Purpose is to avoid calling i915_gem_obj_ggtt_offset from the
interrupt context without the big lock held.

v2: Renamed gtt_start to gtt_offset. (Daniel Vetter)
v3: Cache the VMA instead of address. (Chris Wilson)
Signed-off-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/1452870629-13830-2-git-send-email-tvrtko.ursulin@linux.intel.com

LRC code was calling GEM API like i915_gem_obj_ggtt_offset from
places where the struct_mutex cannot be grabbed (irq handlers).

To avoid that this patch caches some interesting bits and values
in the engine and context structures.

Some usages are also removed where they are not needed like a
few asserts which are either impossible or have been checked
already during engine initialization.

Side benefit is also that interrupt handlers and command
submission stop evaluating invariant conditionals, like what
Gen we are running on, on every interrupt and every command
submitted.

This patch deals with logical ring context id and descriptors
while subsequent patches will deal with the remaining issues.

v2:
 * Cache the VMA instead of the address. (Chris Wilson)
 * Incorporate Dave Gordon's good comments and function name.

v3:
 * Extract ctx descriptor template to a function and group
   functions dealing with ctx descriptor & co together near
   top of the file. (Dave Gordon)
Signed-off-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Gordon <david.s.gordon@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452870629-13830-1-git-send-email-tvrtko.ursulin@linux.intel.com

No need to call ktime_get_raw_ns twice per unlimited wait and can
also elimate a local variable.

v2: Added comment about silencing the compiler warning. (Daniel Vetter)
Signed-off-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NDave Gordon <david.s.gordon@intel.com>
Acked-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452870672-13901-1-git-send-email-tvrtko.ursulin@linux.intel.com

Pull in Dave's drm-next pull request to have a clean base for 4.6.
Also, we need the various atomic state extensions Maarten recently
created.

Conflicts are just adjacent changes that all resolve to nothing in git
diff.
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>

Merge tag 'drm-intel-next-fixes-2016-01-14' of git://anongit.freedesktop.org/drm-intel into drm-next

misc i915 fixes all over the place.

* tag 'drm-intel-next-fixes-2016-01-14' of git://anongit.freedesktop.org/drm-intel:
  drm/i915/gen9: Set PIN_ZONE_4G end to 4GB - 1 page
  drm/i915: Widen return value for reservation_object_wait_timeout_rcu to long.
  drm/i915: intel_hpd_init(): Fix suspend/resume reprobing
  drm/i915: shut up gen8+ SDE irq dmesg noise, again
  drm/i915: Restore inhibiting the load of the default context
  drm/i915: Tune down rpm wakelock debug checks
  drm/i915: Avoid writing relocs with addresses in non-canonical form
  drm/i915: Move Braswell stop_machine GGTT insertion workaround

Since your main drm-next pull isn't out of the door yet I figured I might
as well flush out drm-misc instead of delaying for 4.6. It's really just
random stuff all over, biggest thing probably connector_mask tracking from
Maarten.

* tag 'topic/drm-misc-2016-01-17' of git://anongit.freedesktop.org/drm-intel: (24 commits)
  drm/fb_cma_helper: Remove implicit call to disable_unused_functions
  drm/sysfs: use kobj_to_dev()
  drm/i915: Init power domains early in driver load
  drm: Do not set connector->encoder in drivers
  apple-gmux: Add initial documentation
  drm: move MODULE_PARM_DESC to other file
  drm/edid: index CEA/HDMI mode tables using the VIC
  drm/atomic: Remove drm_atomic_connectors_for_crtc.
  drm/i915: Update connector_mask during readout, v2.
  drm: Remove opencoded drm_gem_object_release_handle()
  drm: Do not set outparam on error during GEM handle allocation
  drm/docs: more leftovers from the big vtable documentation pile
  drm/atomic-helper: Reject legacy flips on a disabled pipe
  drm/atomic: add connector mask to drm_crtc_state.
  drm/tegra: Use __drm_atomic_helper_reset_connector for subclassing connector state, v2.
  drm/atomic: Add __drm_atomic_helper_connector_reset, v2.
  drm/i915: Set connector_state->connector using the helper.
  drm: Use a normal idr allocation for the obj->name
  drm: Only bump object-reference count when adding first handle
  drm: Balance error path for GEM handle allocation
  ...

rotate_pages() doesn't modify the passed in dma addresses, so make
them const.
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452777736-4909-4-git-send-email-ville.syrjala@linux.intel.com

Just for clarity set the type for i915_ggtt_view_normal explicitly.

