Currently object size is returned for the rotated VMA size which can be
bigger than the rotated view itself. Since the binding code pads all
excess size with scratch pages the only minor issue with this is wasting
some GGTT space, but still feels nicer to fix and report the real size.
v2: Rebase for tracking size in bytes instead of pages.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This way data is available as soon as the view is passed into the call chain.
v2: Store size in bytes instead of pages under the appropriate name. (Chris Wilson)
v3: Use uint64_t instead of size_t. (Daniel Vetter)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> (v2)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It is only used in logging and it doesn't need to exist on its own.
Also it was misleading to log view size as object size.
v2: Improve commit message. (Joonas Lahtinen)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
[danvet: s/%lu/%zu/ where needed, reported by 0-day.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With the new DRRS code it kinda sticks out, and we never managed to
get this to work well enough without causing issues. Time to wave
goodbye.
I've decided to keep the logic for programming the reduced clocks
intact, but everything else is gone. If anyone ever wants to resurrect
this we need to redo it all anyway on top of the frontbuffer tracking.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This typo lead to the crtc scaler getting enabled incorrectly and an
evantual state checker mismatch about the scaler_id.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In Indirect context w/a batch buffer,
WaClearSlmSpaceAtContextSwitch
This WA performs writes to scratch page so it must be valid, this check
is performed before initializing the batch with this WA.
v2: s/PIPE_CONTROL_FLUSH_RO_CACHES/PIPE_CONTROL_FLUSH_L3 (Ville)
v3: GTT bit in scratch address should be mbz (Chris)
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Rafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The frontbuffer code gives us accurate information about activity,
let's use it. Again this should avoid unecessary updates when multiple
screens are on.
Also realign function paramaters, I couldn't resist that bit of OCD.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Durgadoss R <durgadoss.r@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The current code tracks business across all pipes, but we're only
really interested in the one pipe DRRS is enabled on. Fairly tiny
optimization, but something I noticed while reading the code. But it
might matter a bit when e.g. showing a video or something only on the
external screen, while the panel is kept static.
Also regroup the code slightly: First compute new bitmasks, then take
appropriate actions.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Durgadoss R <durgadoss.r@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The current code tracks business across all pipes, but we're only
really interested in the one pipe DRRS is enabled on. Fairly tiny
optimization, but something I noticed while reading the code. But it
might matter a bit when e.g. showing a video or something only on the
external screen, while the panel is kept static.
Also regroup the code slightly: First compute new bitmasks, then take
appropriate actions.
Cc: Ramalingam C <ramalingam.c@intel.com>
Cc: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
I was momentarily confused until I've double-checked that these
functions really only compute state and don't update the hardware
state. They once did that, but since Ander's rework of the dpll
computation flow that's no longer the case.
Rename them to avoid further confusion.
Note that the ilk code already follows the compute_dpll naming scheme
for computing the actual register value. DDI code goes with _calc_,
but that is close enough.
Cc: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Useful to figure out whether stuck bits are due to the frontbuffer
tracking code as opposed to individual consumers (who have their own
bitmask tracking).
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Paulo noticed that the fbc frontbuffer tracking flush callback
occasionally gets a call without any bit set. This can happen when we
have to filter flush calls due to e.g. gpu rendering. Filter these
out.
Reported-by: Paulo Zanoni <przanoni@gmail.com>
Cc: Paulo Zanoni <przanoni@gmail.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
The current/old frontbuffer might still have gpu frontbuffer rendering
pending. But once flipped it won't have the corresponding frontbuffer
bits any more and hence the request retire function won't ever clear
the corresponding busy bits. The async flip tracking (with the
flip_prepare and flip_complete functions) already does this, but
somehow I've forgotten to do this for synchronous flips.
Note that we don't track outstanding rendering of the new framebuffer
with busy_bits since all our plane update code waits for previous
rendering to complete before displaying a new buffer. Hence a new
buffer will never be busy.
v2: Drop the spurious inline Ville spotted.
v3: Don't touch flip_bits in the synchronsou frontbuffer_flip
function, noticed by Paulo.
v4: Remove one more inline that slipped through (Paulo).
Reported-by: Paulo Zanoni <przanoni@gmail.com>
Cc: Paulo Zanoni <przanoni@gmail.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Testcase: igt/kms_frontbuffer_tracking/fbc-modesetfrombusy
Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
To initialize WA batch, at the moment we first allocate batch and then check
whether we have any WA to be initialized for the given Gen; if we don't have
any WA then we WARN the user, destroy the batch and return but this is causing
another WARN in cleanup code complaining about sleeping in atomic context.
Till we understand this better and to keep things simpler, bail out early
if we don't have WA.
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Kernel 0-day framework reported warnings with WA batch patches, this patch
fixes those warnings and an additional warning reported in intel_lrc.c file.
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As there is no OLR to check, the check_olr() function is now a no-op and can be
removed.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
A bunch of the low level LRC functions were passing around ringbuf and ctx
pairs. In a few cases, they took the r/c pair and a request as well. This is all
quite messy and unnecesary. The context_queue() call is especially bad since the
fake request code got removed - it takes a request and three extra things that
must be extracted from the request and then it checks them against what it finds
in the request. Removing all the derivable data makes the code much simpler all
round.
This patch updates those functions to just take the request structure.
Note that logical_ring_wait_for_space now takes a request structure but already
had a local request pointer that it uses to scan for something to wait on. To
avoid confusion the local variable has been renamed 'target' (it is searching
for a target request to do something with) and the parameter has been called req
(to guarantee anything accidentally missed gets a compiler error).
v2: Updated commit message re wait_for_space (Tomas Elf review comment).
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The LRC submission code requires a request for tracking purposes. It does not
actually require that request to 'complete' it simply uses it for keeping hold
of reference counts on contexts and such like.
Previously, the fall back path of polling for space in the ring would start by
submitting any outstanding work that was sat in the buffer. This submission was
not done as part of the request that that work was owned by because that would
lead to complications with the request being submitted twice. Instead, a null
request structure was passed in to the submit call and a fake one was created.
That fall back path has long since been obsoleted and has now been removed. Thus
there is never any need to fake up a request structure. This patch removes that
code. A couple of sanity check warnings are added as well, just in case.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Thomas Daniel <thomas.daniel@intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In _i915_add_request(), the request is associated with a userland client.
Specifically it is linked to the 'file' structure and the current user process
is recorded. One problem here is that the current user process is not
necessarily the same as when the request was submitted to the driver. This is
especially true when the GPU scheduler arrives and decouples driver submission
from hardware submission. Note also that it is only in the case where the add
request comes from an execbuff call that there is a client to associate. Any
other add request call is kernel only so does not need to do it.
This patch moves the client association into a separate function. This is then
called from the execbuffer code path itself at a sensible time. It also removes
the now redundant 'file' pointer from the add request parameter list.
An extra cleanup of the client association is also added to the request clean up
code for the eventuality where the request is killed after association but
before being submitted (e.g. due to out of memory error somewhere). Once the
submission has happened, the request is on the request list and the regular
request list removal will clear the association. Note that this still needs to
happen at this point in time because the request might be kept floating around
much longer (due to someone holding a reference count) and the client should not
be worrying about this request after it has been retired.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The outstanding_lazy_request is no longer used anywhere in the driver.
Everything that was looking at it now has a request explicitly passed in from on
high. Everything that was relying upon it behind the scenes is now explicitly
creating/passing/submitting its own private request. Thus the OLR can be
removed.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Much of the driver has now been converted to passing requests around instead of
rings/ringbufs/contexts. Thus the function for retreiving the request from a
ring (i.e. the OLR) is no longer used and can be removed.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that the *_ring_begin() functions no longer call the request allocation
code, it is finally safe for the request allocation code to call *_ring_begin().
This is important to guarantee that the space reserved for the subsequent
i915_add_request() call does actually get reserved.
v2: Renamed functions according to review feedback (Tomas Elf).
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that everything above has been converted to use requests,
intel_logical_ring_begin() can be updated to take a request instead of a
ringbuf/context pair. This also means that it no longer needs to lazily allocate
a request if no-one happens to have done it earlier.
Note that this change makes the execlist signature the same as the legacy
version. Thus the two functions could be merged into a ring->begin() wrapper if
required.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that everything above has been converted to use requests, intel_ring_begin()
can be updated to take a request instead of a ring. This also means that it no
longer needs to lazily allocate a request if no-one happens to have done it
earlier.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated intel_ring_cacheline_align() to take a request instead of a ring.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the various ring->signal() implementations to take a request instead of
a ring. This removes their reliance on the OLR to obtain the seqno value that
should be used for the signal.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the ring->sync_to() implementations to take a request instead of a ring.
