Back in commit a4b2b01523 ("drm/i915: Don't mark an execlists
context-switch when idle") we noticed the presence of late
context-switch interrupts. We were able to filter those out by looking
at whether the ELSP remained active, but in commit beecec9017
("drm/i915/execlists: Preemption!") that became problematic as we now
anticipate receiving a context-switch event for preemption while ELSP
may be empty. To restore the spurious interrupt suppression, add a
counter for the expected number of pending context-switches and skip if
we do not need to handle this interrupt to make forward progress.
v2: Don't forget to switch on for preempt.
v3: Reduce the counter to a on/off boolean tracker. Declare the HW as
active when we first submit, and idle after the final completion event
(with which we confirm the HW says it is idle), and track each source
of activity separately. With a finite number of sources, it should aide
us in debugging which gets stuck.
Fixes: beecec9017 ("drm/i915/execlists: Preemption!")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michal Winiarski <michal.winiarski@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171023213237.26536-3-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
(cherry picked from commit 4a118ecbe9)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
In the next patch, we want to reduce the lock coverage within the
shrinker, and one of the dangerous walks we have is over obj->vma_list.
We are only walking the obj->vma_list in order to check whether it has
been permanently pinned by HW access, typically via use on the scanout.
But we have a couple of other long term pins, the context objects for
which we currently have to check the individual vma pin_count. If we
instead mark these using obj->pin_display, we can forgo the dangerous
and sometimes slow list iteration.
v2: Rearrange code to try and avoid confusion from false associations
due to arrangement of whitespace along with rebasing on obj->pin_global.
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/20171013202621.7276-4-chris@chris-wilson.co.uk
Let GVT-g VM read the CSB and CSB write pointer from virtual HWSP, not all
the host support this feature, need to check the BIT(3) of caps in PVINFO.
v3 : Remove unnecessary comments.
v4 : Separate VM enable patch with GVT-g implementation patch due to code
dependency.
v5 : Use inline for GVT virtual HWSP caps check function.
v6 : Comments refine.
Signed-off-by: Weinan Li <weinan.z.li@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1508039725-1066-1-git-send-email-weinan.z.li@intel.com
There is function to tell how many ports we have, so use it.
We still have direct relationship with array size and port count,
so no harm was done.
Fixes: 76e70087d3 ("drm/i915: Make execlist port count variable")
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171010114857.13108-1-mika.kuoppala@intel.com
Michel Thierry noticed that we were applying WaDisableCtxRestoreArbitration
even to gen9, which does not require the w/a. The rationale is that we
need to enable MI arbitration for execlists to work, and to be safe we
do that before every batch (in addition to every context switch into the
batch). Since this is not clear from the single line comment suggesting
the MI_ARB_ENABLE is solely for the w/a, add a little more detail.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171005191005.13462-1-chris@chris-wilson.co.uk
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
When we write to ELSP, it triggers a context preemption at the earliest
arbitration point (3DPRIMITIVE, some PIPECONTROLs, a few other
operations and the explicit MI_ARB_CHECK). If this is to the same
context, it triggers a LITE_RESTORE where the RING_TAIL is merely
updated (used currently to chain requests from the same context
together, avoiding bubbles). However, if it is to a different context, a
full context-switch is performed and it will start to execute the new
context saving the image of the old for later execution.
Previously we avoided preemption by only submitting a new context when
the old was idle. But now we wish embrace it, and if the new request has
a higher priority than the currently executing request, we write to the
ELSP regardless, thus triggering preemption, but we tell the GPU to
switch to our special preemption context (not the target). In the
context-switch interrupt handler, we know that the previous contexts
have finished execution and so can unwind all the incomplete requests
and compute the new highest priority request to execute.
It would be feasible to avoid the switch-to-idle intermediate by
programming the ELSP with the target context. The difficulty is in
tracking which request that should be whilst maintaining the dependency
change, the error comes in with coalesced requests. As we only track the
most recent request and its priority, we may run into the issue of being
tricked in preempting a high priority request that was followed by a
low priority request from the same context (e.g. for PI); worse still
that earlier request may be our own dependency and the order then broken
by preemption. By injecting the switch-to-idle and then recomputing the
priority queue, we avoid the issue with tracking in-flight coalesced
requests. Having tried the preempt-to-busy approach, and failed to find
a way around the coalesced priority issue, Michal's original proposal to
inject an idle context (based on handling GuC preemption) succeeds.
