Jani Nikula [Wed, 26 Jun 2019 14:36:18 +0000 (17:36 +0300)]
drm/i915: add header search path to subdir Makefiles
With the subdirectories we lost the ability to build individual files on
the command line, for example:
$ make drivers/gpu/drm/i915/display/intel_display.o
This was due to the top level directory missing from header search
path. Add the header search paths to subdir Makefiles.
Note that none of the other options in the top level i915 Makefile are
taken into account when building individual files. Usually this is not a
concern.
Michal Wajdeczko [Wed, 26 Jun 2019 12:38:26 +0000 (12:38 +0000)]
drm/i915: Move OA files to separate folder
OA files look to be auto-generated so we can keep them all in
dedicated subdirectory.
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jani Nikula <jani.nikula@intel.com> Acked-by: Chris Wilson <chris@chris-wilson.co.uk> Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Acked-by: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20190626123826.39760-1-michal.wajdeczko@intel.com
EHL has it own voltage level requirement depending on cd clock.
BSpec: 21809 Cc: Clint Taylor <Clinton.A.Taylor@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190626014053.30541-3-jose.souza@intel.com
v2:
- Limiting maximum CD clock by max_cdclk_freq instead of remove it
from icl_calc_cdclk()(Ville and Jani)
BSpec: 20598 Cc: Clint Taylor <Clinton.A.Taylor@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Jani Nikula <jani.nikula@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190626014053.30541-2-jose.souza@intel.com
180 and 172.8 MHz CD clocks will only be used when audio is not
enabled as state by BSpec and implemented in
intel_crtc_compute_min_cdclk(), CD clock must be at least twice of
Azalia BCLK and BCLK by default is 96 MHz, it could be set to 48 MHz
but we are not reading it.
Chris Wilson [Wed, 26 Jun 2019 15:45:49 +0000 (16:45 +0100)]
drm/i915: Lift intel_engines_resume() to callers
Since the reset path wants to recover the engines itself, it only wants
to reinitialise the hardware using i915_gem_init_hw(). Pull the call to
intel_engines_resume() to the module init/resume path so we can avoid it
during reset.
Chris Wilson [Wed, 26 Jun 2019 15:45:48 +0000 (16:45 +0100)]
drm/i915: Only recover active engines
If we issue a reset to a currently idle engine, leave it idle
afterwards. This is useful to excise a linkage between reset and the
shrinker. When waking the engine, we need to pin the default context
image which we use for overwriting a guilty context -- if the engine is
idle we do not need this pinned image! However, this pinning means that
waking the engine acquires the FS_RECLAIM, and so may trigger the
shrinker. The shrinker itself may need to wait upon the GPU to unbind
and object and so may require services of reset; ergo we should avoid
the engine wake up path.
The danger in skipping the recovery for idle engines is that we leave the
engine with no context defined, which may interfere with the operation of
the power context on some older platforms. In practice, we should only
be resetting an active GPU but it something to look out for on Ironlake
(if memory serves).
Chris Wilson [Wed, 26 Jun 2019 15:45:47 +0000 (16:45 +0100)]
drm/i915: Add a wakeref getter for iff the wakeref is already active
For use in the next patch, we want to acquire a wakeref without having
to wake the device up -- i.e. only acquire the engine wakeref if the
engine is already active.
Ville Syrjälä [Wed, 19 Jun 2019 17:08:42 +0000 (20:08 +0300)]
drm/i915: Initialize drm_driver vblank funcs at compile time
Move the .get_vblank_timestamp() and .get_scanout_position()
initialization to happen at compile time. No point in delaying
it since we always assign the same functions.
Ville Syrjälä [Thu, 20 Jun 2019 10:33:34 +0000 (13:33 +0300)]
drm/i915: Nuke drm_driver irq vfuncs
Stop using the irq vfuncs under drm_driver. That's not going to fly
in a mixed gen environment since the structure is shared between all
the devices.
v2: Allow intel_irq_uninstall() to be called twice due to
intel_modeset_cleanup() calling it as well. Toss in a
FIXME to remind us that this is not great.
