Same as gmc9, basically filter the fault, reroute or handle it.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Felix pointed out that we need this for Navi as well.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Looks like we can't enabled the IH1/IH2 feature for Vega20, make sure
retry faults are handled on a separate ring anyway.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Seems like we won't get the hardware IH1/2 rings on Vega20 working.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Add a soft IH ring implementation similar to the hardware IH1/2.

This can be used if the hardware delegation of interrupts to IH1/2
doesn't work for some reason.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Return early in case of a ratelimit and don't print leading zeros for
the address.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The address space is only 48bit, not 64bit. And the VMHUBs work with
sign extended addresses.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use generic DMA api instead of bus-specific API.
Signed-off-by: NNirmoy Das <nirmoy.das@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use generic DMA api instead of bus-specific API.
Signed-off-by: NNirmoy Das <nirmoy.das@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Update GC golden setting for navy_flounder.
Signed-off-by: NJiansong Chen <Jiansong.Chen@amd.com>
Reviewed-by: NTao Zhou <tao.zhou1@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Update golden setting for sienna_cichlid.
Signed-off-by: NLikun Gao <Likun.Gao@amd.com>
Reviewed-by: NHawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

In preparation to enable -Wimplicit-fallthrough for Clang, fix a couple
of warnings by explicitly adding a break statement instead of letting
the code fall through to the next case, and a fallthrough pseudo-keyword
as a replacement for a /* fall through */ comment,

Notice that Clang doesn't recognize /* fall through */ comments as
implicit fall-through markings.

Link: https://github.com/KSPP/linux/issues/115Signed-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

In preparation to enable -Wimplicit-fallthrough for Clang, fix multiple
warnings by explicitly adding multiple break statements instead of just
letting the code fall through to the next case.

Link: https://github.com/KSPP/linux/issues/115Signed-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

In preparation to enable -Wimplicit-fallthrough for Clang, fix multiple
warnings by explicitly adding multiple fallthrough pseudo-keyword macros,
as replacement for /* fall through */ comments.

Notice that Clang doesn't recognize /* fall through */ comments as
implicit fall-through markings.

Link: https://github.com/KSPP/linux/issues/115Signed-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

In preparation to enable -Wimplicit-fallthrough for Clang, fix multiple
warnings by explicitly adding multiple break statements instead of just
letting the code fall through to the next case.

Link: https://github.com/KSPP/linux/issues/115Signed-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The HDCP feature requires at least one connector attached to the device;
however, some GPUs do not have a physical output, making the HDCP
initialization irrelevant. This patch disables HDCP initialization when
the graphic card does not have output.
Acked-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NRodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

We only need to arbitrate VGA access on VGA compatible devices.
Reviewed-by: NFeifei Xu <Feifei.Xu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Fix check_patch.pl warning:
WARNING: Prefer kmalloc_array over kmalloc with multiply
+bps = kmalloc(align_space * sizeof((*data)->bps), GFP_KERNEL);
WARNING: Prefer kmalloc_array over kmalloc with multiply
+bps_bo = kmalloc(align_space * sizeof((*data)->bps_bo),
GFP_KERNEL);
kmalloc_array has multiply overflow check, which will be safer.
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NBernard Zhao <bernard@vivo.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Fix check_patch.pl warning:
kmalloc_array uses number as first arg, sizeof is generally wrong.
+fences = kmalloc_array(sizeof(void *), id_mgr->num_ids,
GFP_KERNEL);
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NBernard Zhao <bernard@vivo.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Support software ring share memory with vcn firmware.
Signed-off-by: NJames Zhu <James.Zhu@amd.com>
Reviewed-by: NLeo Liu <leo.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Add dec software ring vm functions to support.
Signed-off-by: NJames Zhu <James.Zhu@amd.com>
Reviewed-by: NLeo Liu <leo.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Add vcn software ring decode ring test and decode ib test.
Signed-off-by: NJames Zhu <James.Zhu@amd.com>
Reviewed-by: NLeo Liu <leo.liu@amd.com>
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Add macro, structure and function prototype to
support vcn dec software ring.
Signed-off-by: NJames Zhu <James.Zhu@amd.com>
Reviewed-by: NLeo Liu <leo.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

refactor dec message functions to add dec software ring support.
Signed-off-by: NJames Zhu <James.Zhu@amd.com>
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Default runtime logic not changed.
Provide an alternative runtime method. (set 1 to use BACO; 2 to use BAMACO)
When set reset_method to 4, it will use BACO or BAMACO for gpu reset,
according to runpm value.
Signed-off-by: NLikun Gao <Likun.Gao@amd.com>
Reviewed-by: NKenneth Feng <kenneth.feng@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Some features are still disabled after runtime pm resume. This can take
the hardware back.
Unlike other projects, this doesn't need pptable retransfer.
Signed-off-by: NLikun Gao <Likun.Gao@amd.com>
Reviewed-by: NKenneth Feng <kenneth.feng@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The UVD firmware is copied to cpu addr in uvd_resume, so it
should be used after that. This is to fix a bug introduced by
patch drm/amdgpu: fix SI UVD firmware validate resume fail.
Signed-off-by: NSonny Jiang <sonny.jiang@amd.com>
Reviewed-by: NLeo Liu <leo.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Using KIQ to map GFX queues instead of MMIO for gfx async ring,
add missing doorbell range setting.

