Current implementation of generate_reclaim_entry() overlooks some
differences between the different page implementations: address masking
and compact 64K page handling.
Address masking of each leaf varies depending on the leaf entry size.
generate_reclaim_entry() is using XE_PTE_ADDR_MASK [51:12] for all leaf
entries. For 2MB PTEs, bit 12 (PAT) is part of the flags so the old mask
corrupts the physical address extraction.
64K pages can be represented as PS64 and a compact PT, which the latter
was not handled. Compact pages aren't walked by the unbind walker, so we
separately walk through the compact PT to ensure none of the leaf 64K
PTEs are dropped. Previously, compact PT were causing an abort since it
was considered covered and not descended into.
v2:
- Update 64K entry/unbind walker for 64K compact PT handling. (Matthew)
- Rework calculations of reclamation and address mask size.
- Add new func abstracting the error handling before generating the
reclaim entry.
v3:
- Report finer addr granularity in abort debug print for compact.
(Zongyao)
- Add comments for ADDR_MASK usage. (Zongyao)
- Drop existing phys_addr asserts, the new XE_PAGE_ADDR_MASK clears
bits checked, so redundant asserts. (Sashiko)
- WARN_ON to verify compact pt and edge pt won't be possible.
Fixes: b912138df2 ("drm/xe: Create page reclaim list on unbind")
Assisted-by: Sashiko-Review:gemini-3.1-pro-preview
Cc: stable@vger.kernel.org
Cc: Matthew Auld <matthew.auld@intel.com>
Suggested-by: Zongyao Bai <zongyao.bai@intel.com>
Signed-off-by: Brian Nguyen <brian3.nguyen@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Zongyao Bai <zongyao.bai@intel.com>
Link: https://patch.msgid.link/20260605224257.2194194-2-brian3.nguyen@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
(cherry picked from commit 669252801a)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
The size calculation for the steered register extarray uses only the
geometry DSS mask (g_dss_mask) to determine the number of entries to
allocate:
total = bitmap_weight(gt->fuse_topo.g_dss_mask, ...) * steer_reg_num;
However, the filling loop uses for_each_dss_steering(), which iterates
over for_each_dss(), defined as the union of g_dss_mask and c_dss_mask
(geometry + compute DSS). On platforms with compute-only DSS bits, the
loop writes past the allocated buffer, corrupting adjacent slab objects.
This manifests as list_del corruption and SLUB redzone overwrites during
drm_managed_release on device unbind, since the overflow corrupts the
drmres list_head of neighboring allocations.
Fix by computing the allocation size using the union of both DSS masks,
matching the iteration pattern of for_each_dss_steering().
--
v2:
- use bitmap_weighted_or() (Zhanjun)
Fixes: b170d696c1 ("drm/xe/guc: Add XE_LP steered register lists")
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/8049
Cc: Zhanjun Dong <zhanjun.dong@intel.com>
Cc: stable@vger.kernel.org
Assisted-by: GitHub-Copilot:claude-opus-4.6
Reviewed-by: Zhanjun Dong <zhanjun.dong@intel.com>
Link: https://patch.msgid.link/20260612070401.543305-2-tejas.upadhyay@intel.com
Signed-off-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
(cherry picked from commit 0a78a44f49)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Set the TTM device beneficial_order to 9 (2M), which is the sweet
spot for Xe when attempting reclaim on system memory BOs, as it matches
the large GPU page size. This ensures reclaim is attempted at the most
effective order for the driver.
This fixes an issue where an order-10 (4M) allocation cannot be found
despite an abundance of memory. The 4M allocation triggers reclaim,
unnecessarily evicting the working set and hurting performance. Since
the TTM infrastructure was introduced recently, we are tagging the TTM
patch as the Fixes target, even though this resolves an Xe-side problem.
Fixes: 7e9c548d37 ("drm/ttm: Allow drivers to specify maximum beneficial TTM pool size")
Cc: stable@vger.kernel.org
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://patch.msgid.link/20260611235844.3725147-1-matthew.brost@intel.com
(cherry picked from commit 0d81db90d3)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
When building with 'make M=drivers/gpu/drm/xe modules', kbuild invokes
scripts/Makefile.build with obj=., causing $(obj) to expand to '.'.
Make normalizes './xe_gen_wa_oob' to 'xe_gen_wa_oob' when constructing
the $^ automatic variable (target name normalization), so the recipe
command becomes just 'xe_gen_wa_oob ...' without any path prefix, and
the shell cannot find the tool.