Also updated the tracer to include the request id.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
[danvet: Rebase since I didn't merge the patch which added ->uniq.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the ring->emit_bb_start() implementation to take a request instead of a
ringbuf/context pair.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the various ring->dispatch_execbuffer() implementations to take a
request instead of a ring.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the ring->emit_request() implementation to take a request instead of a
ringbuf/request pair. Also removed its use of the OLR for obtaining the
request's seqno.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the various ring->add_request() implementations to take a request
instead of a ring. This removes their reliance on the OLR to obtain the seqno
value that the request should be tagged with.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the various ring->emit_flush() implementations to take a request instead
of a ringbuf/context pair.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated intel_emit_post_sync_nonzero_flush(), gen7_render_ring_cs_stall_wa() and
gen8_emit_pipe_control() to take requests instead of rings.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the various ring->flush() functions to take a request instead of a ring.
Also updated the tracer to include the request id.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
[danvet: Rebase since I didn't merge the addition of req->uniq.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the switch_mm() code paths to take a request instead of a ring. This
includes the myriad *_mm_switch functions themselves and a bunch of PDP related
helper functions.
v2: Rebased to newer tree.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the *_ring_flush_all_caches() functions to take requests instead of
rings or ringbuf/context pairs.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the *_ring_workarounds_emit() functions to take requests instead of
ring/context pairs.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated *_ring_invalidate_all_caches(), i915_reset_gen7_sol_offsets() and
i915_emit_box() to take request structures instead of ring or ringbuf/context
pairs.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated mi_set_context() to take a request structure instead of a ring and
context pair.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Converted i915_gem_l3_remap() to take a request structure instead of a ring.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that everything above has been converted to use request structures, it is
possible to update the lower level move_to_active() functions to be request
based as well.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that all callers of i915_add_request() have a request pointer to hand, it is
possible to update the add request function to take a request pointer rather
than pulling it out of the OLR.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the display page flip code to do explicit request creation and
submission rather than relying on the OLR and just hoping that the request
actually gets submitted at some random point.
The sequence is now to create a request, queue the work to the ring, assign the
known request to the flip queue work item then actually submit the work and post
the request.
Note that every single flip function used to finish with
'__intel_ring_advance(ring);'. However, immediately after they return there is
now an add request call which will do the advance anyway. Thus the many
duplicate advance calls have been removed.
v2: Updated commit message with comment about advance removal.
v3: The request can now be allocated by the _sync() code earlier on. Thus the
page flip path does not necessarily need to allocate a new request, it may be
able to re-use one.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The overlay update code path to do explicit request creation and submission
rather than relying on the OLR to do the right thing.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The plan is to pass requests around as the basic submission tracking structure
rather than rings and contexts. This patch updates the i915_gem_object_sync()
code path.
v2: Much more complex patch to share a single request between the sync and the
page flip. The _sync() function now supports lazy allocation of the request
structure. That is, if one is passed in then that will be used. If one is not,
then a request will be allocated and passed back out. Note that the _sync() code
does not necessarily require a request. Thus one will only be created until
certain situations. The reason the lazy allocation must be done within the
_sync() code itself is because the decision to need one or not is not really
something that code above can second guess (except in the case where one is
definitely not required because no ring is passed in).
The call chains above _sync() now support passing a request through which most
callers passing in NULL and assuming that no request will be required (because
they also pass in NULL for the ring and therefore can't be generating any ring
code).
The exeception is intel_crtc_page_flip() which now supports having a request
returned from _sync(). If one is, then that request is shared by the page flip
(if the page flip is of a type to need a request). If _sync() does not generate
a request but the page flip does need one, then the page flip path will create
its own request.
v3: Updated comment description to be clearer about 'to_req' parameter (Tomas
Elf review request). Rebased onto newer tree that significantly changed the
synchronisation code.
v4: Updated comments from review feedback (Tomas Elf)
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the two render_state_init() functions to take a request pointer instead
of a ring. This removes their reliance on the OLR.
v2: Rebased to newer tree.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that everything above has been converted to use requests, it is possible to
update init_context() to take a request pointer instead of a ring/context pair.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In execlist mode, context initialisation is deferred until first use of the
given context. This is because execlist mode has per ring context state and thus
many more context storage objects than legacy mode and many are never actually
used. Previously, the initialisation commands were written to the ring and
tagged with some random request structure via the OLR. This seemed to be causing
a null pointer deference bug under certain circumstances (BZ:88865).
This patch adds explicit request creation and submission to the deferred
initialisation code path. Thus removing any reliance on or randomness caused by
the OLR.
Note that it should be possible to move the deferred context creation until even
later - when the context is actually switched to rather than when it is merely
validated. This would allow the initialisation to be done within the request of
the work that is wanting to use the context. Hence, the extra request that is
created, used and retired just for the context init could be removed completely.
However, this is left for a follow up patch.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated do_switch() to take a request pointer instead of a ring/context pair.
v2: Removed some overzealous req-> dereferencing.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that the request is guaranteed to specify the context, it is possible to
update the context switch code to use requests rather than ring and context
pairs. This patch updates i915_switch_context() accordingly.
Also removed the warning that the request's context must match the last context
switch's context. As the context switch now gets the context object from the
request structure, there is no longer any scope for the two to become out of
step.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The final step in removing the OLR from i915_gem_init_hw() is to pass the newly
allocated request structure in to each step rather than passing a ring
structure. This patch updates both i915_ppgtt_init_ring() and
i915_gem_context_enable() to take request pointers.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that a single per ring loop is being done for all the different
intialisation steps in i915_gem_init_hw(), it is possible to add proper request
management as well. The last remaining issue is that the context enable call
eventually ends up within *_render_state_init() and this does its own private
_i915_add_request() call.
This patch adds explicit request creation and submission to the top level loop
and removes the add_request() from deep within the sub-functions.
v2: Updated for removal of batch_obj from add_request call in previous patch.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The render state initialisation code does an explicit i915_add_request() call to
commit the init commands. It was passing in the initialisation batch buffer to
add_request() as the batch object parameter. However, the batch object entry in
the request structure (which is all that parameter is used for) is meant for
keeping track of user generated batch buffers for blame tagging during GPU
hangs.
This patch clears the batch object parameter so that kernel generated batch
buffers are not tagged as being user generated.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The start of day context initialisation code in i915_gem_context_enable() loops
over each ring and calls the legacy switch context or the execlist init context
code as appropriate.
This patch moves the ring looping out of that function in to the top level
caller i915_gem_init_hw(). This means the a single pass can be made over all
rings doing the PPGTT, L3 remap and context initialisation of each ring
altogether.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The i915_gem_init_hw() function calls a bunch of smaller initialisation
functions. Multiple of which have generic sections and per ring sections. This
means multiple passes are done over the rings. Each pass writes data to the ring
which floats around in that ring's OLR until some random point in the future
when an add_request() is done by some random other piece of code.
This patch breaks i915_ppgtt_init_hw() in two with the per ring initialisation
now being done in i915_ppgtt_init_ring(). The ring looping is now done at the
top level in i915_gem_init_hw().
v2: Fix dumb loop variable re-use.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com> (v1)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Added explicit request creation and submission to the GPU idle code path.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In order to explcitly track all GPU work (and completely remove the outstanding
lazy request), it is necessary to add extra i915_add_request() calls to various
places. Some of these do not need the implicit cache flush done as part of the
standard batch buffer submission process.
This patch adds a flag to _add_request() to specify whether the flush is
required or not.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The plan is to pass requests around as the basic submission tracking structure
rather than rings and contexts. This patch updates the
execbuffer_move_to_active() code path.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The plan is to pass requests around as the basic submission tracking structure
rather than rings and contexts. This patch updates the move_to_gpu() code paths.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated a couple of trace points to use the now cached request pointer rather
than extracting it from the ring.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Rather than just having a local request variable in the execbuff code, the
request pointer is now stored in the execbuff params structure. Also added
explicit cleanup of the request (plus wiping the OLR to match) in the error
case. This means that the execbuff code is no longer dependent upon the OLR
keeping track of the request so as to not leak it when things do go wrong. Note
that in the success case, the i915_add_request() at the end of the submission
function will tidy up the request and clear the OLR.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The alloc_request() function does not actually return the newly allocated
request. Instead, it must be pulled from ring->outstanding_lazy_request. This
patch fixes this so that code can create a request and start using it knowing
exactly which request it actually owns.
v2: Updated for new i915_gem_request_alloc() scheme.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Shrunk the parameter list of i915_gem_execbuffer_retire_commands() to a single
structure as everything it requires is available in the execbuff_params object.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The do_execbuf() function takes quite a few parameters. The actual set of
parameters is going to change with the conversion to passing requests around.
Further, it is due to grow massively with the arrival of the GPU scheduler.
This patch simplifies the prototype by passing a parameter structure instead.
Changing the parameter set in the future is then simply a matter of
adding/removing items to the structure.