Ville Syrjälä [Wed, 19 Jun 2019 17:08:39 +0000 (20:08 +0300)]
drm/i915: Fix various tracepoints for gen2
Gen2 doesn't have a frame counter and apparently we no longer provide
a fake .get_vblank_counter() hook for it. That means all tracepoints
calling that hook will oops. Update the tracepoints to use
intel_crtc_get_vblank_counter() which will gracefully fall back to
using the software counter. This is actually a better approach since
we now get (hopefully accurate) frame numbers in the traces.
This also gets rid of the raw driver->get_vblank_counter() calls, which
we need to do in order to switch to the per-crtc vblank vfuncs.
v2: Deal with new tracepoints
v3: Use a distinct variable name for the internal crtc iterator (Chris)
Chris Wilson [Wed, 26 Jun 2019 13:44:33 +0000 (14:44 +0100)]
drm/i915/selftests: Fixup atomic reset checking
We require that the intel_gpu_reset() was atomic, not the whole of
i915_reset() which is guarded by a mutex. However, we do require that
i915_reset_engine() is atomic for use from within the submission tasklet.
Chris Wilson [Wed, 26 Jun 2019 13:44:32 +0000 (14:44 +0100)]
drm/i915/selftests: Drop manual request wakerefs around hangcheck
We no longer need to manually acquire a wakeref for request emission, so
drop the redundant wakerefs, letting us test our wakeref handling more
precisely.
Chris Wilson [Wed, 26 Jun 2019 13:44:31 +0000 (14:44 +0100)]
drm/i915/selftests: Serialise nop reset with retirement
In order for the reset count to be accurate across our selftest, we need
to prevent the background retire worker from modifying our expected
state. To preserve the intent of symmetry, we apply this to both
i915_reset and i915_reset_engine, even though it strictly only affects
i915_reset_engine currently.
Lee Shawn C [Thu, 20 Jun 2019 15:44:16 +0000 (08:44 -0700)]
drm/i915: Check backlight type while doing eDP backlight initializaiton
If LFP backlight type setting from VBT was "VESA eDP AUX Interface".
Driver should check panel capability and try to initialize aux backlight.
No matter i915_modparams.enable_dpcd_backlight was enabled or not.
v2: access dev_priv->vbt.backlight.type directly and remove unused function.
v3: 1. Modify i915.enable_dpcd_backlight type from bool to int and give default
value as 0 (disable).
2. Add a judgement to check LFP backlight type was aux interface or not.
Chris Wilson [Tue, 25 Jun 2019 23:33:48 +0000 (00:33 +0100)]
drm/i915/gt: Drop stale commentary for timeline density
We no longer allocate a contiguous set of timeline ids for all engines
upon creation, so we no longer should assume that the timelines are
densely allocated within a context. Hopefully, the set of fences used
within a workload are still dense enough for us to take advantage of
the compressed radix tree used for the syncmap.
Chris Wilson [Tue, 25 Jun 2019 13:01:14 +0000 (14:01 +0100)]
drm/i915: Rename intel_wakeref_[is]_active
Our general rule is to use is/has as the verb for boolean functions,
rename intel_wakeref_active to intel_wakeref_is_active so the question
being asked is clear.
Robert M. Fosha [Tue, 25 Jun 2019 16:41:07 +0000 (09:41 -0700)]
drm/i915/guc: Add debug capture of GuC exception
Detect GuC firmware load failure due to an exception during execution
in GuC firmware. Output the GuC EIP where exception occurred to dmesg
for GuC debug information.
v2: correct typos, change debug message and error code returned for
GuC exception (Michal)
Signed-off-by: Robert M. Fosha <robert.m.fosha@intel.com> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20190625164107.21512-1-robert.m.fosha@intel.com
Chris Wilson [Tue, 25 Jun 2019 13:01:09 +0000 (14:01 +0100)]
drm/i915/execlists: Convert recursive defer_request() into iterative
As this engine owns the lock around rq->sched.link (for those waiters
submitted to this engine), we can use that link as an element in a local
list. We can thus replace the recursive algorithm with an iterative walk
over the ordered list of waiters.