Fix by replacing $^ with explicit $(obj)/xe_gen_wa_oob and
$(src)/<rules-file> references in both wa_oob recipe commands.
In recipe strings, make does not apply target name normalization, so
$(obj)/xe_gen_wa_oob correctly expands to './xe_gen_wa_oob' and the
shell can execute it. This matches the pattern already used by other
DRM drivers (e.g. radeon's mkregtable).
Fixes: f037e0b78e ("drm/xe: add xe_device_wa infrastructure")
Cc: Matt Atwood <matthew.s.atwood@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: intel-xe@lists.freedesktop.org
Assisted-by: GitHub_Copilot:claude-sonnet-4.6
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patch.msgid.link/20260604074501.172129-1-thomas.hellstrom@linux.intel.com
(cherry picked from commit 3a11a63cc1)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
A job that GuC never scheduled (never started) indicates a GuC
scheduling failure; previously such jobs were silently errored out
instead of triggering a GT reset to recover. Trigger a GT reset and
resubmit them, but only when the queue was not already killed or banned:
an unstarted job on an already banned queue is the ban working as
intended and must neither clear the ban nor kick off a reset, otherwise
a banned userspace queue could be resurrected and spam GT resets.
Kernel queues are always recovered this way and wedge the device once
recovery attempts are exhausted, since kernel work must not silently
fail. A started job that times out on a userspace VM bind queue stays
banned rather than being reset and retried.
The queue is banned early in the timeout handler to signal the G2H
scheduling-done handler so it wakes the disable-scheduling waiter;
without it the waiter sleeps the full 5s timeout. When a reset is
warranted the ban is cleared before rearming so that
guc_exec_queue_start() can resubmit jobs after the GT reset - a
still-banned queue would block resubmission and cause an infinite TDR
loop. The already-banned case is gated out before this point via
skip_timeout_check, so it is unaffected.
v2: (Himal) Do it for any queue type, not just kernel/migration
v3: - (Sashiko and Sanjay): don't clear the ban / GT reset for already
killed/banned queues on unstarted-job timeout
- Update commit message
- (Matt) Add Fixes tag
Fixes: fe05cee4d9 ("drm/xe: Don't short circuit TDR on jobs not started")
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Sanjay Yadav <sanjay.kumar.yadav@intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Assisted-by: GitHub-Copilot:claude-sonnet-4.6
Assisted-by: GitHub-Copilot:claude-opus-4.8
Tested-by: Sanjay Yadav <sanjay.kumar.yadav@intel.com>
Reviewed-by: Sanjay Yadav <sanjay.kumar.yadav@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Link: https://patch.msgid.link/20260610152548.404575-3-rodrigo.vivi@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
(cherry picked from commit b1107d085e)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
xe_range_fence_insert() acquires a reference on fence via
dma_fence_get() and stores it in rfence->fence. It then calls
dma_fence_add_callback() and handles two cases: when the callback
is successfully registered (err == 0) the fence is transferred to
the tree for later cleanup; when the fence is already signaled
(err == -ENOENT) it manually drops the extra reference with
dma_fence_put(fence).
However, dma_fence_add_callback() can fail with other errors
(e.g. -EINVAL) and in that case the code falls through to the free:
label without releasing the acquired reference, leaking it.
Fix the leak by adding an else branch that calls dma_fence_put()
before jumping to free: for any error other than -ENOENT.
Fixes: 845f64bdbf ("drm/xe: Introduce a range-fence utility")
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260610172705.3450560-1-matthew.brost@intel.com
(cherry picked from commit 98c4a42012)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Context-based TLB invalidation currently selects only scheduling-active
exec queues via q->ops->active(). During rebind flows, queues may be
suspended (or transitioning through resume) while still owning valid
translations, causing them to be skipped from invalidation and leading
to missed TLB invalidations on LR rebinds.
The underlying issue is a TOCTOU: q->guc->state bits are flipped lock-free
from enable_scheduling(), disable_scheduling{,_deregister}(), the
suspend/resume sched-msg handlers, handle_sched_done(), and
guc_exec_queue_stop(); nothing in send_tlb_inval_ctx_ppgtt() serializes
against them, so any state-based predicate can race.
Include all the registered queues so that TLB invalidations are not
missed. This is race-free because list membership on vm->exec_queues.list
is stable under vm->exec_queues.lock held by the caller. The performance
impact is expected to be minimal and harmless. If it does turn out to be
a concern, we can come back with a race-safe solution to ignore certain
queues.
