drm/virtio: set up virtqueue sgs before locking

sgs setup does not need to be in the critical section.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205181955.202485-7-olvaffe@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

drm/virtio: unlock object array on errors

We don't propagate erros to the callers.  We have to unlock object
arrays on errors.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205181955.202485-6-olvaffe@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

drm/virtio: no need to pass virtio_gpu_ctrl_hdr

We can get it from vbuf.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205181955.202485-5-olvaffe@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

drm/virtio: add virtio_gpu_vbuf_ctrl_hdr

It is a helper to return the virtio_gpu_ctrl_hdr in a vbuf.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205181955.202485-4-olvaffe@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

drm/virtio: remove incorrect ENOSPC check

The handling of virtqueue_add_sgs ENOSPC error is incorrect because
it can result in out-of-order virtqueue_add_sgs and break fences.
We never get ENOSPC anyway because the caller waits until there is
enough space (the other caller that did not wait was removed in
commit 32d6c2c5b522 ("drm/virtio: Rewrite virtio_gpu_queue_ctrl_buffer
using fenced version.")).  Remove the incorrect and unnecessary error
path.

This also adds a WARN_ON(ret) until we properly handle errors.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Cc: David Riley <davidriley@chromium.org>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205181955.202485-3-olvaffe@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

drm/virtio: fix a wait_event condition

outcnt may be greater than 1 since commit e1218b8c0cc1 ("drm/virtio:
Use vmalloc for command buffer allocations.").

Fixes: e1218b8c0cc1 ("drm/virtio: Use vmalloc for command buffer allocations.")
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Cc: David Riley <davidriley@chromium.org>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205181955.202485-2-olvaffe@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

drm/virtio: ratelimit error logging

Avoid flooding the log in case we screw up badly.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205102552.21409-1-kraxel@redhat.com

drm/vboxvideo: Implement struct drm_mode_config_funcs.mode_valid

The implementation of struct drm_mode_config_funcs.mode_valid verifies
that enough video memory is available for a given display mode.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200203155258.9346-5-tzimmermann@suse.de

drm/hibmc: Implement struct drm_mode_config_funcs.mode_valid

The implementation of struct drm_mode_config_funcs.mode_valid verifies
that enough video memory is available for a given display mode.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200203155258.9346-4-tzimmermann@suse.de

drm/bochs: Implement struct drm_mode_config_funcs.mode_valid

The implementation of struct drm_mode_config_funcs.mode_valid verifies
that enough video memory is available for a given display mode. This
replaces similar functionality in bochs_connector_mode_valid().

v2:
* remove bochs_connector_mode_valid(), which now serves no purpose

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200203155258.9346-3-tzimmermann@suse.de

drm/vram: Add helpers to validate a display mode's memory requirements

Devices with low amount of dedicated video memory may not be able
to use all possible display modes, as the framebuffers may not fit
into VRAM. The new helper function drm_vram_helper_mode_valid()
implements a simple test to sort out all display modes that can
not be used in any case. Drivers should call this function from
struct drm_mode_config_funcs.mode_valid.

The functionality was originally implemented by the ast driver, which
is being converted as well.

v2:
* WARN_ON if VRAM memory manager has not been initialized
* documentation fixes
* unexported drm_vram_helper_mode_valid_internal()

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200203155258.9346-2-tzimmermann@suse.de

drm/virtio: fix vblank handling

virtio has its own commit fail function.  Add the
drm_atomic_helper_fake_vblank() call there.

Fixes: 2a735ad3d211 ("drm/virtio: Remove sending of vblank event")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: http://patchwork.freedesktop.org/patch/msgid/20200205065312.15790-1-kraxel@redhat.com

drm: Remove the dma_alloc_coherent wrapper for internal usage

Internally for "consistent" maps, we create a temporary struct
drm_dma_handle in order to use our own dma_alloc_coherent wrapper then
destroy the temporary wrap. Simplify our logic by removing the temporary
wrapper!

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200202171635.4039044-2-chris@chris-wilson.co.uk

drm: Remove PageReserved manipulation from drm_pci_alloc

drm_pci_alloc/drm_pci_free are very thin wrappers around the core dma
facilities, and we have no special reason within the drm layer to behave
differently. In particular, since

commit de09d31dd38a50fdce106c15abd68432eebbd014
Author: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Date:   Fri Jan 15 16:51:42 2016 -0800

    page-flags: define PG_reserved behavior on compound pages

    As far as I can see there's no users of PG_reserved on compound pages.
    Let's use PF_NO_COMPOUND here.

it has been illegal to combine GFP_COMP with SetPageReserved, so lets
stop doing both and leave the dma layer to its own devices.