EHL can have up to one VECS(video enhancement) engine, so add it to
the device_info.
BSpec: 29152 Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Bob Paauwe <bob.j.paauwe@intel.com> Cc: Matt Roper <matthew.d.roper@intel.com> Cc: Clint Taylor <Clinton.A.Taylor@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190614213749.15870-1-jose.souza@intel.com
We got the wrong offsets (could they have changed?). New values were
computed off an error state by looking up the register offset in the
context image as written by the HW.
Kenneth Graunke [Tue, 25 Jun 2019 09:06:55 +0000 (10:06 +0100)]
drm/i915: Disable SAMPLER_STATE prefetching on all Gen11 steppings.
The Demand Prefetch workaround (binding table prefetching) only applies
to Icelake A0/B0. But the Sampler Prefetch workaround needs to be
applied to all Gen11 steppings, according to a programming note in the
SARCHKMD documentation.
Using the Intel Gallium driver, I have seen intermittent failures in
the dEQP-GLES31.functional.copy_image.non_compressed.* tests. After
applying this workaround, the tests reliably pass.
Chris Wilson [Mon, 24 Jun 2019 09:20:09 +0000 (10:20 +0100)]
drm/i915/execlists: Always clear ring_pause if we do not submit
In the unlikely case (thank you CI!), we may find ourselves wanting to
issue a preemption but having no runnable requests left. In this case,
we set the semaphore before computing the preemption and so must unset
it before forgetting (or else we leave the machine busywaiting until the
next request comes along and so likely hang).
v2: Replace readback with only a wmb after asserting the semaphore
Chris Wilson [Fri, 21 Jun 2019 21:57:33 +0000 (22:57 +0100)]
drm/i915/blt: Remove recursive vma->lock
As we have already plugged the w->dma into the reservation_object, and
have set ourselves up to automatically signal the request and w->dma on
completion, we do not need to export the rq->fence directly and just use
the w->dma fence.
This avoids having to take the reservation_lock inside the worker which
cross-release lockdep would complain about. :)
Chris Wilson [Fri, 21 Jun 2019 18:38:00 +0000 (19:38 +0100)]
drm/i915: Provide an i915_active.acquire callback
If we introduce a callback for i915_active that is only called the first
time we use the i915_active and is symmetrically paired with the
i915_active.retire callback, we can replace the open-coded and
non-atomic implementations -- which will be very fragile (i.e. broken)
upon removing the struct_mutex serialisation.
Chris Wilson [Fri, 21 Jun 2019 18:37:59 +0000 (19:37 +0100)]
drm/i915: Throw away the active object retirement complexity
Remove the accumulated optimisations that we have for i915_vma_retire
and reduce it to the bare essential of tracking the active object
reference. This allows us to only use atomic operations, and so will be
able to avoid the struct_mutex requirement.
The principal loss here is the shrinker MRU bumping, so now if we have
to shrink, we will do so in much more random order and more likely to
try and shrink recently used objects. That is a nuisance, but shrinking
active objects is a second step we try to avoid and will always be a
system-wide performance issue.
The other loss is here is in the automatic pruning of the
reservation_object when idling. This is not as large an issue as upon
reservation_object introduction as now adding new fences into the object
replaces already signaled fences, keeping the array compact. But we do
lose the auto-expiration of stale fences and unused arrays. That may be
a noticeable problem for which we need to re-implement autopruning.
Chris Wilson [Fri, 21 Jun 2019 18:37:58 +0000 (19:37 +0100)]
drm/i915: Track i915_active using debugobjects
Provide runtime asserts and tracking of i915_active via debugobjects.
For example, this should allow us to check that the i915_active is only
active when we expect it to be and is never freed too early.
One consequence is that, for simplicity, we no longer allow i915_active
to be on-stack which only affected the selftests.