Note that the structure does not contain absolutely everything that is passed
in. This is because the intention is to use this structure more extensively
later in this patch series and more especially in the GPU scheduler that is
coming soon. The latter requires hanging on to the structure as the final
hardware submission can be delayed until long after the execbuf IOCTL has
returned to user land. Thus it is unsafe to put anything in the structure that
is local to the IOCTL call itself - such as the 'args' parameter. All entries
must be copies of data or pointers to structures that are reference counted in
some way and guaranteed to exist for the duration of the batch buffer's life.
v2: Rebased to newer tree and updated for changes to the command parser.
Specifically, a code shuffle has required saving the batch start address in the
params structure.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In execlist mode, the context object pointer is written in to the request
structure (and reference counted) at the point of request creation. In legacy
mode, this only happens inside i915_add_request().
This patch updates the legacy code path to match the execlist version. This
allows all the intermediate code between request creation and request submission
to get at the context object given only a request structure. Thus negating the
need to pass context pointers here, there and everywhere.
v2: Moved the context reference so it does not need to be undone if the
get_seqno() fails.
v3: Fixed execlist mode always hitting a warning about invalid last_contexts
(which don't exist in execlist mode).
v4: Updated for new i915_gem_request_alloc() scheme.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Start of explicit request management in the execbuffer code path. This patch
adds a call to allocate a request structure before all the actual hardware work
is done. Thus guaranteeing that all that work is tagged by a known request. At
present, nothing further is done with the request, the rest comes later in the
series.
The only noticable change is that failure to get a request (e.g. due to lack of
memory) will be caught earlier in the sequence. It now occurs right at the start
before any un-undoable work has been done.
v2: Simplified the error handling path.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The i915_add_request() function is called to keep track of work that has been
written to the ring buffer. It adds epilogue commands to track progress (seqno
updates and such), moves the request structure onto the right list and other
such house keeping tasks. However, the work itself has already been written to
the ring and will get executed whether or not the add request call succeeds. So
no matter what goes wrong, there isn't a whole lot of point in failing the call.
At the moment, this is fine(ish). If the add request does bail early on and not
do the housekeeping, the request will still float around in the
ring->outstanding_lazy_request field and be picked up next time. It means
multiple pieces of work will be tagged as the same request and driver can't
actually wait for the first piece of work until something else has been
submitted. But it all sort of hangs together.
This patch series is all about removing the OLR and guaranteeing that each piece
of work gets its own personal request. That means that there is no more
'hoovering up of forgotten requests'. If the request does not get tracked then
it will be leaked. Thus the add request call _must_ not fail. The previous patch
should have already ensured that it _will_ not fail by removing the potential
for running out of ring space. This patch enforces the rule by actually removing
the early exit paths and the return code.
Note that if something does manage to fail and the epilogue commands don't get
written to the ring, the driver will still hang together. The request will be
added to the tracking lists. And as in the old case, any subsequent work will
generate a new seqno which will suffice for marking the old one as complete.
v2: Improved WARNings (Tomas Elf review request).
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It is a bad idea for i915_add_request() to fail. The work will already have been
send to the ring and will be processed, but there will not be any tracking or
management of that work.
The only way the add request call can fail is if it can't write its epilogue
commands to the ring (cache flushing, seqno updates, interrupt signalling). The
reasons for that are mostly down to running out of ring buffer space and the
problems associated with trying to get some more. This patch prevents that
situation from happening in the first place.
When a request is created, it marks sufficient space as reserved for the
epilogue commands. Thus guaranteeing that by the time the epilogue is written,
there will be plenty of space for it. Note that a ring_begin() call is required
to actually reserve the space (and do any potential waiting). However, that is
not currently done at request creation time. This is because the ring_begin()
code can allocate a request. Hence calling begin() from the request allocation
code would lead to infinite recursion! Later patches in this series remove the
need for begin() to do the allocate. At that point, it becomes safe for the
allocate to call begin() and really reserve the space.
Until then, there is a potential for insufficient space to be available at the
point of calling i915_add_request(). However, that would only be in the case
where the request was created and immediately submitted without ever calling
ring_begin() and adding any work to that request. Which should never happen. And
even if it does, and if that request happens to fall down the tiny window of
opportunity for failing due to being out of ring space then does it really
matter because the request wasn't doing anything in the first place?
v2: Updated the 'reserved space too small' warning to include the offending
sizes. Added a 'cancel' operation to clean up when a request is abandoned. Added
re-initialisation of tracking state after a buffer wrap to keep the sanity
checks accurate.
v3: Incremented the reserved size to accommodate Ironlake (after finally
managing to run on an ILK system). Also fixed missing wrap code in LRC mode.
v4: Added extra comment and removed duplicate WARN (feedback from Tomas).
For: VIZ-5115
CC: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Backmerge drm-next because the conflict between Ander's atomic fixes
for 4.2 and Maartens future work are getting to unwielding to handle.
Conflicts:
drivers/gpu/drm/i915/intel_display.c
drivers/gpu/drm/i915/intel_ringbuffer.h
Just always take ours, same as git merge -X ours, but done by hand
because I didn't trust git: It's confusing that it doesn't show any
conflicts in the merge diff at all.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
In Indirect context w/a batch buffer,
+WaFlushCoherentL3CacheLinesAtContextSwitch:bdw
v2: Add LRI commands to set/reset bit that invalidates coherent lines,
update WA to include programming restrictions and exclude CHV as
it is not required (Ville)
v3: Avoid unnecessary read when it can be done by reading register once (Chris).
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Rafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In Indirect and Per context w/a batch buffer,
+WaDisableCtxRestoreArbitration
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Rafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Some of the WA applied using WA batch buffers perform writes to scratch page.
In the current flow WA are initialized before scratch obj is allocated.
This patch reorders intel_init_pipe_control() to have a valid scratch obj
before we initialize WA.
v2: Check for valid scratch page before initializing WA as some of them
perform writes to it.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Some of the WA are to be applied during context save but before restore and
some at the end of context save/restore but before executing the instructions
in the ring, WA batch buffers are created for this purpose and these WA cannot
be applied using normal means. Each context has two registers to load the
offsets of these batch buffers. If they are non-zero, HW understands that it
need to execute these batches.
v1: In this version two separate ring_buffer objects were used to load WA
instructions for indirect and per context batch buffers and they were part
of every context.
v2: Chris suggested to include additional page in context and use it to load
these WA instead of creating separate objects. This will simplify lot of things
as we need not explicity pin/unpin them. Thomas Daniel further pointed that GuC
is planning to use a similar setup to share data between GuC and driver and
WA batch buffers can probably share that page. However after discussions with
Dave who is implementing GuC changes, he suggested to use an independent page
for the reasons - GuC area might grow and these WA are initialized only once and
are not changed afterwards so we can share them share across all contexts.
The page is updated with WA during render ring init. This has an advantage of
not adding more special cases to default_context.
We don't know upfront the number of WA we will applying using these batch buffers.
For this reason the size was fixed earlier but it is not a good idea. To fix this,
the functions that load instructions are modified to report the no of commands
inserted and the size is now calculated after the batch is updated. A macro is
introduced to add commands to these batch buffers which also checks for overflow
and returns error.
We have a full page dedicated for these WA so that should be sufficient for
good number of WA, anything more means we have major issues.
The list for Gen8 is small, same for Gen9 also, maybe few more gets added
going forward but not close to filling entire page. Chris suggested a two-pass
approach but we agreed to go with single page setup as it is a one-off routine
and simpler code wins.
One additional option is offset field which is helpful if we would like to
have multiple batches at different offsets within the page and select them
based on some criteria. This is not a requirement at this point but could
help in future (Dave).
Chris provided some helpful macros and suggestions which further simplified
the code, they will also help in reducing code duplication when WA for
other Gen are added. Add detailed comments explaining restrictions.
Use do {} while(0) for wa_ctx_emit() macro.
(Many thanks to Chris, Dave and Thomas for their reviews and inputs)
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Signed-off-by: Rafael Barbalho <rafael.barbalho@intel.com>
Signed-off-by: Arun Siluvery <arun.siluvery@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If the user disables the GPU reset using the i915.reset parameter and
one occurs, report that we failed to reset the GPU. If we return early,
as we currently do, then we leave all state intact (with a hung GPU)
and clients block forever waiting for their requests to complete.
Testcase: igt/gem_eio
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Mark i915.reset as an unsafe modoption, as discussed with
Chris.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The module pciid list got lost, but somehow most distros seem to
force-load drm drivers early and no one noticed for a while.
Bug introduced in
commit fd930478fb
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri Jun 19 20:27:27 2015 +0100
drm/i915: Remove KMS Kconfig option
Reported-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Damien Lespiau <damien.lespiau@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
fix warning introduced in last -fixes
* tag 'drm-intel-next-fixes-2015-06-22' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Silence compiler warning
Since we only support modesetting by default (disabling modesetting on
the command line prevents i915.ko from loading), having a parameter to
disable modesstting by default is superfluous, i.e. saying
CONFIG_DRM_I915_KMS=n is equivalent to CONFIG_DRM_I915=n.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Veter <daniel.vetter@ffwll.ch>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
On older gen, pre-Ironlake, parts there is no hardwired pin to report
the presence of an LVDS panel. Instead, we have to rely on the VBT to
declare whether the machine has a panel or not. Though notoriously
unreliable, so far we have erred on the side of false-positives and have
required a list of machines which end up falsely reporting a panel as
present. However, we now have reports of false-negatives, machines with
an LVDS that are being ignored due to the VBT not declaring the panel.
