Rob Clark [Tue, 12 Sep 2017 18:23:05 +0000 (14:23 -0400)]
drm/msm: fix _NO_IMPLICIT fencing case
We need to call reservation_object_reserve_shared() in both cases, but
this wasn't happening in the _NO_IMPLICIT submit case.
Fixes: f0a42bb ("drm/msm: submit support for in-fences") Reported-by: Jordan Crouse <jcrouse@codeaurora.org> Signed-off-by: Rob Clark <robdclark@gmail.com>
Keith Packard [Fri, 17 Mar 2017 00:56:28 +0000 (17:56 -0700)]
drm: Add four ioctls for managing drm mode object leases [v7]
drm_mode_create_lease
Creates a lease for a list of drm mode objects, returning an
fd for the new drm_master and a 64-bit identifier for the lessee
drm_mode_list_lesees
List the identifiers of the lessees for a master file
drm_mode_get_lease
List the leased objects for a master file
drm_mode_revoke_lease
Erase the set of objects managed by a lease.
This should suffice to at least create and query leases.
Changes for v2 as suggested by Daniel Vetter <daniel.vetter@ffwll.ch>:
* query ioctls only query the master associated with
the provided file.
* 'mask_lease' value has been removed
* change ioctl has been removed.
Changes for v3 suggested in part by Dave Airlie <airlied@gmail.com>
* Add revoke ioctl.
Changes for v4 suggested by Dave Airlie <airlied@gmail.com>
* Expand on the comment about the magic use of &drm_lease_idr_object
* Pad lease ioctl structures to align on 64-bit boundaries
Changes for v5 suggested by Dave Airlie <airlied@gmail.com>
* Check for non-negative object_id in create_lease to avoid debug
output from the kernel.
Changes for v6 provided by Dave Airlie <airlied@gmail.com>
* For non-universal planes add primary/cursor planes to lease
If we aren't exposing universal planes to this userspace client,
and it requests a lease on a crtc, we should implicitly export the
primary and cursor planes for the crtc.
If the lessee doesn't request universal planes, it will just see
the crtc, but if it does request them it will then see the plane
objects as well.
This also moves the object look ups earlier as a side effect, so
we'd exit the ioctl quicker for non-existant objects.
* Restrict leases to crtc/connector/planes.
This only allows leasing for objects we wish to allow.
Changes for v7 provided by Dave Airlie <airlied@gmail.com>
* Check pad args are 0
* Check create flags and object count are valid.
* Check return from fd allocation
* Refactor lease idr setup and add some simple validation
* Use idr_mutex uniformly (Keith)
Signed-off-by: Keith Packard <keithp@keithp.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Keith Packard [Mon, 10 Apr 2017 04:35:34 +0000 (22:35 -0600)]
drm: Check mode object lease status in all master ioctl paths [v4]
Attempts to modify un-leased objects are rejected with an error.
Information returned about unleased objects is modified to make them
appear unusable and/or disconnected.
Changes for v2 as suggested by Daniel Vetter <daniel.vetter@ffwll.ch>:
* With the change in the __drm_mode_object_find API to pass the
file_priv along, we can now centralize most of the lease-based
access checks in that function.
* A few places skip that API and require in-line checks.
Changes for v3 provided by Dave Airlie <airlied@redhat.com>
* remove support for leasing encoders.
* add support for leasing planes.
Changes for v4
* Only call drm_lease_held if DRIVER_MODESET.
Signed-off-by: Keith Packard <keithp@keithp.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Keith Packard [Wed, 15 Mar 2017 05:26:41 +0000 (22:26 -0700)]
drm: Add drm_object lease infrastructure [v5]
This provides new data structures to hold "lease" information about
drm mode setting objects, and provides for creating new drm_masters
which have access to a subset of the available drm resources.
An 'owner' is a drm_master which is not leasing the objects from
another drm_master, and hence 'owns' them.
A 'lessee' is a drm_master which is leasing objects from some other
drm_master. Each lessee holds the set of objects which it is leasing
from the lessor.
A 'lessor' is a drm_master which is leasing objects to another
drm_master. This is the same as the owner in the current code.
The set of objects any drm_master 'controls' is limited to the set of
objects it leases (for lessees) or all objects (for owners).
Objects not controlled by a drm_master cannot be modified through the
various state manipulating ioctls, and any state reported back to user
space will be edited to make them appear idle and/or unusable. For
instance, connectors always report 'disconnected', while encoders
report no possible crtcs or clones.
The full list of lessees leasing objects from an owner (either
directly, or indirectly through another lessee), can be searched from
an idr in the drm_master of the owner.
Changes for v2 as suggested by Daniel Vetter <daniel.vetter@ffwll.ch>:
* Sub-leasing has been disabled.