Chris Wilson [Fri, 21 Jun 2019 18:37:57 +0000 (19:37 +0100)]
drm/i915: Remove waiting & retiring from shrinker paths
i915_gem_wait_for_idle() and i915_retire_requests() introduce a
dependency on the timeline->mutex. This is problematic as we want to
later perform allocations underneath i915_active.mutex, forming a link
between the shrinker, the timeline and active mutexes. Nip this cycle in
the bud by removing the acquisition of the timeline mutex (i.e.
retiring) from inside the shrinker.
drm/i915/guc: handle GuC messages received with CTB disabled
There is a very small chance of triggering a log flush event when
enabling or disabling CT buffers. Events triggered while CT buffers
are disabled are logged in the SCRATCH_15 register using the same bits
used in the CT message payload. Since our communication channel with
GuC is turned off, we can save the message and handle it after we turn
it back on.
GuC should be idle and not generate more events in the meantime because
we're not talking to it.
v2: clear the mmio register on stop_communication as well (Chris)
drm/i915/guc: reorder enable/disable communication steps
Make sure we always have CT buffers enabled when the interrupts are
enabled, so we can always handle interrupts from GuC. Also move the
setting of the guc->send and guc->handler functions to the GuC
communication control functions for consistency.
The reorder also fixes the onion unwinding of intel_uc_init_hw, because
guc_enable_communication would've left interrupts enabled when failing
to enable CTB.
v2: always retunr the result of ctch_enable() in
intel_guc_ct_enable() (Michal)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110943 Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Reviewed-by: Matthew Brost <matthew.brost@intel.com> Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20190621182123.31368-1-daniele.ceraolospurio@intel.com
Chris Wilson [Fri, 21 Jun 2019 13:16:40 +0000 (14:16 +0100)]
drm/i915/gt: Fixup kerneldoc parameters
drivers/gpu/drm/i915/gt/intel_mocs.c:513: warning: Function parameter or member 'gt' not described in 'intel_mocs_init_l3cc_table'
drivers/gpu/drm/i915/gt/intel_mocs.c:513: warning: Excess function parameter 'dev_priv' description in 'intel_mocs_init_l3cc_table'
intel_vgt_balloon/deballoon, i915_ggtt_probe_hw intel_wopcm_init_hw need
similar treatment
Chris Wilson [Fri, 21 Jun 2019 13:16:39 +0000 (14:16 +0100)]
drm/i915/gt: Rename i915_gt_timelines
Since the anonymous i915_gt became struct intel_gt and encloses
struct i915_gt_timelines, rename i915_gt_timelines to intel_gt_timelines
to match its parentage.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:08:11 +0000 (08:08 +0100)]
drm/i915: Eliminate dual personality of i915_scratch_offset
Scratch vma lives under gt but the API used to work on i915. Make this
consistent by renaming the function to intel_gt_scratch_offset and make
it take struct intel_gt.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:08:05 +0000 (08:08 +0100)]
drm/i915: Compartmentalize i915_gem_init_ggtt
Continuing on the theme of better logical organization of our code, make
the first step towards making the ggtt code better isolated from wider
struct drm_i915_private.
v2:
* Bring the ickle onion unwind back. (Chris)
* Rename to i915_init_ggtt. (Chris)
Tvrtko Ursulin [Fri, 21 Jun 2019 07:08:04 +0000 (08:08 +0100)]
drm/i915: Compartmentalize i915_ggtt_cleanup_hw
Continuing on the theme of better logical organization of our code, make
the first step towards making the ggtt code better isolated from wider
struct drm_i915_private.
v2:
* Cleanup of mm.wc_stash does not need struct_mutex. (Chris)
Continuing on the theme of better logical organization of our code, make
the first step towards making the timeline code better isolated from wider
struct drm_i915_private.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:08:01 +0000 (08:08 +0100)]
drm/i915: Convert i915_gem_flush_ggtt_writes to intel_gt
Having introduced struct intel_gt (named the anonymous structure in i915)
we can start using it to compartmentalize our code better. It makes more
sense logically to have the code internally like this and it will also
help with future split between gt and display in i915.
