Perry Yuan [Tue, 13 Jan 2026 07:04:04 +0000 (15:04 +0800)]
drm/amdgpu: Improve IP discovery checksum failure logging
Enhance the error logging in amdgpu_discovery_verify_checksum() to
print the calculated checksum, the expected checksum, the data size.
This extra context helps quickly identify if the issue is a data
corruption, a partially read binary, or an invalid table header without
requiring additional instrumentation.
Signed-off-by: Perry Yuan <perry.yuan@amd.com> Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jiqian Chen [Wed, 14 Jan 2026 10:06:10 +0000 (18:06 +0800)]
drm/amdgpu: free hw_vm_fence when fail in amdgpu_job_alloc
If drm_sched_job_init fails, hw_vm_fence is not freed currently,
then cause memory leak.
Fixes: db36632ea51e ("drm/amdgpu: clean up and unify hw fence handling") Link: https://lore.kernel.org/amd-gfx/a5a828cb-0e4a-41f0-94c3-df31e5ddad52@amd.com/T/#t Signed-off-by: Jiqian Chen <Jiqian.Chen@amd.com> Reviewed-by: Amos Kong <kongjianjun@gmail.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jesse.Zhang [Tue, 13 Jan 2026 08:13:47 +0000 (16:13 +0800)]
drm/amd/amdgpu: Add independent hang detect work for user queue fence
In error scenarios (e.g., malformed commands), user queue fences may never
be signaled, causing processes to wait indefinitely. To address this while
preserving the requirement of infinite fence waits, implement an independent
timeout detection mechanism:
1. Initialize a hang detect work when creating a user queue (one-time setup)
2. Start the work with queue-type-specific timeout (gfx/compute/sdma) when
the last fence is created via amdgpu_userq_signal_ioctl (per-fence timing)
3. Trigger queue reset logic if the timer expires before the fence is signaled
v2: make timeout per queue type (adev->gfx_timeout vs adev->compute_timeout vs adev->sdma_timeout) to be consistent with kernel queues. (Alex)
v3: The timeout detection must be independent from the fence, e.g. you don't wait for a timeout on the fence
but rather have the timeout start as soon as the fence is initialized. (Christian)
v4: replace the timer with the `hang_detect_work` delayed work.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Jesse Zhang <jesse.zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Thu, 20 Nov 2025 21:43:04 +0000 (16:43 -0500)]
drm/amdkfd: Move gfx9.4.3 and gfx 9.5 MQD to HBM
To reduce queue switch latency further, move MQD to VRAM domain, CP
access MQD and control stack via FB aperture, this requires contiguous
pages.
After MQD is initialized, updated or restored, flush HDP to guarantee
the data is written to HBM and GPU cache is invalidated, then CP will
read the new MQD.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
1. mtk_hdmi_v2: Remove unneeded semicolon
2. Move DP training to hotplug thread
3. Convert legacy DRM logging to drm_* helpers in mtk_crtc.c
4. mtk_dsi: Add support for High Speed (HS) mode
5. Add HDMI support for Mediatek Genio 510/700/1200-EVK and Radxa NIO-12L boards
dt-bindings: phy: mediatek,hdmi-phy: Fix clock output names for MT8195
For all of the HDMI PHYs compatible with the one found on MT8195
the output clock has a different datasheet name and specifically
it is called "hdmi_txpll", differently from the older HDMI PHYs
which output block is called "hdmitx_dig_cts".
Replace clock output name string check by max item number one to allow
the new name on all of the HDMI PHY IPs that are perfectly compatible
with MT8195.
[Louis-Alexis Eyraud: split patch, addressed previous feedback from
mailing list, and reworded description]
drm/mediatek: mtk_dsi: Add support for High Speed (HS) mode
Up until now, the MediaTek DSI Controller has always been working
in Low Power Mode (LPM), as this driver has always ignored the
MIPI_DSI_MSG_USE_LPM flag hence never setting HS mode.
In the current state of the driver the only thing that is needed
to add support for DSI High Speed (HS) transmit is to simply set
the "HSTX" config bit in the configuration register.
Check if flag MIPI_DSI_MSG_USE_LPM is set and, if not, set HSTX.