* BUG_ON for lock checking replaced with lockdep_assert_held
* 'change' ioctl has been removed.
* Leased objects can always be controlled by the lessor; the
'mask_lease' flag has been removed
* Checking for leased status has been simplified, replacing
the drm_lease_check function with drm_lease_held.
Changes in v3, some suggested by Dave Airlie <airlied@gmail.com>
* Add revocation. This allows leases to be effectively revoked by
removing all of the objects they have access to. The lease itself
hangs around as it's hanging off a file.
* Free the leases IDR when the master is destroyed
* _drm_lease_held should look at lessees, not lessor
* Allow non-master files to check for lease status
Changes in v4, suggested by Dave Airlie <airlied@gmail.com>
* Formatting and whitespace changes
Changes in v5 (airlied)
* check DRIVER_MODESET before lease destroy call
* check DRIVER_MODESET for lease revoke (Chris)
* Use idr_mutex uniformly for all lease elements of struct drm_master. (Keith)
Dave Airlie [Tue, 24 Oct 2017 06:51:05 +0000 (16:51 +1000)]
Merge tag 'drm-misc-next-2017-10-20' of git://anongit.freedesktop.org/drm/drm-misc into drm-next
Final drm-misc feature pull for 4.15:
UAPI Changes:
- new madvise ioctl for vc4 (Boris)
Core Changes:
- plane commit tracking fixes (Maarten)
- vgaarb improvements for fancy new platforms (aka ppc64 and arm64) by
Bjorn Helgaas
Driver Changes:
- pile of new panel drivers: Toshiba LT089AC19000, Innolux AT043TN24
- more sun4i work to support A10/A20 Tcon and hdmi outputs
- vc4: fix sleep in irq handler by making it threaded (Eric)
- udl probe/edid read fixes (Robert Tarasov)
And a bunch of misc small cleanups/refactors and doc fixes all over.
* tag 'drm-misc-next-2017-10-20' of git://anongit.freedesktop.org/drm/drm-misc: (32 commits)
drm/vc4: Fix sleeps during the IRQ handler for DSI transactions.
drm/vc4: Add the DRM_IOCTL_VC4_GEM_MADVISE ioctl
drm/panel: simple: add Toshiba LT089AC19000
dma-fence: remove duplicate word in comment
drm/panel: simple: add delays for Innolux AT043TN24
drm/panel: simple: add bus flags for Innolux AT043TN24
drm/panel: simple: fix vertical timings for Innolux AT043TN24
drm/atomic-helper: check that drivers call drm_crtc_vblank_off
drm: some KMS todo ideas
vgaarb: Factor out EFI and fallback default device selection
vgaarb: Select a default VGA device even if there's no legacy VGA
drm/bridge: adv7511: Fix a use after free
drm/sun4i: Add support for A20 display pipeline components
drm/sun4i: Add support for A10 display pipeline components
drm/sun4i: hdmi: Support HDMI controller on A10
drm/sun4i: tcon: Add support for A10 TCON
drm/sun4i: backend: Support output muxing
drm/sun4i: tcon: Move out the tcon0 common setup
drm/sun4i: tcon: Don't rely on encoders to set the TCON mode
drm/sun4i: tcon: Don't rely on encoders to enable the TCON
...
Dave Airlie [Tue, 24 Oct 2017 06:50:45 +0000 (16:50 +1000)]
Merge tag 'drm/tegra/for-4.15-rc1' of git://anongit.freedesktop.org/tegra/linux into drm-next
drm/tegra: Changes for v4.15-rc1
This contains a bit of cleanup and some minor fixes for the host1x and
Tegra DRM drivers. There's also some more preparatory work for Tegra186
support which I'm not quite ready to send upstream because the GPIO
driver needed for HDMI support has been stuck for months, and we can't
do much without it. Hopefully that driver will land in v4.15, which
would mean we could go ahead with Tegra186 display support in v4.16.
* tag 'drm/tegra/for-4.15-rc1' of git://anongit.freedesktop.org/tegra/linux: (21 commits)
drm/tegra: hdmi: Add cec-notifier support
drm/tegra: dc: Perform a complete reset sequence
drm/tegra: dc: Make sure to set the module clock rate
drm/tegra: dc: Simplify atomic plane helper functions
drm/tegra: dc: Move some declarations to dc.h
drm/tegra: vic: Use of_device_get_match_data()
drm/tegra: sor: Use of_device_get_match_data()
drm/tegra: hdmi: Use of_device_get_match_data()
drm/tegra: dc: Use of_device_get_match_data()
drm/tegra: Use u64_to_user_ptr helper
gpu: host1x: Fix incorrect comment for channel_request
gpu: host1x: Disassemble more instructions
gpu: host1x: Improve debug disassembly formatting
gpu: host1x: Enable gather filter
gpu: host1x: Enable Tegra186 syncpoint protection
gpu: host1x: Call of_dma_configure() after setting bus
drm/tegra: Add Tegra186 support for VIC
gpu: host1x: Add Tegra186 support
dt-bindings: host1x: Add Tegra186 information
gpu: host1x: syncpt: Request syncpoints per client
...
