This breaks under the vblank timestamp cleanup patch
by Daniel Vetter. Also it is pointless to return anything
but zero (or any other constant) if the function doesn't
actually query a hw vblank counter. The bogus return of
the current drm vblank counter via direct readout or via
drm_vblank_count() is found in many of the new kms drivers,
but it does exactly nothing different from returning any
arbitrary constant - it's a no operation.
Let's simply return 0 - Easy and fast.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Dave Airlie <airlied@redhat.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Since commit 844b03f277 we make
sure that after vblank irq off, we return the last valid
(vblank count, vblank timestamp) pair to clients, e.g., during
modesets, which is good.
An overlooked side effect of that commit for kms drivers without
support for precise vblank timestamping is that at vblank irq
enable, when we update the vblank counter from the hw counter, we
can't update the corresponding vblank timestamp, so now we have a
totally mismatched timestamp for the new count to confuse clients.
Restore old client visible behaviour from before Linux 3.18, but
zero out the timestamp at vblank counter update (instead of disable
as in original implementation) if we can't generate a meaningful
timestamp immediately for the new vblank counter. This will fix
this regression, so callers know they need to retry again later
if they need a valid timestamp, but at the same time preserves
the improvements made in the commit mentioned above.
v2: Rebased on top of Daniel Vetter's fixup and documentation
patch for timestamp updates. Drop request for stable kernel
backport as this would be more difficult, unless the original
patch would get applied to stable kernels.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Dave Airlie <airlied@redhat.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
For a kms driver to support immediate disable of vblank
irq's reliably without introducing off by one errors or
other mayhem for clients, it must not only support a
hardware vblank counter query, but also high precision
vblank timestamping, so vblank count and timestamp can be
instantaneously reinitialzed to valid values. Additionally
the exposed hardware counter must behave as if it is
incrementing at leading edge of vblank to avoid off by
one errors during reinitialization of the counter while
the display happens to be inside or close to vblank.
Check during drm_vblank_init that a driver which claims to
be capable of vblank_disable_immediate at least supports
high precision timestamping and prevent use of instant
disable if that isn't present as a minimum requirement.
v2: Changed from DRM_ERROR to DRM_INFO and made message
more clear, as suggested by Michel Dänzer.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: Dave Airlie <airlied@redhat.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This was a bit too much cargo-culted, so lets make it solid:
- vblank->count doesn't need to be an atomic, writes are always done
under the protection of dev->vblank_time_lock. Switch to an unsigned
long instead and update comments. Note that atomic_read is just a
normal read of a volatile variable, so no need to audit all the
read-side access specifically.
- The barriers for the vblank counter seqlock weren't complete: The
read-side was missing the first barrier between the counter read and
the timestamp read, it only had a barrier between the ts and the
counter read. We need both.
- Barriers weren't properly documented. Since barriers only work if
you have them on boths sides of the transaction it's prudent to
reference where the other side is. To avoid duplicating the
write-side comment 3 times extract a little store_vblank() helper.
In that helper also assert that we do indeed hold
dev->vblank_time_lock, since in some cases the lock is acquired a
few functions up in the callchain.
Spotted while reviewing a patch from Chris Wilson to add a fastpath to
the vblank_wait ioctl.
v2: Add comment to better explain how store_vblank works, suggested by
Chris.
v3: Peter noticed that as-is the 2nd smp_wmb is redundant with the
implicit barrier in the spin_unlock. But that can only be proven by
auditing all callers and my point in extracting this little helper was
to localize all the locking into just one place. Hence I think that
additional optimization is too risky.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Michel Dänzer <michel@daenzer.net>
Cc: Peter Hurley <peter@hurleysoftware.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-and-tested-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
DRM probe should not repoll a connector if it is already
connected and the DRM_CONNECTOR_POLL_DISCONNECT flag is not set.
Signed-off-by: Josef Holzmayr <holzmayr@rsi-elektrotechnik.de>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
For test 4.2.2.5 to pass per the Link CTS Core 1.2 rev1.1 spec, the source
device must attempt at least 7 times to read the EDID when it receives an
I2C defer. The normal DRM code makes only 7 retries, regardless of whether
or not the response is a native defer or an I2C defer. Test 4.2.2.5 fails
since there are native defers interspersed with the I2C defers which
results in less than 7 EDID read attempts.