Fixes: 6cdaa5346d ("drm/xe: Add context-based invalidation to GuC TLB invalidation backend")
Assisted-by: Claude:claude-opus-4.6
Suggested-by: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Signed-off-by: Tangudu Tilak Tirumalesh <tilak.tirumalesh.tangudu@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260608162745.338725-2-tilak.tirumalesh.tangudu@intel.com
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
(cherry picked from commit aa625e1e9f)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Hardware errors should be logged with HW_ERR prefix. Make them
consistent with existing logs.
Fixes: 01aab7e1c9 ("drm/xe/xe_hw_error: Add support for PVC SoC errors")
Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
Link: https://patch.msgid.link/20260602044919.702209-5-raag.jadav@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
(cherry picked from commit ad60a618c4)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
cleanup_node_param() is not registered for previous node in case of counter
allocation failure, which results in stale memory of previous node that
isn't cleaned up on unwind. Add per node cleanup action which guarantees
cleanup on unwind and also simplifies the cleanup logic.
Fixes: b40db12b54 ("drm/xe/xe_drm_ras: Add support for XE DRM RAS")
Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
Link: https://patch.msgid.link/20260602044919.702209-4-raag.jadav@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
(cherry picked from commit 67fc5543d8)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
cleanup_node_param() is not registered for previous node in case of counter
allocation failure, which results in stale memory of previous node that
isn't cleaned up on unwind. Fix this using drm managed allocation, which is
guaranteed to be cleaned up on unwind.
Fixes: b40db12b54 ("drm/xe/xe_drm_ras: Add support for XE DRM RAS")
Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
Link: https://patch.msgid.link/20260602044919.702209-3-raag.jadav@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
(cherry picked from commit 58d77c77ea)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Return early in submit path when the multi-queue primary exec
queue is suspended to avoid submitting while suspended.
v2: Remove idle_skip_suspend fix as that feature is being
reverted here https://patchwork.freedesktop.org/series/167262/
Fixes: bc5775c592 ("drm/xe/multi_queue: Add GuC interface for multi queue support")
Cc: stable@vger.kernel.org # v7.0+
Assisted-by: GitHub-Copilot:claude-sonnet-4.6
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Signed-off-by: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Link: https://patch.msgid.link/20260603233946.863663-2-niranjana.vishwanathapura@intel.com
(cherry picked from commit b7fb55cc33)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
In the schedule-disable done path for suspend, we
signal the suspend fence before clearing pending_disable.
That wakeup can let suspend_wait complete and resume be queued
immediately. The resume path may then reach enable_scheduling()
while pending_disable is still set and hit the
!exec_queue_pending_disable(q) assertion.
Fix this by clearing pending_disable before signaling
the suspend fence, so any resumed transition observes a
consistent state.
Fixes: 87651f31ae ("drm/xe/guc_submit: fix race around suspend_pending")
Cc: stable@vger.kernel.org # v7.0+
Signed-off-by: Tangudu Tilak Tirumalesh <tilak.tirumalesh.tangudu@intel.com>
Reviewed-by: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patch.msgid.link/20260603065217.3131066-3-tilak.tirumalesh.tangudu@intel.com
(cherry picked from commit 4b1ae138b0)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
This reverts commit 8533051ce9.
The idle-skip optimization bypasses GuC suspend, so the GPU may not
perform the context switch that flushes TLB entries for invalidated
userptr VMAs. In LR/preempt-fence VM mode, this can lead to missed TLB
invalidation and page faults during userptr invalidation tests.
Restore unconditional schedule toggling on suspend so the context-switch
TLB flush is always performed.
This optimization will be reintroduced with a fix that does not skip
suspend in LR/preempt-fence VM mode.
Fixes: 8533051ce9 ("drm/xe: Skip exec queue schedule toggle if queue is idle during suspend")
Cc: stable@vger.kernel.org # v7.0+
Suggested-by: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Signed-off-by: Tangudu Tilak Tirumalesh <tilak.tirumalesh.tangudu@intel.com>
Reviewed-by: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patch.msgid.link/20260603065217.3131066-2-tilak.tirumalesh.tangudu@intel.com
(cherry picked from commit 6a1e7934d9)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
This reverts commit 4e88de313f.
