Chris Wilson
2850748ef8
drm/i915: Pull i915_vma_pin under the vm->mutex
...
Replace the struct_mutex requirement for pinning the i915_vma with the
local vm->mutex instead. Note that the vm->mutex is tainted by the
shrinker (we require unbinding from inside fs-reclaim) and so we cannot
allocate while holding that mutex. Instead we have to preallocate
workers to do allocate and apply the PTE updates after we have we
reserved their slot in the drm_mm (using fences to order the PTE writes
with the GPU work and with later unbind).
In adding the asynchronous vma binding, one subtle requirement is to
avoid coupling the binding fence into the backing object->resv. That is
the asynchronous binding only applies to the vma timeline itself and not
to the pages as that is a more global timeline (the binding of one vma
does not need to be ordered with another vma, nor does the implicit GEM
fencing depend on a vma, only on writes to the backing store). Keeping
the vma binding distinct from the backing store timelines is verified by
a number of async gem_exec_fence and gem_exec_schedule tests. The way we
do this is quite simple, we keep the fence for the vma binding separate
and only wait on it as required, and never add it to the obj->resv
itself.
Another consequence in reducing the locking around the vma is the
destruction of the vma is no longer globally serialised by struct_mutex.
A natural solution would be to add a kref to i915_vma, but that requires
decoupling the reference cycles, possibly by introducing a new
i915_mm_pages object that is own by both obj->mm and vma->pages.
However, we have not taken that route due to the overshadowing lmem/ttm
discussions, and instead play a series of complicated games with
trylocks to (hopefully) ensure that only one destruction path is called!
v2: Add some commentary, and some helpers to reduce patch churn.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20191004134015.13204-4-chris@chris-wilson.co.uk
2019-10-04 15:39:02 +01:00
Chris Wilson
50d16d44cc
drm/i915/selftests: Exercise context switching in parallel
...
We currently test context switching on each engine as a basic stress
test (just verifying that nothing explodes if we execute 2 requests from
different contexts sequentially). What we have not tested is what
happens if we try and do so on all available engines simultaneously,
putting our SW and the HW under the maximal stress.
v2: Clone the set of engines from the first context into the secondary
contexts.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com >
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190930144919.27992-1-chris@chris-wilson.co.uk
2019-09-30 17:38:05 +01:00
Chris Wilson
f269007446
drm/i915/selftests: Remove unused __engines_name()
...
This function was never used and probably will never be used, so remove
it.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190901110431.12393-1-chris@chris-wilson.co.uk
2019-09-03 07:44:04 +01:00
Chris Wilson
c4e6488120
drm/i915/selftests: Try to recycle context allocations
...
igt_ctx_exec allocates a new context for each iteration, keeping them
all allocated until the end. Instead, release the local ctx reference at
the end of each iteration, allowing ourselves to reap those if under
mempressure.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190827161726.3640-2-chris@chris-wilson.co.uk
2019-08-28 08:32:02 +01:00
Chris Wilson
f2085c8e95
drm/i915/selftests: Remove accidental serialization between gpu_fill
...
Upon object creation for live_gem_contexts, we fill the object with
known scratch and flush it out of the CPU cache. Before performing the
GPU fill, we don't need to flush it again and so avoid serialising with
previous fills.
However, we do need some throttling on the internal interfaces if we do
not want to run out of memory!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190827161726.3640-1-chris@chris-wilson.co.uk
2019-08-28 08:32:02 +01:00
Chris Wilson
21b0c32bda
drm/i915/selftests: Markup impossible error pointers
...
If we create a new live_context() we should have a mapping for each
engine. Document that assumption with an assertion.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com >
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Matthew Auld <matthew.auld@intel.com >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190827094933.13778-1-chris@chris-wilson.co.uk
2019-08-27 13:22:01 +01:00
Chris Wilson
75b974a859
drm/i915/selftests: Teach igt_gpu_fill_dw() to take intel_context
...
Avoid having to pass around (ctx, engine) everywhere by passing the
actual intel_context we intend to use. Today we preach this lesson to
igt_gpu_fill_dw and its callers' callers.
The immediate benefit for the GEM selftests is that we aim to use the
GEM context as the control, the source of the engines on which to test
the GEM context.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Matthew Auld <matthew.auld@intel.com >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190823235141.31799-1-chris@chris-wilson.co.uk
2019-08-24 09:10:08 +01:00
Chris Wilson
70d6894d14
drm/i915: Serialize against vma moves
...
Make sure that when submitting requests, we always serialize against
potential vma moves and clflushes.
Time for a i915_request_await_vma() interface!