Keith Packard [Fri, 30 Jun 2017 05:49:31 +0000 (22:49 -0700)]
drm: Add CRTC_GET_SEQUENCE and CRTC_QUEUE_SEQUENCE ioctls [v3]
These provide crtc-id based functions instead of pipe-number, while
also offering higher resolution time (ns) and wider frame count (64)
as required by the Vulkan API.
v2:
* Check for DRIVER_MODESET in new crtc-based vblank ioctls
Failing to check this will oops the driver.
* Ensure vblank interupt is running in crtc_get_sequence ioctl
The sequence and timing values are not correct while the
interrupt is off, so make sure it's running before asking for
them.
* Short-circuit get_sequence if the counter is enabled and accurate
Steal the idea from the code in wait_vblank to avoid the
expense of drm_vblank_get/put
* Return active state of crtc in crtc_get_sequence ioctl
Might be useful for applications that aren't in charge of
modesetting?
* Use drm_crtc_vblank_get/put in new crtc-based vblank sequence ioctls
Daniel Vetter prefers these over the old drm_vblank_put/get
APIs.
* Return s64 ns instead of u64 in new sequence event
Suggested-by: Daniel Vetter <daniel@ffwll.ch> Suggested-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
v3:
* Removed FIRST_PIXEL_OUT_FLAG
* Document that the timestamp in the query and event are
that of the first pixel leaving the display engine for
the display (using the same wording as the Vulkan spec).
Suggested-by: Michel Dänzer <michel@daenzer.net> Acked-by: Dave Airlie <airlied@redhat.com>
[airlied: left->leaves (Michel)]
Signed-off-by: Keith Packard <keithp@keithp.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
Lucas Stach [Fri, 20 Oct 2017 15:46:55 +0000 (17:46 +0200)]
drm/etnaviv: short-circuit perfmon ioctls
The feature implementation isn't stable yet. Reject any attempt to use
the IOCTLs for now. This keeps most of the code in place, so we can stabilize
it in-tree, but keeps userspace from using the feature for now.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Keith Packard [Wed, 5 Jul 2017 21:34:23 +0000 (14:34 -0700)]
drm: Reorganize drm_pending_event to support future event types [v2]
Place drm_event_vblank in a new union that includes that and a bare
drm_event structure. This will allow new members of that union to be
added in the future without changing code related to the existing vbl
event type.
Assignments to the crtc_id field are now done when the event is
allocated, rather than when delievered. This way, delivery doesn't
need to have the crtc ID available.
v2:
* Remove 'dev' argument from create_vblank_event
It wasn't being used anyways, and if we need it in the future,
we can always get it from crtc->dev.
* Check for MODESETTING before looking for crtc in queue_vblank_event
UMS drivers will oops if we try to get a crtc, so make sure
we're modesetting before we try to find a crtc_id to fill into
the event.
Signed-off-by: Keith Packard <keithp@keithp.com> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit dc695b85fde88eca3ef3b03fcd82f15b6bc6e462)
Keith Packard [Thu, 12 Oct 2017 18:57:49 +0000 (11:57 -0700)]
drm: Widen vblank count to 64-bits [v3]
This modifies the datatypes used by the vblank code to provide 64 bits
of vblank count.
The driver interfaces have been left using 32 bits of vblank count;
all of the code necessary to widen that value for the user API was
already included to handle devices returning fewer than 32-bits.
This will provide the necessary datatypes for the Vulkan API.
v2:
* Re-write wait_vblank ioctl to ABSOLUTE sequence
When an application uses the WAIT_VBLANK ioctl with RELATIVE
or NEXTONMISS bits set, the target vblank interval is updated
within the kernel. We need to write that target back to the
ioctl buffer and update the flags bits so that if the wait is
interrupted by a signal, when it is re-started, it will target
precisely the same vblank count as before.
* Leave driver API with 32-bit vblank count
v3:
* Rebase on top of Arnd Bergmann's patch which had
the switch to ktime_t parts.
[airlied: fix conflict with Ville vblank change].