The early GuC FW definition meant for our CI branch was accidentally
merged to the drm-xe-next branch instead. This GuC FW will never be
released to linux-firmware, so we do not want the definition to be
available in the mainline Linux codebase.
Fixes: 4e88de313f ("drm/xe/nvls: Define GuC firmware for NVL-S")
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Julia Filipchuk <julia.filipchuk@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: stable@vger.kernel.org # v7.0+
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patch.msgid.link/20260529193558.185436-11-daniele.ceraolospurio@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
(cherry picked from commit 65b8e0ac86)
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
UAPI Changes:
- amdxdna: Revert read-only user-pointer BO mappings.
- panthor: Add eviction and reclaim info to fdinfo.
Cross-subsystem Changes:
- Convert DMA-buf system and cma heap allocators to module.
Core Changes:
- Cleanup driver misuses of drm/exec.
Driver Changes:
- Add LG LP129WT232166, AM-1280800W8TZQW-T00H, NEC NL6448BC33-70C,
Riverdi RVT70HSLNWCA0 and RVT101HVLNWC00 panels.
- Add support for RZ/T2H SoC to renesas.
- Add cursor plane support to verisilicon.
- Support DVI outputs in ite-it66121 bridge.
- Assorted bugfixes, docbook updates and improvements to ivpu, tegra,
host1x, nouveau.
- Add DSC quirk for ASUS DC301 USB-C dock.
- Use drm client buffer for tegra framebuffer.
- Add support for GA100 to nouveau.
-----BEGIN PGP SIGNATURE-----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=9KcG
-----END PGP SIGNATURE-----
Merge tag 'drm-misc-next-2026-05-28' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next
drm-misc-next for v7.2-rc1:
UAPI Changes:
- amdxdna: Revert read-only user-pointer BO mappings.
- panthor: Add eviction and reclaim info to fdinfo.
Cross-subsystem Changes:
- Convert DMA-buf system and cma heap allocators to module.
Core Changes:
- Cleanup driver misuses of drm/exec.
Driver Changes:
- Add LG LP129WT232166, AM-1280800W8TZQW-T00H, NEC NL6448BC33-70C,
Riverdi RVT70HSLNWCA0 and RVT101HVLNWC00 panels.
- Add support for RZ/T2H SoC to renesas.
- Add cursor plane support to verisilicon.
- Support DVI outputs in ite-it66121 bridge.
- Assorted bugfixes, docbook updates and improvements to ivpu, tegra,
host1x, nouveau.
- Add DSC quirk for ASUS DC301 USB-C dock.
- Use drm client buffer for tegra framebuffer.
- Add support for GA100 to nouveau.
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patch.msgid.link/ef65f43c-becf-473c-a5cc-203fdfecd491@linux.intel.com
- Fix Xe oops in suspend/shutdown when display was disabled (Jani)
Display in general:
- More general refactor towards display separation (Jani)
- Preparation for fix Adaptive-Sync SDP for PR with Link ON + Auxless-ALPM (Ankit)
- PSR related fixes and improvements (Jouni)
- Use polling when irqs are unavailable (Michal)
- Split bandwidth params into platform- and display-IP-specific structs (Gustavo)
- Revert "drm/i915/backlight: Remove try_vesa_interface" (Suraj)
- Casf & scaler refactoring (Michal)
- Add support for pipe background color (Maarten)
- General clean-ups (Maarten)
- Sanitize DP link capability change handling (Imre)
- Multiple BW QGV fixes (Ville)
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEbSBwaO7dZQkcLOKj+mJfZA7rE8oFAmoYRLYACgkQ+mJfZA7r
E8rtVQgArEfJhvSxvoALAQjfeM/mjWHj0AThfCxROed2Z1TO5WN+BB9hXzwGwEqA
EE6TnKhDEQZy22a/t/iNhDDUy1UTYifMbfFAsdVz7NYfXKZpkRdJ4/mZk6MVEcrd
jEbrIw8ZQ7SPTvhVdKn6zWvLzPskrJqzfIS4JFYJn61MepfD+QhOayoUxEU9gTjZ
+SNSHu0lQSpwYwz+0AV65CkDwSbFZDqMLtDd8MxI0b4X8kjX1D9MH+QNu2wYnk7i
9hqcErJ/jH/hMCNHwCIdPh6sfli8qpN7GbAdDcgkA+/E0hsJBkdJe2cSDGxUA/Ws
76P8PjtZ3sGbq7vPL/Db6t0lY1yVJQ==
=cTLp
-----END PGP SIGNATURE-----
Merge tag 'drm-intel-next-2026-05-28' of https://gitlab.freedesktop.org/drm/i915/kernel into drm-next
Xe related:
- Fix Xe oops in suspend/shutdown when display was disabled (Jani)
Display in general:
- More general refactor towards display separation (Jani)
- Preparation for fix Adaptive-Sync SDP for PR with Link ON + Auxless-ALPM (Ankit)
- PSR related fixes and improvements (Jouni)
- Use polling when irqs are unavailable (Michal)
- Split bandwidth params into platform- and display-IP-specific structs (Gustavo)
- Revert "drm/i915/backlight: Remove try_vesa_interface" (Suraj)
- Casf & scaler refactoring (Michal)
- Add support for pipe background color (Maarten)
- General clean-ups (Maarten)
- Sanitize DP link capability change handling (Imre)
- Multiple BW QGV fixes (Ville)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patch.msgid.link/ahhFE5Co5PdZQKG8@intel.com
- drm/xe: Move xe_uc_fw_abi.h to abi/ (Michal Wajdeczko)
- drm/xe: Restore IDLEDLY regiter on engine reset (Balasubramani Vivekanandan)
- drm/xe/pm: Do early initialization in init_early() (Michal Wajdeczko)
- drm/xe/pm: Don't access device in init_early() (Michal Wajdeczko)
- drm/xe: Separate early xe_device initialization (Michal Wajdeczko)
- drm/xe: Move xe->info.devid|revid initialization (Michal Wajdeczko)
- drm/xe: Move xe->info.force_execlist initialization (Michal Wajdeczko)
- drm/xe: Drop unused param from xe_device_create() (Michal Wajdeczko)
- drm/xe: Use raw device ID to find sub-platform descriptor (Michal Wajdeczko)
- drm/xe: Assign queue name in time for drm_sched_init (Tvrtko Ursulin)
- drm/xe/rtp: Implement a structured parser for rule matching (Gustavo Sousa)
- drm/xe/rtp: Fully parse the ruleset (Gustavo Sousa)
- drm/xe/rtp: Extract rule_match_item() (Gustavo Sousa)
- drm/xe/rtp: Do not break parsing when missing context (Gustavo Sousa)
- drm/xe/rtp: Don't short-circuit to false in or-yes case (Gustavo Sousa)
- drm/xe/rtp: Drop rule matching cases from rtp_to_sr_cases and rtp_cases (Gustavo Sousa)
- drm/xe/rtp: Write kunit test cases specific for rule matching (Gustavo Sousa)
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRskUM7w1oG5rx2IZO4FpNVCsYGvwUCahg6XQAKCRC4FpNVCsYG
v1tjAP9yDWHxY1ZLkfg3boEWRrWy5HidWRYXRPEcHFZy9asj9QD/YtTu+ZBEVQuY
p4pyZSYwgDQYFSQI7ajF2yWj2P1KWgs=
=njjy
-----END PGP SIGNATURE-----
Merge tag 'drm-xe-next-2026-05-28' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next
Driver Changes:
- drm/xe: Move xe_uc_fw_abi.h to abi/ (Michal Wajdeczko)
- drm/xe: Restore IDLEDLY regiter on engine reset (Balasubramani Vivekanandan)
- drm/xe/pm: Do early initialization in init_early() (Michal Wajdeczko)
- drm/xe/pm: Don't access device in init_early() (Michal Wajdeczko)
- drm/xe: Separate early xe_device initialization (Michal Wajdeczko)
- drm/xe: Move xe->info.devid|revid initialization (Michal Wajdeczko)
- drm/xe: Move xe->info.force_execlist initialization (Michal Wajdeczko)
- drm/xe: Drop unused param from xe_device_create() (Michal Wajdeczko)
- drm/xe: Use raw device ID to find sub-platform descriptor (Michal Wajdeczko)
- drm/xe: Assign queue name in time for drm_sched_init (Tvrtko Ursulin)
- drm/xe/rtp: Implement a structured parser for rule matching (Gustavo Sousa)
- drm/xe/rtp: Fully parse the ruleset (Gustavo Sousa)
- drm/xe/rtp: Extract rule_match_item() (Gustavo Sousa)
- drm/xe/rtp: Do not break parsing when missing context (Gustavo Sousa)
- drm/xe/rtp: Don't short-circuit to false in or-yes case (Gustavo Sousa)
- drm/xe/rtp: Drop rule matching cases from rtp_to_sr_cases and rtp_cases (Gustavo Sousa)
- drm/xe/rtp: Write kunit test cases specific for rule matching (Gustavo Sousa)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Link: https://patch.msgid.link/ahg7tgWh8djEngjM@fedora
Backmerging to get GEM LRU fixes from commit 379e8f1c ("drm/gem: Make
the GEM LRU lock part of drm_device") and other updates from v7.1-rc5.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Wa_16023105232 programs the register IDLEDLY. The register is reset
whenever the engine is reset. Therefore it should be added to the GuC
save-restore register list for it to be restored after reset.
