Track total time requests have been executing on the hardware.
We add new kernel API to allow software tracking of time GPU
engines are spending executing requests.
Both per-engine and global API is added with the latter also
being exported for use by external users.
v2:
* Squashed with the internal API.
* Dropped static key.
* Made per-engine.
* Store time in monotonic ktime.
v3: Moved stats clearing to disable.
v4:
* Comments.
* Don't export the API just yet.
v5: Whitespace cleanup.
v6:
* Rename ref to active.
* Drop engine aggregate stats for now.
* Account initial busy period after enabling stats.
v7:
* Rebase.
v8:
* Move context in notification after the notifier. (Chris Wilson)
v9:
In cases where stats tracking is getting disabled while there is
an active context on an engine, add up the current value to the
total. This also implies we don't clear the total when tracking
is disabled any longer. There is no real need to do so because
we define the stats as relative while enabled, meaning
comparison between two samples while tracking is enabled is the
valid usage. However, when busy stats will later be plugged into
the perf PMU API, it is beneficial to not reset the total, since
the PMU core likes to do some counter disable/enable cycles on
startup, and while doing so during a single long context
executing on an engine we would lose some accuracy and so make
unit testing more difficult than needs to be.
v10:
* Fix accounting for preemption.
v11:
* Rebase for i915_modparams.enable_execlists removal.
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/20171121181852.16128-5-tvrtko.ursulin@linux.intel.com
From: Chris Wilson <chris@chris-wilson.co.uk>
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
From: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
The first goal is to be able to measure GPU (and invidual ring) busyness
without having to poll registers from userspace. (Which not only incurs
holding the forcewake lock indefinitely, perturbing the system, but also
runs the risk of hanging the machine.) As an alternative we can use the
perf event counter interface to sample the ring registers periodically
and send those results to userspace.
Functionality we are exporting to userspace is via the existing perf PMU
API and can be exercised via the existing tools. For example:
perf stat -a -e i915/rcs0-busy/ -I 1000
Will print the render engine busynnes once per second. All the performance
counters can be enumerated (perf list) and have their unit of measure
correctly reported in sysfs.
v1-v2 (Chris Wilson):
v2: Use a common timer for the ring sampling.
v3: (Tvrtko Ursulin)
* Decouple uAPI from i915 engine ids.
* Complete uAPI defines.
* Refactor some code to helpers for clarity.
* Skip sampling disabled engines.
* Expose counters in sysfs.
* Pass in fake regs to avoid null ptr deref in perf core.
* Convert to class/instance uAPI.
* Use shared driver code for rc6 residency, power and frequency.
v4: (Dmitry Rogozhkin)
* Register PMU with .task_ctx_nr=perf_invalid_context
* Expose cpumask for the PMU with the single CPU in the mask
* Properly support pmu->stop(): it should call pmu->read()
* Properly support pmu->del(): it should call stop(event, PERF_EF_UPDATE)
* Introduce refcounting of event subscriptions.
* Make pmu.busy_stats a refcounter to avoid busy stats going away
with some deleted event.
* Expose cpumask for i915 PMU to avoid multiple events creation of
the same type followed by counter aggregation by perf-stat.
* Track CPUs getting online/offline to migrate perf context. If (likely)
cpumask will initially set CPU0, CONFIG_BOOTPARAM_HOTPLUG_CPU0 will be
needed to see effect of CPU status tracking.
* End result is that only global events are supported and perf stat
works correctly.
* Deny perf driver level sampling - it is prohibited for uncore PMU.
v5: (Tvrtko Ursulin)
* Don't hardcode number of engine samplers.
* Rewrite event ref-counting for correctness and simplicity.
* Store initial counter value when starting already enabled events
to correctly report values to all listeners.
* Fix RC6 residency readout.
* Comments, GPL header.
v6:
* Add missing entry to v4 changelog.
* Fix accounting in CPU hotplug case by copying the approach from
arch/x86/events/intel/cstate.c. (Dmitry Rogozhkin)
v7:
* Log failure message only on failure.
* Remove CPU hotplug notification state on unregister.
v8:
* Fix error unwind on failed registration.
* Checkpatch cleanup.
v9:
* Drop the energy metric, it is available via intel_rapl_perf.
(Ville Syrjälä)
* Use HAS_RC6(p). (Chris Wilson)
* Handle unsupported non-engine events. (Dmitry Rogozhkin)
* Rebase for intel_rc6_residency_ns needing caller managed
runtime pm.
