| Age | Commit message (Collapse) | Author |
|
When a VRAM buffer object is marked as DONTNEED, like in Mesa, the
driver skips migrating its contents to system memory and instead purges
the backing store during eviction (via xe_ttm_bo_purge).
However, xe_evict_flags() still returns tt_placement (XE_PL_TT) for VRAM
BOs even if they were marked DONTNEED. This causes
ttm_bo_handle_move_mem() to always call ttm_bo_populate() to allocate
destination system pages, only for those pages to be immediately freed
right after when xe_bo_move() calls xe_ttm_bo_purge().
Fix this by changing xe_evict_flags() to return sys_placement
(XE_PL_SYSTEM) for DONTNEED BOs. This causes TTM to skip the population
step while still ensuring that TTM calls xe_bo_move(), so
xe_ttm_bo_purge() still triggers.
v2 (Thomas):
- Add a comment to highlight why sys_placement over purge_placement.
Assisted-by: Copilot:gemini-3.1-pro-preview
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Maarten Lankhorst <dev@lankhorst.se>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260723161542.1220276-2-matthew.auld@intel.com
|
|
xe_bo_move() attaches VF CCS read/write batch buffers (BBs) to a BO
after it transitions NULL/SYSTEM -> TT, and detaches them after it
transitions TT -> SYSTEM. Both operations were done synchronously on
the CPU immediately after building the move's copy/clear fence,
without waiting for that fence to signal. This creates two races with
VF migration:
- Attach happens too late relative to the copy job it is meant to
protect. If the copy job is submitted before the CCS BBs are
attached, a VF migration event that pauses execution mid-copy can
observe partially copied CCS metadata without the attach state
needed to correctly save/restore it.
- Detach happens too early relative to the copy job that moves data
out of TT. The CCS BBs are torn down right after the copy fence is
obtained, while the actual blit may still be in flight. A VF
migration event that pauses execution mid-copy can then race the
save/restore path against the still-running blit, and the CCS BBs
it would need to make sense of the paused state have already been
removed.
Fix both races:
- Move the attach call to before the copy/clear job is submitted, so
the CCS BBs are already registered by the time the copy runs. On
attach failure, unwind and bail out of the move. xe_migrate_ccs_rw_copy()
now takes the destination resource explicitly, since bo->ttm.resource
is not updated to the new resource until after the move commits.
- Detach only after explicitly waiting for the copy fence to signal,
instead of tearing down the CCS BBs immediately after obtaining it.
While here, also fix xe_sriov_vf_ccs_attach_bo() to properly unwind and
propagate errors: the per-context loop previously never broke out on
error, silently discarding earlier failures. Unwind by clearing each
attached context directly via xe_migrate_ccs_rw_copy_clear() instead of
reusing xe_sriov_vf_ccs_detach_bo(), which requires both contexts to be
attached before it will clean up either one.
Fixes: 864690cf4dd6 ("drm/xe/vf: Attach and detach CCS copy commands with BO")
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Satyanarayana K V P <satyanarayana.k.v.p@intel.com>
Assisted-by: GitHub_Copilot:claude-sonnet-5
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Acked-by: Satyanarayana K V P <satyanarayana.k.v.p@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260714062440.3421225-1-matthew.brost@intel.com
|
|
Sync some i915/display changes
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
|
|
An imported dma-buf BO is created as a ttm_bo_type_sg BO whose
reservation object is the exporter's dma_buf->resv. The importer,
however, only takes a dma-buf reference after a successful
dma_buf_dynamic_attach(). Until then nothing keeps the exporter alive,
so if the exporter is freed while the BO still references its resv, a
later access to that resv is a use-after-free:
Oops: general protection fault, probably for non-canonical address
0x6b6b6b6b6b6b6b9c
Workqueue: ttm ttm_bo_delayed_delete [ttm]
RIP: 0010:mutex_can_spin_on_owner+0x3f/0xc0
This can be reached on two paths:
- dma_buf_dynamic_attach() fails, or
- ttm_bo_init_reserved() fails during BO creation.
In both cases the BO already has bo->base.resv pointing at the exporter
resv, and sg BOs are always torn down via ttm_bo_delayed_delete(), which
locks bo->base.resv asynchronously - potentially after the exporter has
been freed.
Take the dma-buf reference in xe_bo_init_locked(), before
ttm_bo_init_reserved(), so it also covers a creation failure there, and
release it in xe_ttm_bo_destroy(). The reference is held for the whole
BO lifetime, keeping the shared resv alive on every path.
v2:
- Reworked the fix to avoid creating the imported sg BO before
dma_buf_dynamic_attach() succeeds.
- Attach with importer_priv == NULL and make invalidate_mappings ignore
incomplete imports.
v3:
- Dropped the xe-side reordering approach since importer_priv must be
valid when dma_buf_dynamic_attach() publishes the attachment.
- Per Christian's suggestion on the v1 thread, keyed the check on
import_attach rather than removing the sg guard entirely.
- Fixes both xe and amdgpu in a single TTM patch.
v4:
- Moved import_attach check to after dma_resv_copy_fences() so fences
are copied before returning for successful imports (Thomas).
- Removed exporter-alive claim from commit message (Thomas).
v5:
- Add drm/xe patch to keep imported sg BOs off the LRU before attach
succeeds; the TTM fix alone is not sufficient for xe if the BO is
already LRU-visible. (Thomas)
v4 patch:
https://patchwork.freedesktop.org/patch/736663/?series=169129&rev=2
- Patch 1 (drm/ttm) carries Christian's Reviewed-by from v4.
v6:
- Reworked the fix based on Thomas' suggestion. Instead of the TTM resv
individualization (v1-v5) plus the xe off-LRU/placement handling (v5),
just hold a dma-buf reference for the imported BO lifetime so the
shared resv can never be freed while the BO still references it.
Single xe patch, no TTM change. (Thomas)
- Take the reference in xe_bo_init_locked() before ttm_bo_init_reserved()
so a TTM creation failure is covered too (Thomas).
- Dropped the v5 series (drm/ttm + drm/xe off-LRU); the off-LRU approach
also regressed in CI BAT via ttm_bo_pipeline_gutting() creating a ghost
BO that outlived the exporter.
Link to v5: https://patchwork.freedesktop.org/series/169984/
v7:
- Move changelog above --- so it stays in the commit message.