This patch ignores the VBT setting if the BIOS has already enabled the
LVDS panel (and on Ironlake+ we also have the hardware presence pin).
It fixes the Samsung NP680Z5E-X01FR in the bug report, but is likely to
result in more false-positives, and since we rely on the BIOS to enable
the panel, there are likely different circumstances where the BIOS will
not enable that panel (and so we may see the same machine with and
without a panel all on the whim of the BIOS).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90979
Reported-and-tested-by: lysxia@gmail.com
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Internal requirement for the alignment is that it must be a
power-of-two, so enforce rejection at the user interface to execbuffer
(which allows the caller to specify a stricter-than-expected alignment
criterion).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch doesn't have any functional change, but organize fruntbuffer
invalidate and busy by removing unecesarry signature argument for ring.
It was unsed on mark_fb_busy and only used on fb_obj_invalidate for the
same ORIGIN_CS usage. So let's clean it a bit
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The same dpll p2 divider selection is repeated three times in the
gen2-4 .find_dpll() functions. Factor it out.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We have enough generic hotplug functions sprinkled all over i915_irq.c
to warrant moving them to a file of their own. This should further
underline the distinction between generic code in the new file and
platform specific hotplug and irq code that remains in i915_irq.c.
Add new intel_hpd_init_work to keep work functions static, and rename
get_port_from_pin to intel_hpd_pin_to_port while increasing its
visibility, but keep everything else the same.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We'll have three functions:
intel_hpd_irq_storm_detect for detecting irq storms,
intel_hpd_irq_storm_disable for disabling hotplugs after detected storms,
intel_hpd_irq_storm_reenable_work for re-enabling hotplug.
No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Continue abstracting hotplug storm related functions to clarify the
code. This time, abstract hotplug irq storm related hotplug
disabling. While at it, clean up the loop iterating over connectors for
readability.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The hotplug work function has two loops iterating over connectors, the
first for handling hotplug disabling due to irq storms and the second
for actually handling the hotplug events. Move the debug printing into
the second one, so we can abstract the storm handling better. This may
change the output ordering slightly when there are multiple simultaneous
hotplug events.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The skylake scalers depend on the cdclk freq, but that frequency can
change during a modeset. So when a modeset happens calculate the new
cdclk in the atomic state. With the transitional helpers gone the
cached value can be used in the scaler, and committed after all
crtc's are disabled.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90874
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
All transitional plane helpers are gone, party!
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
By making color key atomic there are no more transitional helpers.
The plane check function will reject the color key when a scaler is
active.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
No need to repeatedly call update_watermarks, or update_fbc.
Down to a single call to update_watermarks in .crtc_enable
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that all planes are added during a modeset we can use the
calculated changes before disabling a plane, and then either commit
or force disable a plane before disabling the crtc.
The code is shared with atomic_begin/flush, except watermark updating
and vblank evasion are not used.
This is needed for proper atomic suspend/resume support.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90868
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Read out the initial state, and add a quirk to force add all planes
to crtc_state->plane_mask during initial commit. This will disable
all planes during the initial modeset.
The initial plane quirk is temporary, and will go away when hardware
readout is fully atomic, and the watermark updates in intel_sprite.c
are removed.
Changes since v1:
- Unset state->visible on !primary planes.
- Do not rely on the plane->crtc pointer in intel_atomic_plane,
instead assume planes are invisible until modeset.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
All the checks in intel_modeset_checks are only useful when a modeset
occurs, because there is nothing to update otherwise.
Same for power/cdclk changes, if there is no modeset they are noops.
Unfortunately intel_modeset_pipe_config still gets called without
modeset, because atomic hw readout isn't done yet.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
To allow them to be used in intel_set_mode.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is probably intended to be be done during vblank evasion.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The idea was good, but planes can have a fb even though
they're disabled. This makes the force argument useless
and always true, because only the commit function updates
state.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
By passing crtc_state to the check_plane functions a lot of duplicated
code can be removed. There are still some transitional helper calls,
they will be removed later.
Changes since v1:
- Revert state->visible changes.
- Use plane->state->crtc instead of plane->crtc.
- Use drm_atomic_get_existing_crtc_state.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
No point in hiding behind big ifs. This will be true most of the time.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This makes it easier to verify that no changes are done when
calling this from crtc instead.
Changes since v1:
- Make intel_wm_need_update static and always check it.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
commit 2c310b9d2859863826c3688c88218d607d5dd19a
Author: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Date: Mon May 18 12:28:52 2015 +0200
drm/i915: Split skl_update_scaler, v4.
It's easier to read separate functions for crtc and plane scaler state.
Changes since v1:
- Update documentation.
Changes since v2:
- Get rid of parameters to skl_update_scaler only used for traces.
This avoids needing to document the other parameters.
Changes since v3:
- Rename scaler_idx to scaler_user.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It saves another loop over all crtc's in the state, and computing
clock is more of a per crtc thing.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The scaler setup may add planes, but since they're unchanged we only
have to wait for primary flips. Also set planes_changed to indicate
at least 1 plane is modified.
Changes since v1:
- Instead of removing planes, do minimal validation needed.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Move the check for encoder cloning here.
Changes since v1:
- Remove was/is crtc_disabled. (mattrope)
- Rename function to intel_crtc_atomic_check. (mattrope)
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Get rid of a whole lot of ternary operators and assign the index
in scaler_id, instead of the id. They're the same thing.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Grabbing crtc state from atomic state is a lot more involved,
and make sure connectors are added before calling this function.
Move check_digital_port_conflicts to intel_modeset_checks,
it's only useful to check it on a modeset.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Tested-by(IVB): Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Silence the following -Wmaybe-uninitialized warnings and make the code
more clear.
drivers/gpu/drm/i915/intel_display.c: In function ‘__intel_set_mode’:
drivers/gpu/drm/i915/intel_display.c:11844:14: warning: ‘crtc_state’ may be used uninitialized in this function [-Wmaybe-uninitialized]
return state->mode_changed || state->active_changed;
^
drivers/gpu/drm/i915/intel_display.c:11854:25: note: ‘crtc_state’ was declared here
struct drm_crtc_state *crtc_state;
^
drivers/gpu/drm/i915/intel_display.c:11868:6: warning: ‘crtc’ may be used uninitialized in this function [-Wmaybe-uninitialized]
if (crtc != intel_encoder->base.crtc)
^
drivers/gpu/drm/i915/intel_display.c:11853:19: note: ‘crtc’ was declared here
struct drm_crtc *crtc;
Reported-by: Chris Wilson <chris@chris-wilson.co.uk>
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
i915 fixes for stuff in next
* tag 'drm-intel-next-fixes-2015-06-18' of git://anongit.freedesktop.org/drm-intel:
drm/i915: Don't set enabled value of all CRTCs when restoring the mode
drm/i915: Don't update staged config during force restore modesets
drm/i915: Don't check modeset state in the hw state force restore path
drm/i915: Add SCRATCH1 and ROW_CHICKEN3 to the register whitelist.
drm/i915: Extend the parser to check register writes against a mask/value pair.
drm/i915: Fix command parser to validate multiple register access with the same command.
drm/i915: Don't skip request retirement if the active list is empty
In order for gen8+ hardware to guarantee that no context switch
takes place during engine reset and that current context is properly
saved, the driver needs to notify and query hw before commencing
with reset.
There are gpu hangs where the engine gets so stuck that it never will
report to be ready for reset. We could proceed with reset anyway, but
with some hangs with skl, the forced gpu reset will result in a system
hang. By inspecting the unreadiness for reset seems to correlate with
the probable system hang.
We will only proceed with reset if all engines report that they
are ready for reset. If root cause for system hang is found and
can be worked around with another means, we can reconsider if
we can reinstate full reset for unreadiness case.
v2: -EIO, Recovery, gen8 (Chris, Tomas, Daniel)
v3: updated commit msg
v4: timeout_ms, simpler error path (Chris)
References: https://bugs.freedesktop.org/show_bug.cgi?id=89959
References: https://bugs.freedesktop.org/show_bug.cgi?id=90854
Testcase: igt/gem_concurrent_blit/prw-blt-overwrite-source-read-rcs-forked
Testcase: igt/gem_concurrent_blit/gtt-blt-overwrite-source-read-rcs-forked
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Tomas Elf <tomas.elf@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Changes for BXT - added a IS_BROXTON check to use the macro related to PPS
registers for BXT.
BXT does not have PP_DIV register. Making changes to handle this.