Reported-by: Taketo Kabe
Bug: https://gitlab.freedesktop.org/drm/intel/issues/1027
Fixes: de09d31dd38a ("page-flags: define PG_reserved behavior on compound pages")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v4.5+
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200202171635.4039044-1-chris@chris-wilson.co.uk

drm/stm: dsi: stm mipi dsi doesn't print error on probe deferral

Change DSI driver to not print an error trace when probe
is deferred for a clock resource.

Signed-off-by: Etienne Carriere <etienne.carriere@st.com>
Reviewed-by: Philippe Cornu <philippe.cornu@st.com>
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@st.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1579602245-7577-1-git-send-email-yannick.fertre@st.com

drm/stm: ltdc: check crtc state before enabling LIE

Following investigations of a hardware bug, the LIE interrupt
can occur while the display controller is not activated.
LIE interrupt (vblank) don't have to be set if the CRTC is not
enabled.

Signed-off-by: Yannick Fertre <yannick.fertre@st.com>
Acked-by: Philippe Cornu <philippe.cornu@st.com>
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@st.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1579601650-7055-1-git-send-email-yannick.fertre@st.com

drm/stm: ltdc: add number of interrupts

The number of interrupts depends on the ltdc version.
Don't try to get interrupt which not exist, avoiding
kernel warning messages.

Signed-off-by: Yannick Fertre <yannick.fertre@st.com>
Acked-by: Philippe Cornu <philippe.cornu@st.com>
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@st.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1579601632-7001-1-git-send-email-yannick.fertre@st.com

MAINTAINERS: Add Thomas as drm-misc co-maintainer

Daniel asked me to serve as co-maintainer of the drm-misc tree.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Maxime Ripard <mripard@kernel.org>
Acked-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200130120643.5759-1-tzimmermann@suse.de

drm/panel: simple: Fix the lt089ac29000 bus_format

The lt089ac29000 panel is an LVDS panel, not a DPI one. Fix the
definition to reflect this fact.

v10:
* Add changelog to the commit message

v8 -> v9:
* No changes

v7:
* New patch

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Suggested-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-12-boris.brezillon@collabora.com

drm/imx: pd: Use bus format/flags provided by the bridge when available

Now that bridges can expose the bus format/flags they expect, we can
use those instead of the relying on the display_info provided by the
connector (which is only valid if the encoder is directly connected
to bridge element driving the panel/display).

We also explicitly expose the bus formats supported by our encoder by
filling encoder->output_bus_caps with proper info.

v10:
* Add changelog to the commit message
* Use kmalloc() instead of kcalloc()
* Add a dev_warn() when unsupported flags are requested

v8 -> v9:
* No changes

v7:
* Add an imx_pd_format_supported() helper (suggested by Philipp)
* Simplify imx_pd_bridge_atomic_get_output_bus_fmts() (suggested by Philipp)
* Simplify imx_pd_bridge_atomic_get_input_bus_fmts()
* Explicitly set the duplicate/destro_state() and reset() hooks

v4 -> v6:
* Patch was not part of the series

v3 (all suggested by Philipp):
* Adjust to match core changes
* Propagate output format to input format
* Pick a default value when output_fmt = _FIXED
* Add missing BGR888 and GBR888 fmts to imx_pd_bus_fmts[]

v2:
* Adjust things to match the new bus-format negotiation infra

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-8-boris.brezillon@collabora.com

drm/bridge: Add the necessary bits to support bus format negotiation

drm_bridge_state is extended to describe the input and output bus
configurations. These bus configurations are exposed through the
drm_bus_cfg struct which encodes the configuration of a physical
bus between two components in an output pipeline, usually between
two bridges, an encoder and a bridge, or a bridge and a connector.

The bus configuration is stored in drm_bridge_state separately for
the input and output buses, as seen from the point of view of each
bridge. The bus configuration of a bridge output is usually identical
to the configuration of the next bridge's input, but may differ if
the signals are modified between the two bridges, for instance by an
inverter on the board. The input and output configurations of a
bridge may differ if the bridge modifies the signals internally,
for instance by performing format conversion, or*modifying signals
polarities.