Dave Airlie [Fri, 16 Jan 2026 03:39:15 +0000 (13:39 +1000)]
Merge tag 'drm-xe-next-2026-01-15' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next
UAPI Changes:
- Remove unused KEEP_ACTIVE flag in the new multi queue uAPI (Niranjana)
- Expose new temperature attributes in HWMON (Karthik)
Driver Changes:
- Force i2c into polling mode when in survivability (Raag)
- Validate preferred system memory placement in xe_svm_range_validate (Brost)
- Adjust page count tracepoints in shrinker (Brost)
- Fix a couple drm_pagemap issues with multi-GPU (Brost)
- Define GuC firmware for NVL-S (Roper)
- Handle GT resume failure (Raag)
- Improve wedged mode handling (Lukasz)
- Add missing newlines to drm_warn messages (Osama)
- Fix WQ_MEM_RECLAIM passed as max_active to alloc_workqueue (Marco)
- Page-reclaim fixes and PRL stats addition (Brian)
- Fix struct guc_lfd_file_header kernel-doc (Jani)
- Allow compressible surfaces to be 1-way coherent (Xin)
- Fix DRM scheduler layering violations in Xe (Brost)
- Minor improvements to MERT code (Michal)
- Privatize struct xe_ggtt_node (Maarten)
- Convert wait for lmem init into an assert (Bala)
- Enable GSC loading and PXP for PTL (Daniele)
- Replace use of system_wq with tlb_inval->timeout_wq (Marco)
- VRAM addr range bit expansion (Fei)
- Cleanup unused header includes (Roper)
Dave Airlie [Fri, 16 Jan 2026 02:57:20 +0000 (12:57 +1000)]
Merge tag 'drm-intel-next-2026-01-15' of https://gitlab.freedesktop.org/drm/i915/kernel into drm-next
Beyond Display:
- Make 'guc_hw_reg_state' static as it isn't exported (Ben)
- Fix doc build on mei related interface header (Jani)
Display related:
- Fix ggtt fb alignment on Xe display (Tvrtko)
- More display clean-up towards deduplication and full separation (Jani)
- Use the consolidated HDMI tables (Suraj)
- Account for DSC slice overhead (Ankit)
- Prepare GVT for display modularization (Ankit, Jani)
- Enable/Disable DC balance along with VRR DSB (Mitul, Ville)
- Protection against unsupported modes in LT PHY (Suraj)
- Display W/a addition and fixes (Gustavo)
- Fix many SPDX identifier comments (Ankit)
- Incorporate Xe3_LPD changes for CD2X divider (Gustavo)
- Clean up link BW/DSC slice config computation (Imre)
Dave Airlie [Fri, 16 Jan 2026 01:03:44 +0000 (11:03 +1000)]
Merge tag 'drm-misc-next-2026-01-15' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next
drm-misc-next for 6.20:
Core Changes:
- atomic: Introduce Gamma/Degamma LUT size check
- gem: Fix a leak in drm_gem_get_unmapped_area
- gpuvm: API sanitation for Rust bindings
- panic: Few corner-cases fixes
Driver Changes:
- Replace system workqueue with percpu equivalent
- amdxdna: Update message buffer allocation requirements, Update
firmware version check
- imagination: Add AM62P support
- ivpu: Implement warm boot flow
- rockchip: Get rid of atomic_check fixups, Add Rockchip RK3506 Support
- rocket: Cleanups
- bridge:
- dw-hdmi-qp: Add support for HPD-less setups
- panel:
- mantix: Various power management related improvements
- new panels: Innolux G150XGE-L05,
- dma-buf:
- cma: Call clear_page instead of memset
Matt Roper [Thu, 15 Jan 2026 03:28:02 +0000 (19:28 -0800)]
drm/xe: Cleanup unused header includes
clangd reports many "unused header" warnings throughout the Xe driver.
Start working to clean this up by removing unnecessary includes in our
.c files and/or replacing them with explicit includes of other headers
that were previously being included indirectly.
By far the most common offender here was unnecessary inclusion of
xe_gt.h. That likely originates from the early days of xe.ko when
xe_mmio did not exist and all register accesses, including those
unrelated to GTs, were done with GT functions.
There's still a lot of additional #include cleanup that can be done in
the headers themselves; that will come as a followup series.
v2:
- Squash the 79-patch series down to a single patch. (MattB)
Ivan Lipski [Tue, 13 Jan 2026 22:29:59 +0000 (17:29 -0500)]
drm/amd/display: Add an hdmi_hpd_debounce_delay_ms module
[Why&How]
Right now, the HDMI HPD filter is enabled by default at 1500ms.
We want to disable it by default, as most modern displays with HDMI do
not require it for DPMS mode.
The HPD can instead be enabled as a driver parameter with a custom delay
value in ms (up to 5000ms).
Fixes: c918e75e1ed9 ("drm/amd/display: Add an HPD filter for HDMI") Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4859 Signed-off-by: Ivan Lipski <ivan.lipski@amd.com> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Wed, 26 Nov 2025 19:35:11 +0000 (14:35 -0500)]
drm/amdkfd: Add domain parameter to alloc kernel BO
To allocate kernel BO from VRAM domain for MQD in the following patch.
No functional change because kernel BO allocate all from GTT domain.
Rename amdgpu_amdkfd_alloc_gtt_mem to amdgpu_amdkfd_alloc_kernel_mem
Rename amdgpu_amdkfd_free_gtt_mem to amdgpu_amdkfd_free_kernel_mem
Rename mem_kfd_mem_obj gtt_mem to mem
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Kent Russell <kent.russell@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu/userq: Fix fence reference leak on queue teardown v2
The user mode queue keeps a pointer to the most recent fence in
userq->last_fence. This pointer holds an extra dma_fence reference.
