Fix IGT test case (kms_atomic_transition)
-DRM sets the mode_changed flag, while we don't need to do a full
modeset.
-We want to override the mode_changed flag in this case
If we don't do this, we will still bypass the modeset in DC. This will
fail to update the new stream_status, causing nullptr at a later stage
when trying to access stream_status"
We now avoid this by discarding the new_stream instead of partially
filling it.
Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use per surface validation hook.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
- Protected by ENABLE_FBC compile flag
- DC part will follow
Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This separation stopped making sense a long time ago.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
also simplify DTN INFO. all we need is a way to have DC log register states
some where we can compare with golden. going through connectivity is overkill.
Signed-off-by: Tony Cheng <tony.cheng@amd.com>
Reviewed-by: Yongqiang Sun <yongqiang.sun@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use dc_validate_stream from crtc atomic_check for single stream
validation.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Wenjing Liu <Wenjing.Liu@amd.com>
Reviewed-by: Wenjing Liu <Wenjing.Liu@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In any drm ioctl call, drm_atomic_state_clear() is called at the end to
destroy the states; even if atomic check fails. Therefore, releasing
states on atomic check failure is incorrect.
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Handling a use case of TEST_ONLY request from DRM where commit is
not goiing to be called. We need to release the allocated dc_validate_context
in this case.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Nothing to validate in DC in this case. Skip it.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Roman Li <Roman.Li@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Attach surface to state.
Remove Create surface from commit.
Propogate any surface creation and initialization error back to atomic_check caller.
clean outdated code in check and commit.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
New surface is created in check only, in duplicate we just
reference exsisting surface and in destroy we release it.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove acrtc->stream and move it into dm_crtc_state.
This allows to get rid of dm_atomic_state->dm_set.
Also reuse streams created in atomic_check during
commit.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reimplement atomic_state_alloc and atomic_state_clear to
release validate_ctx.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Needed by objs in dm_atomic_state
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We really want to use the new private_atomic_state but can't right now
as we have to maintain some backward compatibility to older kernels. For
now let's follow Intel's approach and extend the drm_atomic_state.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Adds some additional information to logs in dc_commit_streams to
better match Dal2.
Also adds a new function, dc_raw_log, that has the same
functionality as dc_conn_log, but does not attach display specific
prefixes to the log output.
Finally, adds a new DC log type, LOG_DTN, that logs to
LogMinor_DispConnect_dtn in DM. This new log type and dc_raw_log
are used to generate clock info in the logs to match Dal2.
Signed-off-by: Corbin McElhanney <corbin.mcelhanney@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This change fixes dcn10 front end reset sequence. Previously we
would reset front end during flip which led to issues
in certain MPO and 4k/5k scenarios. We would also never properly
power gate our front end.
Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
These stubs are initial only since we need to flatten
DC objects (steran at least) to implement deep copy.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Previously, we assumed that allow_modeset=false => page flip. This
assumption breaks when an atomic commit is submitted with allow_modeset
set to false, since the legacy flip code is never called (the legacy
code grabs the vblank reference).
Fix: Move drm_vblank_get() from amdgpu_atomic_helper_page_flip() to
amdgpu_dm_commit_surfaces().
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm_plane_state is subclassed by dm_plane_state, so atomic driver should
use __drm_atomic_helper_plane_destroy_state() to destroy a state.
drm_atomic_helper_plane_destroy_state() frees drm_plane_state which is
allocated inside dm_plane_state, this is problematic.
Fixes: 95ae03a ("drm/amd/display: Create dm_plane_state.")
Signed-off-by: Xiaojie Yuan <Xiaojie.Yuan@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fix typos.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Roman Li <Roman.Li@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
IGT currently does not properly commit changes on planes with multiple
possible CRTC's. Set one valid CRTC for each plane for now, plus one
underlay plane on Carizzo and Stoney that is valid for all CRTCs.
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Switch to wait_for_completion_interruptible_timeout wait
since the lock is called from IOCTL context and can be
interrupted by a signal.
Global lock function might return EDEADLK or EINTR which
is not an error and just singals to user mode to restart
the call.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Due to using dc_commit_surface_to_stream instead of build
stream and surface updates any surface commit today is
evlauted to full. Until we fix this and can corretly
evluate type of surface update, anything which is not page
flip or cursor update will be treted as full update chnage
and global lock will be aquired.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Missing function implementation was leading to EINVAL
in UMD thus not adding MST connector to X topology
and hence not getting set mode for it.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Hook up dc_surface creation/destruction to dm_plane_state.
