drm: bridge: analogix/dp: dynamic parse sync_pol & interlace & dynamic_range

Both hsync/vsync polarity and interlace mode can be parsed from
drm display mode, and dynamic_range and ycbcr_coeff can be judge
by the video code.

But presumably Exynos still relies on the DT properties, so take
good use of mode_fixup() in to achieve the compatibility hacks.

Reviewed-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Tested-by: Caesar Wang <wxt@rock-chips.com>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>

drm: bridge: analogix/dp: remove duplicate configuration of link rate and link count

link_rate and lane_count already configured in analogix_dp_set_link_train(),
so we don't need to config those repeatly after training finished, just
remove them out.

Beside Display Port 1.2 already support 5.4Gbps link rate, the maximum sets
would change from {1.62Gbps, 2.7Gbps} to {1.62Gbps, 2.7Gbps, 5.4Gbps}.

Tested-by: Caesar Wang <wxt@rock-chips.com>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>

drm: bridge: analogix/dp: fix some obvious code style

Fix some obvious alignment problems, like alignment and line
over 80 characters problems, make this easy to be maintained
later.

Acked-by: Jingoo Han <jingoohan1@gmail.com>
Reviewed-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Tested-by: Caesar Wang <wxt@rock-chips.com>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>

drm: bridge: analogix/dp: rename register constants

In the original split we kept the register constants intact to keep the

diff small. Still the constants are Analogix-specific, so rename them now.

Tested-by: Caesar Wang <wxt@rock-chips.com>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>

drm/exynos: dp: rename implementation specific driver part

The core functionality now resides in the generic bridge part so the
exynos-specific implementation details can get a more suitable nameing.

Tested-by: Caesar Wang <wxt@rock-chips.com>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>

drm: bridge: analogix/dp: split exynos dp driver to bridge directory

Split the dp core driver from exynos directory to bridge directory,
and rename the core driver to analogix_dp_*, rename the platform
code to exynos_dp.

Beside the new analogix_dp driver would export six hooks.
"analogix_dp_bind()" and "analogix_dp_unbind()"
"analogix_dp_suspned()" and "analogix_dp_resume()"
"analogix_dp_detect()" and "analogix_dp_get_modes()"

The bind/unbind symbols is used for analogix platform driver to connect
with analogix_dp core driver. And the detect/get_modes is used for analogix
platform driver to init the connector.

They reason why connector need register in helper driver is rockchip drm
haven't implement the atomic API, but Exynos drm have implement it, so
there would need two different connector helper functions, that's why we
leave the connector register in helper driver.

Acked-by: Inki Dae <inki.dae@samsung.com>
Tested-by: Caesar Wang <wxt@rock-chips.com>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Yakir Yang <ykk@rock-chips.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>

Merge branch 'drm-next-4.6' of git://people.freedesktop.org/~agd5f/linux into drm-next

some amd fixes
* 'drm-next-4.6' of git://people.freedesktop.org/~agd5f/linux:
  drm/radeon/mst: cleanup code indentation
  drm/radeon/mst: fix regression in lane/link handling.
  drm/amdgpu: add invalidate_page callback for userptrs
  drm/amdgpu: Revert "remove the userptr rmn->lock"
  drm/amdgpu: clean up path handling for powerplay
  drm/amd/powerplay: fix memory leak of tdp_table

Merge tag 'topic/drm-misc-2016-03-22' of git://anongit.freedesktop.org/drm-intel into drm-next

Bunch of small fixupes all over. Plus a dma-buf patch that Sumit asked me
to cherry-pick since that's the only one he had in his tree.

There's a sparse issue outstanding in the color mgr stuff, but Lionel is
still working on something that actually appeases sparse.

* tag 'topic/drm-misc-2016-03-22' of git://anongit.freedesktop.org/drm-intel:
  dma-buf/fence: fix fence_is_later v2
  dma-buf: Update docs for SYNC ioctl
  drm: remove excess description
  dma-buf, drm, ion: Propagate error code from dma_buf_start_cpu_access()
  drm/atmel-hlcdc: use helper to get crtc state
  drm/atomic: use helper to get crtc state

drm/radeon/mst: cleanup code indentation

This was all sorts of ugly from when I hacked it up,
just clean it up now and remove the extra indents.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon/mst: fix regression in lane/link handling.