Second set of PPS registers have been defined but will be used when VBT
provides a selection between the 2 sets of registers.
v2:
[Jani] Added 2nd set of PPS registers and the macro
Jani's review comments
- remove reference in i915_suspend.c
- Use BXT PP macro
Squashing all PPS related patches into one.
v3: Jani's review comments addressed
- Use pp_ctl instead of pp
- ironlake_get_pp_control() is not required for BXT
- correct the use of && in the print statement
- drop the shift in the print statement
v4: Jani's comments
- modify ironlake_get_pp_control() - dont set unlock key for bxt
v5: Sonika's comments addressed
- check alignment
- move pp_ctrl_reg write (after ironlake_get_pp_control())
to !IS_BROXTON case.
- check before subtracting 1 for t11_t12
Signed-off-by: Vandana Kannan <vandana.kannan@intel.com>
Signed-off-by: A.Sunil Kamath <sunil.kamath@intel.com>
Reviewed-by: Sonika Jindal <sonika.jindal@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We need to call drm_atomic_set_mode_for_crtc() rather than copying the
mode in manually. As of commit
commit 99cf4a29fa
Author: Daniel Stone <daniels@collabora.com>
Date: Mon May 25 19:11:51 2015 +0100
drm/atomic: Add current-mode blob to CRTC state
the helper now also takes care of setting up the mode property blob for
us; if we don't use the helper and never setup the mode blob, this will
also trigger a failure in drm_atomic_crtc_check() when we have the
DRIVER_ATOMIC flag set (i.e., when using the nuclear pageflip support
via i915.nuclear_pageflip kernel command line parameter).
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The code in intel_crtc_restore_mode() sets the enabled value of all the
CRTCs when restoring the mode after a suspend/resume cycle. When more
than one CRTC is enabled, that causes drm_atomic_helper_check_modeset()
to fail if there is more than one pipe enabled, since all but one CRTC
has valid connector data. Instead, set only the enabled value for the
CRTC passed as an argument.
v2: Don't leak atomic state. (Matt)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90468
References: https://bugs.freedesktop.org/show_bug.cgi?id=90396
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
The force restore path relies on the staged config to preserve the
configuration used before a suspend/resume cycle. The update done to it
in intel_modeset_fixup_state() would cause that information to be lost
after the first modeset, making it impossible to restore the modes for
pipes B and C.
References: https://bugs.freedesktop.org/show_bug.cgi?id=90468
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Since the force restore logic will restore the CRTCs state one at a
time, it is possible that the state will be inconsistent until the whole
operation finishes. A call to intel_modeset_check_state() is done once
it's over, so don't check the state multiple times in between. This
regression was introduced in:
commit 7f27126ea3
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date: Wed Nov 5 14:26:06 2014 -0800
drm/i915: factor out compute_config from __intel_set_mode v3
v2: Rename check parameter to force_restore. (Matt)
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=94431
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Leftover from the big purge
commit a561165493
Author: John Harrison <John.C.Harrison@Intel.com>
Date: Thu Mar 5 14:03:03 2015 +0000
drm/i915: Remove ironlake rc6 support
Cc: John Harrison <John.C.Harrison@Intel.com>
Cc: Rob Clark <robdclark@gmail.com>
Reported-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Limit CHV maximum cdclk to 320MHz.
v2: Rebase to the latest
v3: Clean up of if-else tree
Signed-off-by: Mika Kahola <mika.kahola@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This commit is just to make the intentions explicit: on HSW+ these
bits are MBZ, but since we only support plane A and the macro
evaluates to zero when plane A is the parameter, we're not fixing any
bug.
v2:
- Remove useless extra blank like (Chris).
- Init dpfc_ctl in another place (Chris).
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This commit has two main advantages: simplify intel_fbc_update()
and deduplicate the strings.
v2:
- Rebase due to changes on P1.
- set_no_fbc_reason() can now return void (Chris).
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Because we're currently using FBC_UNSUPPORTED_MODE for two different
cases.
This commit will also allow us to write the next one without hiding
information from the user.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We already had a few bugs in the past where FBC was compressing
nothing when it was enabled, which makes the feature quite useless.
Add this information to debugfs so the test suites can check for
regressions in this piece of the code.
Our igt/tests/kms_frontbuffer_tracking already has support for this
message.
v2: - Remove pointless VLV check (Ville).
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The docs don't support the 64k linear scanout alignment we impose
on gen2/3. And it really makes no sense since we have no DSPSURF
register, so the only thing that the hardware will see is the linear
offset which will be just pixel aligned anyway.
There is one case where 64k comes into the picture, and that's FBC.
The start of the line length buffer corresponds to a 64k aligned
address of the uncompressed framebuffer. So if the uncompressed fb is
not 64k aligned, the first actually used entry in the line length
buffer will not be byte 0. There are 32 extra entries in the line
length buffer to account for this extra alignment so we shouldn't
have to worry about it when mapping the uncompressed fb to the GTT.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
VLV/CHV have problems with 4k aligned linear scanout buffers. The VLV
docs got updated at some point to say that we need to align them to
128k, just like we do on gen4.
So far I've seen the problem manifest when the stride is an odd multiple
of 512 bytes, and the surface address meets the following pattern
'(addr & 0xf000) == 0x1000' (also == 0x2000 is problematic on VLV). The
result is a starcase effect (so some pages get dropped maybe?), with a
few pages here and there clearly getting scannout out at the wrong position.
I've not actually been able to reproduce this problem on gen4, so it's
not clear of the issue is any way related to the 128k restrictions
supposedly inherited from gen4. But let's hope the 128k alignment is
sufficient to hide it all.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com>
Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Currently intel_gen4_compute_page_offset() simply picks the closest
page boundary below the linear offset. That however may not be suitably
aligned to satisfy any hardware specific restrictions. So let's make
sure the page boundary we choose is properly aligned.
Also to play it a bit safer lets split the remaining linear offset into
x and y values instead of just x. This should make no difference for
most platforms since we convert the x and y offsets back into a linear
offset before feeding them to the hardware. HSW+ are different however
and use x and y offsets even with linear buffers, so they might have
trouble if either the x or y get too big.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Corrected the documentation on the intel_edp_drrs_flush and
intel_edp_drrs_invalidate.
And accordingly edp_drrs_flush function is modified to restart the idleness
detection after upclocking.
v2: Update kerneldoc
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> (v1)
Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The i915 atomic conversion is a real beast and it's not getting easier
wrangling in a separate branch. I'm might be regretting this, but
right after vacation nothing can burst my little bubble here!
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
In igt, we want to test handling of GPU hangs, both for recovery
purposes and for reporting. However, we don't want to inject a genuine
GPU hang onto a machine that cannot recover and so be permenantly
wedged. Rather than embed heuristics into igt, have the kernel report
exactly when it expects the GPU reset to work.
This can also be usefully extended in future to indicate different
levels of fine-grained resets.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Tim Gore <tim.gore@intel.com>
Cc: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drivers/gpu/drm/i915/i915_debugfs.c: In function ‘i915_runtime_pm_status’:
drivers/gpu/drm/i915/i915_debugfs.c:2528:34: error: ‘struct dev_pm_info’ has no member named ‘usage_count’
atomic_read(&dev->dev->power.usage_count));
Regression from commit a6aaec8be2
Author: Damien Lespiau <damien.lespiau@intel.com>
Date: Thu Jun 4 18:23:58 2015 +0100
drm/i915: Add runtime PM's usage_count in i915_runtime_pm_status
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Damien Lespiau <damien.lespiau@intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Only bit 27 of SCRATCH1 and bit 6 of ROW_CHICKEN3 are allowed to be
set because of security-sensitive bits we don't want userspace to mess
with. On HSW hardware the whitelisted bits control whether atomic
read-modify-write operations are performed on L3 or on GTI, and when
set to L3 (which can be 10x-30x better performing than on GTI,
depending on the application) require great care to avoid a system
hang, so we currently program them to be handled on GTI by default.
Beignet can immediately start taking advantage of this change to
enable L3 atomics. Mesa should eventually switch to L3 atomics too,
but a number of non-trivial changes are still required so it will
continue using GTI atomics for now.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
In some cases it might be unnecessary or dangerous to give userspace
the right to write arbitrary values to some register, even though it
might be desirable to give it control of some of its bits. This patch
extends the register whitelist entries to contain a mask/value pair in
addition to the register offset. For registers with non-zero mask,
any LRM writes and LRI writes where the bits of the immediate given by
the mask don't match the specified value will be rejected.
This will be used in my next patch to grant userspace partial write
access to some sensitive registers.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Until now the software command checker assumed that commands could
read or write at most a single register per packet. This is not
necessarily the case, MI_LOAD_REGISTER_IMM expects a variable-length
list of offset/value pairs and writes them in sequence. The previous
code would only check whether the first entry was valid, effectively
allowing userspace to write unrestricted registers of the MMIO space
by sending a multi-register write with a legal first register, with
potential security implications on Gen6 and 7 hardware.