* Drop HAS_RC6 checks from the read callback since creating those
events will be rejected at init time already.
* Add counter units to sysfs so perf stat output is nicer.
* Cleanup the attribute tables for brevity and readability.
v10:
* Fixed queued accounting.
v11:
* Move intel_engine_lookup_user to intel_engine_cs.c
* Commit update. (Joonas Lahtinen)
v12:
* More accurate sampling. (Chris Wilson)
* Store and report frequency in MHz for better usability from
perf stat.
* Removed metrics: queued, interrupts, rc6 counters.
* Sample engine busyness based on seqno difference only
for less MMIO (and forcewake) on all platforms. (Chris Wilson)
v13:
* Comment spelling, use mul_u32_u32 to work around potential GCC
issue and somne code alignment changes. (Chris Wilson)
v14:
* Rebase.
v15:
* Rebase for RPS refactoring.
v16:
* Use the dynamic slot in the CPU hotplug state machine so that we are
free to setup our state as multi-instance. Previously we were re-using
the CPUHP_AP_PERF_X86_UNCORE_ONLINE slot which is neither used as
multi-instance, nor owned by our driver to start with.
* Register the CPU hotplug handlers after the PMU, otherwise the callback
will get called before the PMU is initialized which can end up in
perf_pmu_migrate_context with an un-initialized base.
* Added workaround for a probable bug in cpuhp core.
v17:
* Remove workaround for the cpuhp bug.
v18:
* Rebase for drm_i915_gem_engine_class getting upstream before us.
v19:
* Rebase. (trivial)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171121181852.16128-2-tvrtko.ursulin@linux.intel.com
Since removing the module parameter to force selection of ringbuffer
emission for gen8, the code is defunct. Remove it.
To put the difference into perspective, a couple of microbenchmarks
(bdw i7-5557u, 20170324):
ring execlists
exec continuous nops on all rings: 1.491us 2.223us
exec sequential nops on each ring: 12.508us 53.682us
single nop + sync: 9.272us 30.291us
vblank_mode=0 glxgears: ~11000fps ~9000fps
Since the earlier submission, gen8 ringbuffer submission has fallen
further and further behind in features. So while ringbuffer may hold the
throughput crown, in terms of interactive latency, execlists is much
better. Alas, we have no convenient metrics for such, other than
demonstrating things we can do with execlists but can not using
legacy ringbuffer submission.
We have made a few improvements to lowlevel execlists throughput,
and ringbuffer currently panics on boot! (bdw i7-5557u, 20171026):
ring execlists
exec continuous nops on all rings: n/a 1.921us
exec sequential nops on each ring: n/a 44.621us
single nop + sync: n/a 21.953us
vblank_mode=0 glxgears: n/a ~18500fps
References: https://bugs.freedesktop.org/show_bug.cgi?id=87725
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Once-upon-a-time-Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171120205504.21892-2-chris@chris-wilson.co.uk
Take a copy of the HW state after a reset upon module loading by
executing a context switch from a blank context to the kernel context,
thus saving the default hw state over the blank context image.
We can then use the default hw state to initialise any future context,
ensuring that each starts with the default view of hw state.
v2: Unmap our default state from the GTT after stealing it from the
context. This should stop us from accidentally overwriting it via the
GTT (and frees up some precious GTT space).
Testcase: igt/gem_ctx_isolation
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110142634.10551-7-chris@chris-wilson.co.uk
In the next few patches, we will have a hard requirement that we emit a
context-switch to the perma-pinned i915->kernel_context (so that we can
save the HW state using that context-switch). As the first context
itself may be classed as a kernel context, we want to be explicit in our
comparison. For an extra-layer of finesse, we can check the last
unretired context on the engine; as well as the last retired context
when idle.
v2: verbose verbosity
v3: Always force the switch, even when the engine is idle, and update
the assert that this happens before suspend.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> #v1
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110142634.10551-2-chris@chris-wilson.co.uk
We want to be able to report back to userspace details about an engine's
class, and in return for userspace to be able to request actions
regarding certain classes of engines. To isolate the uABI from any
variations between hw generations, we define an abstract class for the
engines and internally map onto the hw.
v2: Remove MAX from the uABI; keep it internal if we need it, but don't
let userspace make the mistake of using it themselves.