When the queue is destroyed, we free the fence driver and its xarray,
but we forgot to drop the last_fence reference.
Because of the missing dma_fence_put(), the last fence object can stay
alive when the driver unloads. This leaves an allocated object in the
amdgpu_userq_fence slab cache and triggers
This is visible during driver unload as:
BUG amdgpu_userq_fence: Objects remaining on __kmem_cache_shutdown()
kmem_cache_destroy amdgpu_userq_fence: Slab cache still has objects
Call Trace:
kmem_cache_destroy
amdgpu_userq_fence_slab_fini
amdgpu_exit
__do_sys_delete_module
Fix this by putting userq->last_fence and clearing the pointer during
amdgpu_userq_fence_driver_free().
This makes sure the fence reference is released and the slab cache is
empty when the module exits.
v2: Update to only release userq->last_fence with dma_fence_put()
(Christian)
Fixes: edc762a51c71 ("drm/amdgpu/userq: move some code around") Cc: Alex Deucher <alexander.deucher@amd.com> Cc: Christian König <christian.koenig@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdkfd: No need to suspend whole MES to evict process
Each queue of the process is individually removed and there is not need
to suspend whole mes. Suspending mes stops kernel mode queues also
causing unnecessary timeouts when running mixed work loads
Fixes: 079ae5118e1f ("drm/amdkfd: fix suspend/resume all calls in mes based eviction path") Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4765 Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It’s better to validate VM TLB flushes in the flush‑TLB backend
rather than in the generic VM layer.
Reverting this patch depends on
commit fa7c231fc2b0 ("drm/amdgpu: validate the flush_gpu_tlb_pasid()")
being present in the tree.
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>
Prike Liang [Tue, 6 Jan 2026 09:00:57 +0000 (17:00 +0800)]
drm/amdgpu: validate the flush_gpu_tlb_pasid()
Validate flush_gpu_tlb_pasid() availability before flushing tlb.
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>
Lang Yu [Wed, 17 Dec 2025 03:01:39 +0000 (11:01 +0800)]
drm/amdkfd: Switch to using GC VERSION to decide LDS/Scratch base
Next generation GC IP with 4-level page table needs to use the
same LDS/Scratch base with 5-level page table, use GC VERSION
to decide is more appropriate.
Signed-off-by: Lang Yu <lang.yu@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Yang Wang [Tue, 6 Jan 2026 06:42:40 +0000 (14:42 +0800)]
drm/amd/pm: fix smu overdrive data type wrong issue on smu 14.0.2
resolving the issue of incorrect type definitions potentially causing calculation errors.
Fixes: 54f7f3ca982a ("drm/amdgpu/swm14: Update power limit logic") Signed-off-by: Yang Wang <kevinyang.wang@amd.com> Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Xiaogang Chen [Wed, 14 Jan 2026 02:45:14 +0000 (20:45 -0600)]
drm/amdkfd: kfd driver supports hot unplug/replug amdgpu devices
This patch allows kfd driver function correctly when AMD gpu devices got
unplug/replug at run time.
When an AMD gpu device got unplug kfd driver gracefully terminates existing
kfd processes after stops all queues by sending SIGBUS to user process. After
that user space can still use remaining AMD gpu devices. When all AMD gpu
devices at system got removed kfd driver will not response new requests.
Unplugged AMD gpu devices can be re-plugged. kfd driver will use added devices
to function as usual.
The purpose of this patch is having kfd driver behavior as expected during and
after AMD gpu devices unplug/replug at run time.
Signed-off-by: Xiaogang Chen <Xiaogang.Chen@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Yang Wang [Mon, 12 Jan 2026 09:48:26 +0000 (17:48 +0800)]
drm/amd/pm: fix issue of missing '*' on pp_dpm_xxx nodes
refine the code to fix '*' missing on pp_dpm_xxx series node.
e.g.: missing '*' on navi10 pp_dpm_sclk
$ cat /sys/class/drm/card0/device/pp_dpm_sclk
0: 300Mhz
1: 1930Mhz (missing symbol '*')
Fixes: a08ea4bc7711 ("drm/amd/pm: Add a helper to show dpm table") Signed-off-by: Yang Wang <kevinyang.wang@amd.com> Reviewed-by: Lijo Lazar <lijo.lazar@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Donet Tom [Mon, 12 Jan 2026 14:06:56 +0000 (19:36 +0530)]
drm/amdkfd: Fix GART PTE for non-4K pagesize in svm_migrate_gart_map()
In svm_migrate_gart_map(), while migrating GART mapping, the number of
bytes copied for the GART table only accounts for CPU pages. On non-4K
systems, each CPU page can contain multiple GPU pages, and the GART
requires one 8-byte PTE per GPU page. As a result, an incorrect size was
passed to the DMA, causing only a partial update of the GART table.
Fix this function to work correctly on non-4K page-size systems by
accounting for the number of GPU pages per CPU page when calculating the
number of bytes to be copied.
Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Philip Yang <Philip.Yang@amd.com> Signed-off-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com> Signed-off-by: Donet Tom <donettom@linux.ibm.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>
Donet Tom [Mon, 12 Jan 2026 14:06:55 +0000 (19:36 +0530)]
drm/amdkfd: Fix SVM map/unmap address conversion for non-4k page sizes
SVM range size is tracked using the system page size. The range start and
end are aligned to system page-sized PFNs, so the total SVM range size
equals the total number of pages in the SVM range multiplied by the system
page size.
The SVM range map/unmap functions pass these system page-sized PFN numbers
to amdgpu_vm_update_range(), which expects PFNs based on the GPU page size
(4K). On non-4K page systems, this mismatch causes only part of the SVM
range to be mapped in the GPU page table, while the rest remains unmapped.
If the GPU accesses an unmapped address within the same range, it results
in a GPU page fault.
To fix this, the required conversion has been added in both
svm_range_map_to_gpu() and svm_range_unmap_from_gpu(), ensuring that all
pages in the SVM range are correctly mapped on non-4K systems.
Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Philip Yang <Philip.Yang@amd.com> Signed-off-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com> Signed-off-by: Donet Tom <donettom@linux.ibm.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>
Donet Tom [Mon, 12 Jan 2026 14:06:54 +0000 (19:36 +0530)]
drm/amdkfd: Relax size checking during queue buffer get
HW-supported EOP buffer sizes are 4K and 32K. On systems that do not
use 4K pages, the minimum buffer object (BO) allocation size is
PAGE_SIZE (for example, 64K). During queue buffer acquisition, the driver
currently checks the allocated BO size against the supported EOP buffer
size. Since the allocated BO is larger than the expected size, this check
fails, preventing queue creation.
Relax the strict size validation and allow PAGE_SIZE-sized BOs to be used.
Only the required 4K region of the buffer will be used as the EOP buffer
and avoids queue creation failures on non-4K page systems.
Acked-by: Christian König <christian.koenig@amd.com> Suggested-by: Philip Yang <yangp@amd.com> Signed-off-by: Donet Tom <donettom@linux.ibm.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/amd/display: Initialise backlight level values from hw
Internal backlight levels are initialised from ACPI but the values
are sometimes out of sync with the levels in effect until there has
been a read from hardware (eg triggered by reading from sysfs).
This means that the first drm_commit can cause the levels to be set
to a different value than the actual starting one, which results in
a sudden change in brightness.
This path shows the problem (when the values are out of sync):
This patch calls the backlight ops get_brightness explicitly
at the end of backlight registration to make sure dm->brightness[n]
is in sync with the actual hardware levels.
Fixes: 2fe87f54abdc ("drm/amd/display: Set default brightness according to ACPI") Signed-off-by: Vivek Das Mohapatra <vivek@collabora.com> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Michal Wajdeczko [Mon, 12 Jan 2026 18:37:16 +0000 (19:37 +0100)]
drm/xe/mert: Improve handling of MERT CAT errors
All MERT catastrophic errors but VF's LMTT fault are serious, so
we shouldn't limit our handling only to print debug messages.
Change CATERR message to error level and then declare the device
as wedged to match expectation from the design document. For the
LMTT faults, add a note about adding tracking of this unexpected
VF activity.
While at it, rename register fields defnitions to match the BSpec.
Also drop trailing include guard name from the regs.h file.
drm/rockchip: DRM_ROCKCHIP should depend on ARCH_ROCKCHIP
Rockchip display hardware is only available on Rockchip SoCs. Hence add
a dependency on ARCH_ROCKCHIP, to prevent asking the user about this
driver when configuring a kernel without Rockchip platform support.
Before, this dependency was implicit through a hard dependency on
ROCKCHIP_IOMMU.
Fei Yang [Mon, 12 Jan 2026 22:03:30 +0000 (14:03 -0800)]
drm/xe: vram addr range is expanded to bit[17:8]
The bit field used to be [14:8] with [17:15] marked as SPARE and
defaulted to 0. So, simply expand the read to bit[17:8] assuming
the platforms using only bit[14:8] have zeros in the expanded bits.
Marco Crivellari [Mon, 12 Jan 2026 09:44:06 +0000 (10:44 +0100)]
drm/xe: Replace use of system_wq with tlb_inval->timeout_wq
This patch continues the effort to refactor workqueue APIs, which has begun
with the changes introducing new workqueues and a new alloc_workqueue flag:
commit 128ea9f6ccfb ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566af ("workqueue: Add new WQ_PERCPU flag")
The point of the refactoring is to eventually alter the default behavior of
workqueues to become unbound by default so that their workload placement is
optimized by the scheduler.
Before that to happen, workqueue users must be converted to the better named
new workqueues with no intended behaviour changes:
This way the old obsolete workqueues (system_wq, system_unbound_wq) can be
removed in the future.
After a carefully evaluation, because this is the fence signaling path, we
changed the code in order to use one of the Xe's workqueue.