Suggested-by: Michel Dänzer <michel@daenzer.net> Suggested-by: Daniel Vetter <daniel@ffwll.ch> Signed-off-by: Keith Packard <keithp@keithp.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
(cherry picked from commit 2affbc16983e4fc90960bc7f70e7615f4228199b)
Hans Verkuil [Mon, 11 Sep 2017 12:29:52 +0000 (14:29 +0200)]
drm/tegra: hdmi: Add cec-notifier support
In order to support CEC the HDMI driver has to inform the CEC driver
whenever the physical address changes. So when the EDID is read the
CEC driver has to be informed and whenever the hotplug detect goes
away.
This is done through the cec-notifier framework.
The link between the HDMI driver and the CEC driver is done through
the hdmi-phandle property in the tegra-cec node in the device tree.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Thierry Reding <treding@nvidia.com>
Thierry Reding [Wed, 30 Aug 2017 15:38:39 +0000 (17:38 +0200)]
drm/tegra: dc: Make sure to set the module clock rate
When applying the PLL changes from the computed state object, make sure
to set the rate of the display controller module clock. Failing to do so
can yield to a situation where the parent will be set to the proper
pixel clock, but the module clock will be divided down to the rate that
is happened to be set to before the parent rate change.
Remove the tegra_dc_disable_window() function whose only purpose was to
allow tegra_plane_atomic_update() to also call it. Fix that by shuffling
tegra_plano_atomic_disable() to before tegra_plane_atomic_update().
While at it, also remove the overlay plane helper functions because they
are exactly the same as the primary plane helper functions.
Mikko Perttunen [Thu, 28 Sep 2017 12:50:44 +0000 (15:50 +0300)]
drm/tegra: Use u64_to_user_ptr helper
Use the u64_to_user_ptr helper macro to cast IOCTL argument u64 values
to user pointers instead of writing out the cast manually. Also do
some other cleanup with user pointers to make them stand out more
and look cleaner.
Mikko Perttunen [Thu, 28 Sep 2017 12:50:41 +0000 (15:50 +0300)]
gpu: host1x: Improve debug disassembly formatting
The host1x driver prints out "disassembly" dumps of the command FIFO
and gather contents on submission timeouts. However, the output has
been quite difficult to read with unnecessary newlines and occasional
missing parentheses.
Fix these problems by using pr_cont to remove unnecessary newlines
and by fixing other small issues.
Mikko Perttunen [Thu, 28 Sep 2017 12:50:40 +0000 (15:50 +0300)]
gpu: host1x: Enable gather filter
The gather filter is a feature present on Tegra124 and newer where the
hardware prevents GATHERed command buffers from executing commands
normally reserved for the CDMA pushbuffer which is maintained by the
kernel driver.
This commit enables the gather filter on all supporting hardware.
Mikko Perttunen [Thu, 28 Sep 2017 12:50:39 +0000 (15:50 +0300)]
gpu: host1x: Enable Tegra186 syncpoint protection
Since Tegra186 the Host1x hardware allows syncpoints to be assigned to
specific channels, preventing any other channels from incrementing
them.
Enable this feature where available and assign syncpoints to channels
when submitting a job. Syncpoints are currently never unassigned from
channels since that would require extra work and is unnecessary with
the current channel allocation model.
Mikko Perttunen [Tue, 5 Sep 2017 08:43:05 +0000 (11:43 +0300)]
gpu: host1x: Add Tegra186 support
Add support for the implementation of Host1x present on the Tegra186.
The register space has been shuffled around a little bit, requiring
addition of some chip-specific code sections. Tegra186 also adds
several new features, most importantly the hypervisor, but those are
not yet supported with this commit.
Thierry Reding [Wed, 30 Aug 2017 10:48:31 +0000 (12:48 +0200)]
gpu: host1x: syncpt: Request syncpoints per client
Rather than request syncpoints for a struct device *, request them for a
struct host1x_client *. This is important because subsequent patches are
going to break the assumption that host1x will always be the parent for
devices requesting a syncpoint. It's also a more natural choice because
host1x clients are really the only ones that will know how to deal with
syncpoints.
Note that host1x clients are always guaranteed to be children of host1x,
regardless of their location in the device tree.
Dave Airlie [Fri, 20 Oct 2017 05:54:44 +0000 (15:54 +1000)]
Merge tag 'drm-amdkfd-next-2017-10-18' of git://people.freedesktop.org/~gabbayo/linux into drm-next
This is the amdkfd pull request for 4.15 merge window.
The patches here are relevant only for Kaveri and Carrizo. Still no dGPU patches.
The main goal is to continue alignment with the internal AMD development tree.