- Reorder changelog entries oldest-to-newest. (Thomas)
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/work_items/8023
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Cc: stable@vger.kernel.org
Cc: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Cc: Christian Konig <christian.koenig@amd.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Suggested-by: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Assisted-by: GitHub_Copilot:claude-sonnet-4.6
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Nitin Gote <nitin.r.gote@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260710191027.260160-2-nitin.r.gote@intel.com
|
|
Dropping the _sw_ part from the names was proposed multiple times now and
IIRC people generally agreed with the idea already.
The function requests a fence to signal and triggers some sort of HW
interaction on most backends.
So this is not really software related at all and the callback is already
just named enable_signaling as well.
Just streamline that and use a consistent name everywhere.
Assisted-by: Claude Sonet 4
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Link: https://lore.kernel.org/r/20260624122917.2483-2-christian.koenig@amd.com
|
|
Stolen memory requires physically contiguous allocations for display
scanout and compressed framebuffers. The stolen memory manager was
sharing the gpu_buddy allocator backend with the VRAM manager, but
buddy manages non-contiguous power-of-two blocks making it a poor fit.
Stolen memory also has fundamentally different allocation patterns:
- Allocation sizes are not power-of-two. Since buddy rounds up to the
next power-of-two block size, a ~17MB request can fail even with
~22MB free, because the free space is fragmented across non-fitting
power-of-two blocks.
- Hardware restrictions prevent using the first 4K page of stolen for
certain allocations (e.g., FBC). The display code sets fpfn=1 to
enforce this, but when fpfn != 0, gpu_buddy enables
GPU_BUDDY_RANGE_ALLOCATION mode which disables the try_harder
coalescing path, further reducing allocation success.
This combination caused FBC compressed framebuffer (CFB) allocation
failures on platforms like NVL/PTL. In case of NVL where stolen memory
is ~56MB and the initial plane framebuffer consumes ~34MB at probe time,
leaving ~22MB for subsequent allocations.
Use ttm_range_man_init_nocheck() to set up a drm_mm-backed TTM resource
manager for stolen memory. This reuses the TTM core's ttm_range_manager
callbacks, avoiding duplicate implementations.
Tested on NVL with a 4K DP display: stolen_mm shows a single ~22MB
contiguous free hole after initial plane framebuffer allocation, and
FBC successfully allocates its CFB from that region. The corresponding
IGT was previously skipped and now passes.
v2:
- Clarify that stolen memory requires contiguous allocations (Matt B)
- Properly handle xe_ttm_resource_visible() for stolen instead of
unconditionally returning true (Matt A)
v3:
- Rebase
- Fix xe_display_bo_fbdev_prefer_stolen() to compare in pages, since
ttm_range_manager stores stolen->size in pages not bytes (Matt A)
v4:
- Add kernel-doc for struct xe_ttm_stolen_mgr (Matt B)
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/work_items/7631
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Suggested-by: Matthew Auld <matthew.auld@intel.com>
Assisted-by: GitHub Copilot:claude-sonnet-4.6
Signed-off-by: Sanjay Yadav <sanjay.kumar.yadav@intel.com>
Acked-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Vinod Govindapillai <vinod.govindapillai@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260422125502.3088222-2-sanjay.kumar.yadav@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
|
|
xe_bo_recompute_purgeable_state() walks all VMAs of a BO to determine
whether the BO can be made purgeable. This makes VMA create/destroy and
madvise updates O(n) in the number of mappings.
Replace the walk with BO-local counters protected by the BO dma-resv
lock:
- vma_count tracks the number of VMAs mapping the BO.
- willneed_count tracks active WILLNEED holders, including WILLNEED
VMAs and active dma-buf exports for non-imported BOs.
A DONTNEED BO is promoted back to WILLNEED on a 0->1 transition of
willneed_count. A BO is demoted to DONTNEED on a 1->0 transition only
when it still has VMAs, preserving the previous behaviour where a BO
with no mappings keeps its current madvise state.
PURGED remains terminal, preserving the existing "once purged, always
purged" rule.
Fixes: 4f44961eab84 ("drm/xe/vm: Prevent binding of purged buffer objects")
v2:
- Use early return for imported BOs in all four helpers to avoid
nesting (Matt B).
- Group purgeability state into a purgeable sub-struct on struct
xe_bo (Matt B).
- Reword xe_bo_willneed_put_locked() kernel-doc to explain that a 1->0
transition means all remaining active VMAs are DONTNEED (Matt B).
v3:
- Move DONTNEED/PURGED reject from vma_lock_and_validate() into
xe_vma_create(), gated on attr->purgeable_state == WILLNEED.
Fixes vm_bind bypass and partial-unbind rejection on DONTNEED
BOs (Matt B).
- Drop .check_purged from MAP and REMAP; keep it for PREFETCH and
add a comment why (Matt B).
- Skip BO validation in vma_lock_and_validate() for non-WILLNEED
VMA remnants so cleanup/remap paths do not repopulate
DONTNEED/PURGED BOs.
Suggested-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260506132027.2556046-1-arvind.yadav@intel.com
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
|
|
When XE_BO_FLAG_GGTT_ALL is set without XE_BO_FLAG_GGTT, the function
returns an error without freeing a caller-provided bo, violating the
documented contract that bo is freed on failure.
Add xe_bo_free(bo) before returning the error.
Fixes: 5a3b0df25d6a ("drm/xe: Allow bo mapping on multiple ggtts")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4.6
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260408175255.3402838-3-shuicheng.lin@intel.com
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
|
|
When type is ttm_bo_type_device and aligned_size != size, the function
returns an error without freeing a caller-provided bo, violating the
documented contract that bo is freed on failure.
Add xe_bo_free(bo) before returning the error.
Fixes: 4e03b584143e ("drm/xe/uapi: Reject bo creation of unaligned size")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4.6
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260408175255.3402838-2-shuicheng.lin@intel.com
Signed-off-by: Shuicheng Lin <shuicheng.lin@intel.com>
|
|
Renaming:
The helper function res_to_mem_region is now xe_map_resource_to_region.
Abstraction:
The patch removes instances where block->private was accessed directly
to obtain VRAM region data allowing VRAM manager to own block->private
for future use cases.
