Since we want to do a lockless read of the current active request, and
that request is written to by process_csb also without serialisation, we
need to instruct gcc to take care in reading the pointer itself.
Otherwise, we have observed execlists_active() to report 0x40.
[ 2400.760381] igt/para-4098 1..s. 2376479300us : process_csb: rcs0 cs-irq head=3, tail=4
[ 2400.760826] igt/para-4098 1..s. 2376479303us : process_csb: rcs0 csb[4]: status=0x00000001:0x00000000
[ 2400.761271] igt/para-4098 1..s. 2376479306us : trace_ports: rcs0: promote { b9c59:2622, b9c55:2624 }
[ 2400.761726] igt/para-4097 0d... 2376479311us : __i915_schedule: rcs0: -2147483648->3, inflight:0000000000000040, rq:ffff888208c1e940
which is impossible!
The answer is that as we keep the existing execlists->active pointing
into the array as we copy over that array, the unserialised read may see
a partial pointer value.
Fixes: df40306902 ("drm/i915/execlists: Lift process_csb() out of the irq-off spinlock")
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/20191125094318.1630806-1-chris@chris-wilson.co.uk
Before checking the current i915_active state for the asynchronous work
we submitted, flush any ongoing callback. This ensures that our sampling
is robust and does not sporadically fail due to bad timing as the work
is running on another cpu.
v2: Drop the fence callback sync, retiring under the lock should be good
enough to synchronize with engine_retire() and the
intel_gt_retire_requests() background worker.
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/20191122132404.690440-1-chris@chris-wilson.co.uk
From inside an active timeline in the execbuf ioctl, we may try to
reclaim some space in the GGTT. We need GGTT space for all objects on
!full-ppgtt platforms, and for context images everywhere. However, to
free up space in the GGTT we may need to remove some pinned objects
(e.g. context images) that require flushing the idle barriers to remove.
For this we use the big hammer of intel_gt_wait_for_idle()
However, commit 7936a22dd4 ("drm/i915/gt: Wait for new requests in
intel_gt_retire_requests()") will continue spinning on the wait if a
timeline is active but lacks requests, as is the case during execbuf
reservation. Spinning forever is quite time consuming, so revert that
commit and start again.
In practice, the effect commit 7936a22dd4 was trying to achieve is
accomplished by commit 1683d24c14 ("drm/i915/gt: Move new timelines
to the end of active_list"), so there is no immediate rush to replace
the looping.
Testcase: igt/gem_exec_reloc/basic-range
Fixes: a46bfdc83f ("drm/i915/gt: Wait for new requests in intel_gt_retire_requests()")
References: 1683d24c14 ("drm/i915/gt: Move new timelines to the end of active_list")
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>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191121071044.97798-1-chris@chris-wilson.co.uk
(cherry picked from commit 689122dcc3)
[Joonas: Corrected Fixes: tag ref to match drm-intel-next-fixes]
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Bonded request submission is designed to allow requests to execute in
parallel as laid out by the user. If the master request is already
finished before its bonded pair is submitted, the pair were not destined
to run in parallel and we lose the information about the master engine
to dictate selection of the secondary. If the second request was
required to be run on a particular engine in a virtual set, that should
have been specified, rather than left to the whims of a random
unconnected requests!
In the selftest, I made the mistake of not ensuring the master would
overlap with its bonded pairs, meaning that it could indeed complete
before we submitted the bonds. Those bonds were then free to select any
available engine in their virtual set, and not the one expected by the
test.
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/20191122112152.660743-1-chris@chris-wilson.co.uk
From inside an active timeline in the execbuf ioctl, we may try to
reclaim some space in the GGTT. We need GGTT space for all objects on
!full-ppgtt platforms, and for context images everywhere. However, to
free up space in the GGTT we may need to remove some pinned objects
(e.g. context images) that require flushing the idle barriers to remove.
For this we use the big hammer of intel_gt_wait_for_idle()
However, commit 7936a22dd4 ("drm/i915/gt: Wait for new requests in
intel_gt_retire_requests()") will continue spinning on the wait if a
timeline is active but lacks requests, as is the case during execbuf
reservation. Spinning forever is quite time consuming, so revert that
commit and start again.