The current heuristic for deciding when to preempt are only if the new
request is of higher priority, and has the privileged priority of
greater than 0. Note that the scheduler remains unfair!
v2: Disable for gen8 (bdw/bsw) as we need additional w/a for GPGPU.
Since, the feature is now conditional and not always available when we
have a scheduler, make it known via the HAS_SCHEDULER GETPARAM (now a
capability mask).
v3: Stylistic tweaks.
v4: Appease Joonas with a snippet of kerneldoc, only to fuel to fire of
the preempt vs preempting debate.
Suggested-by: Michal Winiarski <michal.winiarski@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michal Winiarski <michal.winiarski@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Ben Widawsky <benjamin.widawsky@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: Zhi Wang <zhi.a.wang@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003203453.15692-8-chris@chris-wilson.co.uk
With preemption, we will want to "unsubmit" a request, taking it back
from the hw and returning it to the priority sorted execution list. In
order to know where to insert it into that list, we need to remember
its adjust priority (which may change even as it was being executed).
This also affects reset for execlists as we are now unsubmitting the
requests following the reset (rather than directly writing the ELSP for
the inflight contexts). This turns reset into an accidental preemption
point, as after the reset we may choose a different pair of contexts to
submit to hw.
GuC is not updated as this series doesn't add preemption to the GuC
submission, and so it can keep benefiting from the early pruning of the
DFS inside execlists_schedule() for a little longer. We also need to
find a way of reducing the cost of that DFS...
v2: Include priority in error-state
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003203453.15692-6-chris@chris-wilson.co.uk
Let the listener know that the context we just scheduled out was not
complete, and will be scheduled back in at a later point.
v2: Handle CONTEXT_STATUS_PREEMPTED in gvt by aliasing it to
CONTEXT_STATUS_OUT for the moment, gvt can expand upon the difference
later.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: "Zhenyu Wang" <zhenyuw@linux.intel.com>
Cc: "Wang, Zhi A" <zhi.a.wang@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003203453.15692-3-chris@chris-wilson.co.uk
This has been unused since commit afa8ce5b30
("drm/i915: Nuke legacy flip queueing code").
Changes since v1:
- Rebase on top of all the changes to modparams.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
\o/-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171004094416.31306-1-maarten.lankhorst@linux.intel.com
Avoid the repeated rbtree lookup for each request as we unwind them by
tracking the last priolist.
v2: Fix up my unhelpful suggestion of using default_priolist.
Signed-off-by: Michał Winiarski <michal.winiarski@intel.com>
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/20170928193910.17988-4-chris@chris-wilson.co.uk
We use INT_MIN to denote the priority of a request that has not been
submitted to the scheduler; we treat INT_MIN as an invalid priority and
initialise the request to it. Give the value a name so it stands out.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170928193910.17988-3-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com
In the future, we will want to unwind requests following a preemption
point. This requires the same steps as for unwinding upon a reset, so
extract the existing code to a separate function for later use.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170928193910.17988-2-chris@chris-wilson.co.uk
When cancelling requests, also send the notification to any listeners
(gvt) that the request is no longer scheduled on hw. They may require to
keep the in/out exactly balanced, and so the reuse after the reset may
confuse the listener.
Fixes: 221ab9719b ("drm/i915/execlists: Unwind incomplete requests on resets")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: "Zhenyu Wang" <zhenyuw@linux.intel.com>
Cc: "Wang, Zhi A" <zhi.a.wang@intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170926101720.9479-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Since we inherited the context image setup from gen8 which needed a
per-bb workaround (for GPGPU), we are submitting an empty per-bb buffer
on gen9. Now that we can skip adding the buffer to the context image,
remove the dangling per-bb. This slightly improves execution latency,
most notably on an idle engine.