The solution is to add the numer of defers to the retry counter when an I2C
DEFER is returned such that another read attempt will be made. This situation
should normally only occur in compliance testing, however, as a worse case
real-world scenario, it would result in 13 attempts ( 6 native defers, 7 I2C
defers) for a single transaction to complete. The net result is a slightly
slower response to an EDID read that shouldn't significantly impact overall
performance.
V2:
- Added a check on the number of I2C Defers to limit the number
of times that the retries variable will be decremented. This
is to address review feedback regarding possible infinite loops
from misbehaving sink devices.
V3:
- Fixed the limit value to 7 instead of 8 to get the correct retry
count.
- Combined the increment of the defer count into the if-statement
V4:
- Removed i915 tag from subject as the patch is not i915-specific
V5:
- Updated the for-loop to add the number of i2c defers to the retry
counter such that the correct number of retry attempts will be
made
Signed-off-by: Todd Previte <tprevite@gmail.com>
Cc: dri-devel@lists.freedesktop.org
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The debug message is missing a newline at the end and it makes the
logs hard to read when a device defers a lot. Simple 2-character fix
adds the newline at the end.
Signed-off-by: Todd Previte <tprevite@gmail.com>
Cc: dri-devel@lists.freedesktop.org
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It's a silly thing to do and surprises driver writers. Most likely
this did already blow up for exynos.
It's also a silly thing to change plane state when it's off, but fbdev
is silly (it does an unconditional modeset over all planes). And
userspace can be evil. So I think we need this.
With this check in the helpers we can remove the one in i915 code for
the same conditions (becuase ->crtc iff ->fb).
Cc: Gustavo Padovan <gustavo@padovan.org>
Cc: dri-devel@lists.freedesktop.org
Cc: Inki Dae <inki.dae@samsung.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Our legacy SetPlane updates perform integer overflow checking on a
plane's destination rectangle in drm_mode_setplane(), and atomic updates
handled as part of a drm_atomic_state transaction do the same checking
in drm_atomic_plane_check(). However legacy cursor updates that get
routed through universal plane interfaces may bypass this overflow
checking if the driver's .update_plane is serviced by the transitional
plane helpers rather than the full atomic plane helpers.
Move the check for destination rectangle integer overflow from the
drm_mode_setplane() to __setplane_internal() so that it also covers
cursor operations.
This fixes an issue first noticed with i915 commit:
commit ff42e093e9
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Mon Mar 2 16:35:20 2015 +0100
Revert "drm/i915: Switch planes from transitional helpers to full
atomic helpers"
The above revert switched us from full atomic helpers back to the
transitional helpers, and in doing so we lost the overflow checking here
for universal cursor updates. Even though such extreme cursor positions
are unlikely to actually happen in the wild, we still don't want there
to be a change of behavior when drivers switch from transitional helpers
to full helpers.
v2: Move check from setplane ioctl to setplane_internal rather than
adding an additional copy of the checks to the transitional plane
helpers. (Daniel)
Cc: Daniel Vetter <daniel@ffwll.ch>
Testcase: igt/kms_cursor_crc
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84269
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Some non-const pointers were added since the last constification, fix
them.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This saves some typing whenever a iteration over all the connector,
crtc or plane states in the atomic state is written, which happens
quite often.
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch implements the virtual GEM driver with PRIME sharing which
allows vgem to import a gem object from other drivers for the purpose
of mmap-ing them to userspace. The mmap is done using the mmap
operation exported by other drivers.
v2: remove platform_device and do not attach to dma bufs
v3: use drm helpers for get/put pages
v4: correct dumb create pitch
Reviewed-by: Rob Clark <robdclark@gmail.com> (v3)
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org> (v3)
Signed-off-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Zach Reizner <zachr@google.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Final drm-misc pull for 4.0, just various things all over, including a few
more important atomic fixes. btw I didn't pick up the vmwgfx patch from
Ville's series, but one patch has one hunk touching vmwgfx and
Thomas/Jakob didn't get around to ack it. I figured it's simple enough to
be ok though.
