drm/amdgpu: Update SRIOV Exchange Headers for RAS Telemetry Support

The SRIOV PF/VF Data exchange is extended by 64KB for VF RAS Telemetry data.
Add Host RAS Telemetry enable capabilities bitfields.
Add a new VF msg REQ_RAS_ERROR_COUNT, the host response data will be populated
in the RAS Telemetry region.

Signed-off-by: Victor Skvortsov <victor.skvortsov@amd.com>
Reviewed-by: Zhigang Luo <zhigang.luo@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.309

This version brings along the following:

- DML2 fixes
- DP fixes
- DPMS fix
- HPD fixes
- Misc cleanup
- ODM fix
- Replay fix
- SPL fix

Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Adjust VSDB parser for replay feature

At some point, the IEEE ID identification for the replay check in the
AMD EDID was added. However, this check causes the following
out-of-bounds issues when using KASAN:

[   27.804016] BUG: KASAN: slab-out-of-bounds in amdgpu_dm_update_freesync_caps+0xefa/0x17a0 [amdgpu]
[   27.804788] Read of size 1 at addr ffff8881647fdb00 by task systemd-udevd/383

...

[   27.821207] Memory state around the buggy address:
[   27.821215]  ffff8881647fda00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[   27.821224]  ffff8881647fda80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[   27.821234] >ffff8881647fdb00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[   27.821243]                    ^
[   27.821250]  ffff8881647fdb80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[   27.821259]  ffff8881647fdc00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[   27.821268] ==================================================================

This is caused because the ID extraction happens outside of the range of
the edid lenght. This commit addresses this issue by considering the
amd_vsdb_block size.

Cc: ChiaHsuan Chung <chiahsuan.chung@amd.com>
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove unused code

This commit removes a legacy debug_defaults_diags struct.

Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Require minimum VBlank size for stutter optimization

If the nominal VBlank is too small, optimizing for stutter can cause
the prefetch bandwidth to increase drasticaly, resulting in higher
clock and power requirements. Only optimize if it is >3x the stutter
latency.

Reviewed-by: Austin Zheng <austin.zheng@amd.com>
Signed-off-by: Dillon Varone <dillon.varone@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Handle dml allocation failure to avoid crash

[Why]
In the case where a dml allocation fails for any reason, the
current state's dml contexts would no longer be valid. Then
subsequent calls dc_state_copy_internal would shallow copy
invalid memory and if the new state was released, a double
free would occur.

[How]
Reset dml pointers in new_state to NULL and avoid invalid
pointer

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Signed-off-by: Ryan Seto <ryanseto@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Use region6 size in fw_meta_info

[Why]
If driver allocated region6 size is not same as the size in firmware,
dmcub won't enable region6.

[How]
Use region6 size in dmcub_fw_meta instead of a constant value.

Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: JinZe Xu <jinze.xu@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Update SPL Taps Required For Integer Scaling

Number of taps is incorrectly being set when integer scaling is enabled.
Taps required when src_rect != dst_rect previously not considered.
Perform the calculations when integer scaling is enabled.
Set taps to 1 if the scaling ratio is 1:1.

Reviewed-by: Samson Tam <samson.tam@amd.com>
Signed-off-by: Austin Zheng <Austin.Zheng@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: disabling p-state checks for DCN31 and DCN314

[Why]
IGT displays Dmesg warnings which are likely false

[How]
Disabling p-state checks leading to this warning for DCN31 and DCN314

Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Emily Nie <Emily.Nie@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: always blank stream before disable crtc

Garbage will show due to dig is on. So blank stream needed.

Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Fudongwang <Fudong.Wang@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Read DP tunneling support only for DPIA endpoints

Unconditionally reading DP tunneling support results in extraneous
errors messages on certain devices. Fix this by guarding the DPCD read
for DP tunneling support for USB4 DPIA endpoints.

Reviewed-by: Meenakshikumar Somasundaram <meenakshikumar.somasundaram@amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Adding flag for forced MST blocked discovery

[Why]
Need a flag to force MST blocked discovery for certain branch devices.

[How]
Added a flag to force MST blocked discovery in struct dc_panel_patch.

Reviewed-by: PeiChen Huang <peichen.huang@amd.com>
Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Signed-off-by: Meenakshikumar Somasundaram <meenakshikumar.somasundaram@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix Panel Replay not update screen correctly

[Why]
In certain use case such as KDE login screen, there will be no atomic
commit while do the frame update.
If the Panel Replay enabled, it will cause the screen not updated and
looks like system hang.

[How]
Delay few atomic commits before enabled the Panel Replay just like PSR.

Fixes: be64336307a6c ("drm/amd/display: Re-enable panel replay feature")
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3686
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3682
Tested-By: Corey Hickey <bugfood-c@fatooh.org>
Tested-By: James Courtier-Dutton <james.dutton@gmail.com>
Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Tom Chung <chiahsuan.chung@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Change some variable name of psr

Panel Replay feature may also use the same variable with PSR.
Change the variable name and make it not specify for PSR.

Reviewed-by: Leo Li <sunpeng.li@amd.com>
Signed-off-by: Tom Chung <chiahsuan.chung@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Change parameters to fix certain compiler errors

[Why]
String literals must be assigned to const char pointers.

[How]
By adding const keyword to fix compilation errors.

Reviewed-by: Lohita Mudimela <lohita.mudimela@amd.com>
Signed-off-by: Revalla Hari Krishna <Harikrishna.Revalla@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Refactor HPD IRQ error checking flow

[Why]
HPD error status does not cover Replay desync error status
while executing autotests and CTS tests.

[How]
Refactor the checking flow, reporting the HPD error based on
different eDP feature.

Reviewed-by: Robin Chen <robin.chen@amd.com>
Signed-off-by: Leon Huang <Leon.Huang1@amd.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/gfx11: Enable cleaner shader for GFX11.0.0/11.0.2 GPUs

Enable the cleaner shader for GFX11.0.0/11.0.2 GPUs to provide data
isolation between GPU workloads. The cleaner shader is responsible for
clearing the Local Data Store (LDS), Vector General Purpose Registers
(VGPRs), and Scalar General Purpose Registers (SGPRs), which helps
prevent data leakage and ensures accurate computation results.

