The execlists emulation on top of the GuC (used for scheduling and
preemption) depends on the MI_USER_INTERRUPT for its notifications and
tasklet action. As we always employ the irq, there is no advantage in
ever disabling it while we are using the GuC, so allow us to arm the
breadcrumb irq when enabling GuC submission and disarm upon disabling.
The impact should be lessened by the delayed irq disabling we do (we
only disable after receiving an interrupt for which no one was wanting),
but allowing guc to explicitly manage the irq in relation to itself is
simpler and prevents an issue with losing an interrupt for preemption
as it is not coupled to an active request.
Internally, we add a reference counter (breadcrumbs.irq_enabled) as a
simple mechanism to allow GuC to keep the breadcrumb irq enabled. To
improve upon always enabling the irq while guc is selected, we need
to hook into the parking facility of intel_engines so that we only enable
the breadcrumbs while the GT is active (one step better would be to
individually park/unpark each engine).
In effect, this means that we keep the breadcrumb irq always enabled for
the entire duration the guc is busy, whereas before we would try to
switch it off whenever we idled for more than interrupt with no
associated waiters. The difference *should* be negligible in practice!
v2: Stop abusing fence signaling (and its auxiliary data structures) to
enable the breadcrumbs irqs.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Michał Winiarski <michal.winiarski@intel.com>,
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>,
Link: https://patchwork.freedesktop.org/patch/msgid/20171025143943.7661-3-chris@chris-wilson.co.uk
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>
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
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>
The early gen3 machines (i915g/Grantsdale and i915gm/Alviso) share a lot
of characteristics in their MI/GTT blocks with gen2, and in particular
can only use physical addresses in MI_STORE_DATA_IMM. This makes it
incompatible with our usage, so include those two machines in the
blacklist to prevent usage.
v2: Make it easy for gcc and rewrite it as a switch to save some space.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> #v1
Link: https://patchwork.freedesktop.org/patch/msgid/20170906152859.5304-1-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
Typically, there is space available within the ring and if not we have
to wait (by definition a slow path). Rearrange the code to reduce the
number of branches and stack size for the hotpath, accomodating a slight
growth for the wait.
v2: Fix the new assert that packets are not larger than the actual ring.
v3: Make the parameters unsigned as well to make usage.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170504130846.4807-3-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
Or rather it is used only by intel_ring_pin() to extract the
drm_i915_private which we can easily pass in. As this is a relatively
rare operation, save the space in the struct, and as such it is even
break even in the extra code for passing around the parameter:
add/remove: 0/0 grow/shrink: 2/3 up/down: 15/-15 (0)
function old new delta
intel_init_ring_buffer 906 918 +12
execlists_context_pin 1308 1311 +3
mock_engine 407 403 -4
intel_engine_create_ring 367 363 -4
intel_ring_pin 326 319 -7
Total: Before=1261794, After=1261794, chg +0.00%
v2: Reorder intel_init_ring_buffer to keep the ring setup together:
add/remove: 0/0 grow/shrink: 2/3 up/down: 9/-15 (-6)
function old new delta
intel_init_ring_buffer 906 912 +6
execlists_context_pin 1308 1311 +3
mock_engine 407 403 -4
intel_engine_create_ring 367 363 -4
intel_ring_pin 326 319 -7
Total: Before=1261794, After=1261788, chg -0.00%
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170403113426.25707-1-chris@chris-wilson.co.uk
If the request->wa_tail is 0 (because it landed exactly on the end of
the ringbuffer), when we reconstruct request->tail following a reset we
fill in an illegal value (-8 or 0x001ffff8). As a result, RING_HEAD is
never able to catch up with RING_TAIL and the GPU spins endlessly. If
the ring contains a couple of breadcrumbs, even our hangcheck is unable
to catch the busy-looping as the ACTHD and seqno continually advance.
v2: Move the wrap into a common intel_ring_wrap().
Fixes: a3aabe86a3 ("drm/i915/execlists: Reinitialise context image after GPU hang")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: <stable@vger.kernel.org> # v4.10+
Link: http://patchwork.freedesktop.org/patch/msgid/20170327130009.4678-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
GVTg has introduced the context status notifier to schedule the GVTg
workload. At that time, the notifier is bound to GVTg context only,
so GVTg is not aware of host workloads.
Now we are going to improve GVTg's guest workload scheduler policy,
and add Guc emulation support for new Gen graphics. Both these two
features require acknowledgment for all contexts running on hardware.
(But will not alter host workload.) So here try to make some change.