Fixes: 7c53ff050b ("drm/xe: Apply Wa_16023105232")
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patch.msgid.link/20260522163531.1365540-2-balasubramani.vivekanandan@intel.com
Signed-off-by: Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>
There is no need nor gain in splitting mutex or list initializations
between two init functions as all of this is just pure software state
and all this could be done at once.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-8-michal.wajdeczko@intel.com
We should separate software-only state initialization from anything
else that requires access to the device's hardware. Extract d3cold
capability detection into a new function. Add simple kernel-doc for
updated functions here.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-7-michal.wajdeczko@intel.com
We would like to initialize more of the xe_device struct also from
the kunit code, as it should be safe to use most of the generic drm
or xe components without doing any additional tweaks. Separate early
xe initialization code to a new function, so it can be reused.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-6-michal.wajdeczko@intel.com
The xe_info_init_early() is a place where we initialize those of
the xe->info fields that do not require any additional hardware
probes. Move the initialization of the devid/revid also there, but
to avoid breaking the kunit helper, which also calls this function,
keep their initialization separate in sub-function so we can easily
stub it when running the kunit test.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-5-michal.wajdeczko@intel.com
The xe_info_init_early() is a place where we initialize those of
the xe->info fields that do not require any additional hardware
probes. Move the initialization of the force_execlist flag there.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-4-michal.wajdeczko@intel.com
We never used or need anything from the struct pci_device_id there.
And while around, add simple kernel-doc for this function.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-3-michal.wajdeczko@intel.com
We don't need the partially initialized xe_device pointer to
find the sub-platform descriptor, as for the descriptor lookup
only the device ID is required and it can be obtained directly
from the pci_dev.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Raag Jadav <raag.jadav@intel.com>
Link: https://patch.msgid.link/20260526195452.20545-2-michal.wajdeczko@intel.com
It's confusing that intel_display_power.[ch] exposes two groups of
interfaces, one named intel_power_domains_*() and one
intel_display_power_*().
Unify on the latter, based on the file name, but also because it's more
generic. This makes the caller side easier to follow.
Cc: Imre Deak <imre.deak@intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Link: https://patch.msgid.link/8fae4b0e3476aeffb0164215b7e0f0ae1d825f72.1779800132.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
xe_display_flush_cleanup_work() is a bit of an oddball function in xe
display code. There shouldn't be anything this specific or xe
specific. While I'm not sure what the correct refactor for the function
should be, move it to shared display code for starters, next to the
eerily similar but slightly different intel_has_pending_fb_unpin() that
is only called from i915 core.
The main goal here is to unblock some refactors on
for_each_intel_crtc().
v2: Add FIXME comment (Ville)
Acked-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patch.msgid.link/20260525110553.651208-1-jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Currently the queue name is only assigned after the drm scheduler instance
has been created. This loses information with all logging or debug
workqueue facilities so lets re-order things a bit so the name gets
assigned in time.
To be able to assign a GuC ID early we split the allocation into
reservation and publish phases.
First, with the submission state lock held, we reserve the ID in the GuC
ID manager, which serves as an authoritative source of truth. Then we can
drop the lock and reserve entries in the exec queue lookup XArray. This
can be lockless since the NULL entries are invisible both to the kernel
and userspace. Only after the queue has been fully created we replace the
reserved entries with the queue pointer, which can be done locklessly for
single width queues.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patch.msgid.link/20260523103418.61832-1-tvrtko.ursulin@igalia.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Drivers were accessing this drm_exec member directly.
While that may seem harmless, it will require action if
the drm_exec utility is made a subclass of a dma-resv transaction
utility as outlined in the cover-letter.