The function this used changed in
    092c96a8ab9d1bd60ada2ed385cc364ce084180e
    drm/radeon: fix dp link rate selection (v2)

However for MST we should just always train to the
max link/rate. Though we probably need to limit this
for future hw, in theory radeon won't support it.

This fixes my 30" monitor with MST enabled.

Cc: stable@vger.kernel.org # v4.4
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add invalidate_page callback for userptrs

Otherwise we can run into problems with the writeback code.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Revert "remove the userptr rmn->lock"

This reverts commit c02196834456f2d5fad334088b70e98ce4967c34.

In the meantime we moved get_user_pages() outside of the reservation lock,
so that shouldn't be an issue any more

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: clean up path handling for powerplay

Use $(FULL_AMD_PATH) like everything else.

Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: fix memory leak of tdp_table

tdp_table is being leaked on failed allocations of
hwmgr->dyn_state.cac_dtp_table. kfree tdp_table on the error
return path to fix the leak.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

dma-buf/fence: fix fence_is_later v2

A fence is never later than itself. This caused a bunch of overhead for AMDGPU.

v2: simplify check as suggested by Michel.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.co.uk>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

dma-buf: Update docs for SYNC ioctl

Just a bit of wording polish plus mentioning that it can fail and must
be restarted.

Requested by Sumit.

v2: Fix them typos (Hans).

Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tiago Vignatti <tiago.vignatti@intel.com>
Cc: Stéphane Marchesin <marcheu@chromium.org>
Cc: David Herrmann <dh.herrmann@gmail.com>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Daniel Vetter <daniel.vetter@intel.com>
CC: linux-media@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Cc: linaro-mm-sig@lists.linaro.org
Cc: intel-gfx@lists.freedesktop.org
Cc: devel@driverdev.osuosl.org
Cc: Hans Verkuil <hverkuil@xs4all.nl>
Acked-by: Sumit Semwal <sumit.semwal@linaro.org>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

drm/amdgpu: release_pages requires linux/pagemap.h

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Reviewed-by: Christian König <christian.koenig@amd.com.
Signed-off-by: Dave Airlie <airlied@redhat.com>

drm: remove excess description

Description of expected_size doesn't match any parameter of the function
drm_atomic_replace_property_blob. Removing it.

Signed-off-by: Luis de Bethencourt <luisbg@osg.samsung.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1458408156-15990-1-git-send-email-luisbg@osg.samsung.com

dma-buf, drm, ion: Propagate error code from dma_buf_start_cpu_access()

Drivers, especially i915.ko, can fail during the initial migration of a
dma-buf for CPU access. However, the error code from the driver was not
being propagated back to ioctl and so userspace was blissfully ignorant
of the failure. Rendering corruption ensues.

Whilst fixing the ioctl to return the error code from
dma_buf_start_cpu_access(), also do the same for
dma_buf_end_cpu_access().  For most drivers, dma_buf_end_cpu_access()
cannot fail. i915.ko however, as most drivers would, wants to avoid being
uninterruptible (as would be required to guarrantee no failure when
flushing the buffer to the device). As userspace already has to handle
errors from the SYNC_IOCTL, take advantage of this to be able to restart
the syscall across signals.

This fixes a coherency issue for i915.ko as well as reducing the
uninterruptible hold upon its BKL, the struct_mutex.

Fixes commit c11e391da2a8fe973c3c2398452000bed505851e
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Feb 11 20:04:51 2016 -0200

    dma-buf: Add ioctls to allow userspace to flush

Testcase: igt/gem_concurrent_blit/*dmabuf*interruptible
Testcase: igt/prime_mmap_coherency/ioctl-errors
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tiago Vignatti <tiago.vignatti@intel.com>
Cc: Stéphane Marchesin <marcheu@chromium.org>
Cc: David Herrmann <dh.herrmann@gmail.com>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Daniel Vetter <daniel.vetter@intel.com>
CC: linux-media@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Cc: linaro-mm-sig@lists.linaro.org
Cc: intel-gfx@lists.freedesktop.org
Cc: devel@driverdev.osuosl.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1458331359-2634-1-git-send-email-chris@chris-wilson.co.uk

Merge branch 'drm-next-4.6' of git://people.freedesktop.org/~agd5f/linux into drm-next

A few other misc cleanups and bug fixes for 4.6.  Highlights:
- unify endian handling in powerplay
- powerplay fixes
- fix a regression in 4.5 on boards with no display connectors
- fence cleanups and locking fixes
- whitespace cleanups and code refactoring in radeon