v3: s/OTHER/INVALID/
The use of OTHER is ill-defined, so remove it from the uABI as any
future new type of engine can define a class to suit it. But keep a
reserved value for an invalid class, so that we can always
unambiguously express when something doesn't belong to the
classification.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> #v2
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110142634.10551-1-chris@chris-wilson.co.uk
So it appears that commit 5427f20785 ("drm/i915: Bump wait-times for
the final CS interrupt before parking") was a little over optimistic in
its belief that it had successfully waited for all residual activity on
the engines before parking. Numerous sightings in CI since then of
<7>[ 52.542886] [IGT] core_auth: executing
<3>[ 52.561013] [drm:intel_engines_park [i915]] *ERROR* vcs0 is not idle before parking
<7>[ 52.561215] intel_engines_park vcs0
<7>[ 52.561229] intel_engines_park current seqno 98, last 98, hangcheck 0 [-247449 ms], inflight 0
<7>[ 52.561238] intel_engines_park Reset count: 0
<7>[ 52.561266] intel_engines_park Requests:
<7>[ 52.561363] intel_engines_park RING_START: 0x00000000 [0x00000000]
<7>[ 52.561377] intel_engines_park RING_HEAD: 0x00000000 [0x00000000]
<7>[ 52.561390] intel_engines_park RING_TAIL: 0x00000000 [0x00000000]
<7>[ 52.561406] intel_engines_park RING_CTL: 0x00000000
<7>[ 52.561422] intel_engines_park RING_MODE: 0x00000200 [idle]
<7>[ 52.561442] intel_engines_park ACTHD: 0x00000000_00000000
<7>[ 52.561459] intel_engines_park BBADDR: 0x00000000_00000000
<7>[ 52.561474] intel_engines_park Execlist status: 0x00000301 00000000
<7>[ 52.561489] intel_engines_park Execlist CSB read 5 [5 cached], write 5 [5 from hws], interrupt posted? no
<7>[ 52.561500] intel_engines_park ELSP[0] idle
<7>[ 52.561510] intel_engines_park ELSP[1] idle
<7>[ 52.561519] intel_engines_park HW active? 0x0
<7>[ 52.561608] intel_engines_park Idle? yes
<7>[ 52.561617] intel_engines_park
on Braswell, which indicates that the engine just needs that little bit
longer after flushing the tasklet to settle. So give it a few more
milliseconds before declaring an err and applying the emergency brake.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103479
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110112550.28909-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
The workaround for this is described as:
"if RenderSurfaceState.Num_Multisamples > 1, disable RCC clock gating if
RenderSurfaceState.Num_Multisamples == 1, set 0x7010[14] = 1"
Further documentation in the internal bug referenced by the bspec
suggest that any of the above suggestions should suffice to fix the
issue. We are going with disabling RCC clock gating.
Unfortunately, what we are doing doesn't match the name of the
workaround, but at least it matches its description.
This change improves CNL stability by avoiding some of the hangs seen in
the platform.
v2: Only disable RCC clock gating.
Signed-off-by: Rafael Antognolli <rafael.antognolli@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171103183027.5051-1-rafael.antognolli@intel.com
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>
Add another perma-pinned context for using for preemption at any time.
We cannot just reuse the existing kernel context, as first and foremost
we need to ensure that we can preempt the kernel context itself, so
require a distinct context id. Similar to the kernel context, we may
want to interrupt execution and switch to the preempt context at any
time, and so it needs to be permanently pinned and available.
To compensate for yet another permanent allocation, we shrink the
existing context and the new context by reducing their ringbuffer to the
minimum.
v2: Assert that we never allocate a request from the preemption context.
v3: Limit perma-pin to engines that may preempt.
v4: Onion cleanup for early driver death
v5: Onion ordering in main driver cleanup as well.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003203453.15692-4-chris@chris-wilson.co.uk
Supporting fine-granularity preemption levels may require changes in
userspace batch buffer programming. Therefore, we need to fallback to
safe default values, rather that use hardware defaults. Userspace is
still able to enable fine-granularity, since we're whitelisting the
register controlling it in WaEnablePreemptionGranularityControlByUMD.
v2: Extend w/a to cover Cannonlake
v3: Fix commentary to include both fake w/a names.
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/20171003203453.15692-2-chris@chris-wilson.co.uk