This update extends cleaner shader support to GFX11.0.0/11.0.2 GPUs,
previously available for GFX11.0.3. It enhances security by clearing GPU
memory between processes and maintains a consistent GPU state across KGD
and KFD workloads.

Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add documentation for enforce isolation feature

This feature enables process isolation on the graphics engine by
serializing access to it and adding a cleaner shader which clears LDS
(Local Data Store) and GPRs (General Purpose Registers) between jobs.

Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Suggested-by: Asad Kamal <asad.kamal@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Fix wrong usage of INIT_WORK()

In kfd_procfs_show(), the sdma_activity_work_handler is a local variable
and the sdma_activity_work_handler.sdma_activity_work should initialize
with INIT_WORK_ONSTACK() instead of INIT_WORK().

Fixes: 32cb59f31362 ("drm/amdkfd: Track SDMA utilization per process")
Signed-off-by: Yuan Can <yuancan@huawei.com>
Signed-off-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix check in gmc_v9_0_get_vm_pte()

The coherency flags can only be determined when the BO is locked and that
in turn is only guaranteed when the mapping is validated.

Fix the check, move the resource check into the function and add an assert
that the BO is locked.

Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: d1a372af1c3d ("drm/amdgpu: Set MTYPE in PTE based on BO flags")
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: print pp_dpm_mclk in ascending order on SMU v14.0.0

Currently, the pp_dpm_mclk values are reported in descending order
on SMU IP v14.0.0/1/4. Adjust to ascending order for consistency
with other clock interfaces.

Signed-off-by: Tim Huang <tim.huang@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Inform if PCIe based P2P links are not available

Raise an info message in kernel log if PCIe root complex
determines that a AMD GPU device D<i> cannot have P2P
communication with another AMD GPU device D<j>

Signed-off-by: Ramesh Errabolu <Ramesh.Errabolu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix video caps for H264 and HEVC encode maximum size

H264 supports 4096x4096 starting from Polaris.
HEVC also supports 4096x4096, with VCN 3 and newer 8192x4352
is supported.

Signed-off-by: David Rosca <david.rosca@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add sysfs interface for jpeg reset mask

Add the sysfs interface for jpeg:
jpeg_reset_mask

The interface is read-only and show the resets supported by the IP.
For example, full adapter reset (mode1/mode2/BACO/etc),
soft reset, queue reset, and pipe reset.

V2: the sysfs node returns a text string instead of some flags (Christian)
v3: add a generic helper which takes the ring as parameter
    and print the strings in the order they are applied (Christian)

    check amdgpu_gpu_recovery  before creating sysfs file itself,
    and initialize supported_reset_types in IP version files (Lijo)

Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Tim Huang <tim.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add sysfs interface for vpe reset mask

Add the sysfs interface for vpe:
    vpe_reset_mask

The interface is read-only and show the resets supported by the IP.
For example, full adapter reset (mode1/mode2/BACO/etc),
soft reset, queue reset, and pipe reset.

V2: the sysfs node returns a text string instead of some flags (Christian)
v3: add a generic helper which takes the ring as parameter
    and print the strings in the order they are applied (Christian)

    check amdgpu_gpu_recovery  before creating sysfs file itself,
    and initialize supported_reset_types in IP version files (Lijo)

Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Tim Huang <tim.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add sysfs interface for sdma reset mask

Add the sysfs interface for sdma:
sdma_reset_mask

The interface is read-only and show the resets supported by the IP.
For example, full adapter reset (mode1/mode2/BACO/etc),
soft reset, queue reset, and pipe reset.

V2: the sysfs node returns a text string instead of some flags (Christian)
v3: add a generic helper which takes the ring as parameter
   and print the strings in the order they are applied (Christian)

   check amdgpu_gpu_recovery  before creating sysfs file itself,
   and initialize supported_reset_types in IP version files (Lijo)

Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Tim Huang <tim.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Normalize reg offsets on VCN v4.0.3

Remote access to external AIDs isn't possible with VCN RRMT disabled
and it is disabled on SoCs with GC 9.4.4, so use only local offsets.

Signed-off-by: Sathishkumar S <sathishkumar.sundararaju@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Avoid kcq disable during reset

Reset sequence indicates that hardware already ran into a bad state.
Avoid sending unmap queue request to reset KCQ. This will also cover RAS
error scenarios which need a reset to recover, hence remove the check.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix map/unmap queue logic

In current logic, it calls ring_alloc followed by a ring_test. ring_test
in turn will call another ring_alloc. This is illegal usage as a
ring_alloc is expected to be closed properly with a ring_commit. Change
to commit the map/unmap queue packet first followed by a ring_test. Add a
comment about the usage of ring_test.

Also, reorder the current pre-condition checks of job hang or kiq ring
scheduler not ready. Without them being met, it is not useful to attempt
ring or memory allocations.

Fixes tag refers to the original patch which introduced this issue which
then got carried over into newer code.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fixes: 6c10b5cc4eaa ("drm/amdgpu: Remove duplicate code in gfx_v8_0.c")

drm/amdgpu: fix ACA bank count boundary check error

fix ACA bank count boundary check error.

Fixes: f5e4cc8461c4 ("drm/amdgpu: implement RAS ACA driver framework")
Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add sysfs interface for gc reset mask

Add two sysfs interfaces for gfx and compute:
gfx_reset_mask
compute_reset_mask

These interfaces are read-only and show the resets supported by the IP.
For example, full adapter reset (mode1/mode2/BACO/etc),
soft reset, queue reset, and pipe reset.

