As a virtual engine may change the rq->engine to point to the active
request in flight, we need to warn the compiler that an active request's
engine is volatile.
[ 95.017686] write (marked) to 0xffff8881e8386b10 of 8 bytes by interrupt on cpu 2:
[ 95.018123] execlists_dequeue+0x762/0x2150 [i915]
[ 95.018539] __execlists_submission_tasklet+0x48/0x60 [i915]
[ 95.018955] execlists_submission_tasklet+0xd3/0x170 [i915]
[ 95.018986] tasklet_action_common.isra.0+0x42/0xa0
[ 95.019016] __do_softirq+0xd7/0x2cd
[ 95.019043] irq_exit+0xbe/0xe0
[ 95.019068] irq_work_interrupt+0xf/0x20
[ 95.019491] i915_request_retire+0x2c5/0x670 [i915]
[ 95.019937] retire_requests+0xa1/0xf0 [i915]
[ 95.020348] engine_retire+0xa1/0xe0 [i915]
[ 95.020376] process_one_work+0x3b1/0x690
[ 95.020403] worker_thread+0x80/0x670
[ 95.020429] kthread+0x19a/0x1e0
[ 95.020454] ret_from_fork+0x1f/0x30
[ 95.020476]
[ 95.020498] read to 0xffff8881e8386b10 of 8 bytes by task 8909 on cpu 3:
[ 95.020918] __i915_request_commit+0x177/0x220 [i915]
[ 95.021329] i915_gem_do_execbuffer+0x38c4/0x4e50 [i915]
[ 95.021750] i915_gem_execbuffer2_ioctl+0x2c3/0x580 [i915]
[ 95.021784] drm_ioctl_kernel+0xe4/0x120
[ 95.021809] drm_ioctl+0x297/0x4c7
[ 95.021832] ksys_ioctl+0x89/0xb0
[ 95.021865] __x64_sys_ioctl+0x42/0x60
[ 95.021901] do_syscall_64+0x6e/0x2c0
[ 95.021927] entry_SYSCALL_64_after_hwframe+0x44/0xa9
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200310142403.5953-1-chris@chris-wilson.co.uk
Always wait on the start of the signaler request to reduce the problem
of dequeueing the bonded pair too early -- we want both payloads to
start at the same time, with no latency, and yet still allow others to
make full use of the slack in the system. This reduce the amount of time
we spend waiting on the semaphore used to synchronise the start of the
bonded payload.
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/20200306133852.3420322-3-chris@chris-wilson.co.uk
We busywait on an inflight request (one that is currently executing on
HW, and so might complete quickly) prior to setting up an interrupt and
sleeping. The trade off is that we keep an expensive CPU core busy in
order to avoid wake up latency: where that trade off should lie is best
left to the sysadmin.
The busywait mechanism can be compiled out with
./scripts/config --set-val DRM_I915_SPIN_REQUEST 0
The maximum busywait duration can be adjusted per-engine using,
/sys/class/drm/card?/engine/*/ms_busywait_duration_ns
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Steve Carbonari <steven.carbonari@intel.com>
Tested-by: Steve Carbonari <steven.carbonari@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200228131716.3243616-4-chris@chris-wilson.co.uk
We need to be extremely careful inside i915_request_await_start() as it
needs to walk the list of requests in the foreign timeline with very
little protection. As we hold our own timeline mutex, we can not nest
inside the signaler's timeline mutex, so all that remains is our RCU
protection. However, to be safe we need to tell the compiler that we may
be traversing the list only under RCU protection, and furthermore we
need to start declaring requests as elements of the timeline from their
construction.
Fixes: 9ddc8ec027 ("drm/i915: Eliminate the trylock for awaiting an earlier request")
Fixes: 6a79d84840 ("drm/i915: Lock signaler timeline while navigating")
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/20200227085723.1961649-11-chris@chris-wilson.co.uk
Currently on execlists, we use a local hwsp for the kernel_context,
rather than the engine's HWSP, as this is the default for execlists.