Provide an accessor, drm_exec_ticket() to avoid that.
v2:
- Fix amdgpu compile error (Intel CI)
- Update the commit message.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patch.msgid.link/20260520101616.41284-5-thomas.hellstrom@linux.intel.com
The xe driver was using the drm_exec retry pointer directly to
restart the locking loop after out-of-memory errors. This is
relying on undocumented behaviour.
Instead add a drm_exec_retry() macro that can be used in this
situation, and that also warns if the struct drm_exec is
not newly (re-)initialized.
Use that macro in xe.
v2:
- Only allow if the drm_exec context is newly initialized.
(Christian)
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patch.msgid.link/20260520101616.41284-4-thomas.hellstrom@linux.intel.com
Nobody makes any use of it. Possible internal future users can
instead use the _index variable. External users shouldn't use
it since the array it's pointing into is internal drm_exec state.
v2:
- Use a unique id for the loop variable (Christian)
Assisted-by: GitHub Copilot:claude-sonnet-4.6
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patch.msgid.link/20260520101616.41284-2-thomas.hellstrom@linux.intel.com
The current unwritten grammar for RTP rules is as follows:
rules = disjunction;
disjunction = conjunction, { "OR", conjunction };
conjunction = single_rule, { single_rule };
(* the AND operator is implicit *)
single_rule = ? GRAPHICS_VERSION(...), MEDIA_VERSION(...),
FUNC(...), etc ?
While rule_matches() currently works for the grammar above, it doesn't
easily resemble it. Let's replace it with an implementation that is
structured in a way to resemble the grammar.
Such a new implementation, although a bit more verbose, is arguably
easier to reason about and to adapt to any extension we do to the
grammer in the future.
Also take this opportunity to update the kernel-doc for XE_RTP_RULES()
to include the grammar, so that it is not unwritten anymore.
v2:
- Include the grammar in the code documentation. (Matt)
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Violet Monti <violet.monti@intel.com>
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-7-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
The function rule_matches() short-circuits evaluation of the implicit
conjunctions (each substring of rules not containing OR) and the
explicit disjunctions (implicit conjunctions joined by OR). In other
words:
- in a conjunction, once a rule evaluate to false, we skip to the next
OR (if any) to evaluate the next conjunction;
- in a disjunction, once a conjunction evaluates to true, we return
true and skip evaluating all the remaining rules.
While this behavior results in a correct logical value, due to how the
"OR" short-circuiting is implemented, it has the side-effect that rule
set does not get fully "parsed", allowing incomplete constructs like
(rule1, OR) to evaluate to true when rule1 is true. We should treat
such constructs as invalid and treat them the same way we do for stuff
like (OR, rule1).
As such, update rule_matches() to "parse" the whole rule set, and that
while keeping the short-circuit aspect of evaluation. With that, we can
fix the FIXME test cases that cover that behavior.
v2:
- Do not change short-circuit *evaluation* behavior. (Matt)
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Violet Monti <violet.monti@intel.com>
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-6-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
The current logic in rule_matches() mixes individual rule matching
with the logic necessary for handling OR operations. Let's simplify
rule_matches() to focus on the latter by extracting individual rule
matching into a separate function called rule_match_item().
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-5-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
With the current implementation, the RTP framework will cause parsing of
the rule set to be interrupted if one rule requires a context item (gt
or hwe) that is missing (i.e. when the value is NULL).
This is arguably a semantic error instead of a syntactic one, meaning
that RTP should not interrupt parsing the rules. With the current
behavior, we would miss detecting other errors that could appear in the
remaining rules and could also prevent valid rules joined by "OR" from
being evaluated.
Make sure that we do not stop parsing the rule set when detecting
missing context and let's add rtp_rules_test_cases to reflect that.
v2:
- Add "missing-context" in the test case names to indicate that those
are about rules that are missing the necessary context. (Matt)
- Rebase: treat the new match type XE_RTP_MATCH_PLATFORM_STEP in the
same way when the platform is missing step information.
Reviewed-by: Matt Roper <matthew.d.roper@intel.com> # v1
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-4-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
While RTP processing evaluates true on the "yes-or" case (i.e. a
conjunction of rules that evaluate to true followed by an "OR" without
the right hand operand), it does not on the "or-yes" one.
Both cases are considered malformed and could be a result of someone
dropping checks deemed not necessary anymore and forgetting to drop the
superfluous "OR". Nevertheless, we should aim for consistency, and
having the "or-yes" case also evaluating to true while also causing a
warning seems reasonable. So let's do that.