* 'drm-next-4.6' of git://people.freedesktop.org/~agd5f/linux: (35 commits)
  drm/amdgpu/gfx7: add MTYPE definition
  drm/amdgpu: removing BO_VAs shouldn't be interruptible
  drm/amd/powerplay: show uvd/vce power gate enablement for tonga.
  drm/amd/powerplay: show uvd/vce power gate info for fiji
  drm/amdgpu: use sched fence if possible
  drm/amdgpu: move ib.fence to job.fence
  drm/amdgpu: give a fence param to ib_free
  drm/amdgpu: include the right version of gmc header files for iceland
  drm/radeon: fix indentation.
  drm/amd/powerplay: add uvd/vce dpm enabling flag to fix the performance issue for CZ
  drm/amdgpu: switch back to 32bit hw fences v2
  drm/amdgpu: remove amdgpu_fence_is_signaled
  drm/amdgpu: drop the extra fence range check v2
  drm/amdgpu: signal fences directly in amdgpu_fence_process
  drm/amdgpu: cleanup amdgpu_fence_wait_empty v2
  drm/amdgpu: keep all fences in an RCU protected array v2
  drm/amdgpu: add number of hardware submissions to amdgpu_fence_driver_init_ring
  drm/amdgpu: RCU protected amd_sched_fence_release
  drm/amdgpu: RCU protected amdgpu_fence_release
  drm/amdgpu: merge amdgpu_fence_process and _activity
  ...

drm/sti: restore mode_fixup callback

Commit 8a2fa38fddd3 removed the mode_fixup because it was empty,
but 652353e6e561 modified it to call drm_mode_set_crtcinfo()
instead.

Both commits are correct, but the merge of the two kept the nonempty
version without the reference to it, as shown by the gcc warning:

 drm/sti/sti_crtc.c:54:13: error: 'sti_crtc_mode_fixup' defined but not used

This restores the callback pointer to fix the merge.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reverts: 8a2fa38fddd3 ("drm/sti: removed optional dummy crtc mode_fixup function.")
Fixes: 652353e6e561 ("drm/sti: set CRTC modesetting parameters")
Fixes: cf481068cdd4 ("Merge branch '2016-02-26-st-drm-next' of http://git.linaro.org/people/benjamin.gaignard/kernel into drm-next")
Acked-by: Vincent ABRIOU <vincent.abriou@st.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

drm/amdgpu/gfx7: add MTYPE definition

Signed-off-by: Flora Cui <Flora.Cui@amd.com>
Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: removing BO_VAs shouldn't be interruptible

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>

drm/amd/powerplay: show uvd/vce power gate enablement for tonga.

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: show uvd/vce power gate info for fiji

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: use sched fence if possible

when preemption feature lands, the SA bo should rely on sched
fence, because hw fence will be invalid after its job preempted
or skipped.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: move ib.fence to job.fence

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: give a fence param to ib_free

thus amdgpu_ib_free() can hook sched fence to SA manager
in later patches.

BTW:
for amdgpu_free_job(), it should only fence_put() the
fence of the last ib once, so fix it as well in this patch.

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: include the right version of gmc header files for iceland

Signed-off-by: Ken Wang <Qingqing.Wang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

Merge branch '2016-02-26-st-drm-next' of git.linaro.org/people/benjamin.gaignard/kernel into drm-next

Here are sti patches for drm-next.
It brings:
  - The support of the atomic_check for the planes and minor fixes for
planes
  - The support of the vendor specific infoframe for HDMI and the
support of 2 HDMI properties related to the connector
  - The support of the DVO solving panel detection issue and timing issue.
  - The support of debugfs for connectors, encoders, crtcs and planes.

* '2016-02-26-st-drm-next' of http://git.linaro.org/people/benjamin.gaignard/kernel: (36 commits)
  drm/sti: use u32 to store DMA addresses
  drm: sti: remove sti_gem_prime_export hack
  drm/sti: add debugfs fps_show/fps_get mechanism for planes
  drm/sti: add debugfs entries for TVOUT encoders
  drm/sti: add debugfs entries for MIXER crtc
  drm/sti: add debugfs entries for VID plane
  drm/sti: add debugfs entries for HQVDP plane
  drm/sti: add debugfs entries for GDP planes
  drm/sti: add debugfs entries for CURSOR plane
  drm/sti: add debugfs entries for HDA connector
  drm/sti: add debugfs entries for DVO connector
  drm/sti: add debugfs entries for HDMI connector
  drm/sti: add hdmi_mode property for HDMI connector
  drm/sti: add colorspace property to the HDMI connector
  drm/sti: add HDMI vendor specific infoframe
  drm/sti: reset infoframe transmission when HDMI is stopped
  drm/sti: HDMI infoframe transmission mode not take into account
  drm/sti: reset HD DACS when HDA connector is created
  drm/sti: fix dvo data_enable signal
  drm/sti: adjust delay for DVO
  ...

