Platforms that do not support SLPC are exempted from the GuC PC support.
The GuC PC does not get initialized, and neither do its BOs get created.
This causes a problem because the GuC PC debugfs file is still being
created. Whenever the file is attempted to read, it causes a NULL
pointer dereference on the supposed BO of the GuC PC.
So, make the creation of SLPC debugfs files conditional to when SLPC
features are supported.
Fixes: aaab5404b1 ("drm/xe: Introduce GuC PC debugfs")
Suggested-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Tejas Upadhyay <tejas.upadhyay@intel.com>
Reviewed-by: Stuart Summers <stuart.summers@intel.com>
Signed-off-by: Aradhya Bhatia <aradhya.bhatia@intel.com>
Link: https://lore.kernel.org/r/20250516141902.5614-1-aradhya.bhatia@intel.com
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Most H2G messages are FAST_REQ which means no synchronous response is
expected. The messages are sent as fire-and-forget with no tracking.
However, errors can still be returned when something goes unexpectedly
wrong. That leads to confusion due to not being able to match up the
error response to the originating H2G.
So add support for tracking the FAST_REQ H2Gs and matching up an error
response to its originator. This is only enabled in XE_DEBUG builds
given that such errors should never happen in a working system and
there is an overhead for the tracking.
Further, if XE_DEBUG_GUC is enabled then even more memory and time is
used to record the call stack of each H2G and report that with the
error. That makes it much easier to work out where a specific H2G came
from if there are multiple code paths that can send it.
v2: Some re-wording of comments and prints, more consistent use of #if
vs stub functions - review feedback from Daniele & Michal).
v3: Split config change to separate patch, improve a debug print
(review feedback from Michal).
v4: Bunch of minor tweaks (review feedback from Michal).
Original-i915-code: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250512215324.1457009-5-John.C.Harrison@Intel.com
These error codes are not actually used in the driver but it is
extremely useful to have them available to understand error messages.
v2: Add a bunch more error codes and drop 'status' from names (review
feedback by Michal W).
v3: Drop 'SUCCESS' response as meaningless in current API (review
feedback by Michal W).
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250512215324.1457009-3-John.C.Harrison@Intel.com
This commit adds prefetch support for SVM ranges, utilizing the
existing ioctl vm_bind functionality to achieve this.
v2: rebase
v3:
- use xa_for_each() instead of manual loop
- check range is valid and in preferred location before adding to
xarray
- Fix naming conventions
- Fix return condition as -ENODATA instead of -EAGAIN (Matthew Brost)
- Handle sparsely populated cpu vma range (Matthew Brost)
v4:
- fix end address to find next cpu vma in case of -ENOENT
v5:
- Move find next vma logic to drm gpusvm layer
- Avoid mixing declaration and logic
v6:
- Use new function names
- Move eviction logic to prefetch_ranges
v7:
- devmem_only assigned 0
- nit address
v8:
- initialize ctx with 0
Cc: Matthew Brost <matthew.brost@intel.com>
Acked-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250513040228.470682-15-himal.prasad.ghimiray@intel.com
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
The xe_svm_range_validate() function checks if a range is
valid and located in the desired memory region.
xe_svm_range_migrate_to_smem() checks if range have pages in devmem and
migrate them to smem.
v2
- Fix function stub in xe_svm.h
- Fix doc
v3 (Matthew Brost)
- Remove extra new line
- s/range->base.flags.has_devmem_pages/xe_svm_range_in_vram
v4 (Matthew Brost)
- s/xe_svm_range_in_vram/range->base.flags.has_devmem_pages
- Move eviction logic to separate function
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Acked-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://lore.kernel.org/r/20250513040228.470682-12-himal.prasad.ghimiray@intel.com
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Context Timestamp (CTX_TIMESTAMP) in the LRC accumulates the run ticks
of the context, but only gets updated when the context switches out. In
order to check how long a context has been active before it switches
out, two things are required:
(1) Determine if the context is running:
To do so, we program the WA BB to set an initial value for CTX_TIMESTAMP
in the LRC. The value chosen is 1 since 0 is the initial value when the
LRC is initialized. During a query, we just check for this value to
determine if the context is active. If the context switched out, it
would overwrite this location with the actual CTX_TIMESTAMP MMIO value.