* tag 'topic/drm-misc-2015-03-31' of git://anongit.freedesktop.org/drm-intel:
drm: line wrap DRM_IOCTL_DEF* macros
drm/atomic: Don't try to free a NULL state
drm/atomic: Clear crtcs, connectors and planes when clearing state
drm: Rewrite drm_ioctl_flags() to resemble the new drm_ioctl() code
drm: Use max() to make the ioctl alloc size code cleaner
drm: Simplify core vs. drv ioctl handling
drm: Drop ioctl->cmd_drv
drm: Fix DRM_IOCTL_DEF_DRV()
drm/atomic-helpers: Properly avoid full modeset dance
drm: atomic: Allow setting CRTC active property
drm: atomic: Expose CRTC active property
drm: crtc_helper: Update hwmode before mode_set call
drm: mode: Allow NULL modes for equality check
drm: fb_helper: Simplify exit condition
drm: mode: Fix typo in kerneldoc
drm/dp: Print the number of bytes processed for aux nacks
This backmerges 4.0-rc6 due to the recent fixes in rc5/6
- DP link rate refactoring from Ville
- byt/bsw rps tuning from Chris
- kerneldoc for the shrinker code
- more dynamic ppgtt pte work (Michel, Ben, ...)
- vlv dpll code refactoring to prep fro bxt (Imre)
- refactoring the sprite colorkey code (Ville)
- rotated ggtt view support from Tvrtko
- roll out struct drm_atomic_state to prep for atomic update (Ander)
* tag 'drm-intel-next-2015-03-27-merge' of git://anongit.freedesktop.org/drm-intel: (473 commits)
Linux 4.0-rc6
arm64: juno: Fix misleading name of UART reference clock
drm/i915: Update DRIVER_DATE to 20150327
drm/i915: Skip allocating shadow batch for 0-length batches
drm/i915: Handle error to get connector state when staging config
drm/i915: Compare GGTT view structs instead of types
drm/i915: fix simple_return.cocci warnings
drm/i915: Add module param to test the load detect code
drm/i915: Remove usage of encoder->new_crtc from clock computations
drm/i915: Don't look at staged config crtc when changing DRRS state
drm/i915: Convert intel_pipe_will_have_type() to using atomic state
drm/i915: Pass an atomic state to modeset_global_resources() functions
drm/i915: Add dynamic page trace events
drm/i915: Finish gen6/7 dynamic page table allocation
drm/i915: Remove unnecessary gen6_ppgtt_unmap_pages
drm/i915: Fix i915_dma_map_single positive error code
drm/i915: Prevent out of range pt in gen6_for_each_pde
drm/i915: fix definition of the DRM_IOCTL_I915_GET_SPRITE_COLORKEY ioctl
drm/i915: Rip out GET_SPRITE_COLORKEY ioctl
watchdog: imgpdc: Fix default heartbeat
...
Improve the readability and keeps the lines shorter than 80 columns.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
- Allow amdkfd to work with multiple kgd instances. This is in preparation for
AMD's new open source kernel graphic driver (amdgpu), and for the new
AMD APU, Carrizo.
- Convert timestamping to use 64bit time accessors
- Three other minor changes.
* tag 'drm-amdkfd-next-2015-03-25' of git://people.freedesktop.org/~gabbayo/linux:
drm/amdkfd: Add multiple kgd support
drm/amdkfd: Convert timestamping to use 64bit time accessors
drm/amdkfd: add debug prints for process teardown
drm/amdkfd: Remove unused field from struct qcm_process_device
drm/amdkfd: rename fence_wait_timeout
Consistently with other free functions, handle the NULL case without
oopsing.
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJVGHwjAAoJEHm+PkMAQRiG8rcIAJ6cEJ6mbqLpyz5XrGf4yNp0
+wG/QlEpT8rgrxe9wSjB3lfW3kR2Pe69b9fVVCdiklygdkmva5vfmDrVGGzYfe3M
QrFSSlMVBplvh6IiM/L1mVMtr3DSmCO23YZZ9R5b7FoEYatNHRpNWBCBpuXpd4aD
sLuIvO3L/S7LqeOAFkkYWv6AuL9umicmjR8u+nsmCSRJom7At/aJ6R66WIp9vxho
Rn7r6wcUk6B2Q/gYNjdSE8SIwdyKhuBGyvqQ9U9s6Btg9DQfM/b0vG5kw9hqeAq/
9445jqVDP1whA2vz6GjnvltidxrqRvuDPBwzOnFmY5U+KZz4lS3x2mnWAAJ3xWs=
=TqVJ
-----END PGP SIGNATURE-----
Merge tag 'v4.0-rc6' into drm-intel-next
Backmerge Linux 4.0-rc6 because conflicts are (again) getting out of
hand. To make sure we don't lose any bugfixes from the 4.0-rc5-rc6
flurry of patches we've applied them all to -next too.