The following is a summary of the changes:
- Improvements and fixes to suspend/resume code
- Improvements and fixes to process termination code
- Cleanups in the queue unmapping functionality
- Reuse code from amdgpu
* tag 'drm-amdkfd-next-2017-10-18' of git://people.freedesktop.org/~gabbayo/linux:
drm/amdkfd: Improve multiple SDMA queues support per process
drm/amdkfd: Limit queue number per process and device to 127
drm/amdkfd: Clean up process queue management
drm/amdkfd: Compress unnecessary function parameters
drm/amdkfd: Improve process termination handling
drm/amdkfd: Avoid submitting an unnecessary packet to HWS
drm/amdkfd: Fix MQD updates
drm/amdkfd: Pass filter params to unmap_queues_cpsch
drm/amdkfd: move locking outside of unmap_queues_cpsch
drm/amdkfd: Avoid name confusion involved in queue unmapping
drm/amdkfd: Drop _nocpsch suffix from shared functions
drm/amdkfd: Reuse CHIP_* from amdgpu v2
drm/amdkfd: Use VMID bitmap from KGD v2
drm/amdkfd: Adjust dequeue latencies and timeouts
drm/amdkfd: Rectify the jiffies calculation error with milliseconds v2
drm/amdkfd: Fix suspend/resume issue on Carrizo v2
drm/amdkfd: Reorganize kfd resume code
Dave Airlie [Fri, 20 Oct 2017 00:56:10 +0000 (10:56 +1000)]
Merge tag 'drm-intel-next-2017-10-12' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
Last batch of drm/i915 features for v4.15:
- transparent huge pages support (Matthew)
- uapi: I915_PARAM_HAS_SCHEDULER into a capability bitmask (Chris)
- execlists: preemption (Chris)
- scheduler: user defined priorities (Chris)
- execlists optimization (Michał)
- plenty of display fixes (Imre)
- has_ipc fix (Rodrigo)
- platform features definition refactoring (Rodrigo)
- legacy cursor update fix (Maarten)
- fix vblank waits for cursor updates (Maarten)
- reprogram dmc firmware on resume, dmc state fix (Imre)
- remove use_mmio_flip module parameter (Maarten)
- wa fixes (Oscar)
- huc/guc firmware refacoring (Sagar, Michal)
- push encoder specific code to encoder hooks (Jani)
- DP MST fixes (Dhinakaran)
- eDP power sequencing fixes (Manasi)
- selftest updates (Chris, Matthew)
- mmu notifier cpu hotplug deadlock fix (Daniel)
- more VBT parser refactoring (Jani)
- max pipe refactoring (Mika Kahola)
- rc6/rps refactoring and separation (Sagar)
- userptr lockdep fix (Chris)
- tracepoint fixes and defunct tracepoint removal (Chris)
- use rcu instead of abusing stop_machine (Daniel)
- plenty of other fixes all around (Everyone)
* tag 'drm-intel-next-2017-10-12' of git://anongit.freedesktop.org/drm/drm-intel: (145 commits)
drm/i915: Update DRIVER_DATE to 20171012
drm/i915: Simplify intel_sanitize_enable_ppgtt
drm/i915/userptr: Drop struct_mutex before cleanup
drm/i915/dp: limit sink rates based on rate
drm/i915/dp: centralize max source rate conditions more
drm/i915: Allow PCH platforms fall back to BIOS LVDS mode
drm/i915: Reuse normal state readout for LVDS/DVO fixed mode
drm/i915: Use rcu instead of stop_machine in set_wedged
drm/i915: Introduce separate status variable for RC6 and LLC ring frequency setup
drm/i915: Create generic functions to control RC6, RPS
drm/i915: Create generic function to setup LLC ring frequency table
drm/i915: Rename intel_enable_rc6 to intel_rc6_enabled
drm/i915: Name structure in dev_priv that contains RPS/RC6 state as "gt_pm"
drm/i915: Move rps.hw_lock to dev_priv and s/hw_lock/pcu_lock
drm/i915: Name i915_runtime_pm structure in dev_priv as "runtime_pm"
drm/i915: Separate RPS and RC6 handling for CHV
drm/i915: Separate RPS and RC6 handling for VLV
drm/i915: Separate RPS and RC6 handling for BDW
drm/i915: Remove superfluous IS_BDW checks and non-BDW changes from gen8_enable_rps
drm/i915: Separate RPS and RC6 handling for gen6+
...