Bus format negotiation is automated by the core, drivers just have
to implement the ->atomic_get_{output,input}_bus_fmts() hooks if they
want to take part to this negotiation. Negotiation happens in reverse
order, starting from the last element of the chain (the one directly
connected to the display) up to the first element of the chain (the one
connected to the encoder).
During this negotiation all supported formats are tested until we find
one that works, meaning that the formats array should be in decreasing
preference order (assuming the driver has a preference order).

Note that the bus format negotiation works even if some elements in the
chain don't implement the ->atomic_get_{output,input}_bus_fmts() hooks.
In that case, the core advertises only MEDIA_BUS_FMT_FIXED and lets
the previous bridge element decide what to do (most of the time, bridge
drivers will pick a default bus format or extract this piece of
information from somewhere else, like a FW property).

v10:
* Add changelog to the commit message

v9:
* No changes

v8:
* Fix a test in drm_atomic_bridge_chain_select_bus_fmts() (Reported by
  Jonas)

v7:
* Adapt the code to deal with the fact that not all bridges in the
  chain have a bridge state

v5 -> v6:
* No changes

v4:
* Enhance the doc
* Fix typos
* Rename some parameters/fields
* Reword the commit message

v3:
* Fix the commit message (Reported by Laurent)
* Document the fact that bus formats should not be directly modified by
  drivers (Suggested by Laurent)
* Document the fact that format order matters (Suggested by Laurent)
* Propagate bus flags by default
* Document the fact that drivers can tweak bus flags if needed
* Let ->atomic_get_{output,input}_bus_fmts() allocate the bus format
  array (Suggested by Laurent)
* Add a drm_atomic_helper_bridge_propagate_bus_fmt()
* Mandate that bridge drivers return accurate input_fmts even if they
  are known to be the first element in the bridge chain

v2:
* Rework things to support more complex use cases

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
[narmstrong: fixed doc in include/drm/drm_bridge.h:69 fmt->format]
Reviewed-by: Jernej Skrabec <jernej.skrabec@siol.net>
Tested-by: Jonas Karlman <jonas@kwiboo.se>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-7-boris.brezillon@collabora.com

drm/bridge: Add an ->atomic_check() hook

So that bridge drivers have a way to check/reject an atomic operation.
The drm_atomic_bridge_chain_check() (which is just a wrapper around
the ->atomic_check() hook) is called in place of
drm_bridge_chain_mode_fixup() (when ->atomic_check() is not implemented,
the core falls back on ->mode_fixup(), so the behavior should stay
the same for existing bridge drivers).

v10:
* Add changelog to the commit message

v8 -> v9:
* No changes

v7:
* Fix a NULL pointer dereference

v5 -> v6:
* No changes

v4:
* Add R-bs

v3:
* No changes

v2:
* Clarify the fact that ->atomic_check() is replacing ->mode_fixup()

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed by: Jernej Skrabec <jernej.skrabec@siol.net>
Tested-by: Jonas Karlman <jonas@kwiboo.se>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-6-boris.brezillon@collabora.com

drm/bridge: Patch atomic hooks to take a drm_bridge_state

This way the drm_bridge_funcs interface is consistent with the rest of
the subsystem.

The drivers implementing those hooks are patched too.

v10:
* Add changelog to the commit message

v8 -> v9:
* No changes

v7:
* Adjust things to the bridge_state changes

v6:
* Also fixed rcar-du/rcar_lvds.c same as analogix/analogix_dp_core.c

v5:
* No changes

v4:
* Rename func params into old_bridge_state
* Add Laurent's Rb

v3:
* Old state clarification moved to a separate patch

v2:
* Pass the old bridge state

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
[narmstrong: renamed state as old_bridge_state in rcar_lvds_atomic_disable]
Reviewed by: Jernej Skrabec <jernej.skrabec@siol.net>
Tested-by: Jonas Karlman <jonas@kwiboo.se>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-5-boris.brezillon@collabora.com

drm/bridge: analogix: Plug atomic state hooks to the default implementation

This is needed to pass a bridge state to all atomic hooks, if we don't
do that, the core can't duplicate/create bridge states.

v10:
* Add changelog to the commit message

v9:
* Add Neil's R-b
* Move earlier in the series

v8:
* No changes

v7:
* New patch

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-4-boris.brezillon@collabora.com

drm/rcar-du: Plug atomic state hooks to the default implementation

This is needed to pass a bridge state to all atomic hooks, if we don't
do that, the core can't duplicate/create bridge states.

v10:
* Add changelog to the commit message

v9:
* Add Neil's R-b
* Move earlier in the series

v8:
* No changes

v7:
* New patch

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-3-boris.brezillon@collabora.com

drm/bridge: Add a drm_bridge_state object

One of the last remaining objects to not have its atomic state.