In practice, the effect commit 7936a22dd4 was trying to achieve is
accomplished by commit 1683d24c14 ("drm/i915/gt: Move new timelines
to the end of active_list"), so there is no immediate rush to replace
the looping.
Testcase: igt/gem_exec_reloc/basic-range
Fixes: 7936a22dd4 ("drm/i915/gt: Wait for new requests in intel_gt_retire_requests()")
References: 1683d24c14 ("drm/i915/gt: Move new timelines to the end of active_list")
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>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191121071044.97798-1-chris@chris-wilson.co.uk
In commit a79ca656b6 ("drm/i915: Push the wakeref->count deferral to
the backend"), I erroneously concluded that we last modify the engine
inside __i915_request_commit() meaning that we could enable concurrent
submission for userspace as we enqueued this request. However, this
falls into a trap with other users of the engine->kernel_context waking
up and submitting their request before the idle-switch is queued, with
the result that the kernel_context is executed out-of-sequence most
likely upsetting the GPU and certainly ourselves when we try to retire
the out-of-sequence requests.
As such we need to hold onto the effective engine->kernel_context mutex
lock (via the engine pm mutex proxy) until we have finish queuing the
request to the engine.
v2: Serialise against concurrent intel_gt_retire_requests()
v3: Describe the hairy locking scheme with intel_gt_retire_requests()
for future reference.
v4: Combine timeline->lock and engine pm release; it's hairy.
Fixes: a79ca656b6 ("drm/i915: Push the wakeref->count deferral to the backend")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191120165514.3955081-2-chris@chris-wilson.co.uk
The general concept was that intel_timeline.active_count was locked by
the intel_timeline.mutex. The exception was for power management, where
the engine->kernel_context->timeline could be manipulated under the
global wakeref.mutex.
This was quite solid, as we always manipulated the timeline only while
we held an engine wakeref.
And then we started retiring requests outside of struct_mutex, only
using the timelines.active_list and the timeline->mutex. There we
started manipulating intel_timeline.active_count outside of an engine
wakeref, and so introduced a race between __engine_park() and
intel_gt_retire_requests(), a race that could result in the
engine->kernel_context not being added to the active timelines and so
losing requests, which caused us to keep the system permanently powered
up [and unloadable].
The race would be easy to close if we could take the engine wakeref for
the timeline before we retire -- except timelines are not bound to any
engine and so we would need to keep all active engines awake. The
alternative is to guard intel_timeline_enter/intel_timeline_exit for use
outside of the timeline->mutex.
Fixes: e5dadff4b0 ("drm/i915: Protect request retirement with timeline->mutex")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191120165514.3955081-1-chris@chris-wilson.co.uk
In order to avoid some nasty mutex inversions, commit 09c5ab384f
("drm/i915: Keep rings pinned while the context is active") allowed the
intel_ring unpinning to be run concurrently with the next context
pinning it. Thus each step in intel_ring_unpin() needed to be atomic and
ordered in a nice onion with intel_ring_pin() so that the lifetimes
overlapped and were always safe.
Sadly, a few steps in intel_ring_unpin() were overlooked, such as
closing the read/write pointers of the ring and discarding the
intel_ring.vaddr, as these steps were not serialised with
intel_ring_pin() and so could leave the ring in disarray.
Fixes: 09c5ab384f ("drm/i915: Keep rings pinned while the context is active")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191118230254.2615942-6-chris@chris-wilson.co.uk
Our callers fall into two categories, those passing timeout=0 who just
want to flush request retirements and those passing a timeout that need
to wait for submission completion (e.g. intel_gt_wait_for_idle()).
Currently, we only wait for a snapshot of timelines at the start of the
wait (but there was an expectation that new requests would cause timelines
to appear at the end). However, our callers, such as
intel_gt_wait_for_idle() before suspend, do require us to wait for the
power management requests emitted by retirement as well. If we don't,
then it takes an extra second or two for the background worker to flush
the queue and mark the GT as idle.
