To avoid confusion, need to suppress the error message when get
-ERESTARTSYS error code. It is normal when getting interrupted
by signals in the process of a wait for the buffer to become
unreserved. Only propagate to user-mode for further action,
no need to pop up error message.
Signed-off-by: NJerry Zuo <Jerry.Zuo@amd.com>
Reviewed-by: NSun peng Li <Sunpeng.Li@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Regression caused by: Ib98354194d7

Need to check crtc->stream before updating cursor attributes
and position.
Signed-off-by: NJerry Zuo <Jerry.Zuo@amd.com>
Reviewed-by: NRoman Li <Roman.Li@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

	1. Fixed acquire free split pipe bug.
	2. Change return value for dc_add_stream_to_ctx
	   from bool to enum.
	4. Remove redundant apply_ctx_for_surface calling
	5. Unlock pipe after back end programming.
Signed-off-by: NYongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NYongqiang Sun <yongqiang.sun@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Split update_planes_and_stream_state (split Software and Hardware
programming) as the state is already build, so we only need to
program the hardware
Signed-off-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Signed-off-by: NBhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Prevent NULL pointer on new_stream being added to ctx
when added MST connectors cannot be found in existing crtc_state
in the chained mode
Signed-off-by: NJerry Zuo <Jerry.Zuo@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Needs effort to take care of the fake sink scenario
in downstream daisy chain device. Exclude MST from
fake sink feature for now.
Signed-off-by: NJerry Zuo <Jerry.Zuo@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

During system suspend, we:

1. Cache a duplicate of the current DRM atomic state, which calls hooks
   to duplicate our internal states.
2. Call hooks to disable all functionalities.
3. System suspended.

During resume, we attempt to restore the cached state. However, our
interal states are now stale, since step 1 was done before step 2.
i.e. our cached state does not reflect the disabled nature of things.

This change resolves that by destroying all relevant states to reflect
the actual state during resume.
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This was missed in the previous fake sink change. The fake sink
allows us to enable a pipe with a disconnected display. We
shouldn't report it as connected.
Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Not used anywhere.
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Not used outside of that file.
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

To match kernel standards.  No intended functional change.
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

s/amdgpu_dm_find_first_crct_matching_connector/
amdgpu_dm_find_first_crtc_matching_connector/

And while here, make it static.
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Acked-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

in amdgpu_dm_atomic_commit_tail. Just use crtc instead.
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

undersacn -> underscan
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use dm_new_*_state and dm_old_*_state for their respective amdgpu_dm new
and old object states. Helps with readability, and enforces use of new
DRM api (choose either new, or old).
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use new_*_state and old_*_state for their respective new/old DRM object
states.
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

To conform to DRM's new API, we should not be accessing a DRM object's
internal state directly. Rather, the DRM for_each_old/new_* iterators,
and drm_atomic_get_old/new_* interface should be used.

This is an ongoing process. For now, update the DRM-facing atomic
functions, where the atomic state object is given.
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use the correct for_each_new/old_* iterators instead of for_each_*

The following functions were considered:

amdgpu_dm_find_first_crtc_matching_connector: use for_each_new
    - Old from_state_var flag was always choosing the new state

amdgpu_dm_display_resume: use for_each_new
    - drm_atomic_helper_duplicate_state is called during suspend to
      cache the state
    - It sets 'state' within the state triplet to 'new_state'

amdgpu_dm_commit_planes: use for_each_old
    - Called after the state was swapped (via atomic commit tail)

amdgpu_dm_atomic_commit: use for_each_new
    - Called before the state is swapped

amdgpu_dm_atomic_commit_tail: use for_each_old
    - Called after the state was swapped

dm_update_crtcs_state: use for_each_new
    - Called before the state is swapped (via atomic check)

amdgpu_dm_atomic_check: use for_each_new
    - Called before the state is swapped

v2: Split out typo fixes to a new patch.

v3: Say "functions considered" instead of "affected functions". The
    latter implies that changes are made to each.

[airlied: squashed with my hacks]
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This dereference acrtc after freeing it.

Found by the kfree cocci script.
Signed-off-by: NDave Airlie <airlied@redhat.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NDave Airlie <airlied@redhat.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NDave Airlie <airlied@redhat.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NDave Airlie <airlied@redhat.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Abstractions are frowned upon.

cocci script:
virtual context
virtual patch
virtual org
virtual report

@@
expression ptr;
@@

- dm_alloc(ptr)
+ kzalloc(ptr, GFP_KERNEL)

@@
expression ptr, size;
@@

- dm_realloc(ptr, size)
+ krealloc(ptr, size, GFP_KERNEL)

@@
expression ptr;
@@

- dm_free(ptr)
+ kfree(ptr)

v2: use GFP_KERNEL, not GFP_ATOMIC. add cocci script
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Log DC init but default log level to 0 (default for
amdgpu_dc_log) otherwise. Bug reporters can still make
DC more chatty by using the dc_log module param.
    amdgpu.dc_log = 1

v2: Only provide runtime option, no compile time config
Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Linux and Windows often desire different log levels.
Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The problem we're trying to fix is this (and similar):
1) X Desktop with single display
2) VT switch
3) Unplug display
4) VT switch back to X
5) re-plug same display

Before this we'd fail at step 4 when trying to create a dc_stream_state
because of a missing sink. This change will fake a sink in this case.