Merge branch 'drm-next-4.6' of git://people.freedesktop.org/~agd5f/linux into drm-next

A few more fixes and cleanups for 4.6:
- DCE code cleanups
- HDP flush/invalidation fixes
- GPUVM fixes
- switch to drm_vblank_[on|off]
- PX fixes
- misc bug fixes

* 'drm-next-4.6' of git://people.freedesktop.org/~agd5f/linux: (50 commits)
  drm/amdgpu: split pipeline sync out of SDMA vm_flush() as well
  drm/amdgpu: Revert "add mutex for ba_va->valids/invalids"
  drm/amdgpu: Revert "add lock for interval tree in vm"
  drm/amdgpu: Revert "add spin lock to protect freed list in vm (v3)"
  drm/amdgpu: reserve the PD during unmap and remove
  drm/amdgpu: Fix two bugs in amdgpu_vm_bo_split_mapping
  drm/radeon: Don't drop DP 2.7 Ghz link setup on some cards.
  MAINTAINERS: update radeon entry to include amdgpu as well
  drm/amdgpu: disable runtime pm on PX laptops without dGPU power control
  drm/radeon: disable runtime pm on PX laptops without dGPU power control
  drm/amd/amdgpu: Fix indentation in do_set_base() (DCEv8)
  drm/amd/amdgpu: make afmt_init cleanup if alloc fails (DCEv8)
  drm/amd/amdgpu: Move config init flag to bottom of sw_init (DCEv8)
  drm/amd/amdgpu: Don't proceed into audio_fini if audio is disabled (DCEv8)
  drm/amd/amdgpu: Fix identation in do_set_base() (DCEv10)
  drm/amd/amdgpu: Make afmt_init cleanup if alloc fails (DCEv10)
  drm/amd/amdgpu: Move initialized flag to bottom of sw_init (DCEv10)
  drm/amd/amdgpu: Don't proceed in audio_fini if disabled (DCEv10)
  drm/amd/amdgpu: Fix indentation in dce_v11_0_crtc_do_set_base()
  drm/amd/amdgpu: Make afmt_init() cleanup if alloc fails (DCEv11)
  ...

Merge tag 'vmwgfx-next-160316' of git://people.freedesktop.org/~thomash/linux into drm-next

Pull request of 2016-03-16

* tag 'vmwgfx-next-160316' of git://people.freedesktop.org/~thomash/linux:
  drm/vmwgfx: Bump driver minor
  drm/vmwgfx: Allow the UPDATE_LAYOUT ioctl from control nodes
  drm/vmwgfx: Send a hotplug event at master_set
  drm/vmwgfx: Default to explicit crtc placement for screen targets and screen objects
  drm/vmwgfx: Calculate the cursor position based on the crtc gui origin
  drm/vmwgfx: Add connector properties to switch between explicit and implicit placement
  drm/vmwgfx: Add suggested screen x and y connector properties
  drm/vmwgfx: Add implicit framebuffer checks to the screen target code
  drm/vmwgfx: Break out implicit fb code
  drm/vmwgfx: Rework screen target page flips v2
  drm/vmwgfx: Fix screen object page flips for large framebuffers
  drm/vmwgfx: Fix a screen object framebuffer dirty corner case
  drm/vmwgfx: Add DXGenMips support

Merge tag 'drm-intel-next-fixes-2016-03-16' of anongit.freedesktop.org/git/drm-intel into drm-next

* tag 'drm-intel-next-fixes-2016-03-16' of http://anongit.freedesktop.org/git/drm-intel:
  drm/i915: Handle -EDEADLK in drm_atomic_commit from load-detect.

Merge tag 'drm/panel/for-4.6-rc1' of anongit.freedesktop.org/git/tegra/linux into drm-next

drm/panel: Changes for v4.6-rc1

This contains a refactoring of parts of the DSI core to allow creating
DSI devices from non-DSI control busses (i.e. I2C, SPI, ...).

