Big ticket items are hdmi support for 8996 (aka snapdragon 820), and
adreno 430 support.  Also one more small uapi addition to support
timestamp queries.

* 'msm-next' of git://people.freedesktop.org/~robclark/linux: (29 commits)
  drm/msm: rename hdmi symbols
  drm/msm/adreno: remove duplicate adreno_hw_init() call
  drm/msm: add timestamp param
  drm/msm: fix small typo
  drm/msm: grab struct_mutex after allocating submit
  drm/msm: reject submit ioctl if no gpu
  drm/msm/adreno: print details in case of a protect fault interrupt
  drm/msm/adreno: get CP_RPTR from register instead of shadow memory
  drm/msm/adreno: add adreno430 power control
  drm/msm/adreno: support for adreno 430.
  drm/msm: update generated headers
  drm/msm/dsi: fix definition of msm_dsi_pll_28nm_8960_init()
  drm/msm/dsi: Parse DSI lanes via DT
  drm/msm/dsi: Drop VDD regulator for MSM8916
  drm/msm/dsi: Remove incorrect warning on host attach
  drm/msm: Free fb helper resources in msm_unload
  drm/msm/mdp: Detach iommu in mdp4_destroy
  drm/msm: make iommu port names const'ier
  drm/msm/mdp: Use atomic helper to set crtc property
  dt-bindings: msm/hdmi: Add HDMI PHY bindings
  ...

Global symbols in the kernel should be prefixed by the name
of the subsystem and/or driver to avoid conflicts when all
code is built-in.

In this case, function names like 'hdmi_register' or 'hdmi_set_mode'
are way too generic for an MSM specific DRM driver, so I'm renaming
them all to msm_hdmi_* here.

I also rename a lot of the 'static' symbols along with the global
names for consistency, even though those are relatively harmless;
they might only be slightly confusing when they show up in
backtraces.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Rob Clark <robdclark@gmail.com>

Not sure where it came from, but seem unintentional.  And also not
needed on a420, so let's just drop it.
Signed-off-by: Rob Clark <robdclark@gmail.com>

We need this for GL_TIMESTAMP queries.

Note: currently only supported on a4xx.. a3xx doesn't have this
always-on counter.  I think we could emulate it with the one CP
counter that is available, but for now it is of limited usefulness
on a3xx (since we can't seem to do time-elapsed queries in any sane
way with the existing firmware on a3xx, and if you are trying to do
profiling on a tiler you want time-elapsed).  We can add that later
if it becomes useful.
Signed-off-by: Rob Clark <robdclark@gmail.com>

Signed-off-by: Rob Clark <robdclark@gmail.com>

No real need to hold the lock over allocation, and simplifies things
slightly if we change the order.
Signed-off-by: Rob Clark <robdclark@gmail.com>

Existing userspace wouldn't get this far, since getparam ioctl would
have failed and it would have bailed out creating a screen/context.

But all the same, we shouldn't let evil or confused userspace cause a
null ptr deref.
Signed-off-by: Rob Clark <robdclark@gmail.com>

Signed-off-by: Craig Stout <cstout@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

As described in the downstream/kgsl driver:
Sometimes the RPTR shadow memory is unreliable causing timeouts
in adreno_idle().  Read it directly from the register instead.
Signed-off-by: Craig Stout <cstout@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

Signed-off-by: Craig Stout <cstout@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

Signed-off-by: Craig Stout <cstout@chromium.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

Pull in additional regs needed for a430, etc.
Signed-off-by: Rob Clark <robdclark@gmail.com>

This fixes the following build failure:

drivers/gpu/drm/msm/dsi/pll/dsi_pll_28nm.o: In function `msm_dsi_pll_28nm_8960_init':
dsi_pll_28nm.c:(.text+0x1198): multiple definition of `msm_dsi_pll_28nm_8960_init'
drivers/gpu/drm/msm/dsi/pll/dsi_pll.o:dsi_pll.c:(.text+0x0): first defined here
Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
Acked-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

The DSI driver is currently unaware of how the DSI physical data lanes
are mapped to the logical lanes provided by the DSI controller.

Create a DT binding "qcom,data-lane-map" that provides this information
on a given platform.