V2(MattB): Add more detail about patch also adjust variable placement
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260407055107.2782450-2-tejas.upadhyay@intel.com
Signed-off-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
|
|
Encapsulate TTM purgeable flag updates and shrinker page accounting
into helper functions to prevent desynchronization between the TTM
tt->purgeable flag and the shrinker's page bucket counters.
Without these helpers, direct manipulation of xe_ttm_tt->purgeable
risks forgetting to update the corresponding shrinker counters,
leading to incorrect memory pressure calculations.
Update purgeable BO state to PURGED after successful shrinker purge
for DONTNEED BOs.
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260326130843.3545241-11-arvind.yadav@intel.com
|
|
Don't allow new CPU mmaps to BOs marked DONTNEED or PURGED.
DONTNEED BOs can have their contents discarded at any time, making
CPU access undefined behavior. PURGED BOs have no backing store and
are permanently invalid.
Return -EBUSY for DONTNEED BOs (temporary purgeable state) and
-EINVAL for purged BOs (permanent, no backing store).
The mmap offset ioctl now checks the BO's purgeable state before
allowing userspace to establish a new CPU mapping. This prevents
the race where userspace gets a valid offset but the BO is purged
before actual faulting begins.
Existing mmaps (established before DONTNEED) may still work until
pages are purged, at which point CPU faults fail with SIGBUS.
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260326130843.3545241-9-arvind.yadav@intel.com
|
|
Block CPU page faults to buffer objects marked as purgeable (DONTNEED)
or already purged. Once a BO is marked DONTNEED, its contents can be
discarded by the kernel at any time, making access undefined behavior.
Return VM_FAULT_SIGBUS immediately to fail consistently instead of
allowing erratic behavior where access sometimes works (if not yet
purged) and sometimes fails (if purged).
For DONTNEED BOs:
- Block new CPU faults with SIGBUS to prevent undefined behavior.
- Existing CPU PTEs may still work until TLB flush, but new faults
fail immediately.
For PURGED BOs:
- Backing store has been reclaimed, making CPU access invalid.
- Without this check, accessing existing mmap mappings would trigger
xe_bo_fault_migrate() on freed backing store, causing kernel hangs
or crashes.
The purgeable check is added to both CPU fault paths:
- Fastpath (xe_bo_cpu_fault_fastpath): Returns VM_FAULT_SIGBUS
immediately under dma-resv lock, preventing attempts to
migrate/validate DONTNEED/purged pages.
- Slowpath (xe_bo_cpu_fault): Returns -EFAULT under drm_exec lock,
converted to VM_FAULT_SIGBUS.
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260326130843.3545241-5-arvind.yadav@intel.com
|
|
This allows userspace applications to provide memory usage hints to
the kernel for better memory management under pressure:
Add the core implementation for purgeable buffer objects, enabling
memory reclamation of user-designated DONTNEED buffers during eviction.
This patch implements the purge operation and state machine transitions:
Purgeable States (from xe_madv_purgeable_state):
- WILLNEED (0): BO should be retained, actively used
- DONTNEED (1): BO eligible for purging, not currently needed
- PURGED (2): BO backing store reclaimed, permanently invalid
Design Rationale:
- Async TLB invalidation via trigger_rebind (no blocking
xe_vm_invalidate_vma)
- i915 compatibility: retained field, "once purged always purged"
semantics
- Shared BO protection prevents multi-process memory corruption
- Scratch PTE reuse avoids new infrastructure, safe for fault mode
Note: The madvise_purgeable() function is implemented but not hooked
into the IOCTL handler (madvise_funcs[] entry is NULL) to maintain
bisectability. The feature will be enabled in the final patch when all
supporting infrastructure (shrinker, per-VMA tracking) is complete.
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260326130843.3545241-4-arvind.yadav@intel.com
|
|
Rename XE_BO_FLAG_SCANOUT to XE_BO_FLAG_FORCE_WC so that the usage of the
flag can legitimately be expanded to more than just the actual frame-
buffer objects.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Suggested-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patch.msgid.link/20260324084018.20353-2-tvrtko.ursulin@igalia.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
|
|
XA, new pat_index introduced post xe3p_lpg, is memory shared between the
CPU and GPU is treated differently from other GPU memory when the Media
engine is power-gated.
XA is *always* flushed, like at the end-of-submssion (and maybe other
places), just that internally as an optimisation hw doesn't need to make
that a full flush (which will also include XA) when Media is
off/powergated, since it doesn't need to worry about GT caches vs Media
coherency, and only CPU vs GPU coherency, so can make that flush a
targeted XA flush, since stuff tagged with XA now means it's shared with
the CPU. The main implication is that we now need to somehow flush non-XA
before freeing system memory pages, otherwise dirty cachelines could be
flushed after the free (like if Media suddenly turns on and does a full
flush)
V4: Add comments for L2 flush path
V3(Thomas/MattA/MattR): Restrict userptr with non-xa, then no need to
flush manually
V2(MattA): Expand commit description
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://patch.msgid.link/20260305121902.1892593-7-tejas.upadhyay@intel.com
Signed-off-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
|
|
Backmerging to bring in 7.00-rc3. Important ahead GPU SVM merging THP
support.
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
|
|
Implement handling of VM_BIND(..., DECOMPRESS) in xe_vm_bind_ioctl.
Key changes:
- Parse and record per-op intent (op->map.request_decompress) when the
DECOMPRESS flag is present.
- Use xe_pat_index_get_comp_en() helper to check if a PAT index
has compression enabled via the XE2_COMP_EN bit.
- Validate DECOMPRESS preconditions in the ioctl path:
- Only valid for MAP ops.
- The provided pat_index must select the device's "no-compression" PAT.
- Only meaningful on devices with flat CCS and the required XE2+
otherwise return -EOPNOTSUPP.
- Use XE_IOCTL_DBG for uAPI sanity checks.
- Implement xe_bo_decompress():
For VRAM BOs run xe_bo_move_notify(), reserve one fence slot,
schedule xe_migrate_resolve(), and attach the returned fence
with DMA_RESV_USAGE_KERNEL. Non-VRAM cases are silent no-ops.
- Wire scheduling into vma_lock_and_validate() so VM_BIND will schedule
decompression when request_decompress is set.
- Handle fault-mode VMs by performing decompression synchronously during
the bind process, ensuring that the resolve is completed before the bind
finishes.