References: https://bugs.freedesktop.org/show_bug.cgi?id=87725
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170921135444.27330-2-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
The per-context and per-batch workaround buffers are optional, yet we
tell the GPU to execute them even if they contain no instructions. Doing
so incurs the dispatch latency, which we can avoid if we don't ask the
GPU to execute the no-op buffers. Allow ourselves to skip setup of empty
buffer, and then to only enable non-empty buffers in the context image.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170921135444.27330-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
As we emulate execlists on top of the GuC workqueue, it is not
restricted to just 2 ports and we can increase that number arbitrarily
to trade-off queue depth (i.e. scheduling latency) against pipeline
bubbles.
v2: rebase. better commit msg (Chris)
v3: rebase
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20170922124307.10914-5-mika.kuoppala@intel.com
When first execlist entry is processed, we move the port (contents).
Introduce function for this as execlist and guc use this common
operation.
v2: rebase. s/GEM_DEBUG_BUG/GEM_BUG (Chris)
v3: rebase
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20170922124307.10914-4-mika.kuoppala@intel.com
On reset and wedged path, we want to release the requests
that are tied to ports and then mark the ports to be unset.
Introduce a function for this.
v2: rebase
v3: drop local, keep GEM_BUG_ON (Michał, Chris)
v4: rebase
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20170922124307.10914-3-mika.kuoppala@intel.com
Engine's execlist related items have been increasing to
a point where a separate struct is warranted. Carve execlist
specific items to a dedicated struct to add clarity.
v2: add kerneldoc and fix whitespace (Joonas, Chris)
v3: csb_mmio changes, rebase
v4: s/\b(el|execlist)\b/execlists/ (Joonas)
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Acked-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com> (v3)
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> (v3)
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170922124307.10914-1-mika.kuoppala@intel.com
Our global struct with params is named exactly the same way
as new preferred name for the drm_i915_private function parameter.
To avoid such name reuse lets use different name for the global.
v5: pure rename
v6: fix
Credits-to: Coccinelle
@@
identifier n;
@@
(
- i915.n
+ i915_modparams.n
)
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Ville Syrjala <ville.syrjala@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170919193846.38060-1-michal.wajdeczko@intel.com
To create an upper bound on number of GuC workitems, we need to change
the way that requests are being submitted. Rather than submitting each
request as an individual workitem, we can do coalescing in a similar way
we're handlig execlist submission ports. We also need to stop pretending
that we're doing "lite-restore" in GuC submission (we would create a
workitem each time we hit this condition). This allows us to completely
remove the reservation, replacing it with a compile time check.
v2: Also coalesce when replaying on reset (Daniele)
v3: Consistent wq_resv - per-request (Daniele)
v4: Squash removing wq_resv
v5: Reflect i915_guc_submit argument changes in doc
v6: Rebase on top of execlists reset/restart fix (Chris,Michał)
References: https://bugs.freedesktop.org/show_bug.cgi?id=101873
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Jeff McGee <jeff.mcgee@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20170914083216.10192-2-michal.winiarski@intel.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Given the mechanism to unwind and replay requests (designed to support
preemption), we have an alternative to the current method of
resubmitting the ELSP upon reset. Resubmitting ELSP turns out to be more
complicated than expected, due to having to handle lost context-switch
interrupts and so guessing what ELSP we need to resubmit later. Instead,
by unwinding the requests and clearing the ELSP tracking entirely, we
can then just dequeue the first pair of ready requests after resetting,
using the normal submission procedure.
Currently, the unwound requests have maximum priority and so are
guaranteed to be resubmitted upon resume. If we are lucky, we may be
able to coalesce a new request on top!
Suggested-by: Michał Winiarski <michal.winiarski@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170916204414.32762-4-chris@chris-wilson.co.uk
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
In the next patch we will want to reinsert a request not at the end of
the priority queue, but at the front. Here we split insert_request()
into two, the first function retrieves the priority list (for reuse for
unsubmit later) and a wrapper function to insert at the end of that list
and to schedule the tasklet if we were first.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170916204414.32762-3-chris@chris-wilson.co.uk
During a reset, we may skip over completed requests and lost
context-switch interrupts. Following the reset, we may then may end up
with no active requests in the ELSP (and so do not resubmit to restart
the engine), but have a queue of requests ready for execution. This is
unlikely, it requires the last request to complete after the hang is
detected, but not impossible. The outcome of this is that the engine
stalls, possibly leading to full ring and indefinite wait under
struct_mutex, eventually leading to a full driver hang.