Fix it by extending the drm_i915_cmd_descriptor table to represent
multi-register access and making validate_cmd() iterate for all
register offsets present in the command packet.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
I noticed one of those and it turned out we have a few lingering around.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It's handy to have debug message for the "big" events and this one
qualifies IMHO. Also helpful to see what's happening while we're loading
the firwmare and how much time it takes.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In Linux, macros are usually well done and protect their arguments
properly, even avoiding multiple evaluations of the parameters. Extra ()
are really not needed.
Cc: Suketu Shah <suketu.j.shah@intel.com>
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Only bit 27 of SCRATCH1 and bit 6 of ROW_CHICKEN3 are allowed to be
set because of security-sensitive bits we don't want userspace to mess
with. On HSW hardware the whitelisted bits control whether atomic
read-modify-write operations are performed on L3 or on GTI, and when
set to L3 (which can be 10x-30x better performing than on GTI,
depending on the application) require great care to avoid a system
hang, so we currently program them to be handled on GTI by default.
Beignet can immediately start taking advantage of this change to
enable L3 atomics. Mesa should eventually switch to L3 atomics too,
but a number of non-trivial changes are still required so it will
continue using GTI atomics for now.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In some cases it might be unnecessary or dangerous to give userspace
the right to write arbitrary values to some register, even though it
might be desirable to give it control of some of its bits. This patch
extends the register whitelist entries to contain a mask/value pair in
addition to the register offset. For registers with non-zero mask,
any LRM writes and LRI writes where the bits of the immediate given by
the mask don't match the specified value will be rejected.
This will be used in my next patch to grant userspace partial write
access to some sensitive registers.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Until now the software command checker assumed that commands could
read or write at most a single register per packet. This is not
necessarily the case, MI_LOAD_REGISTER_IMM expects a variable-length
list of offset/value pairs and writes them in sequence. The previous
code would only check whether the first entry was valid, effectively
allowing userspace to write unrestricted registers of the MMIO space
by sending a multi-register write with a legal first register, with
potential security implications on Gen6 and 7 hardware.
Fix it by extending the drm_i915_cmd_descriptor table to represent
multi-register access and making validate_cmd() iterate for all
register offsets present in the command packet.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Switch from using 31 PFI credits to 63 PFI credits when cdclk>=czclk on
CHV. The spec lists both 31 and 63 as "suggested" values, but based on
feedback from hardware folks we should actually be using 63. Originally
I picked the 31 basically by flipping a coin.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Clint Taylor <Clinton.A.Taylor@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch fixes the resume from suspend-to-ram on the IBM X30
laptop. The problem is caused by the Bios missing to re-initialize
the iVCH registers, especially the PLL registers.
This patch records the iVCH registers during initialization, and
re-installs this register set when resuming.
Signed-off-by: Thomas Richter <thor@math.tu-berlin.de>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We are always allocating a single page. No need to be verbose so
remove the suffix.
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Free the scratch page if dma mapping fails.
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The divider value to convert from CZ clock rate to ms needs a +1
adjustment on VLV just like on CHV. This matches both the spec and
the accuracy test by pm_rc6_residency.
v2:
- simplify logic checking for the CHV 320MHz special case (Rodrigo)
Testcase: igt/pm_rc6_residency
Signed-off-by: Imre Deak <imre.deak@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=76877
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Currently we're forgetting to double the port clock when using double
clocked modes with 12bpc on HDMI. We're only accounting for the 1.5x
factor due to the 12bpc. So further double the 1.5x port clock when we
have a double clocked mode.
Unfortunately I don't have any displays that support both 12bpc and
double clocked modes, so I was unable to test this.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Account for the pixel multiplier when reading out the HDMI
mode dotclock. Makes the state checked happier on my ILK when using
double clocked modes.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Check that the DIP is enabled on the right port on IBX and VLV/CHV as
we're doing on g4x, and also check for all the infoframe enable bits on
all platforms.
Eventually we should track each infoframe type independently, and also
their contents. This is a small step in that direction as .infoframe_enabled()
return value could be easily turned into a bitmask.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Currently we just disable the GCP infoframe when turning off the port.
That means if the same transcoder is used on a DP port next, we might
end up pushing infoframes over DP, which isn't intended. Just disable
all the infoframes when turning off the port.
Also protect against two ports stomping on each other on g4x due to
the single video DIP instance. Now only the first port to enable
gets to send infoframes.
v2: Rebase
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Follow the procedure listed in Bspec to toggle the port enable bit off
and on when enabling HDMI with 12bpc and pixel repeat on IBX. The old
code didn't actually enable the port before "toggling" the bit back off,
so the whole workaround was essentially a nop.
Also take the opportunity to clarify the code by splitting the gmch
platforms to a separate (much more straightforward) function.
v2: Rebased due to crtc->config changes
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
IBX BSpec says we must specify 8bpc in TRANSCONF for both 8bpc
and 12bpc HDMI output. Do so.
v2: Pass intel_crtc to intel_pipe_has_type()
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When the video timings are suitably aligned so that all different
periods start at phase 0 (ie. none of the periods start mid-pixel)
we can inform the sink about this. Supposedly the sink can then
optimize certain things. Obviously this is only relevant when
outputting >8bpc data since otherwise there are no mid-pixel phases.
v2: Rebased due to crtc->config changes
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
GCP infoframes are required to inform the HDMI sink about the color
depth.
Send the GCP infoframe whenever the sink supports any deep color modes
since such sinks must anyway be capable of receiving them. For sinks
that don't support deep color let's skip the GCP in case it might
confuse the sink, although HDMI 1.4 spec does say all sinks must be
capable of reciving them. In theory we could skip the GCP infoframe
for deep color sinks in 8bpc mode as well since sinks must fall back to
8bpc whenever GCP isn't received for some time.
BSpec says we should disable GCP after disabling the port, so do that as
well.
v2: s/intel_set_gcp_infoframe/intel_hdmi_set_gcp_infoframe/
Rebased due to crtc->config changes
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
[danvet: Resolve conflict with lack of chv phy patches and fixup typo
Chandra spotted.]
Reviewed-by: Chandra Konduru <Chandra.konduru@intel.com>
Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Apparently we can have requests even if though the active list is empty,
so do the request retirement regardless of whether there's anything
on the active list.
The way it happened here is that during suspend intel_ring_idle()
notices the olr hanging around and then proceeds to get rid of it by
adding a request. However since there was nothing on the active lists
i915_gem_retire_requests() didn't clean those up, and so the idle work
never runs, and we leave the GPU "busy" during suspend resulting in a
WARN later.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
CPT/PPT require a specific procedure for enabling 12bpc HDMI. Implement
it, and to keep things neat pull the code into a function.
v2: Rebased due to crtc->config changes
s/HDMI_GC/HDMIUNIT_GC/ to match spec better
Factor out intel_enable_hdmi_audio()
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com>
Reviewed-By: Chandra Konduru <Chandra.konduru@intel.com>
Testecase: igt/kms_render/*
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This was introduced after converting hw readout to atomic,
so it should have been part of the revert too.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90929
Reported-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Tested-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Use a full atomic call instead. intel_crtc_page_flip will still
have to live until async updates are allowed.
This doesn't seem to be a regression from the convert to atomic,
part 3 patch. During GPU reset it fixes the following warning:
------------[ cut here ]------------
WARNING: CPU: 0 PID: 752 at drivers/gpu/drm/drm_crtc.c:5337 drm_mode_page_flip_ioctl+0x27b/0x360()
Modules linked in: i915
CPU: 0 PID: 752 Comm: Xorg Not tainted 4.1.0-rc7-patser+ #4090
Hardware name: NUC5i7RYB, BIOS RYBDWi35.86A.0246.2015.0309.1355 03/09/2015
ffffffff81c90866 ffff8800d87c3ca8 ffffffff817f7d87 0000000080000001
0000000000000000 ffff8800d87c3ce8 ffffffff81084955 ffff880000000000
ffff8800d87c3dc0 ffff8800d93d1208 0000000000000000 ffff8800b7d1f3e0
Call Trace:
[<ffffffff817f7d87>] dump_stack+0x4f/0x7b
[<ffffffff81084955>] warn_slowpath_common+0x85/0xc0
[<ffffffff81084a35>] warn_slowpath_null+0x15/0x20
[<ffffffff8146dffb>] drm_mode_page_flip_ioctl+0x27b/0x360
[<ffffffff8145ccb0>] drm_ioctl+0x1a0/0x6a0
[<ffffffff810b3b41>] ? get_parent_ip+0x11/0x50
[<ffffffff812e5540>] ? avc_has_perm+0x20/0x280
[<ffffffff810b3b41>] ? get_parent_ip+0x11/0x50
[<ffffffff811ea0f8>] do_vfs_ioctl+0x2f8/0x530
[<ffffffff811f6001>] ? expand_files+0x261/0x270
[<ffffffff812e7c16>] ? selinux_file_ioctl+0x56/0x100
[<ffffffff811ea3b1>] SyS_ioctl+0x81/0xa0
[<ffffffff81801b97>] system_call_fastpath+0x12/0x6f
---[ end trace 9ce834560085bd64 ]---
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This should fix fallout caused by making intel_crtc_control
and update_dpms atomic, which became a problem after reverting the
atomic hw readout patch.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90929
Reported-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This reverts commit 3bae26eb2991c00670df377cf6c3bc2b0577e82a.