Note that WA BB runs as the last part of the context restore, so reusing
this LRC location will not clobber anything.
(2) Calculate the time that the context has been active for:
The CTX_TIMESTAMP ticks only when the context is active. If a context is
active, we just use the CTX_TIMESTAMP MMIO as the new value of
utilization. While doing so, we need to read the CTX_TIMESTAMP MMIO
for the specific engine instance. Since we do not know which instance
the context is running on until it is scheduled, we also read the
ENGINE_ID MMIO in the WA BB and store it in the PPHSWP.
Using the above 2 instructions in a WA BB, capture active context
utilization.
v2: (Matt Brost)
- This breaks TDR, fix it by saving the CTX_TIMESTAMP register
"drm/xe: Save CTX_TIMESTAMP mmio value instead of LRC value"
- Drop tile from LRC if using gt
"drm/xe: Save the gt pointer in LRC and drop the tile"
v3:
- Remove helpers for bb_per_ctx_ptr (Matt)
- Add define for context active value (Matt)
- Use 64 bit CTX TIMESTAMP for platforms that support it. For platforms
that don't, live with the rare race. (Matt, Lucas)
- Convert engine id to hwe and get the MMIO value (Lucas)
- Correct commit message on when WA BB runs (Lucas)
v4:
- s/GRAPHICS_VER(...)/xe->info.has_64bit_timestamp/ (Matt)
- Drop support for active utilization on a VF (CI failure)
- In xe_lrc_init ensure the lrc value is 0 to begin with (CI regression)
v5:
- Minor checkpatch fix
- Squash into previous commit and make TDR use 32-bit time
- Update code comment to match commit msg
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4532
Suggested-by: Lucas De Marchi <lucas.demarchi@intel.com>
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://lore.kernel.org/r/20250509161159.2173069-8-umesh.nerlige.ramappa@intel.com
Mixing GPU and CPU atomics does not work unless a strict migration
policy of GPU atomics must be device memory. Enforce a policy of must be
in VRAM with a retry loop of 3 attempts, if retry loop fails abort
fault.
Removing always_migrate_to_vram modparam as we now have real migration
policy.
v2:
- Only retry migration on atomics
- Drop alway migrate modparam
v3:
- Only set vram_only on DGFX (Himal)
- Bail on get_pages failure if vram_only and retry count exceeded (Himal)
- s/vram_only/devmem_only
- Update xe_svm_range_is_valid to accept devmem_only argument
v4:
- Fix logic bug get_pages failure
v5:
- Fix commit message (Himal)
- Mention removing always_migrate_to_vram in commit message (Lucas)
- Fix xe_svm_range_is_valid to check for devmem pages
- Bail on devmem_only && !migrate_devmem (Thomas)
v6:
- Add READ_ONCE barriers for opportunistic checks (Thomas)
- Pair READ_ONCE with WRITE_ONCE (Thomas)
v7:
- Adjust comments (Thomas)
Fixes: 2f118c9491 ("drm/xe: Add SVM VRAM migration")
Cc: stable@vger.kernel.org
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Acked-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://lore.kernel.org/r/20250512135500.1405019-3-matthew.brost@intel.com
This commit adds a new flag, devmem_only, to the drm_gpusvm structure. The
purpose of this flag is to ensure that the get_pages function allocates
memory exclusively from the device's memory. If the allocation from
device memory fails, the function will return an -EFAULT error.
Required for shared CPU and GPU atomics on certain devices.
v3:
- s/vram_only/devmem_only/
Fixes: 99624bdff8 ("drm/gpusvm: Add support for GPU Shared Virtual Memory")
Cc: stable@vger.kernel.org
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250512135500.1405019-2-matthew.brost@intel.com
During post-migration recovery of a VF, it is necessary to update
GGTT references included in messages which are going to be sent
to GuC. GuC will start consuming messages after VF KMD will inform
it about fixups being done; before that, the VF KMD is expected
to update any H2G messages which are already in send buffer but
were not consumed by GuC.