v2: fix typo
Signed-off-by: NJinzhou.Su <Jinzhou.Su@amd.com>
Reviewed-by: NHuang Rui <ray.huang@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This has been confirmed that unload message is not needed from SIENNA_CICHLID in reset.
Otherwise it will cause the fw wrong state after reset and no response for any messages.
Signed-off-by: NKenneth Feng <kenneth.feng@amd.com>
Reviewed-by: NHawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

fix the null pointer issue when runtime pm is triggered.
Signed-off-by: NKenneth Feng <kenneth.feng@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

dpcs reg are missing for dcn302 link encoder regs list, so add them.
Just like dcn3
Signed-off-by: NBhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: NAurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The SI UVD firmware validate key is stored at the end of firmware,
which is changed during resume while playing video. So get the key
at sw_init and store it for fw validate using.
Signed-off-by: NSonny Jiang <sonny.jiang@amd.com>
Reviewed-by: NLeo Liu <leo.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Add device id for dimgrey_cavefish.
Signed-off-by: NTao Zhou <tao.zhou1@amd.com>
Reviewed-by: NJiansong Chen <Jiansong.Chen@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

add device ID for navy_flounder

v2: squash in updates
Signed-off-by: NTao Zhou <tao.zhou1@amd.com>
Reviewed-by: NHawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

[Why]
These comments are helpful in understanding which case each if
statement handles.

[How]
Add comments for state transitions (9 possible cases)
Signed-off-by: NBhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: NZhan Liu <zhan.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

[Why]
We call add display multiple times because DESIRED->ENABLED cannot
happen instantaneously. We can't compare the new_state/old_state to
avoid this because on unplug we set the state to DESIRED and on hotplug
the state will still be DESIRED.

[How]
Add a flag to dm_connector_state to keep track of when to enable or
disable HDCP
Signed-off-by: NBhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: NZhan Liu <zhan.liu@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

There is no need to reset DPM for PPTable uploading on
dimgrey_cavefish and PMFW can handle it, same as navy_flounder.
Signed-off-by: NTao Zhou <tao.zhou1@amd.com>
Reviewed-by: NJiansong Chen <Jiansong.Chen@amd.com>
Acked-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Allocate smu_power->power_context to size of smu_11_0_power_context instead of
smu_11_0_dpm_context.
Signed-off-by: NRyan Taylor <Ryan.Taylor@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Reviewed-by: NEvan Quan <evan.quan@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

We recently improved our display atomic commit and tail sequence to
avoid some issues related to concurrency. One of the major changes
consisted of moving the interrupt disable and the stream release from
our atomic commit to our atomic tail (commit 6d90a208
("drm/amd/display: Move disable interrupt into commit tail")) .
However, the new code introduced inside our commit tail function was
inserted right after the function
drm_atomic_helper_update_legacy_modeset_state(), which has routines for
updating internal data structs related to timestamps. As a result, in
certain conditions, the display module can reach a situation where we
update our constants and, after that, clean it. This situation generates
the following warning:

 amdgpu 0000:03:00.0: drm_WARN_ON_ONCE(drm_drv_uses_atomic_modeset(dev))
 WARNING: CPU: 6 PID: 1269 at drivers/gpu/drm/drm_vblank.c:722
 drm_crtc_vblank_helper_get_vblank_timestamp_internal+0x32b/0x340 [drm]
 ...
 RIP:
 0010:drm_crtc_vblank_helper_get_vblank_timestamp_internal+0x32b/0x340
 [drm]
 ...
 Call Trace:
  ? dc_stream_get_vblank_counter+0x57/0x60 [amdgpu]
  drm_crtc_vblank_helper_get_vblank_timestamp+0x1c/0x20 [drm]
  drm_get_last_vbltimestamp+0xad/0xc0 [drm]
  drm_reset_vblank_timestamp+0x63/0xd0 [drm]
  drm_crtc_vblank_on+0x85/0x150 [drm]
  amdgpu_dm_atomic_commit_tail+0xaf1/0x2330 [amdgpu]
  commit_tail+0x99/0x130 [drm_kms_helper]
  drm_atomic_helper_commit+0x123/0x150 [drm_kms_helper]
  amdgpu_dm_atomic_commit+0x11/0x20 [amdgpu]
  drm_atomic_commit+0x4a/0x50 [drm]
  drm_atomic_helper_set_config+0x7c/0xc0 [drm_kms_helper]
  drm_mode_setcrtc+0x20b/0x7e0 [drm]
  ? tomoyo_path_number_perm+0x6f/0x200
  ? drm_mode_getcrtc+0x190/0x190 [drm]
  drm_ioctl_kernel+0xae/0xf0 [drm]
  drm_ioctl+0x245/0x400 [drm]
  ? drm_mode_getcrtc+0x190/0x190 [drm]
  amdgpu_drm_ioctl+0x4e/0x80 [amdgpu]
  __x64_sys_ioctl+0x91/0xc0
  do_syscall_64+0x38/0x90
  entry_SYSCALL_64_after_hwframe+0x44/0xa9
 ...

For fixing this issue we rely upon a refactor introduced on
drm_atomic_helper_update_legacy_modeset_state ("Remove the timestamping
constant update from drm_atomic_helper_update_legacy_modeset_state()")
which decouples constant values update from
drm_atomic_helper_update_legacy_modeset_state to a new helper.
Basically, this commit uses this new helper and place it right after our
release module to avoid a situation where our CRTC struct gets wrong
values.

Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1373Reviewed-by: NHarry Wentland <harry.wentland@amd.com>
Signed-off-by: NRodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

There are two spelling mistakes in dev_warn messages. Fix these.
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>