Dave Airlie [Fri, 20 Oct 2017 00:47:19 +0000 (10:47 +1000)]
Merge branch 'drm-next-4.15' of git://people.freedesktop.org/~agd5f/linux into drm-next
Last set of features for 4.15. Highlights:
- Add a bo flag to allow buffers to opt out of implicit sync
- Add ctx priority setting interface
- Lots more powerplay cleanups
- Start to plumb through vram lost infrastructure for gpu reset
- ttm support for huge pages
- misc cleanups and bug fixes
* 'drm-next-4.15' of git://people.freedesktop.org/~agd5f/linux: (73 commits)
drm/amd/powerplay: Place the constant on the right side of the test
drm/amd/powerplay: Remove useless variable
drm/amd/powerplay: Don't cast kzalloc() return value
drm/amdgpu: allow GTT overcommit during bind
drm/amdgpu: linear validate first then bind to GART
drm/amd/pp: Fix overflow when setup decf/pix/disp dpm table.
drm/amd/pp: thermal control not enabled on vega10.
drm/amdgpu: busywait KIQ register accessing (v4)
drm/amdgpu: report more amdgpu_fence_info
drm/amdgpu:don't check soft_reset for sriov
drm/amdgpu:fix duplicated setting job's vram_lost
drm/amdgpu:reduce wb to 512 slot
drm/amdgpu: fix regresstion on SR-IOV gpu reset failed
drm/amd/powerplay: Tidy up cz_dpm_powerup_vce()
drm/amd/powerplay: Tidy up cz_dpm_powerdown_vce()
drm/amd/powerplay: Tidy up cz_dpm_update_vce_dpm()
drm/amd/powerplay: Tidy up cz_dpm_update_uvd_dpm()
drm/amd/powerplay: Tidy up cz_dpm_powerup_uvd()
drm/amd/powerplay: Tidy up cz_dpm_powerdown_uvd()
drm/amd/powerplay: Tidy up cz_start_dpm()
...
Eric Anholt [Sat, 14 Oct 2017 00:12:55 +0000 (17:12 -0700)]
drm/vc4: Fix sleeps during the IRQ handler for DSI transactions.
VC4's DSI1 has a bug where the AXI connection is broken for 32-bit
writes from the CPU, so we use the DMA engine to DMA 32-bit values
into registers instead. That sleeps, so we can't do it from the top
half.
As a solution, use an interrupt thread so that all our writes happen
when sleeping is is allowed.
v2: Use IRQF_ONESHOT (suggested by Boris)
v3: Style nitpicks.
Georgiana Chelu [Tue, 17 Oct 2017 20:22:08 +0000 (23:22 +0300)]
drm/amd/powerplay: Remove useless variable
The result variable is initialized at the beginning of the function, but
its value does not change during the function execution. Thus, remove the
variable and return the SUCCESS value, which is 0.
Issue found by coccinelle script:
* Unneeded variable: "result". Return "0"
Path to the cocci script: scripts/coccinelle/misc/returnvar.cocci
Signed-off-by: Georgiana Chelu <georgiana.chelu93@gmail.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Georgiana Chelu [Tue, 17 Oct 2017 20:22:07 +0000 (23:22 +0300)]
drm/amd/powerplay: Don't cast kzalloc() return value
The kzalloc function returns a void pointer and the assignment
operator converts it to the type of pointer it is assigned to.
Therefore, there is no need to cast.
Issue found by alloc_cast.cocci:
* WARNING: casting value returned by memory allocation function
to <struct type> is useless.
Path to the cocci script: scripts/coccinelle/api/alloc/alloc_cast.cocci
Signed-off-by: Georgiana Chelu <georgiana.chelu93@gmail.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 16 Oct 2017 09:18:54 +0000 (11:18 +0200)]
drm/amdgpu: allow GTT overcommit during bind
While binding BOs to GART we need to allow a bit overcommit in the GTT
domain. Otherwise we can never use the full GART space when GART size=GTT size.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 16 Oct 2017 08:32:04 +0000 (10:32 +0200)]
drm/amdgpu: linear validate first then bind to GART
For VM emulation for old UVD/VCE we need to validate the BO with linear
VRAM flag set first and then eventually bind it to GART.
Validating with linear VRAM flag set can move the BO to GART making
UVD/VCE read/write from an unbound GART BO.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> CC: stable@vger.kernel.org Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
pding [Fri, 13 Oct 2017 07:38:35 +0000 (15:38 +0800)]
drm/amdgpu: busywait KIQ register accessing (v4)
Register accessing is performed when IRQ is disabled. Never sleep in
this function.
Known issue: dead sleep in many use cases of index/data registers.
v2:
- wrap polling fence functions.
- don't trigger IRQ for polling in case of wrongly fence signal.
v3:
- handle wrap round gracefully.
- add comments for polling function
v4:
- don't return negative timeout confused with error code
Signed-off-by: pding <Pixel.Ding@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
pding [Thu, 12 Oct 2017 05:53:20 +0000 (13:53 +0800)]
drm/amdgpu: report more amdgpu_fence_info
Only for GFX ring. This can help checking MCBP feature.
The fence at the end of the frame will indicate the completion status.