This is being motivated by our attempt to support runtime bus-format
negotiation between elements of the bridge chain.
This patch just paves the road for such a feature by adding a new
drm_bridge_state object inheriting from drm_private_obj so we can
re-use some of the existing state initialization/tracking logic.

v10:
* Add changelog to the commit message

v9:
* Clarify the fact that the bridge->atomic_reset() and
  {connector,plane,crtc,...}->reset() semantics are different
* Move the drm_atomic_private_obj_init() call back to
  drm_bridge_attach()
* Check the presence of ->atomic_duplicate_state instead of
  ->atomic_reset in drm_atomic_add_encoder_bridges()
* Fix copy&paste errors in the atomic bridge state helpers doc
* Add A-b/R-b tags

v8:
* Move bridge state helpers out of the CONFIG_DEBUGFS section

v7:
* Move helpers, struct-defs, ... to atomic helper files to avoid the
  drm -> drm_kms_helper -> drm circular dep
* Stop providing default implementation for atomic state reset,
  duplicate and destroy hooks (has to do with the helper/core split)
* Drop all R-b/T-b as helpers have now be moved to other places

v6:
* Made helpers private, removed doc and moved them to satisfy dependencies
* Renamed helpers to _default_

v5:
* Re-introduced the helpers from v4

v4:
* Fix the doc
* Kill default helpers (inlined)
* Fix drm_atomic_get_bridge_state() to check for an ERR_PTR()
* Add Neil's R-b

v3:
* No changes

v2:
* Use drm_for_each_bridge_in_chain()
* Rename helpers to be more consistent with the rest of the DRM API
* Improve/fix the doc

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200128135514.108171-2-boris.brezillon@collabora.com

drm/panel: Add Samsung s6e88a0-ams452ef01 panel driver

Add driver for Samsung AMS452EF01 AMOLED panel, which makes
use of their S6E88A0 controller.

Signed-off-by: Michael Srba <Michael.Srba@seznam.cz>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
[fixed checkpatch warnings, added changelog]
Link: https://patchwork.freedesktop.org/patch/msgid/20200130203555.316-2-michael.srba@seznam.cz

dt-bindings: display/panel: add bindings for S6E88A0-AMS452EF01

This patch adds dts bindings for Samsung AMS452EF01 AMOLED panel,
which makes use of their S6E88A0 controller.

Signed-off-by: Michael Srba <Michael.Srba@seznam.cz>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
[fixed syntax and updated example to fix warnings]
Link: https://patchwork.freedesktop.org/patch/msgid/20200130203555.316-1-michael.srba@seznam.cz

drm/panel: simple: Add EDT panel support

EDT ET043080DH6-GP is a 4.3" WQVGA 480x272 RGB LCD panel used on the iWave
Generic SODIMM Development Platform.

Changes in v2:
-added mandatory .connector_type field
-changed the .bus_format MEDIA_BUS_FMT_RGB666_1X18

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Reviewed-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/1580386118-22895-3-git-send-email-marian-cristian.rotariu.rb@bp.renesas.com

dt-bindings: display: Add bindings for EDT panel

Document the Emerging Display Technology Corp. (EDT) ETM043080DH6-GP
display, which is a 480x272 4.3" TFT display with capacitive touchscreen.

Changes in v2:
-modify proper bindings file

Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com>
Reviewed-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/1580386118-22895-2-git-send-email-marian-cristian.rotariu.rb@bp.renesas.com

drm/xen: Explicitly disable automatic sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. This would apply to xen, but xen has
its own vblank logic. To avoid interfering with the atomic helpers,
disable automatic vblank events explicitly.

