Commit ade8a0f598 ("drm/i915: Make all GPU resets atomic") added a
preempt disable section over the hardware reset callback to prepare the
driver for being able to reset from atomic contexts.
In retrospect I can see that the work item at a time was about removing
the struct mutex from the reset path. Code base also briefly entertained
the idea of doing the reset under stop_machine in order to serialize
userspace mmap and temporary glitch in the fence registers (see
eb8d0f5af4 ("drm/i915: Remove GPU reset dependence on struct_mutex"),
but that never materialized and was soon removed in 2caffbf117
("drm/i915: Revoke mmaps and prevent access to fence registers across
reset") and replaced with a SRCU based solution.
As such, as far as I can see, today we still have a requirement that
resets must not sleep (invoked from submission tasklets), but no need to
support invoking them from a truly atomic context.
Given that the preemption section is problematic on RT kernels, since the
uncore lock becomes a sleeping lock and so is invalid in such section,
lets try and remove it. Potential downside is that our short waits on GPU
to complete the reset may get extended if CPU scheduling interferes, but
in practice that probably isn't a deal breaker.
In terms of mechanics, since the preemption disabled block is being
removed we just need to replace a few of the wait_for_atomic macros into
busy looping versions which will work (and not complain) when called from
non-atomic sections.
v2:
* Fix timeouts which are now in us. (Andi)
* Update one comment as a drive by. (Andi)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris.p.wilson@intel.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230926100855.61722-1-tvrtko.ursulin@linux.intel.com
Ideally the busyness worker should take a gt pm wakeref because the
worker only needs to be active while gt is awake. However, the gt_park
path cancels the worker synchronously and this complicates the flow if
the worker is also running at the same time. The cancel waits for the
worker and when the worker releases the wakeref, that would call gt_park
and would lead to a deadlock.
The resolution is to take the global pm wakeref if runtime pm is already
active. If not, we don't need to update the busyness stats as the stats
would already be updated when the gt was parked.
Note:
- We do not requeue the worker if we cannot take a reference to runtime
pm since intel_guc_busyness_unpark would requeue the worker in the
resume path.
- If the gt was parked longer than time taken for GT timestamp to roll
over, we ignore those rollovers since we don't care about tracking the
exact GT time. We only care about roll overs when the gt is active and
running workloads.
- There is a window of time between gt_park and runtime suspend, where
the worker may run. This is acceptable since the worker will not find
any new data to update busyness.
v2: (Daniele)
- Edit commit message and code comment
- Use runtime pm in the worker
- Put runtime pm after enabling the worker
- Use Link tag and add Fixes tag
v3: (Daniele)
- Reword commit and comments and add details
Link: https://gitlab.freedesktop.org/drm/intel/-/issues/7077
Fixes: 77cdd054dd ("drm/i915/pmu: Connect engine busyness stats from GuC to pmu")
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230925192117.2497058-1-umesh.nerlige.ramappa@intel.com
Invalidate instruction and State cache bit using INDIRECT_CTX on
every gpu context switch for gen12.
The goal of this workaround is to actually perform an explicit
invalidation of that cache (by re-writing the register) during every GPU
context switch, which is accomplished via a "workaround batchbuffer"
that's attached to the context via INDIRECT_CTX. (Matt Roper)
Please refer [1] for more reviews and comment on the same patch
[1] https://patchwork.freedesktop.org/series/123377/
v2:
- Remove extra parentheses from the condition (Lucas)
- Align spacing and new line (Lucas)
v3:
- Fix commit message.