v2:
* Keep ggtt flush before fb obj flush. (Chris)
Having made start to better code compartmentalization by introducing
struct intel_gt, continue the theme elsewhere in code by making functions
take parameters take what logically makes most sense for them instead of
the global struct drm_i915_private.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:07:57 +0000 (08:07 +0100)]
drm/i915: Compartmentalize i915_ggtt_init_hw
Having made start to better code compartmentalization by introducing
struct intel_gt, continue the theme elsewhere in code by making functions
take parameters take what logically makes most sense for them instead of
the global struct drm_i915_private.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:07:56 +0000 (08:07 +0100)]
drm/i915: Compartmentalize i915_ggtt_probe_hw
Having made start to better code compartmentalization by introducing
struct intel_gt, continue the theme elsewhere in code by making functions
take parameters take what logically makes most sense for them instead of
the global struct drm_i915_private.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:07:53 +0000 (08:07 +0100)]
drm/i915: Convert i915_gem_init_hw to intel_gt
More removal of implicit dev_priv from using old mmio accessors.
Actually the top level function remains but is split into a part which
writes to i915 and part which operates on intel_gt in order to initialize
the hardware.
GuC and engines are the only odd ones out remaining.
Tvrtko Ursulin [Fri, 21 Jun 2019 07:07:45 +0000 (08:07 +0100)]
drm/i915: Convert i915_gem_init_swizzling to intel_gt
Start using the newly introduced struct intel_gt to fuse together correct
logical init flow with uncore for more removal of implicit dev_priv in
mmio access.
Chris Wilson [Thu, 20 Jun 2019 18:37:05 +0000 (19:37 +0100)]
drm/i915/gtt: Defer address space cleanup to an RCU worker
Enable RCU protection of i915_address_space and its ppgtt superclasses,
and defer its cleanup into a worker executed after an RCU grace period.
In the future we will be able to use the RCU protection to reduce the
locking around VM lookups, but the immediate benefit is being able to
defer the release into a kworker (process context). This is required as
we may need to sleep to reap the WC pages stashed away inside the ppgtt.
Chris Wilson [Thu, 20 Jun 2019 10:24:32 +0000 (11:24 +0100)]
drm/i915/selftests: Use request managed wakerefs
Since commit 14ab88e3a048 ("drm/i915: Invert the GEM wakeref
hierarchy"), the request creation itself took responsibility for
managing the engine/GT wakerefs and so we can remove the redundant grabs
in our selftests.
Chris Wilson [Wed, 19 Jun 2019 20:35:04 +0000 (21:35 +0100)]
drm/i915: Rings are always flushed
Our intel_rings are always flushed as they are continually used to submit
commands to the GPU, and so do not need to be flushed on unpinning. This
avoids pulling in the flush_ggtt_writes locking into our context
unpin, which we want to allow from atomic context (for simplicity).
Chris Wilson [Thu, 20 Jun 2019 14:20:52 +0000 (15:20 +0100)]
drm/i915/execlists: Minimalistic timeslicing
If we have multiple contexts of equal priority pending execution,
activate a timer to demote the currently executing context in favour of
the next in the queue when that timeslice expires. This enforces
fairness between contexts (so long as they allow preemption -- forced
preemption, in the future, will kick those who do not obey) and allows
us to avoid userspace blocking forward progress with e.g. unbounded
MI_SEMAPHORE_WAIT.
For the starting point here, we use the jiffie as our timeslice so that
we should be reasonably efficient wrt frequent CPU wakeups.
Chris Wilson [Thu, 20 Jun 2019 14:20:51 +0000 (15:20 +0100)]
drm/i915/execlists: Preempt-to-busy
When using a global seqno, we required a precise stop-the-workd event to
handle preemption and unwind the global seqno counter. To accomplish
this, we would preempt to a special out-of-band context and wait for the
machine to report that it was idle. Given an idle machine, we could very
precisely see which requests had completed and which we needed to feed
back into the run queue.
However, now that we have scrapped the global seqno, we no longer need
to precisely unwind the global counter and only track requests by their
per-context seqno. This allows us to loosely unwind inflight requests
while scheduling a preemption, with the enormous caveat that the
requests we put back on the run queue are still _inflight_ (until the
preemption request is complete). This makes request tracking much more
messy, as at any point then we can see a completed request that we
believe is not currently scheduled for execution. We also have to be
careful not to rewind RING_TAIL past RING_HEAD on preempting to the
running context, and for this we use a semaphore to prevent completion
of the request before continuing.