Conflicts:
drivers/gpu/drm/i915/intel_display.c
Always take the version from -next, we've already handled all
conflicts with explicit cherrypicking.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Users of the atomic state assume that if the pointer to a crtc, plane or
connector is not NULL in the respective object vector, than the state
for that object in *_states vector also won't be NULL. That assumption
was broken by drm_atomic_state_clear(), which would clear the state
pointer but leave the pointer to the object still set.
This fixes a NULL pointer dereference in i915 caused by the use of
drm_atomic_state_clear().
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Use the same logic when checking for valid ioctl range in
drm_ioctl_flags() that is used in drm_ioctl() to avoid
confusion.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Use max() to make the code to determine the allocation size for
the ioctl data easier to read.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that cmd_drv is gone the handling for core and driver ioctls is
mostly identical, so eliminate the duplication. Also take the
opportunity to simplify the range checks to be less cluttered.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
ioctl->cmd_drv is pointless and we can just as well stick the full ioctl
definition into ioctl->cmd.
Cc: Jakob Bornecrantz <jakob@vmware.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Since
commit 17cabf571e
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed Jan 14 11:20:57 2015 +0000
drm/i915: Trim the command parser allocations
we may then try to allocate a zero-sized object and attempt to extract
its pages. Understandably this fails.
Testcase: igt/gem_exec_nop #ivb,byt,hsw
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The return value of one of the calls to drm_atomic_get_connector_state()
in intel_modeset_stage_output_state() wasn't checked for errors.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
To allow for views where the view type is not defined by the view type only,
like it is in stereo or rotated 90 degree view, change the semantic to require
the whole view structure for comparison when we match a GGTT view.
This allows including parameters like offset to be included in the view which
is useful for eg. partial views.
v3:
- Rely on ggtt_view type being 0 for non-GGTT vma's, which equals to
I915_GGTT_VIEW_NORMAL. (Daniel Vetter)
- Do not use potentially slower comparison when we only want to know if
something is or is not a normal view.
- Rebase on top of rotated view patches. Add rotated view singleton.
- If one view is missing in comparison they're equal only if both are missing.
v4:
- Use comparison helper in obj_to_ggtt_view too. (Tvrtko Ursulin)
- Do WARN_ON if one view is NULL. (Tvrtko Ursulin)
Cc: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drivers/gpu/drm/i915/i915_gem_gtt.c:1349:1-4: WARNING: end returns can be simpified and declaration on line 1347 can be dropped
Simplify a trivial if-return sequence. Possibly combine with a
preceding function call.
Generated by: scripts/coccinelle/misc/simple_return.cocci
CC: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Legacy setCrtc has a nice fastpath for just updating the frontbuffer
when the output routing doesn't change. Which I of course tried to
keep working, except that I fumbled the job: The helpers correctly
compute ->mode_changed, CRTC updates get correctly skipped but
connector functions are called unconditionally.
Fix this.
v2: For the disable sided connector->state->crtc might be NULL.
Instead look at the old_connector_state->crtc, but still use the new
crtc state for that old crtc. Reported by Thierry.
Reported-and-Tested-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk> (v1)
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Gustavo Padovan <gustavo.padovan@collabora.co.uk> (v1)
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
This is useful for writing igts to make sure we don't break this,
without being forced to own a one of these dinosaurs.
Suggested-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Ander Conselvan de Oliveira <conselvan2@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Some of the crtc_compute_clock() still depended on encoder->new_crtc
since they didn't use intel_pipe_will_have_type() and used an open
coded version of that function instead. This patch replaces those with
the appropriate code that checks the atomic state intead.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
[danvet: Separate the if (!connector) continue to facility easier
extraction of a loop iterator for all of these (there's lots more in
i915 and atomic helpers).]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The function intel_dp_set_drrs_state() would decide which pipe to
downclock based on the staged config for the given connector. However,
the result of that function is immediate, and it uses input values from
crtc->config, so it should be looking at the current crtc instead.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Pass a crtc_state to it and find whether the pipe has an encoder of a
given type by looking at the drm_atomic_state the crtc_state points to.