The change is simple:
1. Move the context status notifier head from i915_gem_context to
intel_engine_cs. Which means there is a notifier head per engine
instead of per context. Execlist driver still call notifier for
each context sched-in/out events of current engine.
2. At GVTg side, it binds a notifier_block for each physical engine
at GVTg initialization period. Then GVTg can hear all context
status events.
In this patch, GVTg do nothing for host context event, but later
will add a function there. But in any case, the notifier callback is
a noop if this is no active vGPU.
Since intel_gvt_init() is called at early initialization stage and
require the status notifier head has been initiated, I initiate it in
intel_engine_setup().
v2: remove a redundant newline. (chris)
Fixes: 3c7ba6359d ("drm/i915: Introduce execlist context status change notification")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100232
Signed-off-by: Changbin Du <changbin.du@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Zhi Wang <zhi.a.wang@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/20170313024711.28591-1-changbin.du@intel.com
Acked-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
During reset_all_global_seqno() on seqno rollover, we have to update the
HWS. This causes all in flight requests to be completed, so first we
wait. However, we were only waiting for the requests themselves to be
completed and clearing out the waiter rbtrees - what I had missed was
the extra reference in execlists->port[]. Since commit fe9ae7a3bf
("drm/i915/execlists: Detect an out-of-order context switch") we can
detect when the request is retired before the context switch interrupt
is completed. The impact should be neglible outside of debugging.
Testcase: igt/gem_exec_whisper
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170303121947.20482-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
A significant cost in setting up a wait is the overhead of enabling the
interrupt. As we disable the interrupt whenever the queue of waiters is
empty, if we are frequently waiting on alternating batches, we end up
re-enabling the interrupt on a frequent basis. We do want to disable the
interrupt during normal operations as under high load it may add several
thousand interrupts/s - we have been known in the past to occupy whole
cores with our interrupt handler after accidentally leaving user
interrupts enabled. As a compromise, leave the interrupt enabled until
the next IRQ, or the system is idle. This gives a small window for a
waiter to keep the interrupt active and not be delayed by having to
re-enable the interrupt.
v2: Restore hangcheck/missed-irq detection for continuations
v3: Be more careful restoring the hangcheck timer after reset
v4: Be more careful restoring the fake irq after reset (if required!)
v5: Redo changes to intel_engine_wakeup()
v6: Factor out __intel_engine_wakeup()
v7: Improve commentary for declaring a missed wakeup
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: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170227205850.2828-4-chris@chris-wilson.co.uk
As execlists and other non-semaphore multi-engine devices coordinate
between engines using interrupts, we can shave off a few 10s of
microsecond of scheduling latency by doing the fence signaling from the
interrupt as opposed to a RT kthread. (Realistically the delay adds
about 1% to an individual cross-engine workload.) We only signal the
first fence in order to limit the amount of work we move into the
interrupt handler. We also have to remember that our breadcrumbs may be
unordered with respect to the interrupt and so we still require the
waiter process to perform some heavyweight coherency fixups, as well as
traversing the tree of waiters.
v2: No need for early exit in irq handler - it breaks the flow between
patches and prevents the tracepoint
v3: Restore rcu hold across irq signaling of request
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: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170227205850.2828-2-chris@chris-wilson.co.uk
The two users of the return value from intel_engine_wakeup() are
expecting different results. In the breadcrumbs hangcheck, we are using
it to determine whether wake_up_process() detected the waiter was
currently running (and if so we presume that it hasn't yet missed the
interrupt). However, in the fake_irq path, we are using the return value
as a check as to whether there are any waiters, and so we may
incorrectly stop the fake-irq if that waiter was currently running.
To handle the two different needs, return both bits of information! We
uninline it from the irq path in preparation for the next patch which
makes the irq hotpath special and relegates intel_engine_wakeup() to the
slow fixup paths.
v2: s/ret/result/
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: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170227205850.2828-1-chris@chris-wilson.co.uk
A request is assigned a global seqno only when it is on the hardware
execution queue. The global seqno can be used to maintain a list of
requests on the same engine in retirement order, for example for
constructing a priority queue for waiting. Prior to its execution, or
if it is subsequently removed in the event of preemption, its global
seqno is zero. As both insertion and removal from the execution queue
may operate in IRQ context, it is not guarded by the usual struct_mutex
BKL. Instead those relying on the global seqno must be prepared for its
value to change between reads. Only when the request is complete can
the global seqno be stable (due to the memory barriers on submitting
the commands to the hardware to write the breadcrumb, if the HWS shows
that it has passed the global seqno and the global seqno is unchanged
after the read, it is indeed complete).
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/20170223074422.4125-9-chris@chris-wilson.co.uk