So, a new workqueue named 'timeout_wq' has been added to
'struct xe_tlb_inval' and has been initialized with 'gt->ordered_wq'
changing the system_wq uses with tlb_inval->timeout_wq.
drm/atomic: verify that gamma/degamma LUTs are not too big
The kernel specifies LUT table sizes in a separate property, however it
doesn't enforce it as a maximum. Some drivers implement max size check
on their own in the atomic_check path. Other drivers simply ignore the
issue. Perform LUT size validation in the generic place.
drm/atomic: add max_size check to drm_property_replace_blob_from_id()
The function drm_property_replace_blob_from_id() allows checking whether
the blob size is equal to a predefined value. In case of variable-size
properties (like the gamma / degamma LUTs) we might want to check for
the blob size against the maximum, allowing properties of the size
lesser than the max supported by the hardware. Extend the function in
order to support such checks.
We have a helper to get property values for non-atomic drivers and
another one default property values for atomic drivers. In some cases we
need the ability to get value of immutable property, no matter what kind
of driver it is. Implement new property-related helper,
drm_object_immutable_property_get_value(), which lets the caller to get
the value of the immutable property.
Imre Deak [Mon, 22 Dec 2025 15:35:47 +0000 (17:35 +0200)]
drm/i915/dp: Simplify computing the DSC compressed BPP for DP-MST
The minimum/maximum DSC input (i.e. pipe) and compressed (i.e. link) BPP
limits are computed already in intel_dp_compute_config_limits(), so
there is no need to do this again in
mst_stream_dsc_compute_link_config() called later. Remove the
corresponding alignments from the latter function and use the
precomputed (aligned and within bounds) maximum pipe BPP and the min/max
compressed BPP values instead as-is.
Imre Deak [Mon, 22 Dec 2025 15:35:44 +0000 (17:35 +0200)]
drm/i915/dp: Simplify computing forced DSC BPP for DP-SST
If dsc_compute_compressed_bpp() failed with a forced pipe BPP value
(where the forced pipe BPP value itself is valid within the min/max pipe
BPP limits), the function will also fail when called with the maximum
pipe BPP value: dsc_compute_compressed_bpp() will try all compressed
BPPs below the passed in pipe BPP value and if the function failed with
a given (low) compressed BPP value it will also fail with a compressed
BPP value higher than the one which failed already.
Based on the above remove the logic to retry computing a compressed BPP
value with the maximum pipe BPP value if computing the compressed BPP
failed already with the (lower) forced pipe BPP value.
Imre Deak [Mon, 22 Dec 2025 15:35:43 +0000 (17:35 +0200)]
drm/i915/dp: Simplify computing DSC BPPs for DP-SST
The maximum pipe BPP value (used as the DSC input BPP) has been aligned
already to the corresponding source/sink input BPP capabilities in
intel_dp_compute_config_limits(). So it isn't needed to perform the same
alignment again in intel_dp_dsc_compute_pipe_bpp() called later, this
function can simply use the already aligned maximum pipe BPP value, do
that.
Also, there is no point in trying pipe BPP values lower than the
maximum: this would only make dsc_compute_compressed_bpp() start with a
lower _compressed_ BPP value, but this lower compressed BPP value has
been tried already when dsc_compute_compressed_bpp() was called with the
higher pipe BPP value (i.e. the first dsc_compute_compressed_bpp() call
tries already all the possible compressed BPP values which are all below
the pipe BPP value passed to it). Simplify the function accordingly
trying only the maximum pipe BPP value.
Imre Deak [Mon, 22 Dec 2025 15:35:42 +0000 (17:35 +0200)]
drm/i915/dp: Simplify computing DSC BPPs for eDP
The maximum pipe BPP value (used as the DSC input BPP) has been aligned
already to the corresponding source/sink input BPP capabilities in
intel_dp_compute_config_limits(). So it isn't needed to perform the same
alignment again in intel_edp_dsc_compute_pipe_bpp() called later, this
function can simply use the already aligned maximum pipe BPP value, do
that.
Imre Deak [Mon, 22 Dec 2025 15:35:41 +0000 (17:35 +0200)]
drm/i915/dp: Use helpers to align min/max compressed BPPs
The minimum/maximum compressed BPP values are aligned/bounded in
intel_dp_compute_link_bpp_limits() to the corresponding source limits.
The minimum compressed BPP value doesn't change afterwards, so no need
to align it again, remove that.
The maximum compressed BPP, which depends on the pipe BPP value still
needs to be aligned, since the pipe BPP value could change after the
above limits were computed, via intel_dp_force_dsc_pipe_bpp(). Use the
corresponding helper for this alignment instead of open-coding the same.