V2: the sysfs node returns a text string instead of some flags (Christian)
v3: add a generic helper which takes the ring as parameter
    and print the strings in the order they are applied (Christian)

    check amdgpu_gpu_recovery  before creating sysfs file itself,
    and initialize supported_reset_types in IP version files (Lijo)
v4: Fixing uninitialized variables (Tim)

Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Tim Huang <tim.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix return random value when multiple threads read registers via mes.

The currect code use the address "adev->mes.read_val_ptr" to
store the value read from register via mes.
So when multiple threads read register,
multiple threads have to share the one address,
and overwrite the value each other.

Assign an address by "amdgpu_device_wb_get" to store register value.
each thread will has an address to store register value.

Signed-off-by: chongli2 <chongli2@amd.com>
Reviewed-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: remove gfx 12 trap handler page size cap

GFX 12 does not require a page size cap for the trap handler because
it does not require a CWSR work around like GFX 11 did.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: David Belanger <david.belanger@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add supported NPS modes node

Add sysfs node to show supported NPS mode for the
partition configuration selected using xcp_config

v2: Hide node if dynamic nps switch not supported

v3: Fix removal of files in case of error

Signed-off-by: Asad Kamal <asad.kamal@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Merge tag 'drm-etnaviv-next-2024-11-07' of https://git.pengutronix.de/git/lst/linux into drm-next

- improve handling of DMA address limited systems
- improve GPU hangcheck
- fix address space collision on >= 4K CPU pages
- flush all known writeback caches before memory release
- various code cleanups

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Lucas Stach <l.stach@pengutronix.de>
Link: https://patchwork.freedesktop.org/patch/msgid/c84075a0257e7bee222d008fa3118117422d664e.camel@pengutronix.de

Merge tag 'amd-drm-next-6.13-2024-11-06' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.13-2024-11-06:

amdgpu:
- Misc cleanups
- OLED fixes
- DCN 4.x fixes
- DCN 3.5 fixes
- 8K fixes
- IPS fixes
- DSC fixes
- S3 fix
- KASAN fix
- SMU13 fixes
- fdinfo fixes
- USB-C fixes
- ACPI fix
- Fix dummy page overlapping mappings
- Fix workload profile handling
- Add user control for zero RPM on SMU13
- Cleaner shader updates
- Stop syncing PRT map operations
- Debugfs permissions fixes
- Debugfs bounds check fix
- RAS cleanups
- Enforce isolation updates

amdkfd:
- Add topology cap flag for per queue reset
- Add an interface to query whether KFD queues are present
- Use dynamic allocation for get_cu_occupancy

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241106163904.189108-1-alexander.deucher@amd.com
Signed-off-by: Dave Airlie <airlied@redhat.com>

Merge tag 'exynos-drm-next-for-v6.13-v2' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-next

New feature
- Add a compatible string, "samsung,exynos7870-decon", for supporting
  Exynos7870 SoC Decon device

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Inki Dae <inki.dae@samsung.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241106010943.796273-1-inki.dae@samsung.com

dt-bindings: display: samsung,exynos7-decon: add exynos7870 compatible

Add the compatible string of Exynos7870 to the existing list.

Signed-off-by: Kaustabh Chakraborty <kauschluss@disroot.org>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

Merge tag 'drm-msm-next-2024-11-04' of https://gitlab.freedesktop.org/drm/msm into drm-next

Late updates for v6.13

MDSS:
- cleanup UBWC registers handling

DP:
- Mass-rename the symbols

DPU:
- SSPP handling cleanup
- Move kerneldoc comments from headers to source files
- Misc small fixes

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rob Clark <robdclark@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/CAF6AEGuGL6k3CKXZ0Qv-FTQ589+_PWNtid6i7MmVJLopBm2sYg@mail.gmail.com

Merge tag 'drm-intel-next-2024-11-04' of https://gitlab.freedesktop.org/drm/i915/kernel into drm-next

drm/i915 feature pull #2 for v6.13:

Features and functionality:

- Pantherlake (PTL) Xe3 LPD display enabling for xe driver (Clint, Suraj,
  Dnyaneshwar, Matt, Gustavo, Radhakrishna, Chaitanya, Haridhar, Juha-Pekka, Ravi)
- Enable dbuf overlap detection on Lunarlake and later (Stanislav, Vinod)
- Allow fastset for HDR infoframe changes (Chaitanya)
- Write DP source OUI also for non-eDP sinks (Imre)

Refactoring and cleanups:
- Independent platform identification for display (Jani)
- Display tracepoint fixes and cleanups (Gustavo)
- Share PCI ID headers between i915 and xe drivers (Jani)
- Use x100 version for full version and release checks (Jani)
- Conversions to struct intel_display (Jani, Ville)
- Reuse DP DPCD and AUX macros in gvt instead of duplication (Jani)
- Use string choice helpers (R Sundar, Sai Teja)
- Remove unused underrun detection irq code (Sai Teja)
- Color management debug improvements and other cleanups (Ville)
- Refactor panel fitter code to a separate file (Ville)
- Use try_cmpxchg() instead of open-coding (Uros Bizjak)

Fixes:
- PSR and Panel Replay fixes and workarounds (Jouni)
- Fix panel power during connector detection (Imre)
- Fix connector detection and modeset races (Imre)
- Fix C20 PHY TX MISC configuration (Gustavo)
- Improve panel fitter validity checks (Ville)
- Fix eDP short HPD interrupt handling while runtime suspended (Imre)
- Propagate DP MST DSC BW overhead/slice calculation errors (Imre)
- Stop hotplug polling for eDP connectors (Imre)
- Workaround panels reporting bad link status after PSR enable (Jouni)
- Panel Replay VRR VSC SDP related workaround and refactor (Animesh, Mitul)
- Fix memory leak on eDP init error path (Shuicheng)
- Fix GVT KVMGT Kconfig dependencies (Arnd Bergmann)
- Fix irq function documentation build warning (Rodrigo)
- Add platform check to power management fuse bit read (Clint)
- Revert kstrdup_const() and kfree_const() usage for clarity (Christophe JAILLET)
- Workaround horizontal odd panning issues in display versions 20 and 30 (Nemesa)
- Fix xe drive HDCP GSC firmware check (Suraj)