Other than that there's support for a couple of new panels as well as
a few cleanup patches.

* tag 'drm/panel/for-4.6-rc1' of http://anongit.freedesktop.org/git/tegra/linux:
  drm/bridge: Make (pre/post) enable/disable callbacks optional
  drm/panel: simple: Add URT UMSH-8596MD-xT panels support
  dt-bindings: Add URT UMSH-8596MD-xT panel bindings
  of: Add United Radiant Technology Corporation vendor prefix
  drm/panel: simple: Support for LG lp120up1 panel
  dt-bindings: Add LG lp120up1 panel bindings
  drm/panel: simple: Fix g121x1_l03 hsync/vsync polarity
  drm/dsi: Get DSI host by DT device node
  drm/dsi: Add routine to unregister a DSI device
  drm/dsi: Try to match non-DT DSI devices
  drm/dsi: Use mipi_dsi_device_register_full() for DSI device creation
  drm/dsi: Check for CONFIG_OF when defining of_mipi_dsi_device_add()

Merge tag 'drm/tegra/for-4.6-rc1' of anongit.freedesktop.org/git/tegra/linux into drm-next

drm/tegra: Changes for v4.6-rc1

Only two cleanups this time around. One fixes reference counting of
device tree nodes, the other changes the return value of a function
from an unsigned int to an int to reflect that it will return error
codes.

* tag 'drm/tegra/for-4.6-rc1' of http://anongit.freedesktop.org/git/tegra/linux:
  gpu: host1x: Use a signed return type for do_relocs()
  gpu: host1x: bus: Add missing of_node_put()

drm/radeon: fix indentation.

I hate doing this but it hurts my eyes to go over code that does not
comply with indentation rules. Only thing that is not only space change
is in atom.c all other files are space indentation issues.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: add uvd/vce dpm enabling flag to fix the performance issue for CZ

Set the UVD and VCE DPM flags otherwise UVD and VCE DPM won't get enabled.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu: switch back to 32bit hw fences v2

We don't need to extend them to 64bits any more, so avoid the extra overhead.

v2: update commit message.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: remove amdgpu_fence_is_signaled

It's just overhead to check the fence value
when we signal them directly anyway.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: drop the extra fence range check v2

Amdgpu doesn't support using scratch registers for fences any more.
So we won't see values like 0xdeadbeef as fence value any more.

v2: reschedule timer even if no change detected

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: signal fences directly in amdgpu_fence_process

Because of the scheduler we need to signal all fences immediately
anyway, so try to avoid the waitqueue overhead.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: cleanup amdgpu_fence_wait_empty v2

Just wait for last fence instead of waiting for the sequence manually.

v2: don't use amdgpu_sched_jobs for the mask

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: keep all fences in an RCU protected array v2

Just keep all HW fences in a RCU protected array as a
first step to replace the wait queue.

v2: update commit message, move fixes into separate patch.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: add number of hardware submissions to amdgpu_fence_driver_init_ring

Make this a parameter instead of using the global variable directly.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: RCU protected amd_sched_fence_release

Fences must be freed RCU protected, otherwise the reservation_object_*_rcu()
functions can run into problems.

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

drm/amdgpu: RCU protected amdgpu_fence_release

Fences must be freed RCU protected, otherwise the reservation_object_*_rcu()
functions can run into problems.

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

drm/amdgpu: merge amdgpu_fence_process and _activity

No need to keep the two separate any more.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

drm/amdgpu: cleanup amdgpu_fence_activity

The comment about the loop counter was never valid, even when you have
multiple threads this loop only runs as long as the sequence increases.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>

gpu: host1x: Use a signed return type for do_relocs()

The return type "unsigned int" was used by the do_relocs() function
despite the fact that it will eventually return a negative error code.
Use a signed integer instead to accomodate for error codes.

This issue was detected by using the Coccinelle software.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Thierry Reding <treding@nvidia.com>

gpu: host1x: bus: Add missing of_node_put()

for_each_child_of_node() performs an of_node_get() on each iteration, so
to break out of the loop an of_node_put() is required.