Rename amdgpu_drm_plane_state to dm_plane_state and do
minor cleanups.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
1) Always call flip ISR, for any surface comit, not only for page flip.
2) For events not signaled in flip ISR, handle them in end of commit_tail.
3)Fix race condition between subsequent atomic calls by removing
current state access AFTER drm_atomic_helper_commit_hw_done
was called. After this call a pending next commit will
swap in his own state into objects and proceed.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This code is remanant of pre atomic age when flip was a
standalone IOCTL.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Zeyu Fan <Zeyu.Fan@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Do not broadcast (to DRM) unsupported RGB formats.
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Switch from legacy cursor to DRM cursor plane. Cursor
is not an actual plane but more of a subplane of each
pipe. Bind a DRM cursor plane instance to each CRTC.
Eliminate seperate FB object allocation for cursor and
clean dm_crtc_cursor_set.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Leo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch updates the YUV format supported to
NV12 and NV21 and hence updates the offsets.
BUG: SWDEV-119421
TEST: (On Chromium OS for Stoney Only)
* Executed below tests to see YUV(underlay) & RGB planes on eDP
plane_test --format XR24 --size 500x100 -p --format NV12 --size
500x500
Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use drm_atomic_crtc_needs_modeset instead.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Check if adding surface is failed to prevent NULL pointer deref.
Signed-off-by: Jerry Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Summury of changes:
1: Both in check and commit Connector properties were handled as
part of for_each(crtc) loop while they shoud have been handled
in a dedicated for_each(connector)
loop since they are connector properties. Moved.
2: Removed hacky plane add in amdgpu_dm_connector_atomic_set_property
to force iteration on plane forconnector property. This was
causing double call to commit_surface_for_stream both in crtc loop
and plane loop.
3: Remove middleman DC interface and call dc_commit_surfaces_to_stream
directly to increase code clarity.
Remove it from atomic_commit.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This will be NULL on a new stream. DC handles it gracefully.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
You cannot have modeset and flip in the same call for
same CRTC, in such case it will be set mode and set plane,
not a flip.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use DC interface to query for plane update type
so in case of FULL update you flush any outstanding
commits.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove extra loop we have for page flips and do flips in same loop we do
for surface create/update.
Add documentation for synchronization between commits on different crtcs.
Rename function to have DM prefix.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Today we use special interface for flip because of fear of cuncurency issues
over dc->current_ctx. This should be no longer an issue when flipping on
multiple CRTCs concurently since for fast update (as flip is) no new context
is created and the exsisitng is not destroyed. For full updates case when
removing or adding streams on once CRTC while flipping on another
Adding all current active CRTC's states to the atomic commit in
amdgpu_dm_atomic_check will garntee that any such full update commit
will wait for completion of any outstanding flip.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Problem :
A race between two adjecent page flips makes the earlier one
to release an alocated frame buffer for the subsequent one -
since there are 2 frambuffer swapped back and forth between flips,
the 'new' fb of the later flip is actually the 'previous' fb for the earlier flip.
Fix:
Don't set fb->address = 0 in cleanup hook, this is unnecessery and
erases the newly cached adress that was set in prepare_fb of the second
flip.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Link index is an unnecessery level of inderection when
calling from kernel i2c/aux transfer into DAL.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
There is a hard hang observed during resume from S3 when
the system receives a DP short pulse interrupt. This is
because there are two code paths contending for GPIO
access for AUX channel transactions. One such path is
through amdgpu_dm_display_resume() function which is
invoked from the regular system resume code path. The
other path is through handle_hpd_rx_irq(), which is
invoked in response to system receiving DP short pulse
interrupt. handle_hpd_rx_irq() guards against conflicting
GPIO access using hpd_lock, but the GPIO access from
amdgpu_dm_display_resume() remains unguarded.
This patch makes sure we use hpd_lock inside
amdgpu_dm_display_resume() to avoid race conditions
for GPIO access.
Signed-off-by: Arindam Nath <arindam.nath@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
dc_commit_surfaces_to_stream() function currently
is handle's only one plane at a time.