Merges:
- Backmerge drm-next to get some KVM changes (Rodrigo)
- Fix a build failure originating from previous backmerge (Jani)

Signed-off-by: Dave Airlie <airlied@redhat.com>
# Conflicts:
# drivers/gpu/drm/i915/display/intel_dp_mst.c
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/87h68ni0wd.fsf@intel.com

Merge tag 'mediatek-drm-next-6.13' of https://git.kernel.org/pub/scm/linux/kernel/git/chunkuang.hu/linux into drm-next

Mediatek DRM Next for Linux 6.13

1. Add support for OF graphs
2. Fix child node refcount handling and use scoped

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20241104124103.8041-1-chunkuang.hu@kernel.org

drm/amdgpu: add missing size check in amdgpu_debugfs_gprwave_read()

Avoid a possible buffer overflow if size is larger than 4K.

Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Adjust debugfs eviction and IB access permissions

Users should not be able to run these.

Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Adjust debugfs register access permissions

Regular users shouldn't have read access.

Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: stop syncing PRT map operations

Requested by both Bas and Friedrich. Mapping PTEs as PRT doesn't need to
sync for anything.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Friedrich Vock <friedrich.vock@gmx.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: set the right AMDGPU sg segment limitation

The driver needs to set the correct max_segment_size;
otherwise debug_dma_map_sg() will complain about the
over-mapping of the AMDGPU sg length as following:

WARNING: CPU: 6 PID: 1964 at kernel/dma/debug.c:1178 debug_dma_map_sg+0x2dc/0x370
[  364.049444] Modules linked in: veth amdgpu(OE) amdxcp drm_exec gpu_sched drm_buddy drm_ttm_helper ttm(OE) drm_suballoc_helper drm_display_helper drm_kms_helper i2c_algo_bit rpcsec_gss_krb5 auth_rpcgss nfsv4 nfs lockd grace netfs xt_conntrack xt_MASQUERADE nf_conntrack_netlink xfrm_user xfrm_algo iptable_nat xt_addrtype iptable_filter br_netfilter nvme_fabrics overlay nfnetlink_cttimeout nfnetlink openvswitch nsh nf_conncount nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 libcrc32c bridge stp llc amd_atl intel_rapl_msr intel_rapl_common sunrpc sch_fq_codel snd_hda_codec_realtek snd_hda_codec_generic snd_hda_scodec_component snd_hda_codec_hdmi snd_hda_intel snd_intel_dspcfg edac_mce_amd binfmt_misc snd_hda_codec snd_pci_acp6x snd_hda_core snd_acp_config snd_hwdep snd_soc_acpi kvm_amd snd_pcm kvm snd_seq_midi snd_seq_midi_event crct10dif_pclmul ghash_clmulni_intel sha512_ssse3 snd_rawmidi sha256_ssse3 sha1_ssse3 aesni_intel snd_seq nls_iso8859_1 crypto_simd snd_seq_device cryptd snd_timer rapl input_leds snd
[  364.049532]  ipmi_devintf wmi_bmof ccp serio_raw k10temp sp5100_tco soundcore ipmi_msghandler cm32181 industrialio mac_hid msr parport_pc ppdev lp parport drm efi_pstore ip_tables x_tables pci_stub crc32_pclmul nvme ahci libahci i2c_piix4 r8169 nvme_core i2c_designware_pci realtek i2c_ccgx_ucsi video wmi hid_generic cdc_ether usbnet usbhid hid r8152 mii
[  364.049576] CPU: 6 PID: 1964 Comm: rocminfo Tainted: G           OE      6.10.0-custom #492
[  364.049579] Hardware name: AMD Majolica-RN/Majolica-RN, BIOS RMJ1009A 06/13/2021
[  364.049582] RIP: 0010:debug_dma_map_sg+0x2dc/0x370
[  364.049585] Code: 89 4d b8 e8 36 b1 86 00 8b 4d b8 48 8b 55 b0 44 8b 45 a8 4c 8b 4d a0 48 89 c6 48 c7 c7 00 4b 74 bc 4c 89 4d b8 e8 b4 73 f3 ff <0f> 0b 4c 8b 4d b8 8b 15 c8 2c b8 01 85 d2 0f 85 ee fd ff ff 8b 05
[  364.049588] RSP: 0018:ffff9ca600b57ac0 EFLAGS: 00010286
[  364.049590] RAX: 0000000000000000 RBX: ffff88b7c132b0c8 RCX: 0000000000000027
[  364.049592] RDX: ffff88bb0f521688 RSI: 0000000000000001 RDI: ffff88bb0f521680
[  364.049594] RBP: ffff9ca600b57b20 R08: 000000000000006f R09: ffff9ca600b57930
[  364.049596] R10: ffff9ca600b57928 R11: ffffffffbcb46328 R12: 0000000000000000
[  364.049597] R13: 0000000000000001 R14: ffff88b7c19c0700 R15: ffff88b7c9059800
[  364.049599] FS:  00007fb2d3516e80(0000) GS:ffff88bb0f500000(0000) knlGS:0000000000000000
[  364.049601] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  364.049603] CR2: 000055610bd03598 CR3: 00000001049f6000 CR4: 0000000000350ef0
[  364.049605] Call Trace:
[  364.049607]  <TASK>
[  364.049609]  ? show_regs+0x6d/0x80
[  364.049614]  ? __warn+0x8c/0x140
[  364.049618]  ? debug_dma_map_sg+0x2dc/0x370
[  364.049621]  ? report_bug+0x193/0x1a0
[  364.049627]  ? handle_bug+0x46/0x80
[  364.049631]  ? exc_invalid_op+0x1d/0x80
[  364.049635]  ? asm_exc_invalid_op+0x1f/0x30
[  364.049642]  ? debug_dma_map_sg+0x2dc/0x370
[  364.049647]  __dma_map_sg_attrs+0x90/0xe0
[  364.049651]  dma_map_sgtable+0x25/0x40
[  364.049654]  amdgpu_bo_move+0x59a/0x850 [amdgpu]
[  364.049935]  ? srso_return_thunk+0x5/0x5f
[  364.049939]  ? amdgpu_ttm_tt_populate+0x5d/0xc0 [amdgpu]
[  364.050095]  ttm_bo_handle_move_mem+0xc3/0x180 [ttm]
[  364.050103]  ttm_bo_validate+0xc1/0x160 [ttm]
[  364.050108]  ? amdgpu_ttm_tt_get_user_pages+0xe5/0x1b0 [amdgpu]
[  364.050263]  amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu+0xa12/0xc90 [amdgpu]
[  364.050473]  kfd_ioctl_alloc_memory_of_gpu+0x16b/0x3b0 [amdgpu]
[  364.050680]  kfd_ioctl+0x3c2/0x530 [amdgpu]
[  364.050866]  ? __pfx_kfd_ioctl_alloc_memory_of_gpu+0x10/0x10 [amdgpu]
[  364.051054]  ? srso_return_thunk+0x5/0x5f
[  364.051057]  ? tomoyo_file_ioctl+0x20/0x30
[  364.051063]  __x64_sys_ioctl+0x9c/0xd0
[  364.051068]  x64_sys_call+0x1219/0x20d0
[  364.051073]  do_syscall_64+0x51/0x120
[  364.051077]  entry_SYSCALL_64_after_hwframe+0x76/0x7e
[  364.051081] RIP: 0033:0x7fb2d2f1a94f

Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix DPX valid mode check on GC 9.4.3

For DPX mode, the number of memory partitions supported should be less
than or equal to 2.

Fixes: 1589c82a1085 ("drm/amdgpu: Check memory ranges for valid xcp mode")
Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/gfx11: Add cleaner shader for GFX11.0.3

This commit adds the cleaner shader microcode for GFX11.0.3 GPUs. The
cleaner shader is a piece of GPU code that is used to clear or
initialize certain GPU resources, such as Local Data Share (LDS), Vector
General Purpose Registers (VGPRs), and Scalar General Purpose Registers
(SGPRs).

Clearing these resources is important for ensuring data isolation
between different workloads running on the GPU. Without the cleaner
shader, residual data from a previous workload could potentially be
accessed by a subsequent workload, leading to data leaks and incorrect
computation results.

The cleaner shader microcode is represented as an array of 32-bit words
(`gfx_11_0_3_cleaner_shader_hex`). This array is the binary
representation of the cleaner shader code, which is written in a
low-level GPU instruction set.

When the cleaner shader feature is enabled, the AMDGPU driver loads this
array into a specific location in the GPU memory. The GPU then reads
this memory location to fetch and execute the cleaner shader
instructions.

The cleaner shader is executed automatically by the GPU at the end of
each workload, before the next workload starts. This ensures that all
GPU resources are in a clean state before the start of each workload.

This addition is part of the cleaner shader feature implementation. The
cleaner shader feature helps resource utilization by cleaning up GPU
resources after they are used. It also enhances security and reliability
by preventing data leaks between workloads.

Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: add zero RPM stop temperature OD setting support for SMU13

Together with the feature to enable or disable zero RPM in the last
commit, it also makes sense to expose the OD setting determining under
which temperature the fan should stop if zero RPM is enabled.

Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Wolfgang Müller <wolf@oriole.systems>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/mes: fetch fw version from firmware header

We need this prior to the firmware being loaded so fetch
from the header.

v2: fetch directly from the firmware
v3: store both fw versions

Reviewed-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: add zero RPM OD setting support for SMU13

Whilst we have support for setting fan curves there is no support for
disabling the zero RPM feature. Since the relevant bits are already
present in the OverDriveTable, hook them up to a sysctl setting so users
can influence this behaviour.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3489
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Wolfgang Müller <wolf@oriole.systems>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Merge tag 'exynos-drm-next-for-v6.13' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-next

New feature
- Add Decon driver support for Exynos7870 SoC
  . This patch adds driver data and support for Exynos7870 SoC
    in the Exynos7 Decon driver

Bug fixups for exynos7_drm_decon.c module
- Properly clear channels during bind
  . This patch implements shadow protection/unprotection to clear
    DECON channels properly, preventing kernel panic
- Fix ideal_clk by converting it to HZ
  . This patch corrects the clkdiv values by converting ideal_clk to Hz
    for consistency
- Fix uninitialized crtc reference in functions
  . This patch modifies functions to accept a pointer to
    the decon_context struct to avoid uninitialized references

Cleanups
- Remove unused prototype for crtc
  . This patch removes unused prototypes
      exynos_drm_crtc_wait_pending_update
      exynos_drm_crtc_finish_update
- And just typo fixup

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Inki Dae <inki.dae@samsung.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241104031341.36549-1-inki.dae@samsung.com

Merge tag 'drm-xe-next-2024-10-31' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next

UAPI Changes:
- Define and parse OA sync properties (Ashutosh)

Driver Changes:
- Add caller info to xe_gt_reset_async (Nirmoy)
- A large forcewake rework / cleanup (Himal)
- A g2h response timeout fix (Badal)
- A PTL workaround (Vinay)
- Handle unreliable MMIO reads during forcewake (Shuicheng)
- Ufence user-space access fixes (Nirmoy)
- Annotate flexible arrays (Matthew Brost)
- Enable GuC lite restore (Fei)
- Prevent GuC register capture on VF (Zhanjun)
- Show VFs VRAM / LMEM provisioning summary over debugfs (Michal)
- Parallel queues fix on GT reset (Nirmoy)
- Move reference grabbing to a job's dma-fence (Matt Brost)
- Mark a number of local workqueues WQ_MEM_RECLAIM (Matt Brost)
- OA synchronization support (Ashutosh)
- Capture all available bits of GuC timestamp to GuC log (John)
- Increase readability of guc_info debugfs (John)
- Add a mmio barrier before GGTT invalidate (Matt Brost)
- Don't short-circuit TDR on jobs not started (Matt Brost)