This schedules an in-place GPU resolve (xe_migrate_resolve) for
decompression.
Compute PR: https://github.com/intel/compute-runtime/pull/898
IGT PR: https://patchwork.freedesktop.org/series/157553/
v7: Rebase on latest drm-tip and add compute and igt pr info
v6: (Matt Auld)
- Rebase as xe_pat_index_get_comp_en() is added in separate
patch
- Drop vm param from xe_bo_decompress(), instead of it
extract tile from bo
- Reject decompression on igpu instead of silent skipping
to avoid any failure on Xe2+igpu as xe_device_has_flat_ccs()
can sometimes be false on igpu due some setting in the BIOS
to turn off compression on igpu.
- Nits
v5: (Matt)
- Correct the condition check of xe_pat_index_get_comp_en
v4: (Matt)
- Introduce xe_pat_index_get_comp_en(), which checks
XE2_COMP_EN for the pat_index
- .interruptible should be true, everything else false
v3: (Matt)
- s/xe_bo_schedule_decompress/xe_bo_decompress
- skip the decrompress step if the BO isn't in VRAM
- start/size not required in xe_bo_schedule_decompress
- Use xe_bo_move_notify instead of xe_vm_invalidate_vma
with respect to invalidation.
- Nits
v2:
- Move decompression work out of vm_bind ioctl. (Matt)
- Put that work in a small helper at the BO/migrate layer invoke it
from vma_lock_and_validate which already runs under drm_exec.
- Move lightweight checks to vm_bind_ioctl_check_args (Matthew Auld)
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Acked-by: Michal Mrozek <michal.mrozek@intel.com>
Signed-off-by: Nitin Gote <nitin.r.gote@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260304123758.3050386-8-nitin.r.gote@intel.com
|
|
https://gitlab.freedesktop.org/drm/xe/kernel into drm-next
UAPI Changes:
- restrict multi-lrc to VCS/VECS engines (Xin Wang)
- Introduce a flag to disallow vm overcommit in fault mode (Thomas)
- update used tracking kernel-doc (Auld, Fixes)
- Some bind queue fixes (Auld, Fixes)
Cross-subsystem Changes:
- Split drm_suballoc_new() into SA alloc and init helpers (Satya, Fixes)
- pass pagemap_addr by reference (Arnd, Fixes)
- Revert "drm/pagemap: Disable device-to-device migration" (Thomas)
- Fix unbalanced unlock in drm_gpusvm_scan_mm (Maciej, Fixes)
- Small GPUSVM fixes (Brost, Fixes)
- Fix xe SVM configs (Thomas, Fixes)
Core Changes:
- Fix a hmm_range_fault() livelock / starvation problem (Thomas, Fixes)
Driver Changes:
- Fix leak on xa_store failure (Shuicheng, Fixes)
- Correct implementation of Wa_16025250150 (Roper, Fixes)
- Refactor context init into xe_lrc_ctx_init (Raag)
- Fix GSC proxy cleanup on early initialization failure (Zhanjun)
- Fix exec queue creation during post-migration recovery (Tomasz, Fixes)
- Apply windower hardware filtering setting on Xe3 and Xe3p (Roper)
- Free ctx_restore_mid_bb in release (Shuicheng, Fixes)
- Drop stale MCR steering TODO comment (Roper)
- dGPU memory optimizations (Brost)
- Do not preempt fence signaling CS instructions (Brost, Fixes)
- Revert "drm/xe/compat: Remove unused i915_reg.h from compat header" (Uma)
- Don't expose display modparam if no display support (Wajdeczko)
- Some VRAM flag improvements (Wajdeczko)
- Misc fix for xe_guc_ct.c (Shuicheng, Fixes)
- Remove unused i915_reg.h from compat header (Uma)
- Workaround cleanup & simplification (Roper)
- Add prefetch pagefault support for Xe3p (Varun)
- Fix fs_reclaim deadlock caused by CCS save/restore (Satya, Fixes)
- Cleanup partially initialized sync on parse failure (Shuicheng, Fixes)
- Allow to change VFs VRAM quota using sysfs (Michal)
- Increase GuC log sizes in debug builds (Tomasz)
- Wa_18041344222 changes (Harish)
- Add Wa_14026781792 (Niton)
- Add debugfs facility to catch RTP mistakes (Roper)
- Convert GT stats to per-cpu counters (Brost)
- Prevent unintended VRAM channel creation (Karthik)
- Privatize struct xe_ggtt (Maarten)
- remove unnecessary struct dram_info forward declaration (Jani)
- pagefault refactors (Brost)
- Apply Wa_14024997852 (Arvind)
- Redirect faults to dummy page for wedged device (Raag, Fixes)
- Force EXEC_QUEUE_FLAG_KERNEL for kernel internal VMs (Piotr)
- Stop applying Wa_16018737384 from Xe3 onward (Roper)
- Add new XeCore fuse registers to VF runtime regs (Roper)
- Update xe_device_declare_wedged() error log (Raag)
- Make xe_modparam.force_vram_bar_size signed (Shuicheng, Fixes)
- Avoid reading media version when media GT is disabled (Piotr, Fixes)
- Fix handling of Wa_14019988906 & Wa_14019877138 (Roper, Fixes)
- Basic enabling patches for Xe3p_LPG and NVL-P (Gustavo, Roper, Shekhar)
- Avoid double-adjust in 64-bit reads (Shuicheng, Fixes)
- Allow VF to initialize MCR tables (Wajdeczko)
- Add Wa_14025883347 for GuC DMA failure on reset (Anirban)
- Add bounds check on pat_index to prevent OOB kernel read in madvise (Jia, Fixes)
- Fix the address range assert in ggtt_get_pte helper (Winiarski)
- XeCore fuse register changes (Roper)
- Add more info to powergate_info debugfs (Vinay)
- Separate out GuC RC code (Vinay)
- Fix g2g_test_array indexing (Pallavi)
- Mutual exclusivity between CCS-mode and PF (Nareshkumar, Fixes)
- Some more _types.h cleanups (Wajdeczko)
- Fix sysfs initialization (Wajdeczko, Fixes)
- Drop unnecessary goto in xe_device_create (Roper)
- Disable D3Cold for BMG only on specific platforms (Karthik, Fixes)
- Add sriov.admin_only_pf attribute (Wajdeczko)
- replace old wq(s), add WQ_PERCPU to alloc_workqueue (Marco)
- Make MMIO communication more robust (Wajdeczko)
- Fix warning of kerneldoc (Shuicheng, Fixes)
- Fix topology query pointer advance (Shuicheng, Fixes)
- use entry_dump callbacks for xe2+ PAT dumps (Xin Wang)
