During my tests, everything worked even if the wrong polarity was set.
Still, we should try to set the correct values.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Mask the value before changing it and also select DVI when needed.
DVI was working in cases where the BIOS was setting the correct value
because we were not masking the value before changing it.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Handy for lazy people like me, or when people forget to add the output
of lspci -nn.
v2: Chris Wilson noticed that we have this duplicated already in the
i915_capabilites debugfs file. But there \n as separator looks better,
which would be a bit verbose in dmesg. Abuse the preprocessor to
extract this all.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In commit
commit 20b46e59dd
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Thu Jul 26 11:16:14 2012 +0200
drm/i915: Only set the down rps limit when at the loweset frequency
The computation for the new desired frequency was extracted, but since
the desired frequency was passed-by value, the adjustments didn't
propgate back. Fix this.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Userspace tries to estimate the cost of ring switching based on whether
the GPU and GEM supports semaphores. (If we have multiple rings and no
semaphores, userspace assumes that the cost of switching rings between
batches is exorbitant and will endeavour to keep the next batch on the
active ring - as a coarse approximation to tracking both destination and
source surfaces.) Currently userspace has to guess whether semaphores
exist based on the chipset generation and the module parameter,
i915.semaphores. This is a crude and inaccurate guess as the defaults
internally depend upon other chipset features being enabled or disabled,
nor does it extend well into the future. By exporting a HAS_SEMAPHORES
parameter, we can easily query the driver and obtain an accurate answer.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The requirements for the sync flush to be emitted prior to the render
cache flush is only true for SandyBridge. On IvyBridge and friends we
can just emit the flushes with an inline CS stall.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We should not hit this under any sane conditions, but still, this does not
looks right.
CC: Chris Wilson <chris@chris-wilson.co.uk>
CC: Daniel Vetter <daniel.vetter@ffwll.ch>
CC: stable@vger.kernel.org
Reported-by: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com>
Reviewed-by: Chris Wlison <chris@chris-wilson.co.uk>
Signed-off-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We believe to have squashed all issues around the gen6+ rps interrupt
generation and why the gpu sometimes got stuck. With that cleared up,
there's no user left for the sanitize_pm infrastructure, so let's just
rip it out.
Note that 'intel_reg_write 0xa014 0x13070000' is the w/a if we find
ourselves stuck again.
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The power docs say that when the gt leaves rc6, it is in the lowest
frequency and only about 25 usec later will switch to the frequency
selected in GEN6_RPNSWREQ. If the downclock limit expires in that
window and the down limit is set to the lowest possible frequency, the
hw will not send the down interrupt. Which leads to a too high gpu
clock and wasted power.
Chris Wilson already worked on this with
commit 7b9e0ae6da
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Sat Apr 28 08:56:39 2012 +0100
drm/i915: Always update RPS interrupts thresholds along with
frequency
but got the logic inverted: The current code set the down limit as
long as we haven't reached it. Instead of only once with reached the
lowest frequency.
Note that we can't always set the downclock limit to 0, because
otherwise the hw will keep on bugging us with downclock request irqs
once the lowest level is reached.
For similar reasons also always set the upclock limit, otherwise the
hw might poke us again with interrupts.
v2: Chris Wilson noticed that the limit reg is also computed in
sanitize_pm. To avoid duplication, extract the code into a common
function.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
By selecting the cache level (essentially whether or not the CPU snoops
any updates to the bo, and on more recent machines whether it resides
inside the CPU's last-level-cache) a userspace driver is able to then
manage all of its memory within buffer objects, if it so desires. This
enables the userspace driver to accelerate uploads and more importantly
downloads from the GPU and to able to mix CPU and GPU rendering/activity
efficiently.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Added code comment about where we plan to stuff platform
specific cacheing control bits in the ioctl struct.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Several functions of the GPU have the restriction that differing memory
domains cannot be placed next to each other (as the GPU may prefetch
beyond the end of one domain and hang as it crosses into the other
domain). We use the facility of the drm_mm to mark ranges with a
particular color that corresponds to the cache attributes of those pages
in order to prevent allocating adjacent blocks of differing memory
types.
v2: Rebase ontop of drm_mm coloring v2.
v3: Fix rebinding existing gtt_space and add a verification routine.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Originally I had a macro specifically for DPF support, and Daniel, with
good reason asked me to change it to this. It's not the way I would have
gone (and indeed I didn't), but for now there is no distinction as all
platforms with L3 also have DPF.