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/ZyNvA_vZZYR-1eWE@fedora

Merge tag 'drm-msm-next-2024-10-28' of https://gitlab.freedesktop.org/drm/msm into drm-next

Updates for v6.13

Core:
- Switch to aperture_remove_all_conflicting_devices()
- Simplify msm_disp_state_dump_regs()

DPU:
- Add SA8775P support
- Add (disabled by default) MSM8917, MSM8937, MSM8953 and MSM8996
  support
- Enable support for larger framebuffers (required for X.Org working
  with several outputs)
- Dropped LM_3, LM_4 (MSM8998, SDM845)
- Fixed DSPP_3 routing on SDM845

DP:
- Add SA8775P support

HDMI:
- Mark two arrays as const in MSM8998 HDMI PHY driver

GPU:
- a7xx preemption support
- Adreno A663 support
- Typos fixes, etc
- Fix excessive stack usage in a6xx GMU

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rob Clark <robdclark@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/CAF6AEGt7k8zDHsg2Uzx9apzyQMut8XdLXMQSRNn7WArdPUV5Qw@mail.gmail.com

drm/amd/pm: correct the workload setting

Correct the workload setting in order not to mix the setting
with the end user. Update the workload mask accordingly.

v2: changes as below:
1. the end user can not erase the workload from driver except default workload.
2. always shows the real highest priority workoad to the end user.
3. the real workload mask is combined with driver workload mask and end user workload mask.

v3: apply this to the other ASICs as well.
v4: simplify the code
v5: refine the code based on the review comments.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add compatible NPS mode info

Populate the compatible NPS modes also for providing partition
configuration details through sysfs.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Asad Kamal <asad.kamal@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Skip IP coredump for RAS errors

For RAS errors, source of error is known. Skip the core dump of IP
states.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Group gfx sysfs functions

Make amdgpu_gfx_sysfs_init/fini functions as common entry points for all
gfx related sysfs nodes.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add nps_mode in RAS init_flag

Add nps_mode in RAS init_flag.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Tao Zhou <tao.zhou1@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add amdgpu_sdma_sched_mask debugfs

Userspace wants to run jobs on a specific sdma ring for verification purposes.
This debugfs entry helps to disable or enable submitting jobs to a specific ring.
This entry is populated only if there are at least two or more cores in the sdma ip.

Signed-off-by: Jesse Zhang <jesse.zhang@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Tim Huang <tim.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add amdgpu_gfx_sched_mask and amdgpu_compute_sched_mask debugfs

compute/gfx may have multiple rings on some hardware.
In some cases, userspace wants to run jobs on a specific ring for validation purposes.
This debugfs entry helps to disable or enable submitting jobs to a specific ring.
This entry is populated only if there are at least two or more cores in the gfx/compute ip.

Signed-off-by: Jesse Zhang <jesse.zhang@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Tim Huang <tim.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix dummy_read_page overlapping mappings

Use the dma_map_page_attrs() with DMA_ATTR_SKIP_CPU_SYNC
attribute setting to handle the dummy page overlapping
mappings.

Signed-off-by: Prike Liang <Prike.Liang@amd.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: skip amdgpu_device_cache_pci_state under sriov

Under sriov, host driver will save and restore vf pci cfg space during
reset. And during device init, under sriov, pci_restore_state happens after
fullaccess released, and it can have race condition with mmio protection
enable from host side leading to missing interrupts.

So skip amdgpu_device_cache_pci_state for sriov.

Signed-off-by: Victor Zhao <Victor.Zhao@amd.com>
Acked-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Use dynamic allocation for CU occupancy array in 'kfd_get_cu_occupancy()'

The `kfd_get_cu_occupancy` function previously declared a large
`cu_occupancy` array as a local variable, which could lead to stack
overflows due to excessive stack usage. This commit replaces the static
array allocation with dynamic memory allocation using `kcalloc`,
thereby reducing the stack size.

This change avoids the risk of stack overflows in kernel space,  in
scenarios where `AMDGPU_MAX_QUEUES` is large. The  allocated memory is
freed using `kfree` before the function returns  to prevent memory
leaks.

Fixes the below with gcc W=1:
drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_process.c: In function ‘kfd_get_cu_occupancy’:
drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_process.c:322:1: warning: the frame size of 1056 bytes is larger than 1024 bytes [-Wframe-larger-than=]
  322 | }
      | ^

Fixes: 6ae9e1aba97e ("drm/amdkfd: Update logic for CU occupancy calculations")
Cc: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Cc: Felix Kuehling <felix.kuehling@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Suggested-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Mukul Joshi <mukul.joshi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: always pick the pptable from IFWI

always pick the pptable from IFWI on smu v14.0.2/3

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: prevent NULL pointer dereference if ATIF is not supported

acpi_evaluate_object() may return AE_NOT_FOUND (failure), which
would result in dereferencing buffer.pointer (obj) while being NULL.

Although this case may be unrealistic for the current code, it is
still better to protect against possible bugs.

Bail out also when status is AE_NOT_FOUND.

This fixes 1 FORWARD_NULL issue reported by Coverity
Report: CID 1600951:  Null pointer dereferences  (FORWARD_NULL)

Signed-off-by: Antonio Quartulli <antonio@mandelbit.com>
Fixes: c9b7c809b89f ("drm/amd: Guard against bad data for ATIF ACPI method")
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://lore.kernel.org/r/20241031152848.4716-1-antonio@mandelbit.com
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.308

This version brings along following fixes:
- Prune Invalid Modes for HDMI Output
- SPL Cleanup
- Fix brightness level not retained over reboot
- Remove inaccessible registers from DMU diagnostics

Acked-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Prune Invalid Modes For HDMI Output

[Why]
1. HDMI does not have 6 bpc support. Having 6 bpc pass validation
does not comply with spec.
2. Validate 420 only for native HDMI, but not apply to pcon use
case.
3. Current mode validation log is not readable.