- Fix kernel-doc warning in GuC scheduler ABI header (Chaitanya, Fixes)
- Fix CFI violation in debugfs access (Daniele, Fixes)
- Apply WA_16028005424 to Media (Balasubramani)
- Fix typo in function kernel-doc (Wajdeczko)
- Protect priority against concurrent access (Niranjana)
- Fix nvm aux resource cleanup (Shuicheng, Fixes)
- Fix is_bound() pci_dev lifetime (Shuicheng, Fixes)
- Use CLASS() for forcewake in xe_gt_enable_comp_1wcoh (Shuicheng)
- Reset VF GuC state on fini (Wajdeczko)
- Move _THIS_IP_ usage from xe_vm_create() to dedicated function (Nathan Chancellor, Fixes)
- Unregister drm device on probe error (Shuicheng, Fixes)
- Disable DCC on PTL (Vinay, Fixes)
- Fix Wa_18022495364 (Tvrtko, Fixes)
- Skip address copy for sync-only execs (Shuicheng, Fixes)
- derive mem copy capability from graphics version (Nitin, Fixes)
- Use DRM_BUDDY_CONTIGUOUS_ALLOCATION for contiguous allocations (Sanjay)
- Context based TLB invalidations (Brost)
- Enable multi_queue on xe3p_xpc (Brost, Niranjana)
- Remove check for gt in xe_query (Nakshtra)
- Reduce LRC timestamp stuck message on VFs to notice (Brost, Fixes)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/aaYR5G2MHjOEMXPW@lstrano-desk.jf.intel.com
|
|
Let's merge 7.0-rc1 to start the new drm-misc-next window
Signed-off-by: Maxime Ripard <mripard@kernel.org>
|
|
This was done entirely with mindless brute force, using
git grep -l '\<k[vmz]*alloc_objs*(.*, GFP_KERNEL)' |
xargs sed -i 's/\(alloc_objs*(.*\), GFP_KERNEL)/\1)/'
to convert the new alloc_obj() users that had a simple GFP_KERNEL
argument to just drop that argument.
Note that due to the extreme simplicity of the scripting, any slightly
more complex cases spread over multiple lines would not be triggered:
they definitely exist, but this covers the vast bulk of the cases, and
the resulting diff is also then easier to check automatically.
For the same reason the 'flex' versions will be done as a separate
conversion.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
This is the result of running the Coccinelle script from
scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to
avoid scalar types (which need careful case-by-case checking), and
instead replace kmalloc-family calls that allocate struct or union
object instances:
Single allocations: kmalloc(sizeof(TYPE), ...)
are replaced with: kmalloc_obj(TYPE, ...)
Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...)
are replaced with: kmalloc_objs(TYPE, COUNT, ...)
Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...)
are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...)
(where TYPE may also be *VAR)
The resulting allocations no longer return "void *", instead returning
"TYPE *".
Signed-off-by: Kees Cook <kees@kernel.org>
|
|
As per uapi documentation[1], the prerequisite for wedged device is to
redirected page faults to a dummy page. Follow it.
[1] Documentation/gpu/drm-uapi.rst
v2: Add uapi reference and fixes tag (Matthew Brost)
Fixes: 7bc00751f877 ("drm/xe: Use device wedged event")
Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260212055622.2054991-1-raag.jadav@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
(cherry picked from commit c020fff70d757612933711dd3cc3751d7d782d3c)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
|
|
As per uapi documentation[1], the prerequisite for wedged device is to
redirected page faults to a dummy page. Follow it.
[1] Documentation/gpu/drm-uapi.rst
v2: Add uapi reference and fixes tag (Matthew Brost)
Fixes: 7bc00751f877 ("drm/xe: Use device wedged event")
Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260212055622.2054991-1-raag.jadav@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
|
|
Add platform definition along with device IDs for NVL-P. Here is the
list of device descriptor fields and associated Bspec references:
.dma_mask_size (Bspec 74198)
.has_cached_pt (Bspec 71582)
.has_display (Bspec 74196)
.has_flat_ccs (Bspec 74110)
.has_page_reclaim_hw_assist (Bspec 73451)
.max_gt_per_tile (Bspec 74196)
.va_bits (Bspec 74198)
.vm_max_level (Bspec 59507)
v2:
- Add list of descriptor fields and Bspec references. (Matt)
Signed-off-by: Shekhar Chauhan <shekhar.chauhan@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patch.msgid.link/20260206-nvl-p-upstreaming-v3-12-636e1ad32688@intel.com
Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
|
|
Backmerging to get bug fixes from v6.19-rc7.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
|
|
Along with renaming the .move_notify() callback, rename the corresponding
dma-buf core function. This makes the expected behavior clear to exporters
calling this function.
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20260124-dmabuf-revoke-v5-2-f98fca917e96@nvidia.com
Signed-off-by: Christian König <christian.koenig@amd.com>
|
|
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)
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260115032803.4067824-2-matthew.d.roper@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
|
|
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.
Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20260108101014.579906-14-dev@lankhorst.se
|
|
Previously, compressible surfaces were required to be non-coherent
(allocated as WC) because compression and coherency were mutually
exclusive. Starting with Xe3, hardware supports combining compression
with 1-way coherency, allowing compressible surfaces to be allocated as
WB memory. This provides applications with more efficient memory
allocation by avoiding WC allocation overhead that can cause system
stuttering and memory management challenges.
