Add multi-lrc context registration H2G. In addition a workqueue and
process descriptor are setup during multi-lrc context registration as
these data structures are needed for multi-lrc submission.
v2:
(John Harrison)
- Move GuC specific fields into sub-struct
- Clean up WQ defines
- Add comment explaining math to derive WQ / PD address
v3:
(John Harrison)
- Add PARENT_SCRATCH_SIZE define
- Update comment explaining multi-lrc register
v4:
(John Harrison)
- Move PARENT_SCRATCH_SIZE to common file
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Reviewed-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211014172005.27155-9-matthew.brost@intel.com
Previously, we were storing the ring size in the ring pointer before it
was actually allocated. We would then guard setting the ring size on
checking for CONTEXT_ALLOC_BIT. This is error-prone at best and really
only saves us a few bytes on something that already burns at least 4K.
Instead, this patch adds a new ring_size field and makes everything use
that.
v2 (Daniel Vetter):
- Replace 512 * SZ_4K with SZ_2M
v2 (Jason Ekstrand):
- Rebase on top of page migration code
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210708154835.528166-3-jason@jlekstrand.net
This was done by the following semantic patch:
@@ expression i915; @@
- INTEL_GEN(i915)
+ GRAPHICS_VER(i915)
@@ expression i915; expression E; @@
- INTEL_GEN(i915) >= E
+ GRAPHICS_VER(i915) >= E
@@ expression dev_priv; expression E; @@
- !IS_GEN(dev_priv, E)
+ GRAPHICS_VER(dev_priv) != E
@@ expression dev_priv; expression E; @@
- IS_GEN(dev_priv, E)
+ GRAPHICS_VER(dev_priv) == E
@@
expression dev_priv;
expression from, until;
@@
- IS_GEN_RANGE(dev_priv, from, until)
+ IS_GRAPHICS_VER(dev_priv, from, until)
@def@
expression E;
identifier id =~ "^gen$";
@@
- id = GRAPHICS_VER(E)
+ ver = GRAPHICS_VER(E)
@@
identifier def.id;
@@
- id
+ ver
It also takes care of renaming the variable we assign to GRAPHICS_VER()
so to use "ver" rather than "gen".
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210605155356.4183026-2-lucas.demarchi@intel.com
Driver Changes:
- Prepare for local/device memory support on DG1 by starting
to use it for kernel internal allocations: context, ring
and engine scratch (Matt A, CQ, Abdiel, Imre)
- Sandybridge fix to avoid hard hang on ring resume (Chris)
- Limit imported dma-buf size to int32 (Matt A)
- Double check heartbeat timeout before resetting (Chris)
- Use new tasklet API for execution list (Emil)
- Fix SPDX checkpats warnings (Chris)
- Fixes for various checkpatch warnings (Chris)
- Selftest improvements (Chris)
- Move the defer_request waiter active assertion to correct spot (Chris)
- Make local-memory probing a GT operation (Matt, Tvrtko)
- Protect against request freeing during cancellation on wedging (Chris)
- Retire unexpected starting state error dumping (Chris)
- Distinction of memory regions in debugging (Zbigniew)
- Always flush the submission queue on checking for idle (Chris)
- Consolidate 2big error check to helper (Matt)
- Decrease number of subplatform bits (Tvrtko)
- Remove unused internal request priority levels (Chris)
- Document the unused internal header bits in buddy allocator (Matt)
- Cleanup the region class/instance encoding (Matt)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/YGxksaZGXHnFxlwg@jlahtine-mobl.ger.corp.intel.com
Pull the repeated check for the last active request being completed to a
single spot, when deciding whether or not execlist preemption is
required.
In doing so, we remove the tasklet kick, introduced with the completion
checks in commit 35f3fd8182 ("drm/i915/execlists: Workaround switching
back to a completed context"), if we find the request was completed but
have not yet seen the corresponding CS event. This was devolving into a
busy spin of the tasklet while we waited for the event as the delivery
was not as instantaneous as expected. Under load this is sufficient to
exhaust the tasklet softirq timeslice, and force ksoftirqd. Quite
noticeable overhead for no apparent improvement in latency.
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/20201126140407.31952-2-chris@chris-wilson.co.uk
Since the introduction of preempt-to-busy, requests can complete in the
background, even while they are not on the engine->active.requests list.
As such, the engine->active.request list itself is not in strict
retirement order, and we have to scan the entire list while unwinding to
not miss any. However, if the request is completed we currently leave it
on the list [until retirement], but we could just as simply remove it
and stop treating it as active. We would only have to then traverse it
once while unwinding in quick succession.
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/20201126140407.31952-1-chris@chris-wilson.co.uk
Since preempt-to-busy, we may unsubmit a request while it is still on
the HW and completes asynchronously. That means it may be retired and in
the process destroy the virtual engine (as the user has closed their
context), but that engine may still be holding onto the unsubmitted
compelted request. Therefore we need to potentially cleanup the old
request on destroying the virtual engine. We also have to keep the
virtual_engine alive until after the sibling's execlists_dequeue() have
finished peeking into the virtual engines, for which we serialise with
RCU.
v2: Be paranoid and flush the tasklet as well.
v3: And flush the tasklet before the engines, as the tasklet may
re-attach an rb_node after our removal from the siblings.