[how]
1. Cap 8 bpc for dp-hdmi converter.
2. Validate yuv420 for pcon use case as well,
   if rgb/yuv444 8bpc cannot fit into pcon bw limitation of
   the link from the converter to HDMI sink.
3. Add readable pixel_format and color_depth into debug log.

Reviewed-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Fangzhi Zuo <Jerry.Zuo@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Implement new backlight_level_params structure

[Why]
Implement the new backlight_level_params structure as part of the VBAC
framework, the information in this structure is needed to be passed down
to the DMCUB to identify the backlight control type, to adjust the
backlight of the panel and to perform any required conversions from PWM
to nits or vice versa.

[How]
Modified existing functions to include the new backlight_level_params
structure.

Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Kaitlyn Tse <Kaitlyn.Tse@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: [FW Promotion] Release 0.0.241.0

- Add DPCS health check
- Update USB4 PHY SSC
- Fix FAMS2 SubVP Close to VBlank changes
- Create VESA Aux-based backlight control path
- Fix PSR1 CRC error during CTS test

Acked-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Signed-off-by: Taimur Hassan <Syed.Hassan@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add a missing DCN401 reg definition

Add a mising reg field to the autogenerated header for future use

Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: add inst to dpm_set_vcn_enable

Add an instance parameter to the existing function dpm_set_vcn_enable()
for future implementation. Re-write all pptable functions accordingly.

v2: Remove duplicated dpm_set_vcn_enable() functions in v1. Instead,
adding instance parameter to existing functions.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Suggested-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Don't write DP_MSTM_CTRL after LT

[Why]
Observe after suspend/resme, we can't light up mst monitors under specific
mst hub. The reason is that driver still writes DPCD DP_MSTM_CTRL after LT.
It's forbidden even we write the same value for that dpcd register.

[How]
We already resume the mst branch device dpcd settings during
resume_mst_branch_status(). Leverage drm_dp_mst_topology_queue_probe() to
only probe the topology, not calling drm_dp_mst_topology_mgr_resume() which
will set DP_MSTM_CTRL as well.

Reviewed-by: Jerry Zuo <jerry.zuo@amd.com>
Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Minimize wait for pending updates

[Why/How]
Move the wait for pending updates past prepare_bandwidth if the previous
update was not a full update to reduce the average time it takes to
complete a full update.

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Signed-off-by: Ilya Bakoulin <Ilya.Bakoulin@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: parse umc_info or vram_info based on ASIC

An upstream bug report suggests that there are production dGPUs that are
older than DCN401 but still have a umc_info in VBIOS tables with the
same version as expected for a DCN401 product. Hence, reading this
tables should be guarded with a version check.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3678
Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove otg w/a toggling on HPO interfaces

[why&how]
Adjust otg w/a disable condition to include HPO explicitly rather than
assuming it is implicitly used through DP2.

Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Signed-off-by: Ausef Yousof <Ausef.Yousof@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove hw w/a toggle if on DP2/HPO

[why&how]
Applying a hw w/a only relevant to DIG FIFO causing corruption
using HPO, do not apply the w/a if on DP2/HPO

Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Signed-off-by: Ausef Yousof <Ausef.Yousof@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: SPL cleanup

[Why & How]
Move from pointer to callback to reference callback directly
Missed renaming fixpt functions with spl prefix

Reviewed-by: Navid Assadian <Navid.Assadian@amd.com>
Signed-off-by: Samson Tam <Samson.Tam@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix brightness level not retained over reboot

[Why]
During boot up and resume the DC layer will reset the panel
brightness to fix a flicker issue.

It will cause the dm->actual_brightness is not the current panel
brightness level. (the dm->brightness is the correct panel level)

[How]
Set the backlight level after do the set mode.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Fixes: d9e865826c20 ("drm/amd/display: Simplify brightness initialization")
Reported-by: Mark Herbert <mark.herbert42@gmail.com>
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3655
Reviewed-by: Sun peng Li <sunpeng.li@amd.com>
Signed-off-by: Tom Chung <chiahsuan.chung@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix asserts in SPL during bootup

[Why]
During mode validation, there maybe modes that fail
 max_downscale_src_width check and scaling_quality
 taps are 0.  This will cause an assert to trigger
 in spl_set_filters_data() because taps are 0.

[How]
Move taps calculation for non-adaptive scaling mode
 to separate function and call it
 if max_downscale_src_width fails.  This will
 populate taps if scaling_quality taps are 0.

Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Signed-off-by: Samson Tam <Samson.Tam@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove inaccessible registers from DMU diagnostics

[Why]
SEC_CNTL isn't readable by x86 and can block Z8 entry if read.

[How]
Remove the read.

Reviewed-by: Ovidiu Bunea <ovidiu.bunea@amd.com>
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix rxstatus_msg_sz type narrowing

[Why]
Code reading rxstatus message size was incorrectly assigning it to
uint8_t, despite the value being 10 bits long (lower byte plus lowest
2 bits from upper byte). This caused the highest 2 bits to be ignored,
potentially missing invalid values.

[How]
Change all local variables holding rxstatus message size from uint8_t
to uint16_t, as in mod_hdcp_message_hdcp2::rx_id_list_size.
Replaced untyped HDCP_2_2_HMID_RXSTATUS_MSG_SZ_HI macro with function
hdcp_2_2_hmid_rxstatus_msg_sz(const uint8_t[2]) to encapsulate entire
calculation and return a typed result.
Removed spaces mixed with tabs to fix indentation on modified lines.

Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Signed-off-by: Dominik Kaszewski <dominik.kaszewski@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: Block UHBR Based On USB-C PD Cable ID"

This reverts commit 4f01a68751194d05280d659a65758c09e4af04d6.

[why & how]
The offending commit caused a lighting issue for Samsung Odyssey G9
monitors when connecting via USB-C. The commit was intended to block certain UHBR rates.

Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Signed-off-by: Ausef Yousof <Ausef.Yousof@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: avoid divided by zero

[why]
insert divided by zero protection

Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Signed-off-by: Charlene Liu <Charlene.Liu@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Do Not Fallback To SW Cursor If HW Cursor Required

[Why/How]
Tearing can occur if there is a flip immediate plane and SW cursor.
check_subvp_sw_cursor_fallback_req falls back to SW cursor if the
stream has the potential to use subVP.
Check for fallback not needed if HW cursor is required.
e.g. Fullscreen gaming

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Signed-off-by: Austin Zheng <Austin.Zheng@amd.com>
Signed-off-by: Zaeem Mohamed <zaeem.mohamed@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: use string choice helpers

Use string choice helpers for better readability.

Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Julia Lawall <julia.lawall@inria.fr>
Closes: https://lore.kernel.org/r/202410161814.I6p2Nnux-lkp@intel.com/
Signed-off-by: R Sundar <prosunofficial@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix comment about amdgpu.abmlevel defaults

Since 040fdcde288a ("drm/amdgpu: respect the abmlevel module parameter value
if it is set"), the default value for amdgpu.abmlevel was set to -1, or auto.
However, the comment explaining the default value was not updated to reflect
the change (-1, or auto; not -1, or disabled).

Clarify that the default value (-1) means auto.

Fixes: 040fdcde288a ("drm/amdgpu: respect the abmlevel module parameter value if it is set")
Reported-by: Ruikai Liu <rickliu2000@outlook.com>
Signed-off-by: Mingcong Bai <jeffbai@aosc.io>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Expose special on chip memory pools in fdinfo

In the past these specialized on chip memory pools were reported as system
memory (aka 'cpu') which was not correct and misleading. That has since
been removed so lets make them visible as their own respective memory
regions.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Yunxiang Li <Yunxiang.Li@amd.com>
Cc: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Stop reporting special chip memory pools as CPU memory in fdinfo

So far these specialized on chip memory pools were reported as system
memory (aka 'cpu') which is not correct and misleading. Lets remove that
and consider later making them visible as their own thing.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Suggested-by: Christian König <christian.koenig@amd.com>
Cc: Yunxiang Li <Yunxiang.Li@amd.com>
Cc: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: stop tracking visible memory stats

Since on modern systems all of vram can be made visible anyways, to
simplify the new implementation, drops tracking how much memory is
visible for now. If this is really needed we can add it back on top of
the new implementation, or just report all the BOs as visible.

Signed-off-by: Yunxiang Li <Yunxiang.Li@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: make drm-memory-* report resident memory

The old behavior reports the resident memory usage for this key and the
documentation say so as well. However this was accidentally changed to
include buffers that was evicted.

Fixes: 04bdba46542c ("drm/amdgpu: Use drm_print_memory_stats helper from fdinfo")
Signed-off-by: Yunxiang Li <Yunxiang.Li@amd.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix the memory allocation issue in amdgpu_discovery_get_nps_info()

Fix two issues with memory allocation in amdgpu_discovery_get_nps_info()
for mem_ranges:

 - Add a check for allocation failure to avoid dereferencing a null
   pointer.

 - As suggested by Christophe, use kvcalloc() for memory allocation,
   which checks for multiplication overflow.

Additionally, assign the output parameters nps_type and range_cnt after
the kvcalloc() call to prevent modifying the output parameters in case
of an error return.

Fixes: b194d21b9bcc ("drm/amdgpu: Use NPS ranges from discovery table")
Suggested-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Li Huafei <lihuafei1@huawei.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add ring reset messages

Add messages to make it clear when a per ring reset
happens.  This is helpful for debugging and aligns with
other reset methods.

v2: add ring name in success/fail messages (Lijo)

Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Kent Russell <kent.russell@amd.com> (v1)
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix fairness in enforce isolation handling

Make sure KFD gets a turn when serializing access to
the GC IP.  Currently non-KFD jobs can starve KFD if they
submit often enough.  This patch prevents that by stalling
non-KFD if its time period has elapsed.

v2: fix units
v3: check enablement properly

Acked-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove last parts of timing_trace

Commit c2c2ce1e9623 ("drm/amd/display: Optimize passive update planes.")
removed the last caller of context_timing_trace.
Remove it.

With that gone, no one is now looking at the 'timing_trace' flag, remove
it and all the places that set it.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove unused cm3_helper_translate_curve_to_degamma_hw_format

cm3_helper_translate_curve_to_degamma_hw_format() since it was added in
2020's commit
03f54d7d3448 ("drm/amd/display: Add DCN3 DPP")

Remove it.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Remove unused regamma functions

calculate_user_regamma_coeff() and calculate_user_regamma_ramp() were
added in 2018 in commit
55a01d4023ce ("drm/amd/display: Add user_regamma to color module")

but never used.

Remove them and their helpers.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: add an interface to query whether is KFD is active

Add an interface to query whether KFD has any active queues.

v2: fix build issues

Acked-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/mediatek: Switch to for_each_child_of_node_scoped()

Introduce the scoped variant of the loop to automatically release the
child node when it goes out of scope, which is more robust than the
non-scoped variant, and accounts for new early exits that could be added
in the future.

Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.kernel.org/project/dri-devel/patch/20241011-mtk_drm_drv_memleak-v1-2-2b40c74c8d75@gmail.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>

drm/mediatek: Fix child node refcount handling in early exit

Early exits (goto, break, return) from for_each_child_of_node() required
an explicit call to of_node_put(), which was not introduced with the
break if cnt == MAX_CRTC.

Add the missing of_node_put() before the break.

Cc: stable@vger.kernel.org
Fixes: d761b9450e31 ("drm/mediatek: Add cnt checking for coverity issue")
Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Reviewed-by: Chen-Yu Tsai <wenst@chromium.org>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.kernel.org/project/dri-devel/patch/20241011-mtk_drm_drv_memleak-v1-1-2b40c74c8d75@gmail.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>