Note: The good reasons are that dpf is a l3$ feature (at least on
currrent hw), hence I don't expect one to go without the other.
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: added note]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Basic context support on HSW is no different than previous generations.
The size of the context object changes, but that's about it.
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As suggested by Daniel, rip out the independent timers for device and
crtc busyness and integrate the manual powermanagement of the display
engine into the GEM core and its request tracking. The benefits are that
the code is a lot smaller, fewer moving parts and should fit more neatly
into the overall activity tracking of the driver.
v2: Complete overhaul and removal of the racy timers and workers.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
By moving the function to intel_ringbuffer and currying the appropriate
parameter, hopefully we make the callsites easier to read and
understand.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Otherwise once we use the buffer with a BLT command on gen2/3, we will
always regard future command submissions as continuing the fenced
access. However, now that we flush/invalidate between every batch we can
drop this pessimism.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that we unconditionally flush and invalidate between every batch
buffer, we no longer need the complex logic to decide which domains
require flushing. Remove it and rejoice.
v2 (danvet): Keep around the flip waiting logic. It's gross and
broken, I know, but we can't just kill that thing ... even if we just
keep it around as a reminder that things are broken.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Rely instead on the insertion of the implicit flush before the seqno
breadcrumb.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As the flush is either performed explictly immediately after the
execbuffer dispatch, or before the serialisation of last_fenced_seqno we
can forgo the explict i915_gem_flush_ring().
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is now handled by a global flag to ensure we emit a flush before
the next serialisation point (if we failed to queue one previously).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As we guarantee to emit a flush before emitting the breadcrumb or
the next batchbuffer, there is no further need for the flushing list.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As we always flush the GPU cache prior to emitting the breadcrumb, we no
longer have to worry about the deferred flush causing the
pending_gpu_write to be delayed. So we can instead utilize the known
last_write_seqno to hopefully minimise the wait times.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As we move to lazily clearing the GPU write domain only when the buffer
becomes inactive, this leaves a window of opportunity for
i915_gem_object_pin_to_display_plane() to detect a seemingly
inconsistent value. This function is special as it tries to pipeline the
operation to avoid the stall and so may not retires the buffer and we
may not get the opportunity to clear the write domain. However, we know
all is good, so drop the assertion.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Request preallocation was added to i915_add_request() in order to
support the overlay. However, not all users care and can quite happily
ignore the failure to allocate the request as they will simply repeat
the request in the future.
By pushing the allocation down into i915_add_request(), we can then
remove some rather ugly error handling in the callers.
v2: Nullify request->file_priv otherwise we chase a garbage pointer
when retiring requests.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With the base addresses shifting around, this is easier to handle.
Also move to the real reg offset on vlv.
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The intention is to help select which engine to use for copies with
interoperating clients - such as a GL client making a request to the X
server to perform a SwapBuffers, which may require copying from the
active GL back buffer to the X front buffer.
We choose to report a mask of the active rings to future proof the
interface against any changes which may allow for the object to reside
upon multiple rings.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: bikeshed away the write ring mask and add the explanation
Chris sent in a follow-up mail why we decided to use masks.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The interface's immediate purpose is to do synchronous timestamp queries
as required by GL_TIMESTAMP. The GPU has a register for reading the
timestamp but because that would normally require root access through
libpciaccess, the IOCTL can provide this service instead.
Currently the implementation whitelists only the render ring timestamp
register, because that is the only thing we need to expose at this time.
v2: make size implicit based on the register offset
Add a generation check
Reviewed-by: Eric Anholt <eric@anholt.net>
Cc: Jacek Lawrynowicz <jacek.lawrynowicz@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: fixup the ioctl numerb:]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
I'm planing to merge this next week for 3.7, but I'd like to avoid
stupid conflicts with the exsting userspace when merging the new
reg_read ioctl (which doesn't have userspace yet, but this caching
interface has).
Header extracted from Chris Wilson's patch, but fix up the copy&pasted
comment in the interface struct.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch adds support for the ns2501 DVO, found in some older Fujitsu/Siemens Labtops.
It is in the state of "works for me".
Includes now proper DPMS support. Includes switching between resolutions -
from 640x480 to 1024x768.
Currently assumes that the native display resolution is 1024x768.
The ns2501 seems to be rather critical - if the output PLL is not
running, the chip doesn't seem to be clocked and then doesn't react
on i2c messages. Thus, a quick'n-dirty trick ensures that the DVO
is active before submitting any i2c messages to it. This is
probably to be reviewed.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=17902
Signed-off-by: Thomas Richter <thor@math.tu-berlin.de>
[danvet: fixup whitespace fail.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This will be needed for Haswell, but already has its uses here.