This will not work if multiple planes have to be set to a crtc.
The functionality of dc_commit_surfaces_to_stream() with this patch
is slit into
1. Accumulate and initialise all the surfaces that needs to be
set to a crtc.
2. Update the intialised set of surfaces to the steam in one go.
Hence dc_commit_surfaces_to_stream() is renamed to init_surfaces().
Once all the planes requested by user space are initialised,
dc_commit_surfaces_to_stream() shall sequentially populates *updates,
*flip_addr, *plane_info and *scaling_info for all surfaces.
BUG: SWDEV-119421
TEST: (On Chromium OS for Stoney Only)
* Chromium UI comes up, on both eDP & DP.
* 'new_surface_count' now changes as per user input for e.g for
all below run tests its 2, without this patch for the below
tests it used to be 1
* Executed below tests to see YUV(underlay) & RGB planes on eDP
plane_test --format XR24 --size 500x100 -p --format YV12 --size 500x500
plane_test --format AR24 --size 500x50 -p --format YV12 --size 150x150
plane_test --format AR24 --size 500x50 -p --format YV12 --size 1366x768
Signed-off-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Shirish S <shirish.s@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch populates the YUV surface configurations.
Tests: (On Chromium OS for Stoney Only)
builds without any errors.
Signed-off-by: Shirish S <shirish.s@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This will allow us to carry it from check to commit
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This should make things simpler when we try to rework this later when we
pass validate_context from atomic_check to atomic_commit.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
They could differ between ASIC generations
Signed-off-by: Sylvia Tsai <sylvia.tsai@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Jordan Lazare <Jordan.Lazare@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
- Remove x,y hotspot from dc_cursor_attributes. Only program it
through setPosition.
Signed-off-by: Zeyu Fan <Zeyu.Fan@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Zeyu Fan <Zeyu.Fan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make sure we do not program a hw pageflip inside vblank 'n' iff the
atomic flip is comitted while inside the same vblank 'n'. We must
defer such a flip by one refresh cycle to vblank 'n+1', unless this
is a DRM_MODE_PAGE_FLIP_ASYNC async pageflip, which must always
execute as soon as possible.
Without this, pageflips programmed via X11 GLX_OML_sync_control extensions
glXSwapBuffersMscOML(..., target_msc, ...); call and/or via DRI3/Present
PresentPixmap(..., target_msc, ...); request will complete one vblank
too early whenever target_msc > current_msc + 1, ie. more than 1 vblank
in the future. In such a case, the call of the pageflip ioctl() would be
triggered by a queued drmWaitVblank() vblank event, which itself gets
dispatched inside the vblank one frame before the target_msc vblank.
Testing with this patch does no longer show any problems with
OML_sync_control swap scheduling or flip completion timestamps.
Tested on R9 380 Tonga.
v2: Add acked/r-b by Harry and Michel.
v3: Feedback from Andrey: Must not wait an extra frame for
DRM_MODE_PAGE_FLIP_ASYNC flips.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Acked-by: Harry Wentland <harry.wentland@amd.com> (v1)
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Cc: Harry Wentland <Harry.Wentland@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since DC now uses CRTC_VERTICAL_INTERRUPT0 as VBLANK irq trigger
and vblank interrupts actually happen earliest at start of vblank,
instead of a bit before vblank, we no longer need some of the
fudging logic to deal with too early vblank irq handling (grep for
lb_vblank_lead_lines). This itself fixes a pageflip scheduling
bug in DC, caused by uninitialized use of lb_vblank_lead_lines,
with a wrong startup value of 0. Thanks to the new vblank irq
trigger this value of zero is now actually correct for DC :).
A new problem is that vblank irq's race against pflip irq's,
and as both can fire at first line of vblank, it is no longer
guaranteed that vblank irq handling (therefore -> drm_handle_vblank()
-> drm_update_vblank_count()) executes before pflip irq handling
for a given vblank interval when a pageflip completes. Therefore
the vblank count and timestamps emitted to user-space as part of
the pageflip completion event will be often stale and cause new
timestamping and swap scheduling errors in user-space.
This was observed with large frequency on R9 380 Tonga Pro.