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Matthew Auld <matthew.auld@intel.com >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190819112033.30638-1-chris@chris-wilson.co.uk
2019-08-19 15:25:56 +01:00
Matthew Auld
18851edfa7
drm/i915/selftests: move gpu-write-dw into utils
...
Using the gpu to write to some dword over a number of pages is rather
useful, and we already have two copies of such a thing, and we don't
want a third so move it to utils. There is probably some other stuff
also...
Signed-off-by: Matthew Auld <matthew.auld@intel.com >
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk >
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Link: https://patchwork.freedesktop.org/patch/msgid/20190810105008.14320-1-chris@chris-wilson.co.uk
2019-08-10 13:12:34 +01:00
Chris Wilson
f277bc0c98
drm/i915/selftests: Pass intel_context to igt_spinner
...
Teach igt_spinner to only use our internal structs, decoupling the
interface from the GEM contexts. This makes it easier to avoid
requiring ce->gem_context back references for kernel_context that may
have them in future.
v2: Lift engine lock to verify_wa() caller.
v3: Less than v2, but more so
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190731081126.9139-1-chris@chris-wilson.co.uk
2019-07-31 09:45:27 +01:00
Chris Wilson
f5d974f9d2
drm/i915/gt: Provide a local intel_context.vm
...
Track the currently bound address space used by the HW context. Minor
conversions to use the local intel_context.vm are made, leaving behind
some more surgery required to make intel_context the primary through the
selftests.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190730143209.4549-2-chris@chris-wilson.co.uk
2019-07-30 16:09:35 +01:00
Chris Wilson
c082afac86
drm/i915: Move aliasing_ppgtt underneath its i915_ggtt
...
The aliasing_ppgtt provides a PIN_USER alias for the global gtt, so move
it under the i915_ggtt to simplify later transformations to enable
intel_context.vm.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190730143209.4549-1-chris@chris-wilson.co.uk
2019-07-30 16:09:32 +01:00
Chris Wilson
cb823ed991
drm/i915/gt: Use intel_gt as the primary object for handling resets
...
Having taken the first step in encapsulating the functionality by moving
the related files under gt/, the next step is to start encapsulating by
passing around the relevant structs rather than the global
drm_i915_private. In this step, we pass intel_gt to intel_reset.c
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com >
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190712192953.9187-1-chris@chris-wilson.co.uk
2019-07-12 21:06:56 +01:00
Chris Wilson
bf1315b830
drm/i915/selftests: Ensure we don't clamp a random offset to 32b
...
Specify that we do want a 64b value for sizeof(u32) as we want to
compute the mask of the upper 62bits.
v2: Use round_down() for automatic type promotion
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com >
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190710161413.7115-1-chris@chris-wilson.co.uk
2019-07-11 10:06:37 +01:00
Chris Wilson
8f856c743c
drm/i915/selftests: Be engine agnostic
...
When using MI operations, we do not care which engine we use, so use
them all where possible, and where inconvenient double check we have the
engine we selected at random.
v2: Drop the local copy of engine->sseu to avoid an unchecked deref
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190704212343.6820-1-chris@chris-wilson.co.uk
2019-07-04 22:55:00 +01:00
Chris Wilson
6582f4f613
drm/i915/selftests: Drain the freedlists between exec passes
...
During the context execution tests, we issue a lot of work and discard a
lot of objects without releasing the lock and allowing the background
reaper to free those objects. Insert a small break between each pass to
flush the worker.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190704165317.21060-1-chris@chris-wilson.co.uk
2019-07-04 22:26:31 +01:00
Chris Wilson
63251685c1
drm/i915/selftests: Common live setup/teardown
...
We frequently, but not frequently enough!, remember to flush residual
operations and objects at the end of a live subtest. The purpose is to
cleanup after every subtest, leaving a clean slate for the next subtest,
and perform early detection of leaky state. As this should ideally be
common for all live subtests, pull the task into a common teardown
routine.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190703091726.11690-1-chris@chris-wilson.co.uk
2019-07-03 11:07:57 +01:00
Chris Wilson
9e9539800d
drm/i915: Remove waiting & retiring from shrinker paths
...
i915_gem_wait_for_idle() and i915_retire_requests() introduce a
dependency on the timeline->mutex. This is problematic as we want to
later perform allocations underneath i915_active.mutex, forming a link
between the shrinker, the timeline and active mutexes. Nip this cycle in
the bud by removing the acquisition of the timeline mutex (i.e.
retiring) from inside the shrinker.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190621183801.23252-1-chris@chris-wilson.co.uk
2019-06-21 19:47:48 +01:00
Tvrtko Ursulin
ba4134a419
drm/i915: Save trip via top-level i915 in a few more places
...