Imre Deak [Mon, 22 Dec 2025 15:35:40 +0000 (17:35 +0200)]
drm/i915/dp: Unify detect and compute time DSC mode BW validation
Atm, a DP DSC video mode's required BW vs. the available BW is
determined by calculating the maximum compressed BPP value allowed by
the available BW. Doing that using a closed-form formula as it's done
atm (vs. an iterative way) is problematic, since the overhead of the
required BW itself depends on the BPP value being calculated. Instead of
that calculate the required BW for the minimum compressed BPP value
supported both by the source and the sink and check this BW against the
available BW. This change also aligns the BW calculation during mode
validation with how this is done during state computation, calculating
the required effective data rate with the corresponding BW overhead.
Imre Deak [Mon, 22 Dec 2025 15:35:39 +0000 (17:35 +0200)]
drm/i915/dp: Add intel_dp_mode_valid_with_dsc()
Add intel_dp_mode_valid_with_dsc() and call this for an SST/MST mode
validation to prepare for a follow-up change using a way to verify the
mode's required BW the same way this is done elsewhere during state
computation (which in turn depends on the mode's effective data rate
with the corresponding BW overhead).
Imre Deak [Mon, 22 Dec 2025 15:35:35 +0000 (17:35 +0200)]
drm/i915/dp: Pass intel_output_format to intel_dp_dsc_sink_{min_max}_compressed_bpp()
Prepare for follow-up changes also calling
intel_dp_dsc_min_sink_compressed_bpp() /
intel_dp_dsc_max_sink_compressed_bpp_x16()
without an intel_crtc_state.
While at it remove the stale function declarations from the header file.
Imre Deak [Mon, 22 Dec 2025 15:35:34 +0000 (17:35 +0200)]
drm/i915/dp: Drop intel_dp parameter from intel_dp_compute_config_link_bpp_limits()
The intel_dp pointer can be deducted from the connector pointer, so it's
enough to pass only connector to
intel_dp_compute_config_link_bpp_limits(), do so.
Imre Deak [Mon, 22 Dec 2025 15:35:33 +0000 (17:35 +0200)]
drm/i915/dp: Align min/max compressed BPPs when calculating BPP limits
Align the minimum/maximum DSC compressed BPPs to the corresponding
source compressed BPP limits already when computing the BPP limits. This
alignment is also performed later during state computation, however
there is no reason to initialize the limits to an unaligned/incorrect
value.
Imre Deak [Mon, 22 Dec 2025 15:35:32 +0000 (17:35 +0200)]
drm/i915/dp: Align min/max DSC input BPPs to sink caps
Align the minimum/maximum DSC input BPPs to the corresponding sink DSC
input BPP capability limits already when computing the BPP limits. This
alignment is also performed later during state computation, however
there is no reason to initialize the limits to an unaligned/incorrect
value.
drm/panthor: Implement reading shader_present from nvmem
On some platforms, notably MediaTek MT8196, the shader_present bitmask
in the Mali GPU register for it has cores enabled that may be faulty.
The true shader_present bitmask is found in an efuse instead.
Implement reading shader_present from an nvmem cell if one is present,
falling back to the Mali register if it's absent. The error codes are
trickled up through to the probe function so that probe deferral works.
On the MediaTek MT8196 SoC, the bitmask for which shader cores are
present and functional is not the one in the Mali GPU's registers, but
in an external efuse.
Add the nvmem cell properties to describe such a setup, and make them
required on MT8196.
Abhishek Rajput [Tue, 23 Dec 2025 09:54:34 +0000 (15:24 +0530)]
drm/mediatek: Convert legacy DRM logging to drm_* helpers in mtk_crtc.c
Replace DRM_ERROR() and DRM_DEBUG_DRIVER() calls in
drivers/gpu/drm/mediatek/mtk_crtc.c with the corresponding drm_err()
and drm_dbg_driver() helpers.
The drm_*() logging helpers take a struct drm_device * argument,
allowing the DRM core to prefix log messages with the correct device
name and instance. This is required to correctly distinguish log
messages on systems with multiple GPUs.
This change aligns the Mediatek DRM driver with the DRM TODO item:
"Convert logging to drm_* functions with drm_device parameter".
Uwe Kleine-König [Wed, 10 Dec 2025 08:31:38 +0000 (09:31 +0100)]
host1x: Convert to bus methods
The callbacks .probe(), .remove() and .shutdown() for device_drivers
should go away. So migrate to bus methods. There are two differences
that need addressing:
- The bus remove callback returns void while the driver remove callback
returns int (the actual value is ignored by the core).
- The bus shutdown callback is also called for unbound devices, so an
additional check for dev->driver != NULL is needed.
Uwe Kleine-König [Wed, 10 Dec 2025 08:31:37 +0000 (09:31 +0100)]
host1x: Make remove callback return void
The return value of struct device_driver::remove is ignored by the core
(see device_remove() in drivers/base/dd.c). So it doesn't make sense to
let the host1x remove callback return an int just to ignore it later.
So make the callback return void. All current implementors return 0, so
they are easily converted.
According to the sequence from section 7.3.4 of FT8006P datasheet,
TP_RSTN and RESX should be asserted after disabling AVDD and AVEE and
together with VDDI.