However, seqno wrap requires allocating a new HWSP cacheline, and may
require pinning a new HWSP page in the GGTT. This operation requiring
pinning in the GGTT is not allowed within the kernel_context timeline,
as doing so may require re-entering the kernel_context in order to evict
from the GGTT. As we want to avoid requiring a new HWSP for the
kernel_context, we can use the permanently pinned engine's HWSP instead.
However to do so we must prevent the use of semaphores reading the
kernel_context's HWSP, as the use of semaphores do not support rollover
onto the same cacheline. Fortunately, the kernel_context is mostly
isolated, so unlikely to give benefit to semaphores.
Reported-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200210205722.794180-5-chris@chris-wilson.co.uk
Allocate only an internal intel_context for the kernel_context, forgoing
a global GEM context for internal use as we only require a separate
address space (for our own protection).
Now having weaned GT from requiring ce->gem_context, we can stop
referencing it entirely. This also means we no longer have to create random
and unnecessary GEM contexts for internal use.
GEM contexts are now entirely for tracking GEM clients, and intel_context
the execution environment on the GPU.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andi Shyti <andi.shyti@intel.com>
Acked-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191221160324.1073045-1-chris@chris-wilson.co.uk
We want the bonded request to have the same scheduler properties as its
master so that it is placed at the same depth in the queue. For example,
consider we have requests A, B and B', where B & B' are a bonded pair to
run in parallel on two engines.
A -> B
\- B'
B will run after A and so may be scheduled on an idle engine and wait on
A using a semaphore. B' sees B being executed and so enters the queue on
the same engine as A. As B' did not inherit the semaphore-chain from B,
it may have higher precedence than A and so preempts execution. However,
B' then sits on a semaphore waiting for B, who is waiting for A, who is
blocked by B.
Ergo B' needs to inherit the scheduler properties from B (i.e. the
semaphore chain) so that it is scheduled with the same priority as B and
will not be executed ahead of Bs dependencies.
Furthermore, to prevent the priorities changing via the expose fence on
B', we need to couple in the dependencies for PI. This requires us to
relax our sanity-checks that dependencies are strictly in order.
v2: Synchronise (B, B') execution on all platforms, regardless of using
a scheduler, any no-op syncs should be elided.
Fixes: ee1136908e ("drm/i915/execlists: Virtual engine bonding")
Closes: https://gitlab.freedesktop.org/drm/intel/issues/464
Testcase: igt/gem_exec_balancer/bonded-chain
Testcase: igt/gem_exec_balancer/bonded-semaphore
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/20191210151332.3902215-1-chris@chris-wilson.co.uk
Pull drm updates from Dave Airlie:
"Lots of stuff in here, though it hasn't been too insane this merge
apart from dealing with the security fun.
uapi:
- export different colorspace properties on DP vs HDMI
- new fourcc for ARM 16x16 block format
- syncobj: allow querying last submitted timeline value
- DRM_FORMAT_BIG_ENDIAN defined as unsigned
core:
- allow using gem vma manager in ttm
- connector/encoder/bridge doc fixes
- allow more than 3 encoders for a connector
- displayport mst suspend/resume reprobing support
- vram lazy unmapping, uniform vram mm and gem vram
- edid cleanups + AVI informframe bar info
- displayport helpers - dpcd parser added
dp_cec:
- Allow a connector to be associated with a cec device
ttm:
- pipelining with no_gpu_wait fix
- always keep BOs on the LRU
sched:
- allow free_job routine to sleep
i915:
- Block userptr from mappable GTT
- i915 perf uapi versioning
- OA stream dynamic reconfiguration
- make context persistence optional
- introduce DRM_I915_UNSTABLE Kconfig
- add fake lmem testing under unstable
- BT.2020 support for DP MSA
- struct mutex elimination
- Tigerlake display/PLL/power management improvements
- Jasper Lake PCH support
- refactor PMU for multiple GPUs
- Icelake firmware update
- Split out vga + switcheroo code
amdgpu:
- implement dma-buf import/export without helpers
- vega20 RAS enablement
- DC i2c over aux fixes
- renoir GPU reset
- DC HDCP support
- BACO support for CI/VI asics
- MSI-X support
- Arcturus EEPROM support
- Arcturus VCN encode support
- VCN dynamic powergating on RV/RV2
amdkfd:
- add navi12/14/renoir support to kfd
radeon:
- SI dpm fix ported from amdgpu
- fix bad DMA on ppc platforms
gma500:
- memory leak fixes
qxl:
- convert to new gem mmap
exynos:
- build warning fix
komeda:
- add aclk sysfs attribute
v3d:
- userspace cleanup uapi change
i810:
- fix for underflow in dispatch ioctls
ast:
- refactor show_cursor
mgag200:
- refactor show_cursor
arcgpu:
- encoder finding improvements
mediatek:
- mipi_tx, dsi and partial crtc support for MT8183 SoC
- rotation support
meson:
- add suspend/resume support
omap:
- misc refactors
tegra:
- DisplayPort support for Tegra 210, 186 and 194.