v4:
- Only Gen12 changes are kept and Remove DG2+ condition (Matt Roper)
- Fix the commit message for r-b (Matt Roper)
- Rename the register bit in define
v5:
- Move out this workaround from golden context init (Matt Roper)
- Use INDIRECT_CTX to set bit on each GPU context switch (Matt Roper)
v6:
- Change IP Version base condition for Gen12 (Matt Roper)
- Made imperative form of commit version messages (Suraj)
- s/Added/Add in patch header (Suraj)
v7:
- In version descriptions s/Ropper/Roper (Matt Atwood)
BSpec: 11354
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Suraj Kandpal <suraj.kandpal@intel.com>
Cc: Matt Atwood <matthew.s.atwood@intel.com>
Signed-off-by: Dnyaneshwar Bhadane <dnyaneshwar.bhadane@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230914202000.1069884-1-dnyaneshwar.bhadane@intel.com
The thread_global_copy subtest of the live migrate selftest creates a
large number of threads and waits 10ms for them all to start. This is
not enough time to wait for the threaded tasks to start, as some may
need to wait for additional ring space to be granted. Threads that do
so are at risk of getting stopped (signaled) in the middle of waiting
for additional space, which can result in ERESTARTSYS getting reported
erroneously by i915_request_wait.
Instead of waiting a flat 10ms for the threads to start, wait 10ms per
thread. This grants enough of a buffer for each thread to wait for
additional ring space when needed.
Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230828192852.2894671-2-jonathan.cavitt@intel.com
References to i915_requests may be trapped by userspace inside a
sync_file or dmabuf (dma-resv) and held indefinitely across different
proceses. To counter-act the memory leaks, we try to not to keep
references from the request past their completion.
On the other side on fence release we need to know if rq->engine
is valid and points to hw engine (true for non-virtual requests).
To make it possible extra bit has been added to rq->execution_mask,
for marking virtual engines.
Fixes: bcb9aa45d5 ("Revert "drm/i915: Hold reference to intel_context over life of i915_request"")
Signed-off-by: Chris Wilson <chris.p.wilson@linux.intel.com>
Signed-off-by: Andrzej Hajda <andrzej.hajda@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821153035.3903006-1-andrzej.hajda@intel.com
(cherry picked from commit 280410677a)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
References to i915_requests may be trapped by userspace inside a
sync_file or dmabuf (dma-resv) and held indefinitely across different
proceses. To counter-act the memory leaks, we try to not to keep
references from the request past their completion.
On the other side on fence release we need to know if rq->engine
is valid and points to hw engine (true for non-virtual requests).
To make it possible extra bit has been added to rq->execution_mask,
for marking virtual engines.
Fixes: bcb9aa45d5 ("Revert "drm/i915: Hold reference to intel_context over life of i915_request"")
Signed-off-by: Chris Wilson <chris.p.wilson@linux.intel.com>
Signed-off-by: Andrzej Hajda <andrzej.hajda@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821153035.3903006-1-andrzej.hajda@intel.com
If GuC hits an internal error (and survives long enough to report it
to the KMD), it is basically toast and will stop until a GT reset and
subsequent GuC reload is performed. Previously, the KMD just printed
an error message and then waited for the heartbeat to eventually kick
in and trigger a reset (assuming the heartbeat had not been disabled).
Instead, force the reset immediately to guarantee that it happens and
to eliminate the very long heartbeat delay. The captured error state
is also more likely to be useful if captured at the time of the error
rather than many seconds later.
Note that it is not possible to trigger a reset from with the G2H
handler itself. The reset prepare process involves flushing
outstanding G2H contents. So a deadlock could result. Instead, the G2H
handler queues a worker thread to do the reset asynchronously.
v2: Flush the worker on suspend and shutdown. Add rate limiting to
prevent spam from a totally dead system (review feedback from Daniele).
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230816003957.3572654-1-John.C.Harrison@Intel.com
Many of the IS_METEORLAKE conditions throughout the driver are supposed
to be checks for Xe_LPG and/or Xe_LPM+ IP, not for the MTL platform
specifically. Update those checks to ensure that the code will still
operate properly if/when these IP versions show up on future platforms.
v2:
- Update two more conditions (one for pg_enable, one for MTL HuC
compatibility).
v3:
- Don't change GuC/HuC compatibility check, which sounds like it truly
is specific to the MTL platform. (Gustavo)
- Drop a non-lineage workaround number for the OA timestamp frequency
workaround. (Gustavo)
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821180619.650007-20-matthew.d.roper@intel.com
Stepping-specific media behavior shouldn't be tied to MTL as a platform,
but rather specifically to the Xe_LPM+ IP. Future non-MTL platforms may
re-use this IP and will need to follow the exact same logic and apply
the same workarounds. IS_MTL_MEDIA_STEP() is dropped in favor of
IS_MEDIA_GT_IP_STEP, which checks the media IP version associated with a
specific IP and also ensures that we're operating on the media GT, not
the primary GT.