The implementation adds support for compressed+coherent PAT entry for
the xe3_lpg devices and updates the driver logic to handle the new
compression capabilities.
v2: (Matthew Auld)
- Improved error handling with XE_IOCTL_DBG()
- Enhanced documentation and comments
- Fixed xe_bo_needs_ccs_pages() outdated compression assumptions
v3:
- Improve WB compression support detection by checking PAT table
instead of version check
v4:
- Add XE_CACHE_WB_COMPRESSION, which simplifies the logic.
v5:
- Use U16_MAX for the invalid PAT index. (Matthew Auld)
Bspec: 71582, 59361, 59399
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Xin Wang <x.wang@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20260109093007.546784-1-x.wang@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
|
|
Page accounting can change via the shrinker without calling
xe_ttm_tt_unpopulate(), which normally updates page count tracepoints
through update_global_total_pages. Add a call to
update_global_total_pages when the shrinker successfully shrinks a BO.
v2:
- Don't adjust global accounting when pinning (Stuart)
Cc: stable@vger.kernel.org
Fixes: ce3d39fae3d3 ("drm/xe/bo: add GPU memory trace points")
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Stuart Summers <stuart.summers@intel.com>
Link: https://patch.msgid.link/20260107205732.2267541-1-matthew.brost@intel.com
|
|
When imported dma-bufs are destroyed, TTM is not fully
individualizing the dma-resv, but it *is* copying the fences that
need to be waited for before declaring idle. So in the case where
the bo->resv != bo->_resv we can still drop the preempt-fences, but
make sure we do that on bo->_resv which contains the fence-pointer
copy.
In the case where the copying fails, bo->_resv will typically not
contain any fences pointers at all, so there will be nothing to
drop. In that case, TTM would have ensured all fences that would
have been copied are signaled, including any remaining preempt
fences.
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Fixes: fa0af721bd1f ("drm/ttm: test private resv obj on release/destroy")
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: <stable@vger.kernel.org> # v6.16+
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Tested-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/20251217093441.5073-1-thomas.hellstrom@linux.intel.com
|
|
Introduce DRM_XE_GEM_CREATE_FLAG_NO_COMPRESSION to let userspace
opt out of CCS compression on a per-BO basis. When set, the driver
maps this to XE_BO_FLAG_NO_COMPRESSION, skips CCS metadata
allocation/clearing, and rejects compressed PAT indices at vm_bind.
This avoids extra memory ops and manual CCS state handling for buffers.
To allow userspace to detect at runtime whether the kernel supports this
feature, add DRM_XE_QUERY_CONFIG_FLAG_HAS_NO_COMPRESSION_HINT and expose
it via query_config() on Xe2+ platforms.
Mesa PR: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38425
IGT PR: https://patchwork.freedesktop.org/patch/685180/
v2
- Changed error code from -EINVAL to -EOPNOTSUPP for unsupported flag
usage on pre-Xe2 platforms
- Fixed checkpatch warning in xe_vm.c
- Fixed kernel-doc formatting in xe_drm.h
v3
- Rebase
- Updated commit title and description
- Added UAPI for DRM_XE_QUERY_CONFIG_FLAG_HAS_NO_COMPRESSION_HINT and
exposed it via query_config()
v4
- Rebase
v5
- Included Mesa PR and IGT PR in the commit description
- Used xe_pat_index_get_comp_en() to extract the compression
v6
- Added XE_IOCTL_DBG() checks for argument validation
Suggested-by: Matthew Auld <matthew.auld@intel.com>
Suggested-by: José Roberto de Souza <jose.souza@intel.com>
Acked-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Sanjay Yadav <sanjay.kumar.yadav@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20251204040402.2692921-2-sanjay.kumar.yadav@intel.com
|
|
Backmerging to bring in a needed dependency for the Xe VFIO
driver variant. This should ideally have been done before we
commited that, so we now have a small window in drm-xe-next
where that driver doesn't compile.
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202512030331.I8CveRre-lkp@intel.com/
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
|
|
Introduce device xe_caching_pt flag to selectively turn it on for
supported platforms. It allows to eliminate version check and enable
this feature for the future platforms.
Signed-off-by: Zbigniew Kempczyński <zbigniew.kempczynski@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20251125153732.400766-2-zbigniew.kempczynski@intel.com
|
|
Use scope-based runtime power management in the BO code for consistency
with other parts of the driver.
v2:
- Drop unnecessary 'ret' variable. (Gustavo)
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patch.msgid.link/20251118164338.3572146-50-matthew.d.roper@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
|
|
https://gitlab.freedesktop.org/drm/xe/kernel into drm-next
Driver Changes:
More xe3p support (Harish, Brian, Balasubramani, Matt Roper)
Make panic support work on VRAM for display (Maarten)
Fix stolen size check (Shuicheng)
xe_pci_test update (Gustavo)
VF migration updates (Tomasz)
A couple of fixes around allocation and PM references (Matt Brost)
Migration update for the MEM_COPY instruction (Matt Auld)
Initial CRI support (Balasubramani, Matt Roper)
Use SVM range helpers in PT layer (Matt Brost)
Drop MAX_GT_TYPE_CHARS constant (Matt Roper)
Fix spelling and typos (Sanjay)
Fix VF FLR synchronization between all GTs (Michal)
Add a Workaround (Nitin)
Access VF's register using dedicated MMIO view (Michal)
Signed-off-by: Simona Vetter <simona.vetter@ffwll.ch>
From: Thomas Hellstrom <thomas.hellstrom@linux.intel.com>
Link: https://patch.msgid.link/aQCl9uJxN6CWJ8Vg@fedora
|
|
Corrected various spelling mistakes and typos in multiple
files under the Xe directory. These fixes improve clarity
and maintain consistency in documentation.
v2
- Replaced all instances of "XE" with "Xe" where it referred
to the driver name
- of -> for
- Typical -> Typically
v3
- Revert "Xe" to "XE" for macro prefix reference
Signed-off-by: Sanjay Yadav <sanjay.kumar.yadav@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Stuart Summers <stuart.summers@intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patch.msgid.link/20251023121453.1182035-2-sanjay.kumar.yadav@intel.com
|
|
Most BOs do not care at which offset they will be accessed within
GGTT or PPGTT. The few which do care, should be only created
on PF, and mapped within GGTT. On VFs, mapping at fixed offset
is prohibited, as each VF is granted access to a range of
GGTT address space.
Since fixed addresses of GGTT mapping can only be used on PF,
add an assert which makes sure no attempt of fixed placement
will happen for a driver probed on a VF.
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Tomasz Lis <tomasz.lis@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20251021224817.1593817-4-tomasz.lis@intel.com
|
|
This will make it possible to call from xe_display code.
Reviewed-by: Francois Dugast <francois.dugast@intel.com>
Link: https://lore.kernel.org/r/20251016075701.379023-2-jfalempe@redhat.com
Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
|
|
https://gitlab.freedesktop.org/drm/misc/kernel into drm-next
drm-misc-next for v6.19:
UAPI Changes:
Cross-subsystem Changes:
- fbcon cleanups.