Fixes: 6d06779e86 ("drm/i915: Load balancing across a virtual engine")
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/20201123113717.20500-4-chris@chris-wilson.co.uk
We plan to expand upon the number of available statuses for when we
pretty-print the requests along the timelines, and so need a new set of
flags. We have settled upon:
Unready [U]
- initial status after being submitted, the request is not
ready for execution as it is waiting for external fences
Ready [R]
- all fences the request was waiting on have been signaled,
and the request is now ready for execution and will be
in a backend queue
- a ready request may still need to wait on semaphores
[internal fences]
Ready/virtual [V]
- same as ready, but queued over multiple backends
Executing [E]
- the request has been transferred from the backend queue and
submitted for execution on HW
- a completed request may still be regarded as executing, its
status may not be updated until it is retired and removed
from the lists
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/20201119165616.10834-3-chris@chris-wilson.co.uk
Between events which trigger engine and GPU resets and capturing the error
state we lose information on which engine triggered the reset. Improve
this by passing in the hung engine mask down to error capture.
Result is that the list of engines in user visible "GPU HANG: ecode
<gen>:<engines>:<ecode>, <process>" is now a list of hanging and not just
active engines. Most importantly the displayed process is now the one
which was actually hung.
v2:
* Stub prototype. (Chris)
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/20201104134743.916027-1-tvrtko.ursulin@linux.intel.com
We wrap the timeline on construction of the next request, but there may
still be requests in flight that have not yet finalized the breadcrumb.
(The breadcrumb is delayed as we need engine-local offsets, and for the
virtual engine that is not known until execution.) As such, by the time
we write to the timeline's HWSP offset it may have changed, and we
should use the value we preserved in the request instead.
Though the window is small and infrequent (at full flow we can expect a
timeline's seqno to wrap once every 30 minutes), the impact of writing
the old seqno into the new HWSP is severe: the old requests are never
completed, and the new requests are completed before they are even
submitted.
Fixes: ebece75392 ("drm/i915: Keep timeline HWSP allocated until idle across the system")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.2+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201022064127.10159-1-chris@chris-wilson.co.uk
We may try to preempt the currently executing request, only to find that
after unravelling all the dependencies that the original executing
context is still the earliest in the topological sort and re-submitted
back to HW (if we do detect some change in the ELSP that requires
re-submission). However, due to the way we check for wrap-around during
the unravelling, we mark any context that has been submitted just once
(i.e. with the rq->wa_tail set, but the ring->tail earlier) as
potentially wrapping and requiring a forced restore on resubmission.
This was expected to be not a problem, as it was anticipated that most
unwinding for preemption would result in a context switch and the few
that did not would be lost in the noise. It did not take long for
someone to find one particular workload where the cost of those extra
context restores was measurable.
However, since we know the wa_tail is of fixed size, and we know that a
request must be larger than the wa_tail itself, we can safely maintain
the check for request wrapping and check against a slightly future point
in the ring that includes an expected wa_tail. (That is if the
ring->tail is already set to rq->wa_tail, including another 8 bytes in
the check does not invalidate the incremental wrap detection.)
Fixes: 8ab3a3812a ("drm/i915/gt: Incrementally check for rewinding")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Ramalingam C <ramalingam.c@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.4+
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201002083425.4605-1-chris@chris-wilson.co.uk
When running gem_exec_nop, it floods the system with many requests (with
the goal of userspace submitting faster than the HW can process a single
empty batch). This causes the driver to continually resubmit new
requests onto the end of an active context, a flood of lite-restore
preemptions. If we time this just right, Tigerlake hangs.
Inserting a small delay between the processing of CS events and
submitting the next context, prevents the hang. Naturally it does not
occur with debugging enabled. The suspicion then is that this is related
to the issues with the CS event buffer, and inserting an mmio read of
the CS pointer status appears to be very successful in preventing the
hang. Other registers, or uncached reads, or plain mb, do not prevent
the hang, suggesting that register is key -- but that the hang can be
prevented by a simple udelay, suggests it is just a timing issue like
that encountered by commit 233c1ae3c8 ("drm/i915/gt: Wait for CSB
entries on Tigerlake"). Also note that the hang is not prevented by
applying CTX_DESC_FORCE_RESTORE, or by inserting a delay on the GPU
between requests.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: stable@vger.kernel.org
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20201015195023.32346-1-chris@chris-wilson.co.uk
On Tigerlake, we are seeing a repeat of commit d8f5053117 ("drm/i915/icl:
Forcibly evict stale csb entries") where, presumably, due to a missing
Global Observation Point synchronisation, the write pointer of the CSB
ringbuffer is updated _prior_ to the contents of the ringbuffer. That is
we see the GPU report more context-switch entries for us to parse, but
those entries have not been written, leading us to process stale events,
and eventually report a hung GPU.