v2:
- Switch to the IS_GT_IP_STEP macro.
v3:
- Switch back to long-form IS_MEDIA_GT_IP_STEP. (Jani)
v4:
- Build IS_MEDIA_GT_IP_STEP on top of IS_MEDIA_GT_IP_RANGE and
IS_MEDIA_STEP building blocks and name the parameters from/until
rather than begin/fixed.. (Jani)
v5:
- Tweak macro comment wording. (Gustavo)
- Add a check to catch NULL gt in IS_MEDIA_GT_IP_RANGE; this allows it
to be used safely on i915->media_gt, which may be NULL on some
platforms. (Gustavo)
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821180619.650007-16-matthew.d.roper@intel.com
Several workarounds are guarded by IS_MTL_GRAPHICS_STEP. However none
of these workarounds are actually tied to MTL as a platform; they only
relate to the Xe_LPG graphics IP, regardless of what platform it appears
in. At the moment MTL is the only platform that uses Xe_LPG with IP
versions 12.70 and 12.71, but we can't count on this being true in the
future. Switch these to use a new IS_GFX_GT_IP_STEP() macro instead
that is purely based on IP version. IS_GFX_GT_IP_STEP() is also
GT-based rather than device-based, which will help prevent mistakes
where we accidentally try to apply Xe_LPG graphics workarounds to the
Xe_LPM+ media GT and vice-versa.
v2:
- Switch to a more generic and shorter IS_GT_IP_STEP macro that can be
used for both graphics and media IP (and any other kind of GTs that
show up in the future).
v3:
- Switch back to long-form IS_GFX_GT_IP_STEP macro. (Jani)
- Move macro to intel_gt.h. (Andi)
v4:
- Build IS_GFX_GT_IP_STEP on top of IS_GFX_GT_IP_RANGE and
IS_GRAPHICS_STEP building blocks and name the parameters from/until
rather than begin/fixed. (Jani)
- Fix usage examples in comment.
v5:
- Tweak comment on macro. (Gustavo)
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821180619.650007-15-matthew.d.roper@intel.com
Although some of our Xe_LPG workarounds were already being applied based
on IP version correctly, others were matching on MTL as a base platform,
which is incorrect. Although MTL is the only platform right now that
uses Xe_LPG IP, this may not always be the case. If a future platform
re-uses this graphics IP, the same workarounds should be applied, even
if it isn't a "MTL" platform.
We were also incorrectly applying Xe_LPG workarounds/tuning to the
Xe_LPM+ media IP in one or two places; we should make sure that we don't
try to apply graphics workarounds to the media GT and vice versa where
they don't belong. A new helper macro IS_GT_IP_RANGE() is added to help
ensure this is handled properly -- it checks that the GT matches the IP
type being tested as well as the IP version falling in the proper range.
Note that many of the stepping-based workarounds are still incorrectly
checking for a MTL base platform; that will be remedied in a later
patch.
v2:
- Rework macro into a slightly more generic IS_GT_IP_RANGE() that can
be used for either GFX or MEDIA checks.
v3:
- Switch back to separate macros for gfx and media. (Jani)
- Move macro to intel_gt.h. (Andi)
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230821180619.650007-14-matthew.d.roper@intel.com
DG2 first production steppings were C0 (for DG2-G10), B1 (for DG2-G11),
and A1 (for DG2-G12). Several workarounds that apply onto to
pre-production hardware can be dropped. Furthermore, several
workarounds that apply to all production steppings can have their
conditions simplified to no longer check the GT stepping.
v2:
- Keep Wa_16011777198 in place for now; it will be removed separately
in a follow-up patch to keep review easier.
Bspec: 44477
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Matt Atwood <matthew.s.atwood@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230816214201.534095-10-matthew.d.roper@intel.com