Also, AVEE power down needs to happen at least 150ms after entering
sleep mode.
drm/panel: mantix: Improve power on sequence timings
FP8006P datasheet mentions:
> It is necessary to wait 15msec after releasing RESX before sending
> commands. Also Sleep Out command cannot be sent for 120 msec.
This hasn't been respected by the driver so far, which could interfere
with the LCD init code sequence performed by the controller. In some cases
this leads to VCOM voltage being set to a wrong value, causing "halo"
effects, temporary burn-in around the edges of the screen and degraded
image contrast.
T3 and T4 are counted from when VDDI is enabled. There's no need to add
them when we've already waited more than that in T2 and T2d.
While FT8006P datasheet does not mention a delay between exiting sleep
mode and turning the display on, code provided by the vendor uses 120ms
there and it happens to be the same value as required in newer datasheets
for newer controllers from the same family, so it seems appropriate to
use it here as well.
Karthik Poosa [Mon, 12 Jan 2026 20:35:21 +0000 (02:05 +0530)]
drm/xe/hwmon: Expose individual VRAM channel temperature
Expose individual VRAM temperature attributes.
Update Xe hwmon documentation for this entry.
v2:
- Avoid using default switch case for VRAM individual temperatures.
- Append labels with VRAM channel number.
- Update kernel version in Xe hwmon documentation.
v3:
- Add missing brackets in Xe hwmon documentation from VRAM channel sysfs.
- Reorder BMG_VRAM_TEMPERATURE_N macro in xe_pcode_regs.h.
- Add api to check if VRAM is available on the channel.
v4:
- Improve VRAM label handling to eliminate temp variable by
introducing a dedicated array vram_label in xe_hwmon_thermal_info.
- Remove a magic number.
- Change the label from vram_X to vram_ch_X.
v5:
- Address review comments from Raag.
- Change vram to VRAM in commit title and subject.
- Refactor BMG_VRAM_TEMPERATURE_N macro.
- Refactor is_vram_ch_available().
- Rephrase a comment.
- Check individual VRAM temperature limits in addition to VRAM
availability in xe_hwmon_temp_is_visible. (Raag)
- Move VRAM label change out of this patch.
v6:
- Use in_range() for VRAM_N index check instead of if check. (Raag)
- Minor aesthetic changes.
Karthik Poosa [Mon, 12 Jan 2026 20:35:20 +0000 (02:05 +0530)]
drm/xe/hwmon: Expose GPU PCIe temperature
Expose GPU PCIe average temperature and its limits via hwmon sysfs entry
temp5_xxx.
Update Xe hwmon sysfs documentation for this.
v2: Update kernel version in Xe hwmon documentation. (Raag)
v3:
- Address review comments from Raag.
- Remove redundant debug log.
- Update kernel version in Xe hwmon documentation. (Raag)
v4:
- Address review comments from Raag.
- Group new temperature attributes with existing temperature attributes
as per channel index in Xe hwmon documentation.
- Use TEMP_MASK instead of TEMP_MASK_MAILBOX.
- Add PCIE_SENSOR_MASK which uses REG_FIELD_GET as replacement of
PCIE_SENSOR_SHIFT.
v5:
- Address review comments from Raag.
- Use REG_FIELD_GET to get PCIe temperature.
- Move PCIE_SENSOR_GROUP_ID and PCIE_SENSOR_MASK to xe_pcode_api.h
- Cosmetic change.
Karthik Poosa [Mon, 12 Jan 2026 20:35:19 +0000 (02:05 +0530)]
drm/xe/hwmon: Expose memory controller temperature
Expose GPU memory controller average temperature and its limits under
temp4_xxx.
Update Xe hwmon documentation for this.
v2:
- Rephrase commit message. (Badal)
- Update kernel version in Xe hwmon documentation. (Raag)
v3:
- Update kernel version in Xe hwmon documentation.
- Address review comments from Raag.
- Remove obvious comments.
- Remove redundant debug logs.
- Remove unnecessary checks.
- Avoid magic numbers.
- Add new comments.
- Use temperature sensors count to make memory controller visible.
- Use temperature limits of package for memory controller.
v4:
- Address review comments from Raag.
- Group new temperature attributes with existing temperature attributes
as per channel index in Xe hwmon documentation.
- Use DIV_ROUND_UP to calculate dwords needed for temperature limits.
- Minor aesthetic refinements.
- Remove unused TEMP_MASK_MAILBOX.
v5:
- Use REG_FIELD_GET to get count from READ_THERMAL_DATA output. (Raag)
- Change count print from decimal to hexadecimal.
- Cosmetic changes.
Karthik Poosa [Mon, 12 Jan 2026 20:35:18 +0000 (02:05 +0530)]
drm/xe/hwmon: Expose temperature limits
Read temperature limits using pcode mailbox and expose shutdown
temperature limit as tempX_emergency, critical temperature limit as
tempX_crit and GPU max temperature limit as temp2_max.