- Make drivers depend on FB_TILEBLITTING instead of selecting it,
and hide FB_MODE_HELPERS.
Core Changes:
- More preparations for rust.
- Throttle dirty worker with vblank
- Use drm_for_each_bridge_in_chain_scoped in drm's bridge code and
assorted fixes.
- Ensure drm_client_modeset tests are enabled in UML.
- Rename ttm_bo_put to ttm_bo_fini, as a further step in removing the
TTM bo refcount.
- Add POST_LT_ADJ_REQ training sequence.
- Show list of removed but still allocated bridges.
- Add a simulated vblank interrupt for hardware without it,
and add some helpers to use them in vkms and hypervdrm.
Driver Changes:
- Assorted small fixes, cleanups and updates to host1x, tegra,
panthor, amdxdna, gud, vc4, ssd130x, ivpu, panfrost, panthor,
sysfb, bridge/sn65dsi86, solomon, ast, tidss.
- Convert drivers from using .round_rate() to .determine_rate()
- Add support for KD116N3730A07/A12, chromebook mt8189, JT101TM023,
LQ079L1SX01, raspberrypi 5" panels.
- Improve reclocking on tegra186+ with nouveau.
- Improve runtime pm in amdxdna.
- Add support for HTX_PAI in imx.
- Use a helper to calculate dumb buffer sizes in most drivers.
Signed-off-by: Simona Vetter <simona.vetter@ffwll.ch>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://lore.kernel.org/r/b412fb91-8545-466a-8102-d89c0f2758a7@linux.intel.com
|
|
When using a separate VRAM region for kernel allocations,
some kernel structures, such as context userspace data,
should not reside in the VRAM region dedicated to the kernel.
The VRAM kernel region is intended only for allocations necessary
for driver operation. Allocations created via ioctl are long-lived
and not easily evictable. If this region runs out of space,
there may not be a fallback, which could cause failures.
To prevent this, add a new BO flag that explicitly forces the BO to be
allocated in the general-purpose VRAM region accessible to userspace,
avoiding the kernel-only VRAM region.
v2:
- update commit message (Matthew)
Signed-off-by: Piotr Piórkowski <piotr.piorkowski@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20251003162619.1984236-3-piotr.piorkowski@intel.com
|
|
So far, kernel and userspace allocations have shared the same VRAM region.
However, in some scenarios, it may be necessary to reserve a separate
VRAM area exclusively for kernel allocations.
Let's add preliminary support for such a configuration.
v2:
- replaced for_each_bo_flag_vram with the improved
for_each_set_bo_vram_flag helper (Matthew)
- moved the VRAM flag iteration macro definition into xe_bo.c (Matthew)
- drop unused bo_flgas from bo_vram_flags_to_vram_placement (Matthew)
- use hweight32 helper in __xe_bo_fixed_placement for readability
(Matthew)
v3: remove unnecessary VRAM fixup id
Signed-off-by: Piotr Piórkowski <piotr.piorkowski@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20251003162619.1984236-2-piotr.piorkowski@intel.com
|
|
Before calling ttm_bo_populate() in the CPU fault path of a bo,
we assert that the bo is not being migrated. However, for
local bos we share the reservation object with other local bos
that might be in the process of being migrated. Also some VM
operations may attach USAGE_KERNEL fences to the common
reservation object and trigger false positives from the assert.
So remove the assert and instead wait for bo idle. This may
unnecessarily wait for idle in some cases but since we're
doing this wait later in the fault path anyway we might as
well do it here as well.
This fixes warnings like:
Sep 25 14:56:23 desky kernel: ------------[ cut here ]------------
Sep 25 14:56:23 desky kernel: xe 0000:03:00.0: [drm] Assertion `dma_resv_test_signaled(tbo->base.resv, DMA_RESV_USAGE_KERNEL) || (tbo->ttm && ttm_tt_is_populated(tbo->ttm))` failed!
platform: BATTLEMAGE subplatform: 1
graphics: Xe2_HPG 20.01 step A0
media: Xe2_HPM 13.01 step A1
Sep 25 14:56:23 desky kernel: WARNING: CPU: 6 PID: 24767 at drivers/gpu/drm/xe/xe_bo.c:1748 xe_bo_fault_migrate+0x1bb/0x300 [xe]
Sep 25 14:56:23 desky kernel: Modules linked in: cpuid dm_crypt xt_conntrack nft_chain_nat xt_MASQUERADE nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 bridge stp llc xfrm_user xfr>
Sep 25 14:56:23 desky kernel: snd_soc_sdca snd_seq_midi prime_numbers coretemp snd_seq_midi_event drm_ttm_helper snd_hda_codec drm_buddy drm_exec snd_rawmidi snd_soc_core snd_hda_cor>
Sep 25 14:56:23 desky kernel: CPU: 6 UID: 1000 PID: 24767 Comm: steamwebhelper Tainted: G U W 6.17.0-rc7+ #32 PREEMPT(voluntary)
Sep 25 14:56:23 desky kernel: Tainted: [U]=USER, [W]=WARN
Sep 25 14:56:23 desky kernel: Hardware name: Micro-Star International Co., Ltd. MS-7D36/PRO Z690-P DDR4 (MS-7D36), BIOS A.A1 10/18/2022
Sep 25 14:56:23 desky kernel: RIP: 0010:xe_bo_fault_migrate+0x1bb/0x300 [xe]
Sep 25 14:56:23 desky kernel: Code: fa 64 29 f9 48 c7 c7 40 e0 d3 c1 51 48 c7 c1 c0 e3 d3 c1 52 4c 8b 45 c0 41 50 44 8b 4d c8 4d 89 e0 48 8b 55 a8 e8 25 27 95 ef <0f> 0b 48 83 c4 40 4>
Sep 25 14:56:23 desky kernel: RSP: 0000:ffffae1ca88c7b10 EFLAGS: 00010286
Sep 25 14:56:23 desky kernel: RAX: 0000000000000000 RBX: ffff8d7cfd7e6800 RCX: 0000000000000027
Sep 25 14:56:23 desky kernel: RDX: ffff8d845019cec8 RSI: 0000000000000001 RDI: ffff8d845019cec0
Sep 25 14:56:23 desky kernel: RBP: ffffae1ca88c7bc8 R08: 0000000000000000 R09: 0000000000000000