This patch started as a small patch written patch by Shobhit Kumar,
but it has changed so much that none of its original lines remain.
Credits-to: Shobhit Kumar <shobhit.kumar@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We have some common code that we always run before calling
intel_dp_set_link_train. This common code sets the correct training
patterns to the DP variable. If we add more calls to
intel_dp_set_link_train, we'll also have to duplicate this common
code. So instead of repeating this code whenever we call
intel_dp_set_link_train, we move the code to inside the function: now
we check which training pattern we're going to set and then we set the
DP register according to it.
One of the side-effects of this change is that now we never forget to
mask the training pattern bits before changing them. It looks like
this was working before because we were first masking the bits, then
writing 00, 01 and then 11.
This patch also enables us to use the intel_dp_set_link_train function
when disabling link training: in this case we need to avoid writing
the DP_TRAINING_LANE*_SET AUX commands.
As a bonus, the big intel_dp_{start,complete}_link_train functions
will get smaller and a little bit easier to read.
Version 2 changes:
- Rewrite commit message.
- Also clear the training pattern bits before changing them.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Instead of having a giant if cascade to figure this out according to
the passed-in register. We could do quite a bit more cleaning up and
all by using the port at more places, but I think this should be part
of a bigger rework to introduce a struct intel_digital_port which
would keep track of all these things. I guess this will be part of
some haswell-DP-induced refactoring.
For now this rips out the big cascade, which is what annoyed me so
much.
v2: Add port variable name back for the func decl (I've tried to trick
myself below the 80 char limit).
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Intel hw only has one MUX for encoders, so outputs are either not
cloneable or all in the same group of cloneable outputs. This neatly
simplifies the code and allows us to ditch some ugly if cascades in
the dp and hdmi init code (well, we need these if cascades for other
stuff still, but that can be taken care of in follow-up patches).
Note that this changes two things:
- dvo can now be cloned with sdvo, but dvo is gen2 whereas sdvo is
gen3+, so no problem. Note that the old code had a bug and didn't
allow cloning crt with dvo (but only the other way round).
- sdvo-lvds can now be cloned with sdvo-non-tv. Spec says this won't
work, but the only reason I've found is that you can't use the
panel-fitter (used for lvds upscaling) with anything else. But we
don't use the panel fitter for sdvo-lvds. Imo this part of Bspec is
a) rather confusing b) mostly as a guideline to implementors (i.e.
explicitly stating what is already implicit from the spec, without
always going into the details of why). So I think we can ignore this
- worst case we'll get a bug report from a user with with sdvo-lvds
and sdvo-tmds and have to add that special case back in.
Because sdvo lvds is a bit special explain in comments why sdvo LVDS
outputs can be cloned, but native LVDS and eDP can't be cloned - we
use the panel fitter for the later, but not for sdvo.
Note that this also uncoditionally initializes the panel_vdd work used
by eDP. Trying to be clever doesn't buy us anything (but strange bugs)
and this way we can kill the is_edp check.
v2: Incorporate review from Paulo
- Add in a missing space.
- Pimp comment message to address his concerns.
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Splitting them up between pch and gmch variants just makes it harder
to find things. Especially since the hotplug bits are actually valid
on earlier chips, too.
v2: Fixed the comment as pointed out by Paulo Zanoni.
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When bug hunting, I found the interface to do_switch() overly
complicated and I believe festered the earlier bug. This aims to make
the code a little clearer.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Move the DP structure to shared location so that it can be used from
within the ddi module.
Changes from Paulo:
- Move less code to intel_drv.h
- Remove #include statement
- Replace a tab with a space in train_set
Signed-off-by: Shobhit Kumar <shobhit.kumar@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We now refuse to load on gen6+ if kms is not enabled:
commit 26394d9251
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Mon Mar 26 21:33:18 2012 +0200
drm/i915: refuse to load on gen6+ without kms
Which results in the drm core calling our lastclose function to clean
up the mess, but that one is neatly broken for such failure cases
since kms has been introduced in
commit 79e539453b
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date: Fri Nov 7 14:24:08 2008 -0800
DRM: i915: add mode setting support
Reported-and-tested-by: Paulo Zanoni <przanoni@gmail.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Fixes failures in transform feedback on gen7 because our SOL_RESET
flag was setting the transform feedback offsets in the old context
(occasionally happened to be ours) instead of the new context.
Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
LVDS is the first output where dpms on/off and prepare/commit don't
perfectly match. Now the idea behind this special case seems to be
that for simple resolution changes on the LVDS we don't need to stop
the pipe, because (at least on newer chips) we can adjust the panel
fitter on the fly.
There are a few problems with the current code though:
- We still stop and restart the pipe unconditionally, because the crtc
helper code isn't flexible enough.
- We show some ugly flickering, especially when changing crtcs (this
the crtc helper would actually take into account, but we don't
implement the encoder->get_crtc callback required to make this work
properly).
So it doesn't even work as advertised. I agree that it would be nice
to do resolution changes on LVDS (and also eDP) whithout blacking the
screen where the panel fitter allows to do that. But imo we should
implement this as a special case a few layers up in the mode set code,
akin to how we already detect simple framebuffer changes (and only
update the required registers with ->mode_set_base).
Until this is all in place, make our lives easier and just rip it out.
Also note that this seems to fix actual bugs with enabling the lvds
output, see:
http://lists.freedesktop.org/archives/intel-gfx/2012-July/018614.html
Cc: Takashi Iwai <tiwai@suse.de>
Cc: Giacomo Comes <comes@naic.edu>
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Takashi Iwai <tiwai@suse.de>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We need to check that "ctx" is a valid pointer before dereferencing it.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Otherwise we end up trying to unpin a freed object and BUG.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This prevents a WARN introduced with
commit de2b998552
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Wed Jul 4 22:52:50 2012 +0200
drm/i915: don't return a spurious -EIO from intel_ring_begin
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The issue is that we stale data in the CPU caches, when we come to
swap-out the object, the CPU may short-circuit the reads from those
cacheline and so corrupt the context object.
Secondary, leaving the context object as being marked in the CPU write
domain whilst on the GPU active list is a bad idea and will throw
warnings later.
Note: Thanks to calling set_to_gtt_domain with write = false and not
setting any gpu write domain when putting a context object onto the
active list (when we switch away from it) the set_to_gtt_domain call
won't block.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: Added a note to the commit message and a comment in the code
to explain the clever non-blocking trick.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As we take the struct_mutex lock to access the command-stream, there is
a possibility that we may need to wait for a GPU hang and so should make
the lock both interruptible and error-checking.
References: https://bugs.freedesktop.org/show_bug.cgi?id=50069
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When we enable/disable the CPU backlight registers we can't forget to
enable/disable the PCH backlight registers. Since we're using the CPU
registers we should also unset the override bit.
Fixes a regression on the following commit:
drm/i915: properly enable the blc controller on the right pipe
The commit just deleted the code that sets the PCH registers, so it
was relying on the values set by the BIOS. I told my BIOS to boot on
the DVI monitor instead of the LVDS panel, so I noticed the bug.
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If we drop the breadcrumb request after a batch due to a signal for
example we aim to fix it up at the next opportunity. In this case we
emit a second batchbuffer with no waits upon the first and so no
opportunity to insert the missing request, so we need to emit the
missing flush for coherency. (Note that that invalidating the render
cache is the same as flushing it, so there should have been no
observable corruption.)
Note that beside simply adding the missing flush, avoiding potential
render corruption, this will also fix at least parts of the problem
introduced by some funny interaction of these two commits:
commit de2b998552
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Wed Jul 4 22:52:50 2012 +0200
drm/i915: don't return a spurious -EIO from intel_ring_begin
which allowed intel_ring_begin to return -ERESTARTSYS and
commit cc889e0f6c
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Wed Jun 13 20:45:19 2012 +0200
drm/i915: disable flushing_list/gpu_write_list
which essentially disabled the flushing list.
The issue happens when we submit a batch & emit it, but get
interrupted (thanks to the first patch) while trying to emit the
flush. On the next batch we still assume that the full gpu domain
handling is in effect and hence compute the invalidate&flushing
domains. But thanks to the 2nd patch we totally ignore these and only
invalidate all gpu domains, presuming that any required flushes have
been issued already. Which is wrong and eventually results in us
updating the new write_domain values with the computed
pending_write_domain values, which leaves an object with write_domain
== 0 on the gpu_write_list.
As soon as we try to unbind that object, things blow up.
Fix this by emitting the missing flush according to the new
ring->gpu_caches_dirty flag.
Note that this does _not_ fix all the current cases where we end up
with an object on the flushing_list that can't be flushed.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=52040
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Add bug explanation to commit message.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