While at it fix the indentation fail for i915_ggtt_view_rotated.
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452777736-4909-3-git-send-email-ville.syrjala@linux.intel.com

Don't increment obj->framebuffer_references until we know we actually
managed to create the framebuffer.
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452777736-4909-2-git-send-email-ville.syrjala@linux.intel.com

The drm_fbdev_cma_init function always calls the
drm_helper_disable_unused_functions. Since it's part of the usual probe
process, all the drivers using that helper will end up having their encoder
and CRTC disable functions called at probe if their device has not been
reported as enabled.

This could be fixed by reading out from the registers the current state of
the device if it is enabled, but even that will not handle the case where
the device is actually disabled.

Moreover, the drivers using the atomic modesetting expect that their enable
and disable callback to be called when the device is already enabled or
disabled (respectively).

We can however fix this issue by moving the call to
drm_helper_disable_unused_functions out of drm_fbdev_cma_init and make the
drivers needing it (all the drivers calling drm_fbdev_cma_init and not
using the atomic modesetting) explicitly call it.
Signed-off-by: NMaxime Ripard <maxime.ripard@free-electrons.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452785109-6172-14-git-send-email-maxime.ripard@free-electrons.comAcked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Per DP spec, the source device should fall back to 18 bpp, VESA range
RGB when the sink capability is unknown. Fix the color depth
clamping. 18 bpp color depth should ensure full color range in automatic
mode.

The clamping has been HDMI specific since its introduction in

commit 996a2239
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Fri Apr 19 11:24:34 2013 +0200

    drm/i915: Disable high-bpc on pre-1.4 EDID screens

Cc: stable@vger.kernel.org
Reported-and-tested-by: NDihan Wickremasuriya <nayomal@gmail.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=105331Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: NJani Nikula <jani.nikula@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452695720-7076-1-git-send-email-jani.nikula@intel.com

Two errors in a single line. The size was read from the wrong offset,
and the end index didn't take the five bytes for sequence byte and size
of sequence into account. Fix it all, and break up the calculations a
bit to make it clearer.

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reported-and-tested-by: NMika Kahola <mika.kahola@intel.com>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Fixes: 2a33d934 ("drm/i915/bios: add support for MIPI sequence block v3")
Signed-off-by: NJani Nikula <jani.nikula@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452784327-27258-1-git-send-email-jani.nikula@intel.com

We need to set the DC FLUSH PIPE_CONTROL bit on Gen7+ to guarantee
that writes performed via the HDC are visible in memory.  Fixes an
intermittent failure in a Piglit test that writes to a BO from a
shader using GL atomic counters (implemented as HDC untyped atomics)
and then expects the memory to read back the same value after mapping
it on the CPU.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91298Tested-by: NMark Janes <mark.a.janes@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: NFrancisco Jerez <currojerez@riseup.net>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452740379-3194-1-git-send-email-currojerez@riseup.netSigned-off-by: NJani Nikula <jani.nikula@intel.com>

ilk_program_watermarks() is supposed to merge the active watermarks from
all pipes. Thus we need to use the active config too instead of some
precomputed stuff.

Fixes: aa363136 ("drm/i915: Calculate watermark configuration during atomic check (v2)")
Cc: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NMatt Roper <matthew.d.roper@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452776015-22076-2-git-send-email-ville.syrjala@linux.intel.com

ilk_compute_pipe_wm() assumes as zeroed pipe_wm structure when it
starts. We used to pass such a zeroed struct in, but this got broken
when the pipe_wm structure got embedded in the crtc state.

To fix it without too much fuzz, we need to resort to a memset().