If the frame completed normally, the fence is written to the address
given in the EVENT_WRITE_EOP packet. If preemption occurred in the
previous IB the address is adjusted by 2 DWs. If work submitted in the
frame was reset before completion, the fence address is adjusted by
four DWs. In the case that preemption occurred, and before preemption
completed a reset was initiated, the address will be adjusted with six
DWs
Signed-off-by: pding <Pixel.Ding@amd.com> Reviewed-by: Monk Liu <monk.liu@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Monk Liu [Mon, 16 Oct 2017 12:02:08 +0000 (20:02 +0800)]
drm/amdgpu:fix duplicated setting job's vram_lost
Signed-off-by: Monk Liu <Monk.Liu@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Monk Liu [Tue, 17 Oct 2017 11:23:42 +0000 (19:23 +0800)]
drm/amdgpu:reduce wb to 512 slot
with current WB usage we only use 57 slots, so 512
is extreamly sufficient, and reduce to 512 can
make WB fit into one page.
Signed-off-by: Monk Liu <Monk.Liu@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:33:25 +0000 (13:33 -0400)]
drm/amd/powerplay: Tidy up cz_dpm_powerup_vce()
Use PP_CAP macro.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:32:38 +0000 (13:32 -0400)]
drm/amd/powerplay: Tidy up cz_dpm_powerdown_vce()
Use PP_CAP macro.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:31:49 +0000 (13:31 -0400)]
drm/amd/powerplay: Tidy up cz_dpm_update_vce_dpm()
Use PP_CAP and tidy up indentation.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:30:28 +0000 (13:30 -0400)]
drm/amd/powerplay: Tidy up cz_dpm_update_uvd_dpm()
Use PP_CAP and tidy up indentation.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:28:24 +0000 (13:28 -0400)]
drm/amd/powerplay: Tidy up cz_dpm_powerup_uvd()
Use PP_CAP and simplify enable/disable logic.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:25:42 +0000 (13:25 -0400)]
drm/amd/powerplay: Tidy up cz_dpm_powerdown_uvd()
Use PP_CAP and tidy up indentation.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Mon, 16 Oct 2017 17:23:55 +0000 (13:23 -0400)]
drm/amd/powerplay: Tidy up cz_start_dpm()
Remove unused variables.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jérémy Lefaure [Mon, 16 Oct 2017 02:29:23 +0000 (22:29 -0400)]
drm/amdgpu: use ARRAY_SIZE
Using the ARRAY_SIZE macro improves the readability of the code.
Found with Coccinelle with the following semantic patch:
@r depends on (org || report)@
type T;
T[] E;
position p;
@@
(
(sizeof(E)@p /sizeof(*E))
|
(sizeof(E)@p /sizeof(E[...]))
|
(sizeof(E)@p /sizeof(T))
)
Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Thierry Reding <treding@nvidia.com> Signed-off-by: Jérémy Lefaure <jeremy.lefaure@lse.epita.fr> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Fri, 13 Oct 2017 15:24:31 +0000 (17:24 +0200)]
drm/amdgpu: reserve root PD while releasing it
Otherwise somebody could try to evict it at the same time and try to use
half torn down structures.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Thu, 12 Oct 2017 17:28:42 +0000 (19:28 +0200)]
drm/ttm: fix the fix for huge compound pages
We don't use compound pages at the moment. Take this into account when
freeing them.
Signed-off-by: Christian König <christian.koenig@amd.comd> Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Thu, 12 Oct 2017 20:26:34 +0000 (16:26 -0400)]
drm/amdgpu: bump version for vram lost counter query (v2)
v2: vram -> VRAM in comment
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Thu, 12 Oct 2017 10:16:33 +0000 (12:16 +0200)]
drm/amdgpu: minor CS optimization
We only need to loop over all IBs for old UVD/VCE command stream patching.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Fri, 13 Oct 2017 08:58:15 +0000 (10:58 +0200)]
drm/amd/sched: fix job tear down order v2
Move the trace before we signal the scheduler fence and drop the
scheduler fence reference directly before we free the job.
v2: keep extra s_fence reference
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Liu, Monk <Monk.Liu@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add more generic function amdgpu_copy_ttm_mem_to_mem() that supports
arbitrary copy size, offsets and two BOs (source & dest.).
This is useful for KFD Cross Memory Attach feature where data needs to
be copied from BOs from different processes
v2: Add struct amdgpu_copy_mem and changed amdgpu_copy_ttm_mem_to_mem()
function parameters to use the struct
v3: Minor function name change
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In amdgpu_cs_parser_init() in case of error handling
amdgpu_ctx_put() is called without setting p->ctx to NULL after that,
later amdgpu_cs_parser_fini() also calls amdgpu_ctx_put() again and
mess up the reference count.
