| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Begin pulling the GT setup underneath a single GT umbrella; let intel_gt
take ownership of its engines! As hinted, the complication is the
lifetime of the probed engine versus the active lifetime of the GT
backends. We need to detect the engine layout early and keep it until
the end so that we can sanitize state on takeover and release.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andi Shyti <andi.shyti@intel.com>
Acked-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191222120752.1368352-1-chris@chris-wilson.co.uk
|
|
As we stash a pointer to the HWSP cacheline on the request, when reading
it we only need confirm that the cacheline is still valid by checking
that the request and timeline are still intact.
v2: Protect hwsp_cachline with RCU
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/20191217011659.3092130-1-chris@chris-wilson.co.uk
|
|
The major drawback of commit 7e34f4e4aad3 ("drm/i915/gen8+: Add RC6 CTX
corruption WA") is that it disables RC6 while Skylake (and friends) is
active, and we do not consider the GPU idle until all outstanding
requests have been retired and the engine switched over to the kernel
context. If userspace is idle, this task falls onto our background idle
worker, which only runs roughly once a second, meaning that userspace has
to have been idle for a couple of seconds before we enable RC6 again.
Naturally, this causes us to consume considerably more energy than
before as powersaving is effectively disabled while a display server
(here's looking at you Xorg) is running.
As execlists will get a completion event as each context is completed,
we can use this interrupt to queue a retire worker bound to this engine
to cleanup idle timelines. We will then immediately notice the idle
engine (without userspace intervention or the aid of the background
retire worker) and start parking the GPU. Thus during light workloads,
we will do much more work to idle the GPU faster... Hopefully with
commensurate power saving!
v2: Watch context completions and only look at those local to the engine
when retiring to reduce the amount of excess work we perform.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=112315
References: 7e34f4e4aad3 ("drm/i915/gen8+: Add RC6 CTX corruption WA")
References: 2248a28384fe ("drm/i915/gen8+: Add RC6 CTX corruption WA")
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/20191125105858.1718307-3-chris@chris-wilson.co.uk
|
|
The general concept was that intel_timeline.active_count was locked by
the intel_timeline.mutex. The exception was for power management, where
the engine->kernel_context->timeline could be manipulated under the
global wakeref.mutex.
This was quite solid, as we always manipulated the timeline only while
we held an engine wakeref.
And then we started retiring requests outside of struct_mutex, only
using the timelines.active_list and the timeline->mutex. There we
started manipulating intel_timeline.active_count outside of an engine
wakeref, and so introduced a race between __engine_park() and
intel_gt_retire_requests(), a race that could result in the
engine->kernel_context not being added to the active timelines and so
losing requests, which caused us to keep the system permanently powered
up [and unloadable].
The race would be easy to close if we could take the engine wakeref for
the timeline before we retire -- except timelines are not bound to any
engine and so we would need to keep all active engines awake. The
alternative is to guard intel_timeline_enter/intel_timeline_exit for use
outside of the timeline->mutex.
Fixes: e5dadff4b093 ("drm/i915: Protect request retirement with timeline->mutex")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191120165514.3955081-1-chris@chris-wilson.co.uk
|
|
Forgo the struct_mutex serialisation for i915_active, and interpose its
own mutex handling for active/retire.
This is a multi-layered sleight-of-hand. First, we had to ensure that no
active/retire callbacks accidentally inverted the mutex ordering rules,
nor assumed that they were themselves serialised by struct_mutex. More
challenging though, is the rule over updating elements of the active
rbtree. Instead of the whole i915_active now being serialised by
struct_mutex, allocations/rotations of the tree are serialised by the
i915_active.mutex and individual nodes are serialised by the caller
using the i915_timeline.mutex (we need to use nested spinlocks to
interact with the dma_fence callback lists).
The pain point here is that instead of a single mutex around execbuf, we
now have to take a mutex for active tracker (one for each vma, context,
etc) and a couple of spinlocks for each fence update. The improvement in
fine grained locking allowing for multiple concurrent clients
(eventually!) should be worth it in typical loads.
v2: Add some comments that barely elucidate anything :(
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/20191004134015.13204-6-chris@chris-wilson.co.uk
|
|
The request->timeline is only valid until the request is retired (i.e.
before it is completed). Upon retiring the request, the context may be
unpinned and freed, and along with it the timeline may be freed. We
therefore need to be very careful when chasing rq->timeline that the
pointer does not disappear beneath us. The vast majority of users are in
a protected context, either during request construction or retirement,
where the timeline->mutex is held and the timeline cannot disappear. It
is those few off the beaten path (where we access a second timeline) that
need extra scrutiny -- to be added in the next patch after first adding
the warnings about dangerous access.
One complication, where we cannot use the timeline->mutex itself, is
during request submission onto hardware (under spinlocks). Here, we want
to check on the timeline to finalize the breadcrumb, and so we need to
impose a second rule to ensure that the request->timeline is indeed
valid. As we are submitting the request, it's context and timeline must
be pinned, as it will be used by the hardware. Since it is pinned, we
know the request->timeline must still be valid, and we cannot submit the
idle barrier until after we release the engine->active.lock, ergo while
submitting and holding that spinlock, a second thread cannot release the
timeline.
v2: Don't be lazy inside selftests; hold the timeline->mutex for as long
as we need it, and tidy up acquiring the timeline with a bit of
refactoring (i915_active_add_request)
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/20190919111912.21631-1-chris@chris-wilson.co.uk
|
|
In preparation for removing struct_mutex from around context retirement,
we need to make timeline pinning and unpinning safe. Since multiple
engines/contexts can share a single timeline, we cannot rely on
borrowing the context mutex (otherwise we could state that the timeline
is only pinned/unpinned inside the context pin/unpin and so guarded by
it). However, we only perform a sequence of atomic operations inside the
timeline pin/unpin and the sequence of those operations is safe for a
concurrent unpin / pin, so we can relax the struct_mutex requirement.
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/20190815205709.24285-3-chris@chris-wilson.co.uk
|
|
Lift moving the timeline to/from the active_list on enter/exit in order
to shorten the active tracking span in comparison to the existing
pin/unpin.
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/20190815205709.24285-1-chris@chris-wilson.co.uk
|
|
Move all timeline code under gt and rename to intel_gt prefix.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20190621070811.7006-32-tvrtko.ursulin@linux.intel.com
|
Chris Wilson
Tvrtko Ursulin