The same scenario applies to S3 resume.
Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Stop using amdgpu_connector and roll our own. There is no overlap
with amdgpu.
Signed-off-by: NHarry Wentland <harry.wentland@amd.com>
Reviewed-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

RV1 support only USB-C active DP-2-HDMI dongle. HPD short pulse is
generated only for DP signal.

When processing HPD short pulse, it must be DP active dongle. No need
for I2C-Over-AUX detection.

v2: Add description
Signed-off-by: NHersen Wu <hersenxs.wu@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Rename dc_resource methods from validate_ctx to state
Signed-off-by: NBhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Rename all the dc validate_context to dc_state and
dc current_context to current_state.
Signed-off-by: NJerry Zuo <Jerry.Zuo@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Starting with 4.12 kernel DRM provides page flip flags so we
don't need to have our own copy.
Signed-off-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Noticed while playing “Valley”, which was causing some 8MB of leakage per
second. kmemleak listed many entries looking like this:

    unreferenced object 0xffff8802c2951800 (size 1024):
      comm "Xorg", pid 2982, jiffies 4297410155 (age 392.787s)
      hex dump (first 32 bytes):
        00 50 f9 0c 04 88 ff ff 98 08 00 00 00 00 00 00  .P..............
        80 07 00 00 00 00 00 00 58 00 00 00 2c 00 00 00  ........X...,...
      backtrace:
        [<ffffffff810cd4c3>] create_object+0x13c/0x261
        [<ffffffff815abdc2>] kmemleak_alloc+0x20/0x3c
        [<ffffffff810cad1d>] slab_post_alloc_hook+0x42/0x52
        [<ffffffff810cb8e0>] kmem_cache_alloc+0x67/0x76
        [<ffffffff813bbb54>] dc_create_stream_for_sink+0x24/0x1cf
        [<ffffffff81373aaa>] create_stream_for_sink+0x6f/0x295
        [<ffffffff81373dc2>] dm_update_crtcs_state+0xa6/0x268
        [<ffffffff8137401e>] amdgpu_dm_atomic_check+0x9a/0x314
        [<ffffffff812ac3dd>] drm_atomic_check_only+0x17a/0x42d
        [<ffffffff812ac6a3>] drm_atomic_commit+0x13/0x4b
        [<ffffffff812ad1a5>] drm_atomic_connector_commit_dpms+0xcb/0xe8
        [<ffffffff812b1238>] drm_mode_obj_set_property_ioctl+0xe6/0x1e3
        [<ffffffff812b027b>] drm_mode_connector_property_set_ioctl+0x2b/0x2d
        [<ffffffff8129f427>] drm_ioctl_kernel+0x64/0x9d
        [<ffffffff8129f6a2>] drm_ioctl+0x230/0x316
        [<ffffffff812ca4d3>] amdgpu_drm_ioctl+0x4b/0x7d

v2: also handle break statements.
Signed-off-by: NDarren Salt <devspam@moreofthesa.me.uk>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Split into update crtcs and update plane functions.
Signed-off-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Introduce add/remove plane to/from context.
Make DC wrapper to use them in WIndows/Diags.
Use them in dc_update_surface_to_stream.
Call add/remove plane from Linux DM.

Remove dc_validation_set from dc_validate_global_state interface
and by this remove clean Linux DM from using it.
Signed-off-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NBhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Previously atomic_check assumed that dc_state is allocated and filled
in. This is not the case when we hit EDEADLK and have to backup and
retry.

We could change atomic_state_clear but that revealed some other
problems. For now let's create and construct dc_state in atomic_check.
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

For less repetition and easy debugging.
Signed-off-by: NLeo (Sunpeng) Li <sunpeng.li@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Until now new context would start as empty, then populated
with exsisting pipes + new. Now we start with duplication
of existing context and then add/delete from the context
pipes as needed.

This allows to do a per stream resource
population, start discarding dc_validation_set
and by this brings DC closer to to DRM.

v2: Add some fixes and rebase.
Signed-off-by: NAndrey Grodzovsky <Andrey.Grodzovsky@amd.com>
Reviewed-by: NTony Cheng <Tony.Cheng@amd.com>
Acked-by: NHarry Wentland <Harry.Wentland@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>