The MSM DSI controller is restricted in terms of what all mappings
it can support. The lane polarity is fixed for all the lanes, the clock
lanes are fixed, and the data lanes can be swapped among each other only
for a few combinations. Apply these restrictions when we parse the DT
data.

Cc: devicetree@vger.kernel.org
Cc: Rob Herring <robh@kernel.org>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Acked-by: Rob Herring <robh@kernel.org>

VDD regulator input was specified for MSM8916. It turns our that this
regulator is used for the display panels used on MSM8916 platforms, but
not the DSI controller itself. Drop this regulator from the list.
Reported-by: Vinay Simha <vinaysimha@inforcecomputing.com>
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

With the implementation of of_graph parsing, it isn't any longer
necessary for msm_host->device node to be same as dsi->dev.of_node. This
only holds true when the connected device is also a child of the dsi_host.

In the case of external bridge chips belonging to a different control
bus, these are guaranteed to be different.
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

We have a msm_fbev_free function to uninit fb_helper stuff, but we aren't
using it. Call it in msm_unload.
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

attach_dev gets called in mdp4_kms_init, but there is no corresponding
detach_dev called in the error path or in the kms driver unload path.

Detach and destroy mmu in mdp4_destroy.
Signed-off-by: Sricharan R <sricharan@codeaurora.org>
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>

Signed-off-by: Rob Clark <robdclark@gmail.com>

Assign drm_atomic_helper_crtc_set_property helper to MDP4 and MDP5
crtcs' set_property ops. This replaces the custom funcs that
returned an error even for standard crtc properties.
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Rob Clark <robdclark@gmail.com>

Add HDMI PHY bindings. Update the example to use HDMI PHY.

Added a missing power-domains property in the HDMI core bindings. Also,
simplified HDMI TX's DT node name in the example.

Cc: devicetree@vger.kernel.org
Cc: Rob Herring <robh@kernel.org>
Signed-off-by: Archit Taneja <architt@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Acked-by: Rob Herring <robh@kernel.org>

imx-drm vblank IRQ control, fence support, and of endpoint helpers

- Add and make use of drm_of_active_endpoint helpers
- Silence a noisy dev_info into a dev_dbg
- Stop touching primary fb on pageflips
- Track flip state explicitly
- Keep GEM buffer objects referenced while scanout is active
- Implement fence sync by deferring flips to a workqueue for
  dma-bufs with pending fences
- Actually disable vblank IRQs while they are not needed

* tag 'imx-drm-next-20160301' of git://git.pengutronix.de/git/pza/linux:
  drm/imx: only enable vblank IRQs when needed
  drm/imx: implement fence sync
  drm/imx: keep GEM object referenced as long as scanout is active
  drm/imx: track flip state explicitly
  drm/imx: don't touch primary fb on pageflip
  drm/imx: ipuv3 plane: Replace dev_info with dev_dbg if a plane's CRTC changes
  gpu: ipu-v3: ipu-dc: Simplify display controller microcode setup
  drm/rockchip: remove rockchip_drm_encoder_get_mux_id
  drm/imx: remove imx_drm_encoder_get_mux_id
  drm: add drm_of_encoder_active_endpoint helpers

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

   Summary:
   - Add Exynos5420 SoC support to FIMD driver.
     . This patch makes MIC(Mobile Image Compressor) IP to be bypassed in default
       in case of Exynos5420 and later. The Display pipe line configuraion for
       Exynos DRM driver will be considered through of graph concept later.
   - Add Exynos5422 SoC support to MIPI-DSI driver.
     . Exynos5422 SoC is similar to Exynos5433 SoC but software reset is different
       each other so this patch consideres the difference.
   - Get more precise clock divider value of FIMD controller.
     . This patch changes DIV_ROUND_CLOSEST macro to be used instead of DIV_ROUND_UP.
   - Refactor Exynos DRM device and driver registeration.
     . This patch makes Exynos DRM driver to be easy-to-read and at the same time,
       cleans it up by removing #ifdef ~ #endif things.
   - Configure DMA-mapping address space common to Exynos DRM devices in more generic
     without any hacks.
   - some fixups and cleanups.