Update Xe hwmon documentation with above entries.
v2:
- Resolve a documentation warning.
- Address below review comments from Raag.
- Update date and kernel version in Xe hwmon documentation.
- Remove explicit disable of has_mbx_thermal_info for unsupported
platforms.
- Remove unnecessary default case in switches.
- Remove obvious comments.
- Use TEMP_LIMIT_MAX to compute number of dwords needed in
xe_hwmon_thermal_info.
- Remove THERMAL_LIMITS_DWORDS macro.
- Use has_mbx_thermal_info for checking thermal mailbox support.
v3:
- Address below minor comments. (Raag)
- Group new temperature attributes with existing temperature attributes
as per channel index in Xe hwmon documentation.
- Rename enums of xe_temp_limit to improve clarity.
- Use DIV_ROUND_UP to calculate dwords needed for temperature limits.
- Use return instead of breaks in xe_hwmon_temp_read.
- Minor aesthetic refinements.
v4:
- Remove a redundant break. (Raag)
- Update drm_dbg to drm_warn to inform user of unavailability for
thermal mailbox on expected platforms.
drm/xe/gsc: Make GSC FW load optional for newer platforms
On newer platforms GSC FW is only required for content protection
features, so the core driver features work perfectly fine without it
(and we did in fact not enable it to start with on PTL). Therefore, we
can selectively enable the GSC only if the FW is found on disk, without
failing if it is not found.
Note that this means that the FW can now be enabled (i.e., we're looking
for it) but not available (i.e., we haven't found it), so checks on FW
support should use the latter state to decide whether to go on or not.
As part of the rework, the message for FW not found has been cleaned up
to be more readable.
While at it, drop the comment about xe_uc_fw_init() since the code has
been reworked and the statement no longer applies.
drm/xe/device: Convert wait for lmem init into an assert
Prior to lmem init check, driver is waiting for the pcode uncore_init
status. uncore_init status will be flagged after the complete boot and
initialization of the SoC by the pcode. uncore_init confirms that lmem
init and mmio unblock has been already completed.
It makes no sense to check for lmem init after the pcode uncore_init
check. So change the wait for lmem init check into an assert which
confirms lmem init is set.
A careful inspection of __xe_ggtt_insert_bo_at() shows that
the ggtt_node can always be seen as inserted from xe_bo.c
due to the way error handling is performed.
The checks are also a little bit too paranoid, since we
never create a bo with ggtt_node[id] initialised but not
inserted into the GGTT, which can be seen by looking at
__xe_ggtt_insert_bo_at()
Additionally, the size of the GGTT is never bigger than 4 GB,
so adding a check at that level is incorrect.
drm/xe: Convert xe_fb_pin to use a callback for insertion into GGTT
The rotation details belong in xe_fb_pin.c, while the operations involving
GGTT belong to xe_ggtt.c. As directly locking xe_ggtt etc results in
exposing all of xe_ggtt details anyway, create a special function that
allocates a ggtt_node, and allow display to populate it using a callback
as a compromise.
drm/xe: Start using ggtt->start in preparation of balloon removal
Instead of having ggtt->size point to the end of ggtt, have ggtt->size
be the actual size of the GGTT, and introduce ggtt->start to point to
the beginning of GGTT.
This will allow a massive cleanup of GGTT in case of SRIOV-VF.
drm/xe/mert: Use local mert variable to simplify the code
There is no need to always refer to MERT data using tile pointer.
Use of local mert pointer will simplify the code and make it look
like other existing MERT function.
Michal Wajdeczko [Sun, 11 Jan 2026 21:38:47 +0000 (22:38 +0100)]
drm/xe/mert: Always refer to MERT using xe_device
There is only one MERT instance and while it is located on the root
tile, it is safer to refer to it using xe_device rather than xe_tile.
This will also allow to align signature with other MERT function.
Matthew Brost [Sat, 10 Jan 2026 01:27:39 +0000 (17:27 -0800)]
drm/xe: Avoid toggling schedule state to check LRC timestamp in TDR
We now have proper infrastructure to accurately check the LRC timestamp
without toggling the scheduling state for non-VFs. For VFs, it is still
possible to get an inaccurate view if the context is on hardware. We
guard against free-running contexts on VFs by banning jobs whose
timestamps are not moving. In addition, VFs have a timeslice quantum
that naturally triggers context switches when more than one VF is
running, thus updating the LRC timestamp.
For multi-queue, it is desirable to avoid scheduling toggling in the TDR
because this scheduling state is shared among many queues. Furthermore,
this change simplifies the GuC state machine. The trade-off for VF cases
seems worthwhile.
v5:
- Add xe_lrc_timestamp helper (Umesh)
v6:
- Reduce number of tries on stuck timestamp (VF testing)
- Convert job timestamp save to a memory copy (VF testing)
v7:
- Save ctx timestamp to LRC when start VF job (VF testing)
projects / thirdparty / kernel / linux.git / log