- IOMMU-backed DMA API fixes
panfrost:
- fix lockdep issue
- simplify devfreq integration
rcar-du:
- R8A774B1 SoC support
- fixes for H2 ES2.0
sun4i:
- vcc-dsi regulator support
virtio-gpu:
- vmexit vs spinlock fix
- move to gem shmem helpers
- handle large command buffers with cma"
* tag 'drm-next-2019-11-27' of git://anongit.freedesktop.org/drm/drm: (1855 commits)
drm/amdgpu: invalidate mmhub semaphore workaround in gmc9/gmc10
drm/amdgpu: initialize vm_inv_eng0_sem for gfxhub and mmhub
drm/amd/amdgpu/sriov skip RLCG s/r list for arcturus VF.
drm/amd/amdgpu/sriov temporarily skip ras,dtm,hdcp for arcturus VF
drm/amdgpu/gfx10: re-init clear state buffer after gpu reset
merge fix for "ftrace: Rework event_create_dir()"
drm/amdgpu: Update Arcturus golden registers
drm/amdgpu/gfx10: fix out-of-bound mqd_backup array access
drm/amdgpu/gfx10: explicitly wait for cp idle after halt/unhalt
Revert "drm/amd/display: enable S/G for RAVEN chip"
drm/amdgpu: disable gfxoff on original raven
drm/amdgpu: remove experimental flag for Navi14
drm/amdgpu: disable gfxoff when using register read interface
drm/amdgpu/powerplay: properly set PP_GFXOFF_MASK (v2)
drm/amdgpu: fix bad DMA from INTERRUPT_CNTL2
drm/radeon: fix bad DMA from INTERRUPT_CNTL2
drm/amd/display: Fix debugfs on MST connectors
drm/amdgpu/nv: add asic func for fetching vbios from rom directly
drm/amdgpu: put flush_delayed_work at first
drm/amdgpu/vcn2.5: fix the enc loop with hw fini
...
Pull locking updates from Ingo Molnar:
"The main changes in this cycle were:
- A comprehensive rewrite of the robust/PI futex code's exit handling
to fix various exit races. (Thomas Gleixner et al)
- Rework the generic REFCOUNT_FULL implementation using
atomic_fetch_* operations so that the performance impact of the
cmpxchg() loops is mitigated for common refcount operations.