Alternatively, we can solve this by unsubmitting the incomplete requests
and just kickstarting the tasklet. Michał has patches for that, which I
initially disliked due to the extra complexity, but the complexity of
this "simple" restart is growing...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170916204414.32762-1-chris@chris-wilson.co.uk
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
When wedging the hw, we want to mark all in-flight requests as -EIO.
This is made slightly more complex by execlists who store the ready but
not yet submitted-to-hw requests on a private queue (an rbtree
priolist). Call into execlists to cancel not only the ELSP tracking for
the submitted requests, but also the queue of unsubmitted requests.
v2: Move the majority of engine_set_wedged to the backends (both legacy
ringbuffer and execlists handling their own lists).
Reported-by: Michał Winiarski <michal.winiarski@intel.com>
Testcase: igt/gem_eio/in-flight-contexts
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170915173100.26470-1-chris@chris-wilson.co.uk
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
The engine also provides a mirror of the CSB write pointer in the HWSP,
but not of our read pointer. To take advantage of this we need to
remember where we read up to on the last interrupt and continue off from
there. This poses a problem following a reset, as we don't know where
the hw will start writing from, and due to the use of power contexts we
cannot perform that query during the reset itself. So we continue the
current modus operandi of delaying the first read of the context-status
read/write pointers until after the first interrupt. With this we should
now have eliminated all uncached mmio reads in handling the
context-status interrupt, though we still have the uncached mmio writes
for submitting new work, and many uncached mmio reads in the global
interrupt handler itself. Still a step in the right direction towards
reducing our resubmit latency, although it appears lost in the noise!
v2: Cannonlake moved the CSB write index
v3: Include the sw/hwsp state in debugfs/i915_engine_info
v4: Also revert to using CSB mmio for GVT-g
v5: Prevent the compiler reloading tail (Mika)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: Zhi Wang <zhi.a.wang@intel.com>
Acked-by: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170913085605.18299-6-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
The engine provides a mirror of the CSB in the HWSP. If we use the
cacheable reads from the HWSP, we can shave off a few mmio reads per
context-switch interrupt (which are quite frequent!). Just removing a
couple of mmio is not enough to actually reduce any latency, but a small
reduction in overall cpu usage.
Much appreciation for Ben dropping the bombshell that the CSB was in the
HWSP and for Michel in digging out the details.
v2: Don't be lazy, add the defines for the indices.
v3: Include the HWSP in debugfs/i915_engine_info
v4: Check for GVT-g, it currently depends on intercepting CSB mmio
v5: Fixup GVT-g mmio path
v6: Disable HWSP if VT-d is active as the iommu adds unpredictable
memory latency. (Mika)
v7: Also markup the CSB read with READ_ONCE() as it may still be an mmio
read and we want to stop the compiler from issuing a later (v.slow) reload.
Suggested-by: Ben Widawsky <benjamin.widawsky@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: Zhi Wang <zhi.a.wang@intel.com>
Acked-by: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170913133534.26927-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
On gen8+ we're currently using the PPHWSP of the kernel ctx as the
global HWSP. However, when the kernel ctx gets submitted (e.g. from
__intel_autoenable_gt_powersave) the HW will use that page as both
HWSP and PPHWSP. This causes a conflict in the register arena of the
HWSP, i.e. dword indices below 0x30. We don't current utilize this arena,
but in the following patches we will take advantage of the cached
register state for handling execlist's context status interrupt.
To avoid the conflict, instead of re-using the PPHWSP of the kernel
ctx we can allocate a separate page for the HWSP like what happens for
pre-execlists platform.
v2: Add a use-case for the register arena of the HWSP.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1499357440-34688-1-git-send-email-daniele.ceraolospurio@intel.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170913085605.18299-3-chris@chris-wilson.co.uk
Not only the context image consist of two parts (the PPHWSP, and the
logical context state), but we also allocate a header at the start of
for sharing data with GuC. Thus every lrc looks like this:
| [guc] | [hwsp] [logical state] |
|<- our header ->|<- context image ->|
So far, we have oversimplified whenever we use each of these parts of the
context, just because the GuC header happens to be in page 0, and the
(PP)HWSP is in page 1. But this had led to using the same define for more
than one meaning (as a page index in the lrc and as 1 page).