However, this effect appears to be much more severe than we previously
saw on Icelake (though it might be best if we try the same approach
there as well and measure), and Bruce suggested the good idea of resetting
the CSB entry after use so that we can detect when it has been updated by
the GPU. By instrumenting how long that may be, we can set a reliable
upper bound for how long we should wait for:
513 late, avg of 61 retries (590 ns), max of 1061 retries (10099 ns)
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2045
References: d8f5053117 ("drm/i915/icl: Forcibly evict stale csb entries")
References: HSDES#22011327657, HSDES#1508287568
Suggested-by: Bruce Chang <yu.bruce.chang@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Bruce Chang <yu.bruce.chang@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: stable@vger.kernel.org # v5.4
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200915134923.30088-2-chris@chris-wilson.co.uk
To implement preempt-to-busy (and so efficient timeslicing and best utilization
of the hardware submission ports) we let the GPU run asynchronously in respect
to the ELSP submission queue. This created challenges in keeping and accessing
the driver state mirroring the asynchronous GPU execution.
The latest occurence of this was spotted by KCSAN:
[ 1413.563200] BUG: KCSAN: data-race in __await_execution+0x217/0x370 [i915]
[ 1413.563221]
[ 1413.563236] race at unknown origin, with read to 0xffff88885bb6c478 of 8 bytes by task 9654 on cpu 1:
[ 1413.563548] __await_execution+0x217/0x370 [i915]
[ 1413.563891] i915_request_await_dma_fence+0x4eb/0x6a0 [i915]
[ 1413.564235] i915_request_await_object+0x421/0x490 [i915]
[ 1413.564577] i915_gem_do_execbuffer+0x29b7/0x3c40 [i915]
[ 1413.564967] i915_gem_execbuffer2_ioctl+0x22f/0x5c0 [i915]
[ 1413.564998] drm_ioctl_kernel+0x156/0x1b0
[ 1413.565022] drm_ioctl+0x2ff/0x480
[ 1413.565046] __x64_sys_ioctl+0x87/0xd0
[ 1413.565069] do_syscall_64+0x4d/0x80
[ 1413.565094] entry_SYSCALL_64_after_hwframe+0x44/0xa9
To complicate matters, we have to both avoid the read tearing of *active and
avoid any write tearing as perform the pending[] -> inflight[] promotion of the
execlists.
This is because we cannot rely on the memcpy doing u64 aligned copies on all
kernels/platforms and so we opt to open-code it with explicit WRITE_ONCE
annotations to satisfy KCSAN.
v2: When in doubt, write the same comment again.
v3: Expanded commit message.
Fixes: b55230e5e8 ("drm/i915: Check for awaits on still currently executing requests")
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/20200716142207.13003-1-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
[Joonas: Rebased and reordered into drm-intel-gt-next branch]
[Joonas: Added expanded commit message from Tvrtko and Chris]
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
On the virtual engines, we only use the intel_breadcrumbs for tracking
signaling of stale breadcrumbs from the irq_workers. They do not have
any associated interrupt handling, active requests are passed to a
physical engine and associated breadcrumb interrupt handler. This causes
issues for us as we need to ensure that we do not actually try and
enable interrupts and the powermanagement required for them on the
virtual engine, as they will never be disabled. Instead, let's
specify the physical engine used for interrupt handler on a particular
breadcrumb.
v2: Drop b->irq_armed = true mocking for no interrupt HW
Fixes: 4fe6abb8f5 ("drm/i915/gt: Ignore irq enabling on the virtual engines")
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/20200731154834.8378-4-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
One more complication of preempt-to-busy with respect to the virtual
engine is that we may have retired the last request along the virtual
engine at the same time as preparing to submit the completed request to
a new engine. That submit will be shortcircuited, but not before we have
updated the context with the new register offsets and marked the virtual
engine as bound to the new engine (by calling swap on ve->siblings[]).
As we may have just retired the completed request, we may also be in the
middle of calling virtual_context_exit() to turn off the power management
associated with the virtual engine, and that in turn walks the
ve->siblings[]. If we happen to call swap() on the array as we walk, we
will call intel_engine_pm_put() twice on the same engine.
In this patch, we prevent this by only updating the bound engine after a
successful submission which weeds out the already completed requests.
Alternatively, we could walk a non-volatile array for the pm, such as
using the engine->mask. The small advantage to performing the update
after the submit is that we then only have to do a swap for active
requests.
Fixes: 22b7a426bb ("drm/i915/execlists: Preempt-to-busy")
References: 6d06779e86 ("drm/i915: Load balancing across a virtual engine"
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: "Nayana, Venkata Ramana" <venkata.ramana.nayana@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200731154834.8378-3-chris@chris-wilson.co.uk
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