Seems it introduces regressions for 3 different reasons, oh boy..
In bug #90868 as I can see the atomic state will be restored on
resume without the planes being set up properly. Because plane
setup here requires the atomic state, we'll have to settle
for committing atomic planes first.
In bug #90861 the failure appears to affect mostly DP devices,
and happens because reading out the atomic state prevents a modeset
on boot, which would require better hw state readout.
In bug #90874 it's shown that cdclk should be part of the atomic
state, so only performing a single modeset during resume excarbated
the issue.
It's better to fix those issues first, and then commit this patch,
so do that temporarily.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90868
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90861
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90874
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Ander Conselvan de Oliveira <conselvan2@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This reverts commit 490f400db5d886fc28566af69b02f6497f31be4b.
We're not ready yet to make it atomic, we calculate some state in
advance, but without atomic plane support atomic the hw readout will
fail.
It's required to revert this commit to revert the atomic hw
state readout patch.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90868
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90861
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Ander Conselvan de Oliveira <conselvan2@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
crtc->active will be gone eventually, and this check should be just as good.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This is a small behavioral change because it leaves DVO_2X_MODE
set between crtc_disable and crtc_enable. This is probably harmless
though and if not should be fixed by calculating 2x mode before
enable/disable pll.
This is needed because intel_crtc->active will be removed eventually.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This needs to be done last after all modesets have been calculated.
A modeset first disables all crtc's, so any crtc that undergoes a
modeset counts as inactive.
If no modeset's done, or > 1 crtc's stay w/a doesn't apply.
Apply workaround on the first crtc if 1 crtc stays active.
Apply workaround on the second crtc if no crtc was active.
Changes since v1:
- Use intel_crtc->atomic as a place to put hsw_workaround_pipe.
- Make sure quirk only applies to haswell.
- Use first loop to iterate over newly enabled crtc's only.
This increases readability.
Changes since v2:
- Move hsw_workaround_pipe back to crtc_state.
Changes since v3:
- Return errors from haswell_mode_set_planes_workaround.
Changes since v4:
- Clean up commit message.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
crtc->config is updated to always contain to the active crtc_state
and only differs from crtc_state during crtc_disable. It will
eventually be removed, so start with some low hanging fruit.
For crtc->active the situation is the same; it will be removed
eventually. Instead use crtc->state->active.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
intel_crtc->config will be removed eventually, so use crtc->hwmode.
drm_atomic_helper_update_legacy_modeset_state updates hwmode,
but crtc->active will eventually be gone too. Set dotclock to zero
to indicate the crtc is inactive.
Changes since v1:
- With the hwmode update in drm*update_legacy_modeset_state removed,
intel_modeset_update_state has to assign it instead.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This is a preparation for passing crtc state to the helpers.
When converting all users of crtc->config to use the old or
new state it's easier to find regressions when swap_state is
done first.
If crtc->config is swapped at the same place as swap_state
bugs will never be found.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Calculate all state using a normal transition, but afterwards fudge
crtc->state->active back to its old value. This should still allow
state restore in setup_hw_state to work properly.
Calling intel_set_mode will cause intel_display_set_init_power to be
called, make sure init_power gets set again afterwards.
Changes since v1:
- Fix to compile with v2 of the patch that adds intel_display_suspend.
- Add intel_display_set_init_power.
- Set return value to int to allow error checking.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Assume the callers lock everything with drm_modeset_lock_all.
This change had to be done after converting suspend/resume to
use atomic_state so the atomic state is preserved, otherwise
all transitional state is erased.
Now all callers of .crtc_enable and .crtc_disable go through
atomic modeset! :-D
Changes since v1:
- Only check for crtc_state->active in valleyview_modeset_global_pipes.
- Only check for crtc_state->active in modeset_update_crtc_power_domains.
Changes since v2:
- Rework on top of the changed patch order.
Changes since v3:
- Rename intel_crtc_toggle in description to *_control
- Change return value to int.
- Do not add plane state, should be done implicitly already.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
To make this work we load the new hardware state into the
atomic_state, then swap it with the sw state.
This lets us change the force restore path in setup_hw_state()
to use a single call to intel_mode_set() to restore all the
previous state.
As a nice bonus this kills off encoder->new_encoder,
connector->new_enabled and crtc->new_enabled. They were used only
to restore the state after a modeset.
Changes since v1:
- Make sure all possible planes are added with their crtc set,
so they will be turned off on first modeset.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
It makes more sense there, since these are computation steps that can
fail.
Changes since v1:
- Rename __intel_set_mode_checks to intel_modeset_checks (Matt Roper)
- Move intel_modeset_checks to before check_planes, so it won't
have to be moved later.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Repeated calls to begin_crtc_commit can cause warnings like this:
[ 169.127746] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:616
[ 169.127835] in_atomic(): 0, irqs_disabled(): 1, pid: 1947, name: kms_flip
[ 169.127840] 3 locks held by kms_flip/1947:
[ 169.127843] #0: (&dev->mode_config.mutex){+.+.+.}, at: [<ffffffff814774bc>] __drm_modeset_lock_all+0x9c/0x130
[ 169.127860] #1: (crtc_ww_class_acquire){+.+.+.}, at: [<ffffffff814774cd>] __drm_modeset_lock_all+0xad/0x130
[ 169.127870] #2: (crtc_ww_class_mutex){+.+.+.}, at: [<ffffffff81477178>] drm_modeset_lock+0x38/0x110
[ 169.127879] irq event stamp: 665690
[ 169.127882] hardirqs last enabled at (665689): [<ffffffff817ffdb5>] _raw_spin_unlock_irqrestore+0x55/0x70
[ 169.127889] hardirqs last disabled at (665690): [<ffffffffc0197a23>] intel_pipe_update_start+0x113/0x5c0 [i915]
[ 169.127936] softirqs last enabled at (665470): [<ffffffff8108a766>] __do_softirq+0x236/0x650
[ 169.127942] softirqs last disabled at (665465): [<ffffffff8108ae75>] irq_exit+0xc5/0xd0
[ 169.127951] CPU: 1 PID: 1947 Comm: kms_flip Not tainted 4.1.0-rc4-patser+ #4039
[ 169.127954] Hardware name: LENOVO 2349AV8/2349AV8, BIOS G1ETA5WW (2.65 ) 04/15/2014
[ 169.127957] ffff8800c49036f0 ffff8800cde5fa28 ffffffff817f6907 0000000080000001
[ 169.127964] 0000000000000000 ffff8800cde5fa58 ffffffff810aebed 0000000000000046
[ 169.127970] ffffffff81c5d518 0000000000000268 0000000000000000 ffff8800cde5fa88
[ 169.127981] Call Trace:
[ 169.127992] [<ffffffff817f6907>] dump_stack+0x4f/0x7b
[ 169.128001] [<ffffffff810aebed>] ___might_sleep+0x16d/0x270
[ 169.128008] [<ffffffff810aed38>] __might_sleep+0x48/0x90
[ 169.128017] [<ffffffff817fc359>] mutex_lock_nested+0x29/0x410
[ 169.128073] [<ffffffffc01635f0>] ? vgpu_write64+0x220/0x220 [i915]
[ 169.128138] [<ffffffffc017fddf>] ? ironlake_update_primary_plane+0x2ff/0x410 [i915]
[ 169.128198] [<ffffffffc0190e75>] intel_frontbuffer_flush+0x25/0x70 [i915]
[ 169.128253] [<ffffffffc01831ac>] intel_finish_crtc_commit+0x4c/0x180 [i915]
[ 169.128279] [<ffffffffc00784ac>] drm_atomic_helper_commit_planes+0x12c/0x240 [drm_kms_helper]
[ 169.128338] [<ffffffffc0184264>] __intel_set_mode+0x684/0x830 [i915]
[ 169.128378] [<ffffffffc018a84a>] intel_crtc_set_config+0x49a/0x620 [i915]
[ 169.128385] [<ffffffff817fdd39>] ? mutex_unlock+0x9/0x10
[ 169.128391] [<ffffffff81467b69>] drm_mode_set_config_internal+0x69/0x120
[ 169.128398] [<ffffffff8119b547>] ? might_fault+0x57/0xb0
[ 169.128403] [<ffffffff8146bf93>] drm_mode_setcrtc+0x253/0x620
[ 169.128409] [<ffffffff8145c600>] drm_ioctl+0x1a0/0x6a0
[ 169.128415] [<ffffffff810b3b41>] ? get_parent_ip+0x11/0x50
[ 169.128424] [<ffffffff811e9ab8>] do_vfs_ioctl+0x2f8/0x530
[ 169.128429] [<ffffffff810d0fcd>] ? trace_hardirqs_on+0xd/0x10
[ 169.128435] [<ffffffff812e7676>] ? selinux_file_ioctl+0x56/0x100
[ 169.128439] [<ffffffff811e9d71>] SyS_ioctl+0x81/0xa0
[ 169.128445] [<ffffffff81800697>] system_call_fastpath+0x12/0x6f
Solve it by using the newly introduced drm_atomic_helper_commit_planes_on_crtc.