Fixes: 4e0963c7 ("drm/i915: Calculate pipe watermarks into CRTC state (v3)")
Cc: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NMatt Roper <matthew.d.roper@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452776015-22076-1-git-send-email-ville.syrjala@linux.intel.com

Add a common function to return "on" or "off" string based on the
argument, and drop the local versions of it.

This is the onoff version of

commit 42a8ca4c
Author: Jani Nikula <jani.nikula@intel.com>
Date:   Thu Aug 27 16:23:30 2015 +0300

    drm/i915: add yesno utility function
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: NJani Nikula <jani.nikula@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1452768814-29787-1-git-send-email-jani.nikula@intel.com

this fixes the build on arm.
Signed-off-by: NDave Airlie <airlied@redhat.com>

Merge branch 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos into drm-next

 This 2nd pull request includes the following,
   - add configurable plane support and relevant cleanups.
   - fixup kernel panic issue at drm releasing.
   - remove unnecessary codes.

   This has been delayed to resolve a critical issue - which incurrs
   a kernel panic when driver is released - and review it.

* 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos:
  drm/exynos: fix kernel panic issue at drm releasing
  drm/exynos: crtc: do not wait for the scanout completion
  drm/exynos: mixer: properly update all planes on the same vblank event
  drm/exynos: crtc: rework atomic_{begin,flush}
  drm/exynos: mixer: unify a check for video-processor window
  drm/exynos: mixer: also allow ARGB1555 and ARGB4444
  drm/exynos: mixer: refactor layer setup
  drm/exynos: mixer: remove all static blending setup
  drm/exynos: mixer: set window priority based on zpos
  drm/exynos: make zpos property configurable
  drm/exynos: rename zpos to index

etnaviv fixes.

* 'drm-etnaviv-fixes' of git://git.pengutronix.de/git/lst/linux:
  drm/etnaviv: fix workaround for GC500
  drm/etnaviv: unlock on error in etnaviv_gem_get_iova()

Merge branch 'drm-rockchip-next-2016-01-06' of https://github.com/markyzq/kernel-drm-rockchip into drm-next

new rockchip bits.

* 'drm-rockchip-next-2016-01-06' of https://github.com/markyzq/kernel-drm-rockchip:
  drm: rockchip: Support Synopsys DW MIPI DSI
  Documentation: dt-bindings: Add bindings for rk3288 DW MIPI DSI driver
  drm/rockchip: return a true clock rate to adjusted_mode

When the framebuffer is a vmwgfx dma buffer and a proxy surface is
created, the vmw_kms_update_proxy() function requires that the proxy
surface width and the framebuffer pitch are compatible, otherwise
display corruption occurs as seen in gnome-shell/native with software
3D. Since the framebuffer pitch is determined by user-space, allocate
a proxy surface the width of which is based on the framebuffer pitch
rather than on the framebuffer width.

Cc: <stable@vger.kernel.org>
Reported-by: NRaphael Hertzog <buxy@kali.org>
Tested-by: NMati Aharoni <muts@kali.org>
Signed-off-by: NThomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: NBrian Paul <brianp@vmware.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

With CONFIG_SMP=n and CONFIG_DEBUG_SPINLOCK=y the vmwgfx kernel module
would unconditionally throw a bug when checking for a held spinlock
in the command buffer code. Fix this by using a lockdep check.

Cc: <stable@vger.kernel.org>
Reported-and-tested-by: NTetsuo Handa <penguin-kernel@i-love-sakura.ne.jp>
Signed-off-by: NThomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: NSinclair Yeh <syeh@vmware.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

The new debugfs initialization code fails to build when CONFIG_DEBUG_FS
is disabled:

In file included from /git/arm-soc/drivers/gpu/drm/nouveau/nouveau_drm.c:57:0:
drivers/gpu/drm/nouveau/nouveau_debugfs.h: In function 'nouveau_debugfs_init':
drivers/gpu/drm/nouveau/nouveau_debugfs.h:37:29: error: parameter name omitted
 nouveau_debugfs_init(struct nouveau_drm *)