Found using Coccinelle. The semantic patch used for this is as follows:

// <smpl>
@@
expression e;
local idexpression n;
@@

 for_each_child_of_node(..., n) {
   ... when != of_node_put(n)
       when != e = n
(
   return n;
|
+  of_node_put(n);
?  return ...;
)
   ...
 }
// </smpl>

Signed-off-by: Amitoj Kaur Chawla <amitoj1606@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>

drm/vmwgfx: Bump driver minor

signals availability of resolutionKMS support

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Sinclar Yeh <syeh@vmware.com>

nouveau: fix nv40_perfctr_next() cleanup regression

gcc-6 warns about code in the nouveau driver that is obviously silly:

drivers/gpu/drm/nouveau/nvkm/engine/pm/nv40.c: In function 'nv40_perfctr_next':
drivers/gpu/drm/nouveau/nvkm/engine/pm/nv40.c:62:19: warning: self-comparison always evaluats to false [-Wtautological-compare]
  if (pm->sequence != pm->sequence) {

The behavior was accidentally introduced in a patch described as "This is
purely preparation for upcoming commits, there should be no code changes here.".
As far as I can tell, that was true for the rest of that patch except for
this one function, which has been changed to a NOP.

This patch restores the original behavior.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 8c1aeaa13954 ("drm/nouveau/pm: cosmetic changes")
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Merge tag 'topic/drm-misc-2016-03-14' of git://anongit.freedesktop.org/drm-intel into drm-next

* tag 'topic/drm-misc-2016-03-14' of git://anongit.freedesktop.org/drm-intel: (27 commits)
  drm: atomic helper: do not unreference error pointer
  drm/edid: Extract SADs properly from multiple audio data blocks
  drm: fix blob pointer check
  drm: introduce pipe color correction properties
  drm/atomic: Clean up update_connector_routing.
  drm/atomic: Clean up steal_encoder, v2.
  drm/atomic: Handle encoder assignment conflicts in a separate check, v3.
  drm/atomic: Handle encoder stealing from set_config better.
  drm/atomic: Always call steal_encoder, v2.
  drm/ast: removed optional dummy crtc mode_fixup function.
  drm/bochs: removed optional dummy crtc mode_fixup function.
  drm/fsl-dcu: removed optional dummy crtc mode_fixup function.
  drm/virtio: removed optional dummy crtc mode_fixup function.
  drm/nouveau/dispnv04: removed optional dummy crtc mode_fixup function.
  drm/atmel-hlcdc: remove optional dummy crtc mode_fixup function.
  drm/sti: removed optional dummy crtc mode_fixup function.
  drm/shmobile: removed optional dummy crtc mode_fixup function.
  drm/msm/mdp: removed optional dummy crtc mode_fixup function.
  drm/omapdrm: removed optional dummy crtc mode_fixup function.
  drm/rcar-du: removed optional dummy crtc mode_fixup function.
  ...

Merge tag 'drm-amdkfd-next-fixes-2016-03-15' of git://people.freedesktop.org/~gabbayo/linux into drm-next

* tag 'drm-amdkfd-next-fixes-2016-03-15' of git://people.freedesktop.org/~gabbayo/linux:
  drm/amdkfd: uninitialized variable in dbgdev_wave_control_set_registers()

drm/omap: fix panel/encoder probes

The recent changes which removed platform data support from panels &
encoders had a few mistakes, causing probes of DVI connector and DSI
command mode panels to fail every time due to missing '!'. Fix the
if()s.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Reported-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

drm/atmel-hlcdc: use helper to get crtc state

DRM core provide helper to access crtc state.

Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1458045988-25246-1-git-send-email-a.hajda@samsung.com

drm/atomic: use helper to get crtc state

DRM core provide helper to access crtc state.

Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1458045960-25193-1-git-send-email-a.hajda@samsung.com

drm/amdkfd: uninitialized variable in dbgdev_wave_control_set_registers()

At the end of the function we expect "status" to be zero, but it's
either -EINVAL or uninitialized.

Fixes: 788bf83db301 ('drm/amdkfd: Add wave control operation to debugger')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

Merge tag 'drm-vc4-next-2016-03-14' of github.com:anholt/linux into drm-next

This pull request covers what's left for 4.6.  Notably, it includes a
significant 3D performance improvement and a fix to HDMI hotplug
detection for the Pi2/3.

* tag 'drm-vc4-next-2016-03-14' of github.com:anholt/linux:
  drm/vc4: Recognize a more specific compatible string for V3D.
  dt-bindings: Add binding docs for V3D.
  drm/vc4: Return -EFAULT on copy_from_user() failure
  drm/vc4: Respect GPIO_ACTIVE_LOW on HDMI HPD if set in the devicetree.
  drm/vc4: Let gpiolib know that we're OK with sleeping for HPD.
  drm/vc4: improve throughput by pipelining binning and rendering jobs

drm/amdgpu: move fence structure into amdgpu_fence.c

No need to have that in the header file any more.