Fixes: 7e80576266 ("drm/i915: Drop struct_mutex from around i915_retire_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/20191114225736.616885-1-chris@chris-wilson.co.uk
(cherry picked from commit 7936a22dd4)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Pull drm fixes from Dave Airlie:
"Here is this weeks non-intel hw vuln fixes pull. Three drivers, all
small fixes.
i915:
- MOCS table fixes for EHL and TGL
- Update Display's rawclock on resume
- GVT's dmabuf reference drop fix
amdgpu:
- Fix a potential crash in firmware parsing
sun4i:
- One fix to the dotclock dividers range for sun4i"
* tag 'drm-fixes-2019-11-15' of git://anongit.freedesktop.org/drm/drm:
drm/amdgpu: fix null pointer deref in firmware header printing
drm/i915/tgl: MOCS table update
Revert "drm/i915/ehl: Update MOCS table for EHL"
drm/sun4i: tcon: Set min division of TCON0_DCLK to 1.
drm/i915: update rawclk also on resume
drm/i915/gvt: fix dropping obj reference twice
Our callers fall into two categories, those passing timeout=0 who just
want to flush request retirements and those passing a timeout that need
to wait for submission completion (e.g. intel_gt_wait_for_idle()).
Currently, we only wait for a snapshot of timelines at the start of the
wait (but there was an expectation that new requests would cause timelines
to appear at the end). However, our callers, such as
intel_gt_wait_for_idle() before suspend, do require us to wait for the
power management requests emitted by retirement as well. If we don't,
then it takes an extra second or two for the background worker to flush
the queue and mark the GT as idle.
Fixes: 7e80576266 ("drm/i915: Drop struct_mutex from around i915_retire_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/20191114225736.616885-1-chris@chris-wilson.co.uk
Backmerge to get dfce90259d ("Backmerge i915 security patches from
commit 'ea0b163b13ff' into drm-next") and thus 100d46bd72 ("Merge
Intel Gen8/Gen9 graphics fixes from Jon Bloomfield.").
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
- PMU "Frequency" is reported as accumulated cycles
- Avoid OOPS in dumb_create IOCTL when no CRTCs
- Mitigation for userptr put_pages deadlock with trylock_page
- Fix to avoid freeing heartbeat request too early
- Fix LRC coherency issue
- Fix Bugzilla #112212: Avoid screen corruption on MST
- Error path fix to unlock context on failed context VM SETPARAM
- Always consider holding preemption a privileged op in perf/OA
- Preload LUTs if the hw isn't currently using them to avoid color flash on VLV/CHV
- Protect context while grabbing its name for the request
- Don't resize aliasing ppGTT size
- Smaller fixes picked by tooling
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191114085213.GA6440@jlahtine-desk.ger.corp.intel.com
We repeatedly (and more so in future) use the same looping construct
over the mocs definition table to setup the register state. Refactor the
loop construct into a reusable macro.
add/remove: 2/1 grow/shrink: 1/2 up/down: 113/-330 (-217)
Function old new delta
intel_mocs_init_engine.cold - 71 +71
offset - 28 +28
__func__ 17273 17287 +14
intel_mocs_init 143 113 -30
mocs_register.isra 91 - -91
intel_mocs_init_engine 503 294 -209
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/20191112223600.30993-3-chris@chris-wilson.co.uk
This backmerges the branch that ended up in Linus' tree. It removes
all the changes for the rc6 patches from Linus' tree in favour of
a patch that is based on a large refactor that occured.
Otherwise it all looks good.
Signed-off-by: Dave Airlie <airlied@redhat.com>
In some circumstances the RC6 context can get corrupted. We can detect
this and take the required action, that is disable RC6 and runtime PM.
The HW recovers from the corrupted state after a system suspend/resume
cycle, so detect the recovery and re-enable RC6 and runtime PM.
v2: rebase (Mika)
v3:
- Move intel_suspend_gt_powersave() to the end of the GEM suspend
sequence.
- Add commit message.
v4:
- Rebased on intel_uncore_forcewake_put(i915->uncore, ...) API
change.
v5:
- Rebased on latest upstream gt_pm refactoring.
v6:
- s/i915_rc6_/intel_rc6_/
- Don't return a value from i915_rc6_ctx_wa_check().
v7:
- Rebased on latest gt rc6 refactoring.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
[airlied: pull this later version of this patch into drm-next
to make resolving the conflict mess easier.]
Signed-off-by: Dave Airlie <airlied@redhat.com>