For gt related operations it makes more logical sense to stay in the realm
of gt instead of dereferencing via driver i915.
This patch handles a few of the easy ones with work requiring more
refactoring still outstanding.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk >
Link: https://patchwork.freedesktop.org/patch/msgid/20190621070811.7006-30-tvrtko.ursulin@linux.intel.com
2019-06-21 13:48:48 +01:00
Chris Wilson
60a0933bff
drm/i915/selftests: Use request managed wakerefs
...
Since commit 79ffac859979ffac8599chris@chris-wilson.co.uk >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190620102432.31580-1-chris@chris-wilson.co.uk
2019-06-20 19:13:26 +01:00
Chris Wilson
2f5309452d
drm/i915: Stop passing I915_WAIT_LOCKED to i915_request_wait()
...
Since commit eb8d0f5af4eb8d0f5af4chris@chris-wilson.co.uk >
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190618074153.16055-3-chris@chris-wilson.co.uk
2019-06-19 12:58:38 +01:00
Daniele Ceraolo Spurio
c447ff7db3
drm/i915: update with_intel_runtime_pm to use the rpm structure
...
Matching the underlying get/put functions.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com >
Cc: Imre Deak <imre.deak@intel.com >
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk >
Acked-by: Imre Deak <imre.deak@intel.com >
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Link: https://patchwork.freedesktop.org/patch/msgid/20190613232156.34940-8-daniele.ceraolospurio@intel.com
2019-06-14 15:58:33 +01:00
Daniele Ceraolo Spurio
d858d5695f
drm/i915: update rpm_get/put to use the rpm structure
...
The functions where internally already only using the structure, so we
need to just flip the interface.
v2: rebase
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com >
Cc: Imre Deak <imre.deak@intel.com >
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk >
Acked-by: Imre Deak <imre.deak@intel.com >
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Link: https://patchwork.freedesktop.org/patch/msgid/20190613232156.34940-7-daniele.ceraolospurio@intel.com
2019-06-14 15:58:33 +01:00
Chris Wilson
a09507682e
drm/i915: Discard some redundant cache domain flushes
...
Since commit a679f58d05chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.william.auld@gmail.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190614111053.25615-1-chris@chris-wilson.co.uk
2019-06-14 12:16:30 +01:00
Chris Wilson
e568ac3874
drm/i915: Pull kref into i915_address_space
...
Make the kref common to both derived structs (i915_ggtt and i915_ppgtt)
so that we can safely reference count an abstract ctx->vm address space.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190611091238.15808-1-chris@chris-wilson.co.uk
2019-06-11 11:44:24 +01:00
Chris Wilson
1fe2d6f94f
drm/i915: Skip context_barrier emission for unused contexts
...
The intent was to skip unused HW contexts by checking ce->state.
However, this only works for execlists where the ppGTT pointers is
stored inside the HW context. For gen7, the ppGTT is alongside the
logical state and must be updated on all active engines but, crucially,
only on active engines. As we need different checks, and to keep
context_barrier_task() agnostic, pass in the predicate.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110836
Fixes: 62c8e42345chris@chris-wilson.co.uk >
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com >
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190604152408.24468-1-chris@chris-wilson.co.uk
2019-06-06 13:35:16 +01:00
Chris Wilson
c017cf6b1a
drm/i915: Drop the deferred active reference
...
An old optimisation to reduce the number of atomics per batch sadly
relies on struct_mutex for coordination. In order to remove struct_mutex
from serialising object/context closing, always taking and releasing an
active reference on first use / last use greatly simplifies the locking.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190528092956.14910-15-chris@chris-wilson.co.uk
2019-05-28 12:45:29 +01:00
Chris Wilson
6951e5893b
drm/i915: Move GEM object domain management from struct_mutex to local
...
Use the per-object local lock to control the cache domain of the
individual GEM objects, not struct_mutex. This is a huge leap forward
for us in terms of object-level synchronisation; execbuffers are
coordinated using the ww_mutex and pread/pwrite is finally fully
serialised again.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Matthew Auld <matthew.auld@intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190528092956.14910-10-chris@chris-wilson.co.uk
2019-05-28 12:45:29 +01:00
Chris Wilson
10be98a77c
drm/i915: Move more GEM objects under gem/
...
Continuing the theme of separating out the GEM clutter.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk >
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com >
Link: https://patchwork.freedesktop.org/patch/msgid/20190528092956.14910-8-chris@chris-wilson.co.uk
2019-05-28 12:45:29 +01:00