Only a small subset of messages allowed for VFs have GGTT references
in them. This patch adds the functionality to parse the CTB send
ring buffer and shift addresses contained within.
While fixing the CTB content, ct->lock is not taken. This means
the only barrier taken remains GGTT address lock - which is ok,
because only requests with GGTT addresses matter, but it also means
tail changes can happen during the CTB fixups execution (which may
be ignored as any new messages will not have anything to fix).
The GGTT address locking will be introduced in a future series.
v2: removed storing shift as that's now done in VMA nodes patch;
macros to inlines; warns to asserts; log messages fixes (Michal)
v3: removed inline keywords, enums for offsets in CTB messages,
less error messages, if return unused then made functs void (Michal)
v4: update the cached head before starting fixups
v5: removed/updated comments, wrapped lines, converted assert into
error, enums for offsets to separate patch, reused xe_map_rd
v6: define xe_map_*_array() macros, support CTB wrap which divides
a message, updated comments, moved one function to an earlier patch
v7: renamed few functions, wider use on previously introduced helper,
separate cases in parsing messges, documented a static funct
v8: Introduced more helpers, fixed coding style mistakes
v9: Move xe_map*() functs to macros, add asserts, add debug print
v10: Errors in place of some asserts, style fixes
v11: Fixed invalid conditionals, added debug-only local pointer
v12: Removed redundant __maybe_unused
Signed-off-by: Tomasz Lis <tomasz.lis@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250512114018.361843-5-tomasz.lis@intel.com
Some GuC messages are constructed with incrementing dword counter
rather than referencing specific DWORDs, as described in GuC interface
specification.
This change introduces the definitions of DWORD numbers for parameters
which will need to be referenced in a CTB parser to be added in a
following patch. To ensure correctness of these DWORDs, verification
in form of asserts was added to the message construction code.
v2: Renamed enum members, added ones for single context registration,
modified asserts to check values rather than indexes.
v3: Reordered assert args to take less lines
v4: Added lengths
v5: Renamed MULTI_LRC_MSG_LEN to MULTI_LRC_MSG_MIN_LEN
Suggested-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Tomasz Lis <tomasz.lis@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250512114018.361843-4-tomasz.lis@intel.com
We have only one GGTT for all IOV functions, with each VF having assigned
a range of addresses for its use. After migration, a VF can receive a
different range of addresses than it had initially.
This implements shifting GGTT addresses within drm_mm nodes, so that
VMAs stay valid after migration. This will make the driver use new
addresses when accessing GGTT from the moment the shifting ends.
By taking the ggtt->lock for the period of VMA fixups, this change
also adds constraint on that mutex. Any locks used during the recovery
cannot ever wait for hardware response - because after migration,
the hardware will not do anything until fixups are finished.
v2: Moved some functs to xe_ggtt.c; moved shift computation to just
after querying; improved documentation; switched some warns to asserts;
skipping fixups when GGTT shift eq 0; iterating through tiles (Michal)
v3: Updated kerneldocs, removed unused funct, properly allocate
balloning nodes if non existent
v4: Re-used ballooning functions from VF init, used bool in place of
standard error codes
v5: Renamed one function
v6: Subject tag change, several kerneldocs updated, some functions
renamed, some moved, added several asserts, shuffled declarations
of variables, revealed more detail in high level functions
v7: Fixed typos, added `_locked` suffix to some functs, improved
readability of asserts, removed unneeded conditional
v8: Moved one function, removed implementation detail from kerneldoc,
added asserts
v9: Code shuffling without much change, and one param rename
v10: Minor error path change, added printing the shift via debugfs
Signed-off-by: Tomasz Lis <tomasz.lis@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250512114018.361843-3-tomasz.lis@intel.com
The balloon nodes, which are used to fill areas of GGTT inaccessible
for a specific VF, were allocated and inserted into GGTT within one
function. To be able to re-use that insertion code during VF
migration recovery, we need to split it.