Until recently i9xx_get_refclk() used to be called indirectly from
vlv_force_pll_on() with a dummy crtc_state. That dummy crtc state is not
converted to be part of a full drm atomic state, so add a WARN in case
someone decides to call that again with a such dummy state. This was
removed in
commit 9cbe40c15a
Author: Vijay Purushothaman <vijay.a.purushothaman@linux.intel.com>
Date: Thu Mar 5 19:33:08 2015 +0530
drm/i915: Update prop, int co-eff and gain threshold for CHV
v2: Warn if there is no connectors for a given crtc. (Daniel)
Replace comment i9xx_get_refclk() with a WARN_ON(). (Ander)
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
[danvet: Add commit reference for when i9xx_get_refclk was removed.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Follow up patches will convert some functions called from there to use
the atomic state, instead of directly accessing the new or current
config. This patch just changes the parameters, but shouldn't have any
functional changes.
Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan.de.oliveira@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Traces for page directories and tables allocation and map.
v2: Removed references to teardown.
v3: bitmap_scnprintf has been deprecated.
v4: Replace bitmap_scnprintf with scnprintf correctly, and get right
range lengths. (Mika)
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This patch continues on the idea from "Track GEN6 page table usage".
From here on, in the steady state, PDEs are all pointing to the scratch
page table (as recommended in the spec). When an object is allocated in
the VA range, the code will determine if we need to allocate a page for
the page table. Similarly when the object is destroyed, we will remove,
and free the page table pointing the PDE back to the scratch page.
Following patches will work to unify the code a bit as we bring in GEN8
support. GEN6 and GEN8 are different enough that I had a hard time to
get to this point with as much common code as I do.
The aliasing PPGTT must pre-allocate all of the page tables. There are a
few reasons for this. Two trivial ones: aliasing ppgtt goes through the
ggtt paths, so it's hard to maintain, we currently do not restore the
default context (assuming the previous force reload is indeed
necessary). Most importantly though, the only way (it seems from
empirical evidence) to invalidate the CS TLBs on non-render ring is to
either use ring sync (which requires actually stopping the rings in
order to synchronize when the sync completes vs. where you are in
execution), or to reload DCLV. Since without full PPGTT we do not ever
reload the DCLV register, there is no good way to achieve this. The
simplest solution is just to not support dynamic page table
creation/destruction in the aliasing PPGTT.
We could always reload DCLV, but this seems like quite a bit of excess
overhead only to save at most 2MB-4k of memory for the aliasing PPGTT
page tables.
v2: Make the page table bitmap declared inside the function (Chris)
Simplify the way scratching address space works.
Move the alloc/teardown tracepoints up a level in the call stack so that
both all implementations get the trace.
v3: Updated trace event to spit out a name
v4: Aliasing ppgtt is now initialized differently (in setup global gtt)
v5: Rebase to latest code. Also removed unnecessary aliasing ppgtt check
for trace, as it is no longer possible after the PPGTT cleanup patch series
of a couple of months ago (Daniel).
v6: Implement changes from code review (Daniel):
- allocate/teardown_va_range calls added.
- Add a scratch page allocation helper (only need the address).
- Move trace events to a new patch.
- Use updated mark_tlbs_dirty.
- Moved pt preallocation for aliasing ppgtt into gen6_ppgtt_init.
v7: teardown_va_range removed (Daniel).
In init, gen6_ppgtt_clear_range call is only needed for aliasing ppgtt.
v8: Rebase after s/page_tables/page_table/.
v9: Remove unnecessary scratch flag in page_table struct, future patches
can just compare against ppgtt->scratch_pt, and alloc_pt_scratch becomes
redundant. Initialize scratch_pt and pt. (Mika)
v10: Clean up aliasing ppgtt init error path and prevent leaking the
ppgtt obj when init fails. (Mika)
Updated commit author. (Daniel)
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v4+)
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We are already unmapping them in gen6_ppgtt_free. This function became
redundant since commit 06fda602db
("drm/i915: Create page table allocators").
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
i915_dma_map_single relies on dma_mapping_error, which returns positive
error codes. Found by static checker.
Introduced by commit 678d96fbb3
("drm/i915: Track GEN6 page table usage").
v2: Return negative error code and renamed commit title. (Dan)
v3: Missing reported-by tag (Daniel)
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Found by static analysis tool, this was harmless as the pt was not
used out of scope though.
Introduced by commit 678d96fbb3
("drm/i915: Track GEN6 page table usage").
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>