With these performance improvements the generic implementation of
refcount_t should be good enough for everybody - and this got
confirmed by performance testing, so remove ARCH_HAS_REFCOUNT and
REFCOUNT_FULL entirely, leaving the generic implementation enabled
unconditionally. (Will Deacon)
- Other misc changes, fixes, cleanups"
* 'locking-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (27 commits)
lkdtm: Remove references to CONFIG_REFCOUNT_FULL
locking/refcount: Remove unused 'refcount_error_report()' function
locking/refcount: Consolidate implementations of refcount_t
locking/refcount: Consolidate REFCOUNT_{MAX,SATURATED} definitions
locking/refcount: Move saturation warnings out of line
locking/refcount: Improve performance of generic REFCOUNT_FULL code
locking/refcount: Move the bulk of the REFCOUNT_FULL implementation into the <linux/refcount.h> header
locking/refcount: Remove unused refcount_*_checked() variants
locking/refcount: Ensure integer operands are treated as signed
locking/refcount: Define constants for saturation and max refcount values
futex: Prevent exit livelock
futex: Provide distinct return value when owner is exiting
futex: Add mutex around futex exit
futex: Provide state handling for exec() as well
futex: Sanitize exit state handling
futex: Mark the begin of futex exit explicitly
futex: Set task::futex_state to DEAD right after handling futex exit
futex: Split futex_mm_release() for exit/exec
exit/exec: Seperate mm_release()
futex: Replace PF_EXITPIDONE with a state
...
As we start peeking into requests for longer and longer, e.g.
incorporating use of spinlocks when only protected by an
rcu_read_lock(), we need to be careful in how we reset the request when
recycling and need to preserve any barriers that may still be in use as
the request is reset for reuse.
Quoting Linus Torvalds:
> If there is refcounting going on then why use SLAB_TYPESAFE_BY_RCU?
.. because the object can be accessed (by RCU) after the refcount has
gone down to zero, and the thing has been released.
That's the whole and only point of SLAB_TYPESAFE_BY_RCU.
That flag basically says:
"I may end up accessing this object *after* it has been free'd,
because there may be RCU lookups in flight"
This has nothing to do with constructors. It's ok if the object gets
reused as an object of the same type and does *not* get
re-initialized, because we're perfectly fine seeing old stale data.
What it guarantees is that the slab isn't shared with any other kind
of object, _and_ that the underlying pages are free'd after an RCU
quiescent period (so the pages aren't shared with another kind of
object either during an RCU walk).
And it doesn't necessarily have to have a constructor, because the
thing that a RCU walk will care about is
(a) guaranteed to be an object that *has* been on some RCU list (so
it's not a "new" object)
(b) the RCU walk needs to have logic to verify that it's still the
*same* object and hasn't been re-used as something else.
In contrast, a SLAB_TYPESAFE_BY_RCU memory gets free'd and re-used
immediately, but because it gets reused as the same kind of object,
the RCU walker can "know" what parts have meaning for re-use, in a way
it couidn't if the re-use was random.
That said, it *is* subtle, and people should be careful.
> So the re-use might initialize the fields lazily, not necessarily using a ctor.
If you have a well-defined refcount, and use "atomic_inc_not_zero()"
to guard the speculative RCU access section, and use
"atomic_dec_and_test()" in the freeing section, then you should be
safe wrt new allocations.
If you have a completely new allocation that has "random stale
content", you know that it cannot be on the RCU list, so there is no
speculative access that can ever see that random content.
So the only case you need to worry about is a re-use allocation, and
you know that the refcount will start out as zero even if you don't
have a constructor.
So you can think of the refcount itself as always having a zero
constructor, *BUT* you need to be careful with ordering.
In particular, whoever does the allocation needs to then set the
refcount to a non-zero value *after* it has initialized all the other
fields. And in particular, it needs to make sure that it uses the
proper memory ordering to do so.
NOTE! One thing to be very worried about is that re-initializing
whatever RCU lists means that now the RCU walker may be walking on the
wrong list so the walker may do the right thing for this particular
entry, but it may miss walking *other* entries. So then you can get
spurious lookup failures, because the RCU walker never walked all the
way to the end of the right list. That ends up being a much more
subtle bug.
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/20191122094924.629690-1-chris@chris-wilson.co.uk
The locks (active.lock and rq->lock) need to be taken with disabled
interrupts. This is done in i915_request_retire() by disabling the
interrupts independently of the locks itself.
While local_irq_disable()+spin_lock() equals spin_lock_irq() on vanilla
it does not on PREEMPT_RT.
Chris Wilson confirmed that local_irq_disable() was just introduced as
an optimisation to avoid enabling/disabling interrupts during
lock/unlock combo.