Fix this by enforcing a vblank count+timestamp update right
before emitting the pageflip completion event from the pflip
irq handler. The logic in core drm_update_vblank_count() makes
sure that no redundant or conflicting updates happen, iow. the
call turns into a no-op if it wasn't needed for that vblank,
burning a few microseconds of cpu time though.
Successfully tested on AMD R9 380 "Tonga Pro" (VI/DCE 10)
with DC enabled on the current DC staging branch. Independent
measurement of pageflip completion timing with special hardware
measurement equipment now confirms correct pageflip timestamps
and counts in the pageflip completion events.
v2: Review comments by Michel, drop outdated paragraph
about problem already fixed in 2nd patch of the series.
Add acked/r-b by Harry and Michel.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Acked-by: Harry Wentland <harry.wentland@amd.com> (v1)
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
DRM_IOCTL_MODE_GETCONNECTOR fails with EINVAL on enabling DRIVER_ATOMIC
With this DRM_IOCTL_MODE_GETCONNECTOR returns all the connector properties.
freesync_property and freesync_capable_property return 0 currently.
TESTS(On Chromium OS on Stoney Only)
* Builds without compilation errors.
* 'atomictest' proceeds after applying patch and fails with vblank event
timed out.
* Chromium OS ui comes up.
Signed-off-by: Pratik Vishwakarma <Pratik.Vishwakarma@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We still rely on legacy properties. Fallback on legacy properties until
we get to pull these into some atomic state.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
4.9 kernel will always add the assigned crtc to possible_crtcs on a
plane. This is no longer the case on newer kernels. Make sure we allow
any plane on any crtc.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Jordan Lazare <Jordan.Lazare@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Roman Li <Roman.Li@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Avoid enabling CRTC_VERTICAL_INTERRUPT0 twice on resume.
It's enabled once from within manage_dm_interrupts in mode set
and another explicitly from amdgpu_dm_irq_resume_late.
Seems it lead to CRTC hang.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Allow Linux to use DRM provided EDID read functioality
by moving DAL edid implementation to module hence
removing this code from DC by this cleaning up DC
code for upstream.
v2: Removing ddc_service. No more need for it.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
I2C_M_RD was translated to write instead of read.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We can keep a 1:1 relation between a link and a physical
connector and hence skip the iteration. This function
is used in context of only physical connetors.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch introduces amdgpu_drm_plane_state
structure, which subclasses drm_plane_state and
holds data suitable for configuring hardware.
It switches reset(), atomic_duplicate_state()
& atomic_destroy_state() functions to new internal
implementation, earlier they were pointing to
drm core functions.
TESTS(On Chromium OS on Stoney Only)
* Builds without compilation errors.
* 'plane_test' passes for XR24 format
based Overlay plane.
* Chromium OS ui comes up.
Signed-off-by: Shirish S <shirish.s@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Current design has per-crtc-plane model.
As a result, for asic's that support underlay,
are unable to expose it to user space for modesetting.
To enable this, the drm driver intialisation now runs
for number of surfaces instead of stream/crtc.
This patch plumbs surface capabilities to drm framework
so that it can be effectively used by user space.
Tests: (On Chromium OS for Stoney Only)
* 'modetest -p' now shows additional plane
with YUV capabilities in case of CZ and ST.
* 'plane_test' fails with below error:
[drm:amdgpu_dm_connector_atomic_set_property [amdgpu]] *ERROR* Unsupported screen depth 0
as ther is no support for YUYV
* Checked multimonitor display works fine
Signed-off-by: Shirish S <shirish.s@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Register ISR hnadler on the new interrupt.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This will return back MAX cursor size for given ASIC.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Jordan Lazare <Jordan.Lazare@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
ret value in amdgpu_dm_atomic_check was not rest to EINVAL
after drm_atomic_add_affected_planes and by this making
any subsequent validation failure pass when returning
to atomic_check.
v2: Add WARN_ON print for dc_commit_streams in
amdgpu_dm_atomic_commit_tail since this should never
fail.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Jordan Lazare <Jordan.Lazare@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This wires DCE12 support into DC and enables it.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Jordan Lazare <Jordan.Lazare@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This function makes a lot of assumptions to try and deduce
tha an atomic_commit came from FLIP ioctl. Those assumptions
are not explicit contract with DRM API and might lead to wrong
result as the code changes. Instead use drm_atomic_state.allow_modeset
flag which explicitly is set to false only in atomic flip
helper function, otherwise this flag is always true.
