A frame trigger with posted start may have two frames
in wait state due to irq disable on that CPU. In such
case, frame_done count can reach till 2. Allowing count
only till 1, can cause the release_fence trigger miss
and a buffer is held by DPU driver.
Change-Id: I42c10b064ebcaff136591975f3010c11f99a0731
Signed-off-by: Dhaval Patel <[email protected]>
Return ERR_PTR with -ENODEV when the USBPD module is not
enabled. This allows the DP driver to handle the failure
correctly and prevents any potential NULL pointer
dereferencing in the clean up path.
Change-Id: I4fbc559276bebe770a27935243a59c10c1bb5491
Signed-off-by: Tatenda Chipeperekwa <[email protected]>
Treating hw version as a signed int results in a negative
number when major version is more than 7, leading to errors
when comparing versions in the driver. Fix this by making
sure all HW version definitions and variables are unsigned.
Change-Id: Ic77183f85e5408092e05817cf95df5b0f0fcac75
Signed-off-by: Nilaan Gunabalachandran <[email protected]>
SDE driver started supporting multirect mode with rect_1 only
configuration. In such case, master plane can not trigger
pipe fetch halt independently. This change removes the pipe
fetch halt check completely because it was only done for
master plane without buffer flip usecase. AXI fetch halt
provides similar functionality during idle power collapse
and suspend-resume.
Change-Id: I79d9d0eac2de95f1bb88561c7cc259e0cc4b2ca4
Signed-off-by: Dhaval Patel <[email protected]>
Address out of range issues due to faulty and missing
array size checks.
Change-Id: I2aefb1b8d4015a73ea87a64182e090247f5a9da0
Signed-off-by: Christopher Braga <[email protected]>
Change IGC version to 4.0 to clearly indicate that support
for LUT 257 in DSPP is now supported and expected.
Change-Id: Ie3bb5b0150bdf6c332f86d0ae416b4f6fc42e70f
Signed-off-by: Christopher Braga <[email protected]>
When CONFIG_DEBUG_FS is disabled skip creating root directory.
Change-Id: I97bed925392b781b73687164ca55e6cb09f951fc
Signed-off-by: Rajat Gupta <[email protected]>
Currently DP driver determines if a mode is DSC capable
based on a DTSI entry and the required number of DSC
to support it. This approach does not scale when there
is an overlap in DSC requirement between DSI displays
and external DP display, thus causing one of the display to
report modes that cannot be supported.
This change compares the resources reserved for DP driver
calculated at initialization time and the currently available
ones to determine the correct number of resources that DP driver
can use. It also adds DSC and topology filtering logic and moves
DSC hardware specific from DP driver to SDE driver.
Change-Id: I8e601de33422b7c6d786826f7bfe152c4af8a6b5
Signed-off-by: Amine Najahi <[email protected]>
Setup vig pipe makes an early return if csc and qseed are
not supported. This change moves addition of vig formats
before intializing qseed and csc.
Change-Id: I2eb651f7bbd81757a9de23501fda51a510d0e673
Signed-off-by: santosh <[email protected]>
Modify the size of SID's passed as an argument during
scm call as per client requirement.
Change-Id: Idd3bb57a8f9e0a4e7eb6a23d96bfa5b68510063a
Signed-off-by: Jayaprakash <[email protected]>
Remove mutex unlock from msm_gem_get_iova_locked
as acquiring and releasing the lock are handled
in the caller function.
Change-Id: Ie4a8408d3e6308d19eb26aafee8dfe15bcb682da
Signed-off-by: Krishna Manikandan <[email protected]>
During transition to cwb, kthread initialisation has to
be done for the corresponding encoder off work so that
the correct worker thread is used for this particular
work. There can be scenarios where a cwb commit is
received after a writeback session and the worker
associated with the off work is still assigned to
old crtc's worker resulting in a mismatch when
this work is queued. Add support to handle such
scenarios.
Change-Id: I6080025e799977827f4d0f4ab7eb93c6644f981e
Signed-off-by: Krishna Manikandan <[email protected]>
Enable dim layer event log to know the composition of all
blend stages in a commit from xlogs.
Change-Id: I387fd888f06f0c61d2042896459119d614b3f60b
Signed-off-by: Raviteja Tamatam <[email protected]>
This change checks max limit while allocating encoders
during display init.
Change-Id: I48736667c2b83f916f09f9e81cb1b24a9fef215a
Signed-off-by: Yashwanth <[email protected]>
In targets where ubwc is not supported, atomic check
should fail and return a error value if the input format
is ubwc.
Change-Id: I21a40f510cc852e64fbcc05a5fb4848da4b4faaa
Signed-off-by: Yashwanth <[email protected]>
framebuffer support is required in some use cases, so add
framebuffer emulation for KMS driver.
Change-Id: Icf0bbcd59ca430478d2bef4df033988d0338f2bb
Signed-off-by: Ray Zhang <[email protected]>
This change adds dt based support for pm qos irq
latency instead of using hardcoded value.
Change-Id: I9f67ed1092eefac193a409773f841350532bc722
Signed-off-by: Yashwanth <[email protected]>
Currently during rm_release, drm_conn is available only if connector
and encoder are bound together. If a test_only commit occurs
before a non-blocking modeset commit, this bound is uncertain.
The resources allocated by test_only commit will not be released
and leads to resource allocation failures for the consequent
commits. Add changes to modify this by checking for the encoder
attached to sde_connector which happens during bootup.
Change-Id: Icb1d9595efda8a104ba527175b2dbc6a2c856d44
Signed-off-by: Jayaprakash <[email protected]>
Primary and secondary dsi displays are built-in displays
and they are supported during all concurrency usecases
without resource allocation failure. DP mode filter
logic should provide supported mode information based
on free mdp resources after dsi resource assignment.