v5:
* update comment
v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-16-tzimmermann@suse.de

drm/virtio: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-15-tzimmermann@suse.de

drm/vboxvideo: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-14-tzimmermann@suse.de

drm/udl: Don't set struct drm_crtc_state.no_vblank explicitly

As udl does not initialize vblanking, atomic helpers initialize the
value of struct drm_crtc_state.no_vblank to be true. No need to set
it from within the driver.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-13-tzimmermann@suse.de

drm/st7586: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-12-tzimmermann@suse.de

drm/repaper: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-11-tzimmermann@suse.de

drm/qxl: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-10-tzimmermann@suse.de

drm/mipi-dbi: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-9-tzimmermann@suse.de

drm/ili9225: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-8-tzimmermann@suse.de

drm/gm12u320: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-7-tzimmermann@suse.de

drm/cirrus: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-6-tzimmermann@suse.de

drm/bochs: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-5-tzimmermann@suse.de

drm/ast: Don't set struct drm_crtc_state.no_vblank explicitly

As ast does not initialize vblanking, atomic helpers initialize the
value of struct drm_crtc_state.no_vblank to be true. No need to set
it from within the driver.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-4-tzimmermann@suse.de

drm/arc: Remove sending of vblank event

The atomic helpers automatically send out fake VBLANK events if no
vblanking has been initialized. Remove the sending code from the
driver.

v4:
* separate commit from core vblank changes

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-3-tzimmermann@suse.de

drm: Initialize struct drm_crtc_state.no_vblank from device settings

At the end of a commit, atomic helpers can generate a fake VBLANK event
automatically. Originally implemented for writeback connectors, the
functionality can be used by any driver and/or hardware without proper
VBLANK interrupt.

The patch updates the documentation to make this behaviour official:
settings struct drm_crtc_state.no_vblank to true enables automatic
generation of fake VBLANK events.

The new interface drm_dev_has_vblank() returns true if vblanking has
been initialized for a device, or false otherwise. Atomic helpers use
this function when initializing no_vblank in the CRTC state in
drm_atomic_helper_check_modeset(). If vblanking has been initialized
for a device, no_blank is disabled. Otherwise it's enabled. Hence,
atomic helpers will automatically send out fake VBLANK events with any
driver that did not initialize vblanking.

v5:
* more precise documentation and commit message
v4:
* replace drm_crtc_has_vblank() with drm_dev_has_vblank()
* add drm_dev_has_vblank() in this patch
* move driver changes into separate patches
v3:
* squash all related changes patches into this patch

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200129120531.6891-2-tzimmermann@suse.de

drm/auth: Drop master_create/destroy hooks

vmwgfx stopped using them.

With the drm device model that we've slowly evolved over the past few
years master status essentially controls access to display resources,
and nothing else. Since that's a pure access permission check drivers
should have no need at all to track additional state on a per file
basis.

Aside: For cleanup and restoring kernel-internal clients the grand
plan is to move everyone over to drm_client and
drm_master_internal_acquire/release, like the generic fbdev code
already does. That should get rid of most ->lastclose implementations,
and I think also subsumes any processing vmwgfx does in
master_set/drop.

Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Cc: "Thomas Hellström (VMware)" <thomas_os@shipmail.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200127100203.1299322-1-daniel.vetter@ffwll.ch

drm/crc: Actually allow to change the crc source

Oops.

Fixes: 9edbf1fa600a ("drm: Add API for capturing frame CRCs")
Cc: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Cc: Emil Velikov <emil.velikov@collabora.com>
Cc: Benjamin Gaignard <benjamin.gaignard@linaro.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190821203835.18314-1-daniel.vetter@ffwll.ch

drm/zte: plane_state->fb iff plane_state->crtc

Checking both is one too much, so wrap a WARN_ON around it to stope
the copypasta.

Reviewed-by: Rodrigo Siqueira <rodrigosiqueira@gmail.com>
Tested-by: Rodrigo Siqueira <rodrigosiqueira@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Cc: Shawn Guo <shawnguo@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20191213172612.1514842-10-daniel.vetter@ffwll.ch

drm/vkms: plane_state->fb iff plane_state->crtc

Checking both is one too much, so wrap a WARN_ON around it to stope
the copypasta.

Reviewed-by: Rodrigo Siqueira <rodrigosiqueira@gmail.com>
Tested-by: Rodrigo Siqueira <rodrigosiqueira@gmail.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Cc: Rodrigo Siqueira <rodrigosiqueiramelo@gmail.com>
Cc: Haneen Mohammed <hamohammed.sa@gmail.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20191213172612.1514842-9-daniel.vetter@ffwll.ch

drm/vc4: plane_state->fb iff plane_state->crtc

Checking both is one too much, so wrap a WARN_ON around it to stope
the copypasta.

Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191213172612.1514842-7-daniel.vetter@ffwll.ch

drm/rockchip: plane_state->fb iff plane_state->crtc

Checking both is one too much, so wrap a WARN_ON around it to stope
the copypasta.

Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Cc: Sandy Huang <hjc@rock-chips.com>
Cc: "Heiko Stübner" <heiko@sntech.de>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-rockchip@lists.infradead.org
Link: https://patchwork.freedesktop.org/patch/msgid/20191213172612.1514842-6-daniel.vetter@ffwll.ch

drm/imx: plane_state->fb iff plane_state->crtc

Checking both is one too much, so wrap a WARN_ON around it to stope
the copypasta.

Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Cc: Philipp Zabel <p.zabel@pengutronix.de>
Cc: Shawn Guo <shawnguo@kernel.org>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Pengutronix Kernel Team <kernel@pengutronix.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: NXP Linux Team <linux-imx@nxp.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: https://patchwork.freedesktop.org/patch/msgid/20191213172612.1514842-4-daniel.vetter@ffwll.ch

drm/todo: Add item for the plane->atomic_check confusion

It's frankly a mess, and the confusion around plane_state->crtc/fb
that I fixed up in this series is the least of the problems. Add a
todo as a future note of how this could be done a lot better, and with
a lot less driver confusion.

Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191213172612.1514842-1-daniel.vetter@ffwll.ch

MAINTAINERS: add entry for tidss

Add entry for tidss DRM driver.

Version history:

v2: no change

v3: - Move tidss entry after omapdrm
    - Add "T:     git git://anongit.freedesktop.org/drm/drm-misc"

v4: no change

v5: no change

v6: no change

v7: no change

v8: - Add Reviewed-by: Benoit Parrot <bparrot@ti.com>

v9: - Add Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>

Signed-off-by: Jyri Sarha <jsarha@ti.com>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/c22f7aa31af09cd4adcca6b1a6b0721d9429f055.1580129724.git.jsarha@ti.com

drm/tidss: New driver for TI Keystone platform Display SubSystem

This patch adds a new DRM driver for Texas Instruments DSS IPs used on
Texas Instruments Keystone K2G, AM65x, and J721e SoCs. The new DSS IP is
a major change to the older DSS IP versions, which are supported by
the omapdrm driver. While on higher level the Keystone DSS resembles
the older DSS versions, the registers are completely different and the
internal pipelines differ a lot.

DSS IP found on K2G is an "ultra-light" version, and has only a single
plane and a single output. The K3 DSS IPs are found on AM65x and J721E
SoCs. AM65x DSS has two video ports, one full video plane, and another
"lite" plane without scaling support. J721E has 4 video ports, 2 video
planes and 2 lite planes. AM65x DSS has also an integrated OLDI (LVDS)
output.

Version history:

v2: - rebased on top of drm-next-2019-11-27
    - sort all include lines in all files
    - remove all include <drm/drmP.h>
    - remove select "select VIDEOMODE_HELPERS"
    - call dispc_vp_setup() later in tidss_crtc_atomic_flush() (there is no
      to call it in new modeset case as it is also called in vp_enable())
    - change probe sequence and drm_device allocation (follow example in
      drm_drv.c)
    - use __maybe_unused instead of #ifdef for pm functions
    - remove "struct drm_fbdev_cma *fbdev;" from driver data
    - check panel connector type before connecting it

v3: no change

v4: no change

v5: - remove fifo underflow irq handling, it is not an error and
      it should be used for debug purposes only
    - memory tuning, prefetch plane fifo up to high-threshold value to
      minimize possibility of underflows.

v6: - Check CTM and gamma support from dispc_features when creating crtc
    - Implement CTM support for k2g and fix k3 CTM implementation
    - Remove gamma property persistence and always write color properties
      in a new modeset

v7: - Fix checkpatch.pl --strict issues
    - Rebase on top of drm-misc-next-2020-01-10

v8: - Remove idle debug prints from dispc_init()
    - Add Reviewed-by: Benoit Parrot <bparrot@ti.com>

v9: - Rename dispc_write_irqenable() to dispc_set_irqenable() to avoid
      conflict exported omapfb function with same name
    - Add Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>

Co-developed-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Jyri Sarha <jsarha@ti.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/925fbfad58ff828e8e07fdff7073a0ee65750c3d.1580129724.git.jsarha@ti.com

dt-bindings: display: ti,j721e-dss: Add dt-schema yaml binding

Add dt-schema yaml bindig for J721E DSS, J721E version TI Keystone
Display SubSystem.