P.S When we get to do actual atomic IOCTL this will have to be revised
with the atomic call behavior.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Headless mode set needs to be synchronized against outstanding nonblocking
commits. This achieved by building atomic state and commiting it.
v2: rebase on 4.11
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Modify amdgpu_dm_atomic_comit to implement
atomic_comit_tail hook.
Unify Buffer objects allocation and dealocation
for surface updates and page flips.
Simplify wait for fences and target_vbank logic
for non blockiing commit.
Remove hacky update surface to page flip synchronization
we had and rely on atomic framework synchronization logic.
v2:
Add state->allow_modeset as indicator of page flip call.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Don't tell DRM otherwise.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
No need to assign them dynamically. This is much more readable.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Remove a false assumption that a cached EDID will be present whenever
the connector is in a connected state as this will only be true for
logical MST ports. For physical ports the EDID will never be cached,
which will cause them to always appear as disconnected.
This reverts commit 4ff8a88888de271bfb7750b2a5c68163848e2bf1
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
They might call commit with ACTION_NOTHING without a mode.
We shouldn't crash but simply skip updating stream scaling
settings since scaling obviously didn't change without a
provided mode.
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This parameter name is misleading. It's previous
meaning confuses with ASYNC (immidiate flip)
which is totatlly different thing then nonblocking
commit.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Jordan Lazare <Jordan.Lazare@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Switch from VUPDATE to VBLANK.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use new functions so flip failures can be gracefully handled
v2:
Avoid -EINVAL returned from amdgpu_crtc_prepare_flip in some
error cases, it is not allowed according to expected
return values for atomic_commit hook.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
dc_target does not fit well into DRM framework so removed it.
This will prevent the driver from leveraging the pipe-split
code for tiled displays, so will have to be handled at a higher
level. Most places that used dc_target now directly use dc_stream
instead.
Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Since during suspend, the cursor registers are cleared,
once the system resumes back, the cursor remains disabled.
cursor_set_attributes() only sets the cursor attributes
along with cursor size and surface address, but does not
enable the cursor back on.
We need to save the current cursor location so that
we can resume back to the same location. This is done
in dm_crtc_cursor_move(), where we save the current
cursor location into cursor_x and cursor_y. Later during
resume we use these same values to set the cursor
position along with cursor attributes.
Signed-off-by: Arindam Nath <arindam.nath@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Problem: MST topology discovery was started before fbdev
init causing kernel panic on boot.
Fix: Defer MST to late init hook, after fbdev strucutres
already initilized.
v2: Add explicit late_init hook function in DAL.
v3:
Change signature of detect_mst_link_for_all_connectors
from void* to drm_dev*
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For dpms usecase we need to set surface transfer function
regardless of gamma set flag.
Signed-off-by: Roman Li <Roman.Li@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The dc_gamma in dc_surface will be programmed to the input
LUT if provided. If dc_gamma is not provided in dc_surface
regamma may be used to emulate gamma.
Some refactor and cleanup included as well.
Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
While I'm sure this is useful I think we should bring it back later.
It's usage of pid/tgid is incorrect, you have to get/put
pid/tgids not store them away.
Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Refactor part 3 - Moving output transfer function from surface to stream
Split HWSS to program degamma and regamma separately.
Degamma should be dependent on input transfer function.
And Regamma should depend on the desired output transfer function.
Signed-off-by: Anthony Koo <anthony.koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v2: agd: squash in dm fix, rebase
Signed-off-by: Jordan Lazare <Jordan.Lazare@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Retain a dc_sink pointer until a new physical pointer
arrives in case of new display connected.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Some functions like dm_bandwidth_update are placeholders
for future implementations. Right now, they simply
print messages like "DM_NOT_IMPL: dm_bandwidth_update".
This message even though informational, sometimes can
create some confusion for users.
Since these functions are only skeletons for future, we
simply remove the rather un-unecessary messages from these
functions.
Signed-off-by: Arindam Nath <arindam.nath@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Now that the mc_access functions are gone, we no longer
need separate structs for all the different dce families
in dm.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Supported DCE versions: 8.0, 10.0, 11.0, 11.2
v2: rebase against 4.11
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