Change-Id: I3a9637a91ea1ffcc31997e25caff7f13605283ac
Signed-off-by: Dhaval Patel <[email protected]>
Set the aborted flag during module initialization so the default
behavior is to error out on aux transactions. The flag will be
reset when there is a hot plug connect and a sink device is
actually present.
Change-Id: I9e31b9293e817ed951ba5464d403e7178615418f
Signed-off-by: Yuan Zhao <[email protected]>
This change addresses out of range and null checks in
sde driver.
Change-Id: I4ee82760ce3ee7053c336e49ec9eaae8b4c31b1e
Signed-off-by: Narendra Muppalla <[email protected]>
In current driver if client did not reset system cache
crtc property, system cache write state is on forever and
breaks the system cache feature. This change restricts
entering into cache write state only if it's commit right
after idle notify. This change also adds event logs to
capture system cache feature state changes.
Change-Id: Ie46fc9113f752ed8989dab99301690a13003b00b
Signed-off-by: Prabhanjan Kandula <[email protected]>
Allow clients to reset the hdr meta data property even when
connector does not support hdr. This is required for shell stop
case cleanup when dp connector might be in disconnected state.
Change-Id: Ie0db61530216f1fab6c952933425259d32cfc834
Signed-off-by: Abhijit Kulkarni <[email protected]>
This change allows the configuration where virtual plane is
staged without staging the corresponding master plane. Hw allows
such configuration and additionally during shell stop cases,
framework removes each staged plane one by one which requires
this configuration to be supported.
Change-Id: I85ce571dbbc148c069c2cf35cba55ad740d52b5d
Signed-off-by: Abhijit Kulkarni <[email protected]>
Lots of dsi FIFO error logs were printed into serial in
dsi ISR. It will make device stuck or deadlock. Remove
these logs from dsi ISR, and enable the log in FIFO error
workqueue handler.
Change-Id: I0e9b2312cb76d345ec5a9b9628c52b47d5163fde
Signed-off-by: Yuan Zhao <[email protected]>
This is a common case that seamless transition was rejected
when active/connectors changed, so just print an info log here.
Change-Id: Iada9ad84af1c072ccdbc97f5bf438de6a032f45f
Signed-off-by: Yuan Zhao <[email protected]>
The DRM framework has added support to check for and calculate the
correct checksum for test cases in which the EDID is corrupted e.g.
test 4.2.2.6. Update the implementation to use the correct checksum
that is available as a field in the DP connector structure.
Change-Id: Ia9f65926e2c0fb3444aa87a4ac97850fbb4d70f3
Signed-off-by: Tatenda Chipeperekwa <[email protected]>
Add support for all 4LM topologies in new DCE encoder framework.
This change also aligns with the new way of checking topology
information.
Change-Id: I5358d60634070bdb26059056db884ad4161c073e
Signed-off-by: Amine Najahi <[email protected]>
Add smmu cache hint at during the msm gem prime import
to ensure memory is cacheable. Ensure sys cache feature
is added to all sspp, not just vig.
Change-Id: Icc10468ad8d3e7c6aabd437776cde99eb627375a
Signed-off-by: Nilaan Gunabalachandran <[email protected]>
Register for USB cable notifications so we can manually trigger
a disconnect event if the USB cable is disconnected during a
simulation.
Change-Id: I08a9313bcde3e74f2a76804e934423e0fc2d6ab6
Signed-off-by: Tatenda Chipeperekwa <[email protected]>
Use lower frame threshold time for high refresh rates. For resfreh rates
greater than 120 FPS the threshold time is 500us.
Change-Id: Ie5a37c0eec6c2c2cbf90e039eb6b15bc7edd2492
Signed-off-by: Satya Rama Aditya Pinapala <[email protected]>
Vsync event timer wakeup was designed to reduced the
interrupt latency and trigger retire fence without delay.
This is fixed by avoiding CPU power collapse where MDSS
interrupt is scheduled. This change avoids extra CPU
wakeup.
Change-Id: Iadaf0e2b84fb079bbc64d9201230df54f8dbe8c1
Signed-off-by: Dhaval Patel <[email protected]>
Video mode display keeps the MDP clocks in ON state but
disables irq during static screen to avoid cpu wakeup.
In such case, CPU pm_qos vote should also be removed
to allow LPM transition. This change triggers the
pm_qos vote based on mdp interrupt enable counts
instead of runtime_pm callback. It works for
multi-display concurrency also.
Change-Id: I7a60f3f593e409269e00abd7499c4a5756035615
Signed-off-by: Dhaval Patel <[email protected]>
This change adds support for enabling WB maximum linewidth
based on color format is linear or UWBC.
Change-Id: Icc71eb14b3156e06036a4a82029d9d7a5c89e909
Signed-off-by: Prabhanjan Kandula <[email protected]>
Some sinks need more time to stabilize after synchronization and
before it can handle an HDCP authentication request.
This change adds the delay for better interoperability.
Change-Id: I4fffb9024f782dd68e3a883208cb8e9d0c5350a1
Signed-off-by: poyewu <[email protected]>
This change resets the bandwidth vote after lastclose is called
and all the custom properties are reset. This is required so that
on next power commit when user space driver does not apply any
bandwidth vote, driver applies correct vote. Additionally this
change ensures that aggregate bandwidth vote never exceeds the
max threshold.
Change-Id: I95ca761b381709850b630ecd6647b45968a5041e
Signed-off-by: Abhijit Kulkarni <[email protected]>
