sequence to avoid incorrect configuration. Instead get
resource address from cmd-db driver.
Change-Id: If2cc8378c1c00304033d726fdb8bd4e8690ef110
Signed-off-by: Hareesh Gundu <hareeshg@codeaurora.org>
Add clock IDs for GCC/DISPCC/GPUCC for clients to be able to
request for clocks from these clock controllers.
Change-Id: I601af44801668ae76d5cd181051ae9ba7b0c6dc0
Signed-off-by: Saurabh Sahu <sausah@codeaurora.org>
This change is a performance optimization. Instead of sending a DRI
deassert command at the end of a Data Ready DRI, the FW shall get a FIFO
empty internal signal. FW sets the don't deassert bit in relevant cases,
like Data Ready DRI, or SYSASSERT DRI.
Change-Id: I810e891a82b80c256857987b446c2162bacc399e
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
Data read over SPI was guaranteed to be in a full burst size. This change
enables FW to raise the Data Ready DRI earlier, before the full burst is
ready. The driver must handle the case where data size reported by the HW
is less than a full burst, and then execute another iteration until the
whole burst is read.
This change causes a performance optimization, since the FW can raise the
Data Ready DRI earlier. The whole burst data is usually in place when the
driver handles it.
Change-Id: I98966ec5c1ef614c3c145e99bb3327febe1ebe74
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
Return -ENODEV in case a change_mode command is issued when the system
is in SYS_ASSERT state. This guarantees that the user space application
is aware of the bad state and waits for the FW_READY event.
Change-Id: If26dc115078a3cd1af5d7efa93225019c1105184
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
Impose less restrictions on state changes, as required by the new system
design.
Change-Id: I9ea6d2fcf2bde4dd013cf7c494dcb5c0c4948fd7
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
In case user space application requests for data size which is not a
multiple of the burst size and there is enough data in the drivers internal
buffer, a partial burst may be read by the application, which may cause
loss of burst boundary.
This patch fixes the above. In case the application supplies a buffer which
is smaller than the burst size, an error is returned to the application.
Change-Id: I5233476d99937dd1a13ceaec625588414a1dc04a
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
In case sys-assert DRI is pending, there is no need to process other
pending DRIs.
Change-Id: I3f1422d75ae5d966c855213031b3524c46d7f5d5
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
Prevent state changes during DRI handling, which may cause the driver to
get out of synchronization. Prevent data read from user space and from
kernel space during change_mode processing.
Change-Id: I2ee277c8894996119d13acbe87a5163e4064923c
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
change_mode command sometimes fail. In such a case, a proper error code
should be returned to the calling user space application. Otherwise, the
application tries to read data, which is not available.
Change-Id: I03778cace54fd98847a099bbeaf7884cfbaac586
Signed-off-by: Gidon Studinski <gidons@codeaurora.org>
Signed-off-by: Alexei Avshalom Lazar <ailizaro@codeaurora.org>