This patch adds defines for the GuC shared page, the PPHWSP page and the
start of the logical state. It also updated the places where the old
define was being used. Since we are not changing the size (or format) of
the context, there are no functional changes.
v2: Use PPHWSP index for hws again.
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: intel-gvt-dev@lists.freedesktop.org
Link: http://patchwork.freedesktop.org/patch/msgid/20170712193032.27080-1-michel.thierry@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/20170913085605.18299-1-chris@chris-wilson.co.uk
The context descriptor is stored inside the per-engine context state, as
we only need to compute it once and access it frequently. However,
currently only intel_lrc.c has easy access, but i915_guc_submission.c
would like to frequently read it as well, and more so only ever needs
the lower 32bits. Make it an inline as the compiler should be able to
retrieve the value in less instructions than it takes to do the function
call:
add/remove: 0/1 grow/shrink: 1/0 up/down: 8/-45 (-37)
function old new delta
i915_guc_submit 621 629 +8
intel_lr_context_descriptor 45 - -45
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20170912214905.21987-1-chris@chris-wilson.co.uk
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Let's inherit workarounds from previous platforms that
according to wa_database and BSpec are still valid for
Cannonlake.
v2: Add missed workarounds.
v3: Rebase
v4: Remove bad chunk that was added to rc6 disable. (Ander)
Also remove A0 W/a that are not needed anymore.
v5: Rebase on top of CFL.
v6: Remove empty gen9_init_perctx_bb and gen9_init_indirectctx_bb
since they don't carry any gen10 related W/a. (by Oscar).
Also Remove A0 exclusive workaround.
v7: Remove more A0 exclusive workarounds. As pointed out by Oscar
many workarounds were changed to be A0 only so let's remove
them.
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170815231651.975-1-rodrigo.vivi@intel.com
During a global reset, we disable the irq. As we disable the irq, the
hardware may be raising a GT interrupt that we then ignore, leaving it
pending in the GTIIR. After the reset, we then re-enable the irq,
triggering the pending interrupt. However, that interrupt was for the
stale state from before the reset, and the contents of the CSB buffer
are now invalid.
v2: Add a comment to make it clear that the double clear is purely my
paranoia.
Reported-by: "Dong, Chuanxiao" <chuanxiao.dong@intel.com>
Fixes: 821ed7df6e ("drm/i915: Update reset path to fix incomplete requests")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: "Dong, Chuanxiao" <chuanxiao.dong@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Michal Winiarski <michal.winiarski@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170807121919.30165-1-chris@chris-wilson.co.uk
Link: https://patchwork.freedesktop.org/patch/msgid/20170818090509.5363-1-chris@chris-wilson.co.uk
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
When doing a GPU reset, the CSB register will be trashed and we will
lose any context-switch notifications that happened since the tasklet
was disabled. If we find that all requests on this engine were
completed, we want to make sure that the ELSP tracker is similarly empty
so that we do not feed back in the completed requests upon recovering
from the reset.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170721123238.16428-4-chris@chris-wilson.co.uk
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Enables access to OA unit metrics for BDW, CHV, SKL and BXT which all
share (more-or-less) the same OA unit design.
Of particular note in comparison to Haswell: some OA unit HW config
state has become per-context state and as a consequence it is somewhat
more complicated to manage synchronous state changes from the cpu while
there's no guarantee of what context (if any) is currently actively
running on the gpu.
The periodic sampling frequency which can be particularly useful for
system-wide analysis (as opposed to command stream synchronised
MI_REPORT_PERF_COUNT commands) is perhaps the most surprising state to
have become per-context save and restored (while the OABUFFER
destination is still a shared, system-wide resource).
This support for gen8+ takes care to consider a number of timing
challenges involved in synchronously updating per-context state
primarily by programming all config state from the cpu and updating all
current and saved contexts synchronously while the OA unit is still
disabled.
The driver intentionally avoids depending on command streamer
programming to update OA state considering the lack of synchronization
between the automatic loading of OACTXCONTROL state (that includes the
periodic sampling state and enable state) on context restore and the
parsing of any general purpose BB the driver can control. I.e. this
implementation is careful to avoid the possibility of a context restore
temporarily enabling any out-of-date periodic sampling state. In
addition to the risk of transiently-out-of-date state being loaded
automatically; there are also internal HW latencies involved in the
loading of MUX configurations which would be difficult to account for
from the command streamer (and we only want to enable the unit when once
the MUX configuration is complete).