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom St Denis [Thu, 12 Oct 2017 11:25:08 +0000 (07:25 -0400)]
drm/ttm: Fix unused variables with huge page support
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 9 Oct 2017 15:53:06 +0000 (17:53 +0200)]
drm/amdgpu: add VRAM lost query
Allows userspace to figure out if VRAM was lost.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 9 Oct 2017 13:51:10 +0000 (15:51 +0200)]
drm/amdgpu: set -ECANCELED when dropping jobs
And return from the wait functions the fence error code.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 9 Oct 2017 13:18:43 +0000 (15:18 +0200)]
drm/amdgpu: move the VRAM lost counter per context
Instead of per device track the VRAM lost per context and return ECANCELED
instead of ENODEV.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 9 Oct 2017 13:04:41 +0000 (15:04 +0200)]
drm/amdgpu: keep copy of VRAM lost counter in job
Instead of reading the current counter from fpriv.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 9 Oct 2017 12:45:09 +0000 (14:45 +0200)]
drm/amdgpu: partial revert VRAM lost handling v2
Keep blocking the CS, but revert everything else. Mapping BOs and info IOCTL
are harmless and can still happen even when VRAM content ist lost.
Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Thu, 5 Oct 2017 12:27:34 +0000 (14:27 +0200)]
drm/ttm: add transparent huge page support for wc or uc allocations v2
Add a new huge page pool and try to allocate from it when it makes sense.
v2: avoid compound pages for now
Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Thu, 21 Sep 2017 09:28:25 +0000 (11:28 +0200)]
drm/ttm: move more logic into ttm_page_pool_get_pages
Make it easier to add huge page pool.
Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Wed, 20 Sep 2017 13:06:12 +0000 (15:06 +0200)]
drm/ttm: add transparent huge page support for cached allocations v2
Try to allocate huge pages when it makes sense.
v2: avoid compound pages for now
Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Felix Kuehling <Felix.Kuehling@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Mon, 9 Oct 2017 12:34:13 +0000 (14:34 +0200)]
drm/ttm: don't use compound pages for now
We need to figure out first how to correctly map them into the CPU page tables.
bug: https://bugs.freedesktop.org/show_bug.cgi?id=103138 Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Ken Wang [Fri, 29 Sep 2017 07:41:43 +0000 (15:41 +0800)]
drm/amdgpu: correct reference clock value on vega10
Old value from bringup was wrong.
Cc: stable@vger.kernel.org Signed-off-by: Ken Wang <Ken.Wang@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Horace Chen [Mon, 9 Oct 2017 08:17:16 +0000 (16:17 +0800)]
drm/amdgpu: SR-IOV data exchange between PF&VF
SR-IOV need to exchange some data between PF&VF through shared VRAM
PF will copy some necessary firmware and information to the shared
VRAM. It also requires some information from VF. PF will send a
key through mailbox2 to help guest calculate checksum so that it can
verify whether the data is correct.
So check the data on the specified offset of the shared VRAM, if the
checksum is right, read values from it and write some VF information
next to the data from PF.
Signed-off-by: Horace Chen <horace.chen@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: Move old fence waiting before reservation lock is aquired v2
Helps avoiding deadlock during GPU reset.
Added mutex to amdgpu_ctx to preserve order of fences on a ring.
v2:
Put waiting logic in a function in a seperate function in amdgpu_ctx.c
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: Refactor amdgpu_cs_ib_vm_chunk and amdgpu_cs_ib_fill.
This enables old fence waiting before reservation lock is aquired
which in turn is part of a bigger solution to deadlock happening
when gpu reset with VRAM recovery accures during intensive rendering.
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Colin Ian King [Wed, 11 Oct 2017 09:21:11 +0000 (10:21 +0100)]
drm/amdgpu: make function uvd_v6_0_enc_get_destroy_msg static
The function uvd_v6_0_enc_get_destroy_msg is local to the source and
does not need to be in global scope, so make it static.
Cleans up sparse warning:
symbol 'uvd_v6_0_enc_get_destroy_msg' was not declared. Should it be
static?
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Boris Brezillon [Thu, 19 Oct 2017 12:57:48 +0000 (14:57 +0200)]
drm/vc4: Add the DRM_IOCTL_VC4_GEM_MADVISE ioctl
This ioctl will allow us to purge inactive userspace buffers when the
system is running out of contiguous memory.
For now, the purge logic is rather dumb in that it does not try to
release only the amount of BO needed to meet the last CMA alloc request
but instead purges all objects placed in the purgeable pool as soon as
we experience a CMA allocation failure.
Note that the in-kernel BO cache is always purged before the purgeable
cache because those objects are known to be unused while objects marked
as purgeable by a userspace application/library might have to be
restored when they are marked back as unpurgeable, which can be
expensive.
projects / linux.git / log