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>

drm/amdgpu: remove amdgpu_fence_wait_next

Not used any more.

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>

drm/amdgpu: remove amdgpu_ring_from_fence

Not used any more.

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>

drm/amdgpu: stop using the ring index in the SA

The ring index will always collide as hash into the fence list, so use
the context number instead. That can still cause collisions, but they
are less likely than using ring indices.

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>

drm/amdgpu: stop waiting on UVD messages before mapping them

amdgpu_bo_kmap() now always waits for moves to finish.

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>

drm/amdgpu: always wait before kmap a BO

When a BO is currently moving we otherwise would blindly
access the new location without checking.

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>

drm/radeon: refactor SI tiling table initialization

Simplify the control flow of si_tiling_mode_table_init() similar to how
it was done in gfx_v7_0.c and gfx_v8_0.c.

Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: refactor CIK tiling table initialization

Simplify the control flow of cik_tiling_mode_table_init() similar to how
it was done in gfx_v7_0.c and gfx_v8_0.c.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: allow write access to mapped userptrs

With the updated MMU notifier we should also be able to
handle the writeback case correctly.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: use pp_endian.h for Tonga

Drop local versions of these macros.

Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Ken Wang <Qingqing.Wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: use pp_endian.h for Fiji

Drop local versions of these macros.

Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Ken Wang <Qingqing.Wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: add a common pp endian header

To replace the duplicated versions of this in all asic
variants.

Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Ken Wang <Qingqing.Wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/powerplay: mv avfs status to smumgr.h

Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: rework fbdev handling on chips with no connectors

Move all the logic to radeon_fb.c and add checks to functions
called frome elsewhere.

bug:
https://bugzilla.kernel.org/show_bug.cgi?id=112781

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/vmwgfx: Allow the UPDATE_LAYOUT ioctl from control nodes

On vmware there is a daemon telling the KMS system about the GUI layout.
Typically it talks to the X server but in the absence of an X server or if
there are multiple, it wants to talk directly to the vmwgfx kernel module.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>

drm/vmwgfx: Send a hotplug event at master_set

Make sure drm clients (mostly the X server) are communicated the current
layout when switched in.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Default to explicit crtc placement for screen targets and screen objects

Enables using multiple framebuffers. For legacy display units,
explicit crtc placement is not supported due to hardware limitations.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Calculate the cursor position based on the crtc gui origin

Base the cursor position on the coordinate of the crtc origin in the
gui coordinate system rather than in the framebuffer coordinate system.

With explicit placement, these may differ (for example when two crtcs
scan out of the same framebuffer location).

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Add connector properties to switch between explicit and implicit placement

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Add suggested screen x and y connector properties

Introduced by qxl, add these properties as a generic way to tell a
display manager about the GUI layout.

Also add the hotplug_mode_update_property which advises display managers to
reread the mode list on a hotplug event.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>

drm/vmwgfx: Add implicit framebuffer checks to the screen target code

Just like for screen objects, make sure we use only a single framebuffer
for implicit placement.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Break out implicit fb code

Preparation for supporting explicit fbs for screen objects and screen
targets.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Rework screen target page flips v2

Gnome-Shell / Wayland assumes that page-flips can be done on a crtc
regardless of framebuffer size and the crtc position within the
framebuffer.

Therefore rework the screen target code to correctly handle changes in
framebuffer size and content_fb_type. Also make sure that we update
the screen target correctly when the content_fb_type is not
SAME_AS_DISPLAY.

This commit breaks out the framebuffer binding code from crtc_set so it
can be used both from page_flip() and crtc_set() and reworks those
functions a bit to be more robust.

v2: Address review comments by Sinclair Yeh.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Fix screen object page flips for large framebuffers

For page flips the framebuffer may be much larger than the crtc
scanout area and may be attached to multiple crtcs.
When flipping a crtc, make sure we dirty only that crtc's area of the
framebuffer.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Fix a screen object framebuffer dirty corner case

If there are no cliprects for a particular crtc, an invalid command would
have been generated. If that's the case, instead ditch the generated
command sequence.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>

drm/vmwgfx: Add DXGenMips support

Add support for DXGenMips command.