Since the Gen8+ OA unit design no longer supports clock gating the unit
off for a single given context (which effectively stopped any progress
of counters while any other context was running) and instead supports
tagging OA reports with a context ID for filtering on the CPU, it means
we can no longer hide the system-wide progress of counters from a
non-privileged application only interested in metrics for its own
context. Although we could theoretically try and subtract the progress
of other contexts before forwarding reports via read() we aren't in a
position to filter reports captured via MI_REPORT_PERF_COUNT commands.
As a result, for Gen8+, we always require the
dev.i915.perf_stream_paranoid to be unset for any access to OA metrics
if not root.
v5: Drain submitted requests when enabling metric set to ensure no
lite-restore erases the context image we just updated (Lionel)
v6: In addition to drain, switch to kernel context & update all
context in place (Chris)
v7: Add missing mutex_unlock() if switching to kernel context fails
(Matthew)
v8: Simplify OA period/flex-eu-counters programming by using the
batchbuffer instead of modifying ctx-image (Lionel)
v9: Back to updating the context image (due to erroneous testing,
batchbuffer programming the OA unit doesn't actually work)
(Lionel)
Pin context before updating context image (Chris)
Drop MMIO programming now that we switch to a kernel context with
right values in initial context image (Chris)
v10: Just pin_map the contexts we want to modify or let the
configuration happen on first use (Chris)
v11: Update kernel context OA config through the batchbuffer rather
than on the fly ctx-image update (Lionel)
v12: Rework OA context registers update again by swithing away from
user contexts and reconfiguring the kernel context through the
batchbuffer and updating all the other contexts' context image.
Also take care to lock slice/subslice configuration when OA is
on. (Lionel)
v13: Request rpcs updates on all engine when updating the OA config
(Lionel)
v14: Drop any kind of rpcs management now that we monitor sseu
configuration changes in a later patch (Lionel)
Remove usleep after programming the NOA configs on Gen8+, this
doesn't seem to be needed (Lionel)
v15: Respect coding style for block comments (Chris)
v16: Add missing i915_add_request() in case we fail to emit OA
configuration (Matthew)
Signed-off-by: Robert Bragg <robert@sixbynine.org>
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com> \o/
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Indirect Context Offset Pointer has changed for Cannonlake.
INDIRECT_CTX_OFFSET[15:6] valid value for CNL is 19h per Spec.
v2: rebased to intel_lr_indirect_ctx_offset
v3: Commit message added per Tvrtko request.
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1496781040-20888-9-git-send-email-rodrigo.vivi@intel.com
If we do not require to perform priority bumping, and we haven't yet
submitted the request, we can update its priority in situ and skip
acquiring the engine locks -- thus avoiding any contention between us
and submit/execute.
v2: Remove the stack element from the list if we can do the early
assignment.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170517121007.27224-10-chris@chris-wilson.co.uk
The i915_priolist are allocated within an atomic context on a path where
we wish to minimise latency. If we use a dedicated kmem_cache, we have
the advantage of a local freelist from which to service new requests
that should keep the latency impact of an allocation small. Though
currently we expect the majority of requests to be at default priority
(and so hit the preallocate priolist), once userspace starts using
priorities they are likely to use many fine grained policies improving
the utilisation of a private slab.
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: http://patchwork.freedesktop.org/patch/msgid/20170517121007.27224-9-chris@chris-wilson.co.uk
All the requests at the same priority are executed in FIFO order. They
do not need to be stored in the rbtree themselves, as they are a simple
list within a level. If we move the requests at one priority into a list,
we can then reduce the rbtree to the set of priorities. This should keep
the height of the rbtree small, as the number of active priorities can not
exceed the number of active requests and should be typically only a few.
Currently, we have ~2k possible different priority levels, that may
increase to allow even more fine grained selection. Allocating those in
advance seems a waste (and may be impossible), so we opt for allocating
upon first use, and freeing after its requests are depleted. To avoid
the possibility of an allocation failure causing us to lose a request,
we preallocate the default priority (0) and bump any request to that
priority if we fail to allocate it the appropriate plist. Having a
request (that is ready to run, so not leading to corruption) execute
out-of-order is better than leaking the request (and its dependency
tree) entirely.