* 'exynos-drm-next' of git://git.kernel.org/pub/scm/linux/kernel/git/daeinki/drm-exynos: (26 commits)
  drm/exynos/dsi: use core helper to create DSI packet
  drm/exynos: use real device for DMA-mapping operations
  drm/exynos: refactor driver and device registration code
  drm/exynos: use arch independent types in uapi header
  drm/exynos: remove platform data structures and include/drm/exynos_drm.h
  drm/exynos/fimc: remove unused camera interface polarization code
  drm/exynos: remove struct exynos_drm_panel_info
  drm/exynos: add exynos5420 support for fimd
  drm/exynos: use DIV_ROUND_CLOSEST to find the closest div
  drm/exynos: remove incorrect ccflags from Makefile
  drm/exynos/decon: make irq handler static
  drm/exynos/hdmi: remove unused variable
  drm/exynos/dsi: constify read only structures
  drm/exynos/dsi: replace registry access macros with functions
  drm/exynos: support exynos5422 mipi-dsi
  drm/exynos/decon: fix disable clocks order
  drm/exynos: fix incorrect cpu address for dma_mmap_attrs()
  drm/exynos: exynos5433_decon: fix wrong state in decon_vblank_enable
  drm/exynos: exynos5433_decon: fix wrong state assignment in decon_enable
  drm/exynos: dsi: restore support for drm bridge
  ...

Core provides generic helper to create DSI packet, use it instead of
custom code.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

This patch changes device pointer provided to all calls to DMA-mapping
subsystem from the virtual exynos-drm 'device' to the real device pointer
of one of the CRTC devices (decon, fimd or mixer). This way no more hacks
will be needed to configure proper DMA-mapping address space on the common
virtual exynos-drm device. This change also removes the need for some
hacks in IOMMU related code. It also finally solves the problem of Exynos
DRM driver not working on ARM64 architecture, which provides noop-based
DMA-mapping operations for virtual platform devices.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

This patch refactors driver and device registration by moving all drivers
to the common array. This way additional flags can be added later for
new features. #ifdef-based code has been replaced by IS_ENABLED() macro
usage.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

User API structs should not use types which size/alignment/padding depends
on architecture. The patch fixes it for all structures except
drm_exynos_g2d_userptr, as g2d related stuff seems to be more complicated
and will be reviewed/adjusted later.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

Platform data structures are not used for long time so the whole header
file can be safely removed.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

Polarization was never configured for DRM-FIMC device, so fimc_set_polarity
function did nothing. In fact DRM does not use camera interface so there
is no point in configuring it.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

struct exynos_drm_panel_info is not used anymore, except exynos_dp,
which can integrate useful fields directly into its context.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

This patch adds a exynos5420 driver data to support mic_bypass
option to bypass the mic from display out path.
The mic(Mobile image compressor) compresses RGB data from fimd
and send the compressed data to the mipi dsi.
The bypass option can be founded from system register and the bit
is 11. The option bit has been introduced since exynos5420. The
only difference between exynos5250 and exynos5420/exynos5422 is
existence of the bit. Until the MIC is defined and enabled from
device tree, the bypass mic will be default option.
Signed-off-by: Chanho Park <chanho61.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

This patch uses DIV_ROUND_CLOSEST instead of DIV_ROUND_UP
The DIV_ROUND_CLOSEST can be used to find the closest integer
value when we divide some integers.
Signed-off-by: Chanho Park <chanho61.park@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

Include directories are provided by core already, adding them in driver
is redundand and causes warnings in case of out-of-tree build.

v2:
    - fixed include in exynos_drm_iommu.c
    - typo in commit message
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

The function is used only locally.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

The variable is unused for long time.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

All global variables are read only.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

Functions are preferred over macros as more type-safe.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

This patch supports mipi dsi for exynos5422. The dsi register
offsets of the exynos5422 are similar with exynos5433. However,
the values of the registers are quite different from the
exynos5433. For example, the exynos5422 uses sw reset like
previous chips.
Signed-off-by: Chanho Park <chanho61.park@samsung.com>
Reviewed-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

Decon requires that clocks should be disabled in reverse order. Otherwise
system hangs.
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>

dma_mmap_attrs() should be called with cpu address returned by
dma_alloc_attrs(). Existing code however passed pages array base as cpu
address. This worked only by a pure luck on ARM architecture. This patch
fixes this issue.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>