Enable/disable interrupts as part of the locking instruction.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20191017161352.e5z3ugse7gxl5ari@linutronix.de
If the system is being slow and userspace is racing ahead of the GPU and
finds itself waiting for the GPU to catch up, before the process sleeps
give the tasklet a kick, bypassing ksoftirqd. If the system is
overloaded, then ksoftirqd may be delayed incurring additional latency
to our user.
This should not be a frequent problem, but in the past we have observed
several hundred millisecond delays before ksoftirqd services an
interrupt, so burn a few cycles to lend a helping hand.
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/20191015132606.14349-1-chris@chris-wilson.co.uk
Since commit e2144503bf ("drm/i915: Prevent bonded requests from
overtaking each other on preemption") we have restricted requests to run
on their chosen engine across preemption events. We can take this
restriction into account to know that we will want to resubmit those
requests onto the same physical engine, and so can shortcircuit the
virtual engine selection process and keep the request on the same
engine during unwind.
References: e2144503bf ("drm/i915: Prevent bonded requests from overtaking each other on preemption")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Ramlingam C <ramalingam.c@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191013203012.25208-1-chris@chris-wilson.co.uk
If we are asked to submit a completed request, just move it onto the
active-list without modifying it's payload. If we try to emit the
modified payload of a completed request, we risk racing with the
ring->head update during retirement which may advance the head past our
breadcrumb and so we generate a warning for the emission being behind
the RING_HEAD.
v2: Commentary for the sneaky, shared responsibility between functions.
v3: Spelling mistakes and bonus assertion
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/20190923110056.15176-3-chris@chris-wilson.co.uk
(cherry picked from commit c0bb487dc1)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Forgo the struct_mutex serialisation for i915_active, and interpose its
own mutex handling for active/retire.
This is a multi-layered sleight-of-hand. First, we had to ensure that no
active/retire callbacks accidentally inverted the mutex ordering rules,
nor assumed that they were themselves serialised by struct_mutex. More
challenging though, is the rule over updating elements of the active
rbtree. Instead of the whole i915_active now being serialised by
struct_mutex, allocations/rotations of the tree are serialised by the
i915_active.mutex and individual nodes are serialised by the caller
using the i915_timeline.mutex (we need to use nested spinlocks to
interact with the dma_fence callback lists).
The pain point here is that instead of a single mutex around execbuf, we
now have to take a mutex for active tracker (one for each vma, context,
etc) and a couple of spinlocks for each fence update. The improvement in
fine grained locking allowing for multiple concurrent clients
(eventually!) should be worth it in typical loads.
v2: Add some comments that barely elucidate anything :(
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191004134015.13204-6-chris@chris-wilson.co.uk
If we are asked to submit a completed request, just move it onto the
active-list without modifying it's payload. If we try to emit the
modified payload of a completed request, we risk racing with the
ring->head update during retirement which may advance the head past our
breadcrumb and so we generate a warning for the emission being behind
the RING_HEAD.
v2: Commentary for the sneaky, shared responsibility between functions.
v3: Spelling mistakes and bonus assertion
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/20190923110056.15176-3-chris@chris-wilson.co.uk
As we need to take a walk back along the signaler timeline to find the
fence before upon which we want to wait, we need to lock that timeline
to prevent it being modified as we walk. Similarly, we also need to
acquire a reference to the earlier fence while it still exists!
Though we lack the correct locking today, we are saved by the
overarching struct_mutex -- but that protection is being removed.
v2: Tvrtko made me realise I was being lax and using annotations to
ignore the AB-BA deadlock from the timeline overlap. As it would be
possible to construct a second request that was using a semaphore from the
same timeline as ourselves, we could quite easily end up in a situation
where we deadlocked in our mutex waits. Avoid that by using a trylock
and falling back to a normal dma-fence await if contended.
v3: Eek, the signal->timeline is volatile and must be carefully
dereferenced to ensure it is valid.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190919111912.21631-2-chris@chris-wilson.co.uk