To accomplish this feat, we change how we track requests scheduled to
the HW. Instead of appending our requests onto a single list as we
submit, we track each submission to ELSP as its own block. Then upon
receiving the CS preemption event, we promote the pending block to the
inflight block (discarding what was previously being tracked). As normal
CS completion events arrive, we then remove stale entries from the
inflight tracker.
v2: Be a tinge paranoid and ensure we flush the write into the HWS page
for the GPU semaphore to pick in a timely fashion.
drm/i915/gvt: decouple check_vgpu() from uncore_init()
With multiple uncore to initialize (GT vs Display), it makes little
sense to have the vgpu_check inside uncore_init(). We also have
a catch-22 scenario where the uncore is required to read the vgpu
capabilities while the vgpu capabilities are required to decide if
we need to initialize forcewake support. To remove this circular
dependency, we can perform the required MMIO access by mmapping just
the vgtif shared page in mmio space and use raw accessors.
We'd like to introduce a display uncore with no forcewake domains, so
let's avoid wasting memory and be ready to allocate only what we need.
Even without multiple uncore, we still don't need all the domains on all
gens.
v2: avoid hidden control flow, improve checks (Tvrtko), fix IVB special
case, add failure injection point
drm/i915: skip forcewake actions on forcewake-less uncore
We always call some of the setup/cleanup functions for forcewake, even
if the feature is not actually available. Skipping these operations if
forcewake is not available saves us some operations on older gens and
prepares us for having a forcewake-less display uncore.
v2: do not make suspend/resume functions forcewake-specific (Chris,
Tvrtko), use GEM_BUG_ON in internal forcewake-only functions (Tvrtko)
drm/i915: use vfuncs for reg_read/write_fw_domains
Instead of going through the if-else chain every time, let's save the
function in the uncore structure. Note that the new functions are
purposely not used from the reg read/write functions to keep the
inlining there.
While at it, use the new macro to call the old ones to clean the code a
bit.
v2: Rename macros for no-forcewake function assignment (Tvrtko)
Chris Wilson [Wed, 19 Jun 2019 17:01:35 +0000 (18:01 +0100)]
drm/i915: Keep rings pinned while the context is active
Remember to keep the rings pinned as well as the context image until the
GPU is no longer active.
v2: Introduce a ring->pin_count primarily to hide the
mock_ring that doesn't fit into the normal GGTT vma picture.
v3: Order is important in teardown, ringbuffer submission needs to drop
the pin count on the engine->kernel_context before it can gleefully free
its ring.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110946 Fixes: 66a3baea852f ("drm/i915: Keep contexts pinned until after the next kernel context switch") 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/20190619170135.15281-1-chris@chris-wilson.co.uk
Matt Roper [Tue, 18 Jun 2019 17:51:31 +0000 (10:51 -0700)]
drm/i915/ehl: Allow combo PHY A to drive a third external display
EHL has a mux on combo PHY A that allows it to be driven either by an
internal display (DDI-A or DSI DPHY) or by an external display (DDI-D).
This is a motherboard design decision that can not be changed on the
fly. Unfortunately there are no strap registers that allow us to detect
the board configuration directly, so let's use the VBT to try to figure
it out and program the mux accordingly.
For now if we run across a broken VBT that tries to claim that PHY A
is attached to both internal and external displays at the same time,
we'll resolve the conflict in favor of the internal display. To help
debug these kind of bad VBT's, let's also add a quick DRM_DEBUG message
during child device parsing so that it's easier to understand these
cases if they show up in bug reports.
v2:
- Confirmed that VBT's dvo port refers to the DDI and not the PHY.