This fixes the prototypes so we can build it again.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Fixes: b126a200 ("drm/nouveau/debugfs: we need a ctrl object for debugfs")
Signed-off-by: NDave Airlie <airlied@redhat.com>

v2: remove unneeded pci check
Signed-off-by: NKarol Herbst <nouveau@karolherbst.de>
Signed-off-by: NDave Airlie <airlied@redhat.com>

A few more misc things for radeon and amdgpu for 4.5:
- TTM fixes for imported buffers
- amdgpu fixes to avoid -ENOMEM in CS ioctl
- CZ UVD and VCE clock force options for debugging video issues
- A couple of ACP prerequisites
- Misc fixes

* 'drm-next-4.5' of git://people.freedesktop.org/~agd5f/linux:
  drm/amdgpu: validate duplicates first
  drm/amdgpu: move VM page tables to the LRU end on CS v2
  drm/ttm: add ttm_bo_move_to_lru_tail function v2
  drm/ttm: fix adding foreign BOs to the swap LRU
  drm/ttm: fix adding foreign BOs to the LRU during init v2
  drm/radeon: use kobj_to_dev()
  drm/amdgpu: use kobj_to_dev()
  drm/amdgpu/cz: force vce clocks when sclks are forced
  drm/amdgpu/cz: force uvd clocks when sclks are forced
  drm/amdgpu/cz: add code to enable forcing VCE clocks
  drm/amdgpu/cz: add code to enable forcing UVD clocks
  drm/amdgpu: fix lost sync_to if scheduler is enabled.
  drm/amd/powerplay: fix static checker warning for return meaningless value.
  drm/amdgpu: add irq domain support
  drm/amdgpu/cgs: add an interface to access PCI resources

Most VM BOs end up in the duplicates list, validate it
first make -ENOMEM less likely.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NChunming Zhou <David1.Zhou@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This makes it less likely to run into an ENOMEM because
VM page tables are evicted last.

v2: move the BOs in the LRU tail after validation
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

This allows the drivers to move a BO to the end of the LRU
without removing and adding it again.

v2: Make it more robust, handle pinned and swapable BOs as well.
Reviewed-by: NThomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

It doesn't make any sense to try to swap out imported BOs.
Reviewed-by: NThomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

If we import a BO with an external reservation object we don't
reserve/unreserve it. So we never add it to the LRU causing a possible
denial of service.

v2: fix typo in commit message
Reviewed-by: NThomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use kobj_to_dev() instead of open-coding it.
Signed-off-by: NGeliang Tang <geliangtang@163.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Use kobj_to_dev() instead of open-coding it.
Signed-off-by: NGeliang Tang <geliangtang@163.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

VCE DPM works similarly to SCLK DPM.  Add a similar interface
for VCE for forcing the VCE clocks.
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

UVD DPM works similarly to SCLK DPM.  Add a similar interface
for UVD for forcing the UVD clocks.
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Pull the code to determine the surface alignment for both linear and
tiled surfaces into a separate function intel_surf_alignment(). This
will be used not only for the vma alignment but actually aligning
the plane SURF once SKL+ starts using intel_compute_page_offset()
(since SKL+ needs >4K alignment for tiled surfaces too).
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452625717-9713-8-git-send-email-ville.syrjala@linux.intel.com

Since intel_gen4_compute_page_offset() can now handle tiling formats
all the way down to gen2, rename it to intel_compute_tile_offset().
Not that we actually use it on gen2/3 since there's no DSPSURF etc.
registers which would take a page aligned address.

v2: s/page/tile/ (Daniel)
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452625717-9713-7-git-send-email-ville.syrjala@linux.intel.com

Make intel_gen4_compute_page_offset() ready for other tiling formats
besied X-tile by getting the tile dimensions through
intel_tile_{size,width,height}().
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452625717-9713-6-git-send-email-ville.syrjala@linux.intel.com

Use the actual tile size as to compute stuff in
intel_fill_fb_ggtt_view() instead of assuming it's PAGE_SIZE. I suppose
it doesn't matter since we don't use the results on gen2 platforms
where the tile size is 2k.

v2: Update due to CbCr plane
Signed-off-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1452625717-9713-5-git-send-email-ville.syrjala@linux.intel.com