There should be a benefit to reducing execlists_dequeue() to principally
using a simple list (and reducing the frequency of both rbtree iteration
and balancing on erase) but for typical workloads, request coalescing
should be small enough that we don't notice any change. The main gain is
from improving PI calls to schedule, and the explicit list within a
level should make request unwinding simpler (we just need to insert at
the head of the list rather than the tail and not have to make the
rbtree search more complicated).
v2: Avoid use-after-free when deleting a depleted priolist
v3: Michał found the solution to handling the allocation failure
gracefully. If we disable all priority scheduling following the
allocation failure, those requests will be executed in fifo and we will
ensure that this request and its dependencies are in strict fifo (even
when it doesn't realise it is only a single list). Normal scheduling is
restored once we know the device is idle, until the next failure!
Suggested-by: Michał Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170517121007.27224-8-chris@chris-wilson.co.uk
add/remove: 1/1 grow/shrink: 5/4 up/down: 391/-578 (-187)
function old new delta
execlists_submit_ports 262 471 +209
port_assign.isra - 136 +136
capture 6344 6359 +15
reset_common_ring 438 452 +14
execlists_submit_request 228 238 +10
gen8_init_common_ring 334 341 +7
intel_engine_is_idle 106 105 -1
i915_engine_info 2314 2290 -24
__i915_gem_set_wedged_BKL 485 411 -74
intel_lrc_irq_handler 1789 1604 -185
execlists_update_context 294 - -294
The most important change there is the improve to the
intel_lrc_irq_handler and excclist_submit_ports (net improvement since
execlists_update_context is now inlined).
v2: Use the port_api() for guc as well (even though currently we do not
pack any counters in there, yet) and hide all port->request_count inside
the helpers.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170517121007.27224-5-chris@chris-wilson.co.uk
During execlist_context_deferred_alloc() we presumed that the context is
uninitialised (we only just allocated the state object for it!) and
chose to optimise away the later call to engine->init_context() if
engine->init_context were NULL. This breaks with GVT's contexts that are
marked as pre-initialised to avoid us annoyingly calling
engine->init_context(). The fix is to not override ce->initialised if it
is already true.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chuanxiao Dong <chuanxiao.dong@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1494497262-24855-1-git-send-email-chuanxiao.dong@intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Since unifying ringbuffer/execlist submission to use
engine->pin_context, we ensure that the intel_ring is available before
we start constructing the request. We can therefore move the assignment
of the request->ring to the central i915_gem_request_alloc() and not
require it in every engine->request_alloc() callback. Another small step
towards simplification (of the core, but at a cost of handling error
pointers in less important callers of engine->pin_context).
v2: Rearrange a few branches to reduce impact of PTR_ERR() on gcc's code
generation.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170504093308.4137-1-chris@chris-wilson.co.uk
Pre-calculate engine context size based on engine class and device
generation and store it in the engine instance.
v2:
- Squash and get rid of hw_context_size (Chris)
v3:
- Move after MMIO init for probing on Gen7 and 8 (Chris)
- Retained rounding (Tvrtko)
v4:
- Rebase for deferred legacy context allocation
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: intel-gvt-dev@lists.freedesktop.org
Acked-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
We need to keep track of the last location we ask the hw to read up to
(RING_TAIL) separately from our last write location into the ring, so
that in the event of a GPU reset we do not tell the HW to proceed into
a partially written request (which can happen if that request is waiting
for an external signal before being executed).
v2: Refactor intel_ring_reset() (Mika)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100144
Testcase: igt/gem_exec_fence/await-hang
Fixes: 821ed7df6e ("drm/i915: Update reset path to fix incomplete requests")
Fixes: d55ac5bf97 ("drm/i915: Defer transfer onto execution timeline to actual hw submission")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170425130049.26147-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
As we now share the execlist_port[] tracking for both execlists/guc, we
can reset the inflight count on both and report which requests are being
restarted.
Suggested-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170425103835.31871-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
We indirectly hold the runtime-pm for the intel_lrc_irq_handler() by
virtue of dev_priv->gt.awake keeping a wakeref whilst the requests are
busy. As this is not obvious from the code, add a comment.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170411175850.2470-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