Thus we can check more explicitly for (ddi_d && !(ddi_a || dsi)). If
a bad VBT contradicts itself, let internal display win. (Ville)
v3:
- Switch condition from !IS_ICELAKE to IS_ELKHARTLAKE. Although the
convention is usually to assume that future platforms will inherit
all current platform behavior, this feels more like a one-platform
quirk. (Ville)
- Update commit message to describe what we do if/when we encounter
broken VBT's, and note that the new debug print during child device
parsing is intentional.
Chris Wilson [Tue, 18 Jun 2019 07:41:31 +0000 (08:41 +0100)]
drm/i915: Flush the execution-callbacks on retiring
In the unlikely case the request completes while we regard it as not even
executing on the GPU (see the next patch!), we have to flush any pending
execution callbacks at retirement and ensure that we do not add any
more.
Chris Wilson [Wed, 19 Jun 2019 11:23:37 +0000 (12:23 +0100)]
drm/i915: Signal fence completion from i915_request_wait
With the upcoming change to automanaged i915_active, the intent is that
whenever we wait on the set of active fences, they are signaled and
collected. The requirement is that all successful returns from
i915_request_wait() signal the fence, so fixup the one remaining path
where we may return before the interrupt has been run.
Chris Wilson [Tue, 18 Jun 2019 07:41:30 +0000 (08:41 +0100)]
drm/i915: Stop passing I915_WAIT_LOCKED to i915_request_wait()
Since commit f879ea8f2683 ("drm/i915: Remove GPU reset dependence on
struct_mutex"), the I915_WAIT_LOCKED flags passed to i915_request_wait()
has been defunct. Now go ahead and remove it from all callers.
Chris Wilson [Tue, 18 Jun 2019 11:07:36 +0000 (12:07 +0100)]
drm/i915/execlists: Detect cross-contamination with GuC
The process_csb routine from execlists_submission is incompatible with
the GuC backend. Add a warning to detect if we accidentally end up in
the wrong spot.
Chris Wilson [Tue, 18 Jun 2019 07:41:35 +0000 (08:41 +0100)]
drm/i915: Make the semaphore saturation mask global
The idea behind keeping the saturation mask local to a context backfired
spectacularly. The premise with the local mask was that we would be more
proactive in attempting to use semaphores after each time the context
idled, and that all new contexts would attempt to use semaphores
ignoring the current state of the system. This turns out to be horribly
optimistic. If the system state is still oversaturated and the existing
workloads have all stopped using semaphores, the new workloads would
attempt to use semaphores and be deprioritised behind real work. The
new contexts would not switch off using semaphores until their initial
batch of low priority work had completed. Given sufficient backload load
of equal user priority, this would completely starve the new work of any
GPU time.
To compensate, remove the local tracking in favour of keeping it as
global state on the engine -- once the system is saturated and
semaphores are disabled, everyone stops attempting to use semaphores
until the system is idle again. One of the reason for preferring local
context tracking was that it worked with virtual engines, so for
switching to global state we could either do a complete check of all the
virtual siblings or simply disable semaphores for those requests. This
takes the simpler approach of disabling semaphores on virtual engines.
The downside is that the decision that the engine is saturated is a
local measure -- we are only checking whether or not this context was
scheduled in a timely fashion, it may be legitimately delayed due to user
priorities. We still have the same dilemma though, that we do not want
to employ the semaphore poll unless it will be used.
v2: Explain why we need to assume the worst wrt virtual engines.
To do frontbuffer tracking we are depending on Display WA #0884 to
exit PSR when there is a frontbuffer modification but according to
user reports a write to CURSURFLIVE do not cause PSR to exit in older
gens so lets force a PSR exit.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110799 Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: José Roberto de Souza <jose.souza@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Tested-by: Thomas Rohwer <trohwer85@gmail.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190617195154.30292-1-jose.souza@intel.com
Chris Wilson [Tue, 18 Jun 2019 16:19:51 +0000 (17:19 +0100)]
drm/i915: Don't dereference request if it may have been retired when printing
This has caught me out on countless occasions, when we retrieve a pointer
from the submission/execlists backend, it does not carry a reference to
the context or ring. Those are only pinned while the request is active,
so if we see the request is already completed, it may be in the process
of being retired and those pointers defunct.
projects / kernel.git / log