Signed-off-by: Charmaine Lee <charmainel@vmware.com>
Reviewed-by: Sinclair Yeh <syeh@vmware.com>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>

drm/i915: Handle -EDEADLK in drm_atomic_commit from load-detect.

CI runs with DEBUG_WW_MUTEX_SLOWPATH, so -EDEADLK occurs a lot more.
Handle the case where drm_atomic_commit fails with -EDEADLK correctly.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/56D3FEF1.6070306@linux.intel.com
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
(cherry picked from commit 3ba86073edcbe2be53d9862d5a3098f0ebf8ae9a)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>

drm/vc4: Recognize a more specific compatible string for V3D.

The Raspberry Pi Foundation's firmware updates are shipping device
trees using the old string, so we'll keep recognizing that as this rev
of V3D.  Still, we should use a more specific name in the upstream DT
to clarify which board is being supported, in case we do other revs of
V3D in the future.

Signed-off-by: Eric Anholt <eric@anholt.net>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>

dt-bindings: Add binding docs for V3D.

This was missed in the upstreaming process.

Signed-off-by: Eric Anholt <eric@anholt.net>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>

Merge branch 'linux-4.6' of git://github.com/skeggsb/linux into drm-next

- GM20x secure boot support (hence, acceleration, finally \o/)
- GM200 support
- GM20B clock driver
- Support for power sensors on some GPUs
- Various other fixes all over the place

* 'linux-4.6' of git://github.com/skeggsb/linux: (95 commits)
  drm/nouveau/clk/gm20b: add basic driver
  drm/nouveau/clk/gk20a: share reusable structures/functions
  drm/nouveau/clk/gk20a: set lowest frequency during init()
  drm/nouveau/clk/gk20a: split gk20a_clk_new()
  drm/nouveau/clk/gk20a: abstract pl_to_div
  drm/nouveau/clk/gk20a: put mnp values into their own struct
  drm/nouveau/clk/gk20a: emit parent rate as debug message
  drm/nouveau/clk/gk20a: only restore divider to 1:1 if needed
  drm/nouveau/clk/gk20a: only compute n_lo if needed
  drm/nouveau/clk/gk20a: fix VCO bit mask
  drm/nouveau/clk/gk20a: rename enable/disable functions
  drm/nouveau/clk/gk20a: reorganize variables in gk20a_pllg_calc_mnp()
  drm/nouveau/clk/gk20a: convert parameters to Khz
  drm/nouveau/volt: add GM20B driver
  drm/nouveau/volt/gk20a: split constructor
  drm/nouveau/volt/gk20a: share reusable members & functions
  drm/nouveau/ce/gm107: expose MaxwellDmaCopyA
  drm/nouveau/fifo/gm107: KeplerChannelGpfifoB, and 2048 channels
  drm/nouveau/fifo/gk110: expose KeplerChannelGpfifoB
  drm/nouveau/fifo/gk104: submit NOP after all PBDMA_INTR_0, not just DEVICE
  ...

drm/nouveau/clk/gm20b: add basic driver

Add a basic clock driver that reuses the GK20A logic.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: share reusable structures/functions

Make functions/structures that the GM20B driver will reuse public.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: set lowest frequency during init()

Err on the safe side by setting the lowest frequency (and thus voltage)
during device init.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: split gk20a_clk_new()

This allows to instanciate drivers that use the same logic as gk20a with
different parameters.

Add a constructor function to allow other chips that inherit from this
clock to easily initialize its members

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: abstract pl_to_div

pl_to_div may be done differently depending on the chip. Abstract this
operation so the same logic can be reused for them as well.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: put mnp values into their own struct

This allows us to read them using one single function and will be handy
to the GM20B driver.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: emit parent rate as debug message

Most users are probably not interested in this information.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: only restore divider to 1:1 if needed

Only restore the 1:1 divider if it is not set already. Also use the
proper masks for this operation and add a second write as done in the
Android code.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: only compute n_lo if needed

n_lo is used if we are going to slide. Compute it only if that condition
succeeds to avoid confusion about future usage of this computation.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: fix VCO bit mask

Fix the mask specified to switch to VCO mode was given as an (incorrect)
immediate value. Although the side-effect happens to be the same, this
is clearly incorrect.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/clk/gk20a: rename enable/disable functions

gk20a_pllg_disable() is only used in the context of gk20a_clk_fini().
Move its body there and rename _gk20a_pllg_enable() and
_gk20a_pllg_disable() to non-underscored versions.

Signed-off-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>