This patch separates allocation (init/fini functs) from the insertion
of balloons (balloon/deballoon functs). Locks are also moved to ensure
calls from post-migration recovery worker will not cause a deadlock.
v2: Moved declarations to proper header
v3: Rephrased description, introduced "_locked" versions of some
functs, more lockdep checks, some functions renamed, altered error
handling, added missing kerneldocs.
v4: Suffixed more functs with `_locked`, moved lockdep asserts,
fixed finalization in error path, added asserts
v5: Renamed another few functs, used xe_ggtt_node_allocated(),
moved lockdep back again to avoid null dereference, added
asserts, improved comments
v6: Changed params of cleanup_ggtt()
Signed-off-by: Tomasz Lis <tomasz.lis@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250512114018.361843-2-tomasz.lis@intel.com
The xe buffer object shrinker name is visible in the
<debugfs>/shrinker directory and most if not all other shinkers
follow a naming convention that looks like
<subsystem>-<driver>_<objects>:<unique>
Follow the same convention for xe, changing the name to
drm-xe_gem:<unique>.
Other shrinkers typically use the device node for <unique> but
since drm drivers typically don't have a single unique device-
node, instead use the unique name in the drm device.
Fixes: 00c8efc318 ("drm/xe: Add a shrinker for xe bos")
Cc: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Francois Dugast <francois.dugast@intel.com>
Link: https://lore.kernel.org/r/20250508112931.3347-1-thomas.hellstrom@linux.intel.com
The workqueue used for the reset worker is marked as WQ_MEM_RECLAIM,
while the GSC one isn't (and can't be as we need to do memory
allocations in the gsc worker). Therefore, we can't flush the latter
from the former.
The reason why we had such a flush was to avoid interrupting either
the GSC FW load or in progress GSC proxy operations. GSC proxy
operations fall into 2 categories:
1) GSC proxy init: this only happens once immediately after GSC FW load
and does not support being interrupted. The only way to recover from
an interruption of the proxy init is to do an FLR and re-load the GSC.
2) GSC proxy request: this can happen in response to a request that
the driver sends to the GSC. If this is interrupted, the GSC FW will
timeout and the driver request will be failed, but overall the GSC
will keep working fine.
Flushing the work allowed us to avoid interruption in both cases (unless
the hang came from the GSC engine itself, in which case we're toast
anyway). However, a failure on a proxy request is tolerable if we're in
a scenario where we're triggering a GT reset (i.e., something is already
gone pretty wrong), so what we really need to avoid is interrupting
the init flow, which we can do by polling on the register that reports
when the proxy init is complete (as that ensure us that all the load and
init operations have been completed).
Note that during suspend we still want to do a flush of the worker to
make sure it completes any operations involving the HW before the power
is cut.
v2: fix spelling in commit msg, rename waiter function (Julia)
Fixes: dd0e89e5ed ("drm/xe/gsc: GSC FW load")
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4830
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: John Harrison <John.C.Harrison@Intel.com>
Cc: Alan Previn <alan.previn.teres.alexis@intel.com>
Cc: <stable@vger.kernel.org> # v6.8+
Reviewed-by: Julia Filipchuk <julia.filipchuk@intel.com>
Link: https://lore.kernel.org/r/20250502155104.2201469-1-daniele.ceraolospurio@intel.com
The device core dumps are copied in 1.5GB chunks, which leads to a
link-time error on 32-bit builds because of the 64-bit division not
getting trivially turned into mask and shift operations:
ERROR: modpost: "__moddi3" [drivers/gpu/drm/xe/xe.ko] undefined!
On top of this, I noticed that the ALIGN_DOWN() usage here cannot
work because that is only defined for power-of-two alignments.
Change ALIGN_DOWN into an explicit div_u64_rem() that avoids the
link error and hopefully produces the right results.
Doing a 1.5GB kvmalloc() does seem a bit suspicious as well, e.g.
this will clearly fail on any 32-bit platform and is also likely
to run out of memory on 64-bit systems under memory pressure, so
using a much smaller power-of-two chunk size might be a good idea
instead.
v2:
- Always call div_u64_rem (Matt)
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202504251238.JsNgFeFc-lkp@intel.com/
Fixes: c4a2e5f865 ("drm/xe: Add devcoredump chunking")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://lore.kernel.org/r/20250501012545.1045247-1-matthew.brost@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