Sep 25 14:56:23 desky kernel: R10: 0000000000000000 R11: 0000000000000004 R12: ffffffffc1db1faa
Sep 25 14:56:23 desky kernel: R13: ffffffffc1db2ab4 R14: 0000000000000001 R15: ffffae1ca88c7bd8
Sep 25 14:56:23 desky kernel: FS: 00007fb1baf31940(0000) GS:ffff8d849c870000(0000) knlGS:0000000000000000
Sep 25 14:56:23 desky kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Sep 25 14:56:23 desky kernel: CR2: 00007fb1b2860020 CR3: 00000001705a9004 CR4: 0000000000772ef0
Sep 25 14:56:23 desky kernel: PKRU: 55555558
Sep 25 14:56:23 desky kernel: Call Trace:
Sep 25 14:56:23 desky kernel: <TASK>
Sep 25 14:56:23 desky kernel: xe_bo_cpu_fault_fastpath+0x11e/0x220 [xe]
Sep 25 14:56:23 desky kernel: xe_bo_cpu_fault+0x84/0x410 [xe]
Sep 25 14:56:23 desky kernel: ? __x64_sys_mmap+0x33/0x50
Sep 25 14:56:23 desky kernel: ? x64_sys_call+0x1b2e/0x20d0
Sep 25 14:56:23 desky kernel: ? do_syscall_64+0x9d/0x1f0
Sep 25 14:56:23 desky kernel: ? __check_object_size+0x4a/0x2e0
Sep 25 14:56:23 desky kernel: __do_fault+0x36/0x190
Sep 25 14:56:23 desky kernel: do_fault+0xcf/0x570
Sep 25 14:56:23 desky kernel: __handle_mm_fault+0x92b/0xfe0
Sep 25 14:56:23 desky kernel: ? ktime_get_mono_fast_ns+0x39/0xd0
Sep 25 14:56:23 desky kernel: handle_mm_fault+0x164/0x2c0
Sep 25 14:56:23 desky kernel: do_user_addr_fault+0x2cb/0x840
Sep 25 14:56:23 desky kernel: exc_page_fault+0x75/0x180
Sep 25 14:56:23 desky kernel: asm_exc_page_fault+0x27/0x30
Sep 25 14:56:23 desky kernel: RIP: 0033:0x7fb1bc388bb7
Sep 25 14:56:23 desky kernel: Code: 48 ff c7 48 01 fe 48 8d 54 11 80 0f 1f 84 00 00 00 00 00 c5 fe 6f 0e c5 fe 6f 56 20 c5 fe 6f 5e 40 c5 fe 6f 66 60 48 83 ee 80 <c5> fd 7f 0f c5 fd 7>
Sep 25 14:56:23 desky kernel: RSP: 002b:00007ffd7814fad8 EFLAGS: 00010207
Sep 25 14:56:23 desky kernel: RAX: 00007fb1b2860000 RBX: 0000000000000690 RCX: 00007fb1b2860000
Sep 25 14:56:23 desky kernel: RDX: 00007fb1b2860610 RSI: 0000556eda79f4c0 RDI: 00007fb1b2860020
Sep 25 14:56:23 desky kernel: RBP: 00007ffd7814fb60 R08: 0000000000000000 R09: 000000012be0e000
Sep 25 14:56:23 desky kernel: R10: 00007fb1b2860000 R11: 0000000000000246 R12: 0000556edd39a240
Sep 25 14:56:23 desky kernel: R13: 00007fb1b2dcb010 R14: 0000556eda79f420 R15: 0000000000000000
Sep 25 14:56:23 desky kernel: </TASK>
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/5250
Fixes: c2ae94cf8cd8 ("drm/xe: Convert the CPU fault handler for exhaustive eviction")
Cc: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250929112649.6131-1-thomas.hellstrom@linux.intel.com
|
|
Call drm_mode_size_dumb() to compute dumb-buffer scanline pitch
and buffer size. Align the pitch to a multiple of 8. Align the
buffer size according to hardware requirements.
Xe's internal calculation allowed for 64-bit wide buffer sizes, but
the ioctl's internal checks always verified against 32-bit wide limits.
Hance, it is safe to limit the driver code to 32-bit calculations as
well.
v3:
- mention 32-bit calculation in commit description (Matthew)
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Acked-by: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://lore.kernel.org/r/20250821081918.79786-24-tzimmermann@suse.de
|
|
Give TTM BOs a separate cleanup function.
No funktional change, but the next step in removing the TTM BO reference
counting and replacing it with the GEM object reference counting.
v2: move the code around a bit to make it clearer what's happening
v3: fix nouveau_bo_fini as well
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Acked-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://lore.kernel.org/r/20250909144311.1927-1-christian.koenig@amd.com
|
|
If a bo is created via xe_managed_bo_create_pin_map() then it cannot be
freed by the driver using xe_bo_unpin_map_no_vm(), or indeed any other
existing function. The DRM layer will still have a pointer stashed
away for later freeing, causing a invalid memory access on driver
unload. So add a helper for releasing the DRM action as well.
v2: Drop 'xe' parameter (review feedbak from Michal W)
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://lore.kernel.org/r/20250910210237.603576-4-John.C.Harrison@Intel.com
|
|
This is a backmerge of Linux 6.17-rc6, needed for msm,
also requested by misc.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
Pinned suspend eviction and preparation for eviction validates
system memory for eviction buffers. Do that under a
validation exclusive lock to avoid interfering with other
processes validating system graphics memory.
v2:
- Avoid gotos from within xe_validation_guard().
- Adapt to signature change of xe_validation_guard().
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250908101246.65025-14-thomas.hellstrom@linux.intel.com
|
|
A common pattern is to create a locked bo, pin it without mapping
and then unlock it. Add a function to do that, which internally
uses xe_validation_guard().
With that we can remove xe_bo_create_locked_range() and add
exhaustive eviction to stolen, pf_provision_vf_lmem and
psmi_alloc_object.
v4:
- New patch after reorganization.
v5:
- Replace DRM_XE_GEM_CPU_CACHING_WB with 0. (CI)
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250908101246.65025-13-thomas.hellstrom@linux.intel.com
|