The problem here was that the drm_atomic_helper_commit_planes() helper
we were using was basically designed to do
begin_crtc_commit(crtc #1)
begin_crtc_commit(crtc #2)
...
commit all planes
finish_crtc_commit(crtc #1)
finish_crtc_commit(crtc #2)
The problem here is that since our hardware relies on vblank evasion,
our CRTC 'begin' function waits until we're out of the danger zone in
which register writes might wind up straddling the vblank, then disables
interrupts; our 'finish' function re-enables interrupts after the
registers have been written. The expectation is that the operations between
'begin' and 'end' must be performed without sleeping (since interrupts
are disabled) and should happen as quickly as possible. By clumping all
of the 'begin' calls together, we introducing a couple problems:
* Subsequent 'begin' invocations might sleep (which is illegal)
* The first 'begin' ensured that we were far enough from the vblank that
we could write our registers safely and ensure they all fell within
the same frame. Adding extra delay waiting for subsequent CRTC's
wasn't accounted for and could put us back into the 'danger zone' for
CRTC #1.
This commit solves the problem by using a new helper that allows an
order of operations like:
for each crtc {
begin_crtc_commit(crtc) // sleep (maybe), then disable interrupts
commit planes for this specific CRTC
end_crtc_commit(crtc) // reenable interrupts
}
so that sleeps will only be performed while interrupts are enabled and
we can be sure that registers for a CRTC will be written immediately
once we know we're in the safe zone.
The crtc->config->base.crtc update may seem unrelated, but the helper
will use it to obtain the crtc for the state. Without the update it
will dereference NULL and crash.
Changes since v1:
- Use Matt Roper's commit message.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
And update crtc->config to point to the new state. There is no point
in swapping only part of the state when the rest of the state
should be untouched.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Now that we can subclass drm_atomic_state we can also use it to keep
track of all the pll settings. atomic_state is a better place to hold
all shared state than keeping pll->new_config everywhere.
Changes since v1:
- Assert connection_mutex is held.
Changes since v2:
- Fix swapped arguments to kzalloc for intel_atomic_state_alloc. (Jani Nikula)
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
The primary plane can still be configured when crtc is off,
furthermore this is also a noop now that affected planes are
added on modesets.
Changes since v1:
- Move commit so no frontbuffer_bits warnings are generated.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Compute new pipe_configs for all crtcs in the atomic state. The commit
part of the mode set (__intel_set_mode()) is already enabled to support
multiple pipes, the only thing missing was calculating a new pipe_config
for every crtc.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This should be much cleaner, with the same effects.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This can happen when turning off a sprite plane. Because the crtc state
is not yet always swapped correctly and transitional helpers are used
the crtc state cannot be relied on.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Atomic planes updates rely on having a accurate plane_mask.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Add missing calls to drm_atomic_add_affected_*. This is needed
to convert to atomic planes. When converting to atomic all planes
are needed on modeset. For good measure make sure all connectors
are added too.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
crtc_state->enable means a crtc is configured, but it may be turned
off for dpms. Until the commit "use intel_crtc_control everywhere"
crtc_state->active was not updated on crtc off, but now
crtc_state->active should be used for tracking whether a crtc is
scanning out or not.
A few commits from now dpms will be handled by calling
intel_set_mode with a different value for crtc_state->active,
which causes a crtc to turn on or off.
At this point crtc->active should mirror crtc_state->active,
so some paranoia from the crtc_disable functions can be removed.
intel_set_mode_setup_plls still checks for ->enable, because all
resources that are needed have to be calculated, else
dpms changes may not succeed.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
A follow up patch will make intel_modeset_compute_config() deal with
multiple crtcs, so move crtc specific stuff into the lower level crtc
specific function.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
With the use of drm_atomic_helper_update_legacy_modeset_state the
last user of modeset_crtc is removed from this function.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Now that the helper is exported there's no need to duplicate
this code any more.
Changes since v1:
- move intel_modeset_update_staged_output_state call to the right place.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Having a single path for everything makes it a lot easier to keep
crtc_state->active in sync with intel_crtc->active.
A crtc cannot be changed to active when not enabled, because it means
no mode is set and no connectors are connected.
This should also make intel_crtc->active match crtc_state->active.
Changes since v1:
- Reworded commit message, there's no intel_crtc_toggle.
Changes since v2:
- Change some callers of intel_crtc_control to intel_display_suspend.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This is a function used to disable all crtc's. This makes it clearer
to distinguish between when mode needs to be preserved and when
it can be trashed.
Changes since v1:
- Copy power changes from intel_crtc_control.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Now that the dpll updates are (mostly) atomic, the .off() code is a noop,
and intel_crtc_disable does mostly the same as intel_modeset_update_state.
Move all logic for connectors_active and setting dpms to that function.
Changes since v1:
- Move drm_atomic_helper_swap_state up.
Changes since v2:
- Split out intel_put_shared_dpll removal.
Changes since v3:
- Rebase on top of latest drm-intel.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Now that the pll updates are staged the put_shared_dpll function
consists only of checks that are done in check_shared_dpll_state
after a modeset too.
The changes to pll->config are overwritten by
intel_shared_dpll_commit, so this entire function is a noop.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
According to bspec the DDI PHY vswing scale value is "don't care" in
case the scale enable bit [27] is clear. But this doesn't seem to be
correct. The scale value seems to also matter if the scale mode bit
[26] is set. So both bit 26 and 27 depend on the value. Setting the
scale value to 0 while either bit is set results in a failed modeset on
HDMI (sink reports no signal).
After reset the scale value is 0x98, but according to the spec we have
to program it to 0x9a. So for consistency program it always to 0x9a
regardless of the scale enable bit.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Tested-by: Matt Roper <matthew.d.roper@intel.com>
Acked-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Ville's and Mika's cdclk series was in flight at the same time as the
SKL S3 patches so we were missing that update.
intel_update_max_cdclk() and intel_update_cdclk() had to be moved up a
bit to avoid forward declarations.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
intel_update_cdclk() will already display the boot CDCLK for DDI
platforms, no need to repeat there.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
We can operate with DPLL0 off with CDCLK backed by the 24Mhz reference
clock, and that's a supported configuration. Don't warn when notice
DPLL0 is off then.
We still have a separate warn at boot if cdclk is disabled (because we
don't currently try to handle the case (that shouldn't happen on SKL as
far as I know) where we boot with display not initialized.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Add support for changing cdclk frequency during runtime on BDW.
Also with IPS enabled the actual pixel rate mustn't exceed 95% of cdclk,
so take that into account when computing the max pixel rate.
v2: Grab rps.hw_lock around sandybridge_pcode_write()
v3: Rebase due to power well vs. .global_resources() reordering
v4: Rebased to the latest
v5: Rebased to the latest
v6: Patch order shuffle so that Broadwell CD clock change is
applied before the patch for Haswell CD clock change
v7: Fix for patch style problems
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Mika Kahola <mika.kahola@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
We need to tell BDW ULT and ULX apart.
v2: Rebased to the latest
v3: Rebased to the latest
v4: Fix for patch style problems
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Mika Kahola <mika.kahola@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Bspec says we shouldn't enable IPS on BDW when the pipe pixel rate
exceeds 95% of the core display clock. Apparently this can cause
underruns.
There's no similar restriction listed for HSW, so leave that one alone
for now.
v2: Add pipe_config_supports_ips() (Chris)
v3: Compare against the max cdclk insted of the current cdclk
v4: Rebased to the latest
v5: Rebased to the latest
v6: Fix for patch style problems
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83497
Tested-by: Timo Aaltonen <tjaalton@ubuntu.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Mika Kahola <mika.kahola@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Keep the cdclk maximum supported frequency around in dev_priv so that we
can verify certain things against it before actually changing the cdclk
frequency.
For now only VLV/CHV have support changing cdclk frequency, so other
plarforms get to assume cdclk is fixed.
v2: Rebased to the latest
v3: Rebased to the latest
v4: Fix for patch style problems
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Mika Kahola <mika.kahola@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Rather than reading out the current cdclk value use the cached value we
have tucked away in dev_priv.
v2: Rebased to the latest
v3: Rebased to the latest
v4: Fix for patch style problems
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Mika Kahola <mika.kahola@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