The "or-yes" pattern being evaluated to false comes from the fact that
that we unconditionally short-circuit upon finding XE_RTP_MATCH_OR on
the outer loop. We should only do that if the preceding conjunction of
rules evaluated to true (meaning that rcount must be non-zero) and
continue the evaluation otherwise.
Do that and also add extra test cases to validate the short-circuiting
behavior.
Notice that some of the new test cases have a "FIXME" comment, which
comes from the fact that we are unable to detect syntax errors after the
short-circuit point. That is going to be fixed in a follow-up change.
Link: https://lore.kernel.org/intel-xe/871pfw4lo9.fsf@intel.com/
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Violet Monti <violet.monti@intel.com>
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-3-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
The kunit test cases for the RTP framework are currently separated into
three groups:
(1) rtp_rules_cases:
Those to verify rule matching logic.
(2) rtp_to_sr_cases:
Those to verify generation of save/restore tables from RTP tables.
(3) rtp_cases
Those to verify processing of RTP tables without save/restore action
associated, which are used for OOB workarounds.
Today we have some cases in (2) and (3) that are actually meant to
verify rule matching logic. Now that we have (1), let's cleanup (2)
and (3) so that they become focused on their main objectives.
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-2-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
The kunit test cases for the RTP framework are currently separated into
those that validate xe_rtp_process_to_sr() and those that validate
xe_rtp_process(). In both of them, we also have mixed stuff to validate
rule matching functionality, which should rather be done in a separate
test case group.
Let's create such a group, specific for validating rule matching, and
also add an initial set of cases. In an upcoming change, we will do a
cleanup of the other groups by migrating those cases intended for rule
matching to this new group.
v2:
- s/no-yes-or-no-yes/no-yes-or-yes-no/ (Matt)
- Drop leftover include of <kunit/test.h>.
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patch.msgid.link/20260522-rtp-rule-parser-v3-1-0c51039899f4@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
In xe_oa_stream_open_ioctl(), when param.exec_q->width > 1 the
function returns -EOPNOTSUPP directly, skipping the existing
err_exec_q cleanup path. The exec_queue reference obtained by
xe_exec_queue_lookup() is leaked.
The exec queue holds a reference on the xe_file, which is only
dropped during queue teardown. The leaked lookup ref is not on
the file's exec_queue xarray, so file close cannot release it.
This keeps both the exec queue and the file private state pinned
indefinitely.
Jump to err_exec_q instead of returning directly so the reference
is released.
Fixes: f0ed39830e ("xe/oa: Fix query mode of operation for OAR/OAC")
Assisted-by: Claude:claude-opus-4.6
Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Link: https://patch.msgid.link/20260514203210.593488-1-shuicheng.lin@intel.com
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
Since addition of the MSI-X support, we mostly rely on the offset
calculations done by XE_MEMIRQ_STATUS_OFFSET. We don't use this
separate map pointing to the first status page anymore.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-10-michal.wajdeczko@intel.com
It is used occasionally and iosys_map_wr() helper takes an offset
parameter anyway. There is no extra benefit to keep a separate map.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-9-michal.wajdeczko@intel.com
When using MSI-X, engines report their source/status on separate
MEMIRQ pages, so we need to dump additional source pages, not just
the first one.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Ilia Levi <ilia.levi@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-8-michal.wajdeczko@intel.com
Instead printing static offset values, print number of allocated
pages and the actual GGTT addresses of the page zero source and
status and address of the common mask vector.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-7-michal.wajdeczko@intel.com
When using MSI-X, we don't have to allocate the largest possible
buffer to accommodate all potential engine instances. Loop through
available engines, find highest engine instance and reduce buffer
size to avoid memory waste.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Ilia Levi <ilia.levi@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-6-michal.wajdeczko@intel.com
We can now drop repeated calculations of the actual IRQ page used
by the engines from our memory based interrupt handler and other
functions.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-5-michal.wajdeczko@intel.com
Introduce and use simple macro to calculate exact location of the
status vector to avoid inline calculation. Fix type for the GuC
source and status MEMIRQ addresses.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-4-michal.wajdeczko@intel.com
There is no need to expose the macros describing memory-based
interrupts page layouts in the .h file as we only use them in
the private code. Move them to the .c file near the kernel-doc.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Ilia Levi <ilia.levi@intel.com>
Link: https://patch.msgid.link/20260518192547.600-3-michal.wajdeczko@intel.com