Version history:

v2: no change

v3: - reg-names: "wp" -> "wb"
    - Add ports node
    - Add includes to dts example
    - reindent dts example

v4: - Add descriptions to reg, clocks, and interrupts properties
    - Remove minItems when its value is the same as maxItems value

v5: - itemize reg, clocks and interrupts properties' descriptions
    - there is no "vp" reg-name, only "wb" for write back

v6: - Add Reviewed-by: from Rob Herring <robh@kernel.org> and
      Benoit Parrot <bparrot@ti.com>

v7: no change

v8: no change

v9: no change

Signed-off-by: Jyri Sarha <jsarha@ti.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/fcba52837808853a4b2bf729bd33bd986cbdadf4.1580129724.git.jsarha@ti.com

dt-bindings: display: ti,am65x-dss: Add dt-schema yaml binding

Add dt-schema yaml bindig for AM65x DSS, AM65x version TI Keystone
Display SubSystem.

Version history:

v2: no change

v3: - Add ports node
    - use allOf in ti,am65x-oldi-io-ctrl to add both $ref and maxItems
    - Add includes to dts example
    - reindent dts example

v4: - Add descriptions to reg and clocks properties
    - Remove minItems when its value is the same as maxItems value

v5: - itemize reg and clocks properties' descriptions

v6: - Add Reviewed-by: from Rob Herring <robh@kernel.org> and
      Benoit Parrot <bparrot@ti.com>

v7: no change

v8: no change

Signed-off-by: Jyri Sarha <jsarha@ti.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1799656fb8cc42ca698fc29bbc80c1bae442d6d3.1580129724.git.jsarha@ti.com

dt-bindings: display: ti,k2g-dss: Add dt-schema yaml binding

Add dt-schema yaml bindig for K2G DSS, an ultra-light version of TI
Keystone Display SubSystem.

Version history:

v2: no change

v3: - Add ports node
    - Add includes to dts example
    - reindent dts example

v4: - Add descriptions to reg and clocks properties
    - Remove minItems when its value is the same as maxItems value
    - Remove ports node

v5: - itemize reg and clocks properties' descriptions

v6: - Add Reviewed-by: from Rob Herring <robh@kernel.org> and
      Benoit Parrot <bparrot@ti.com>

v7: no change

v8: no change

v9: - Remove ports-node from the dts example

Signed-off-by: Jyri Sarha <jsarha@ti.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Link: https://patchwork.freedesktop.org/patch/msgid/270297321f0768c10e241d289e3ac10e39cf12a9.1580129724.git.jsarha@ti.com

drm/lima: increase driver version to 1.1

Increase driver version for mesa driver to identify
the support of new heap buffer interface.

Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Tested-by: Andreas Baierl <ichgeh@imkreisrum.de>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-6-yuq825@gmail.com

drm/lima: recover task by enlarging heap buffer

Increase heap buffer backup memory when GP receive PLBU
out of memory interrupt, then resume the task.

Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Tested-by: Andreas Baierl <ichgeh@imkreisrum.de>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-5-yuq825@gmail.com

drm/lima: support heap buffer creation

heap buffer is used as output of GP and input of PP for
Mali Utgard GPU. Size of heap buffer depends on the task
so is a runtime variable.

Previously we just create a large enough buffer as heap
buffer. Now we add a heap buffer type to be able to
increase the backup memory dynamically when GP fail due
to lack of heap memory.

Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Tested-by: Andreas Baierl <ichgeh@imkreisrum.de>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-4-yuq825@gmail.com

drm/lima: add lima_vm_map_bo

For dynamically mapping added backup memory of lima_bo to vm.
This is a preparation for adding heap buffer support.

Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Tested-by: Andreas Baierl <ichgeh@imkreisrum.de>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-3-yuq825@gmail.com

drm/lima: update register info

From Mali r10p0 kernel driver source code.

Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Tested-by: Andreas Baierl <ichgeh@imkreisrum.de>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116131157.13346-2-yuq825@gmail.com

drm/debugfs: also take per device driver features into account

Use drm_core_check_all_features() to ensure both the driver features and
the per-device driver features are taken into account when registering
debugfs files.

v3:
- files[i].driver_features == 0 actually means "don't care"

v2:
- use drm_core_check_all_features()

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200123124801.14958-2-jani.nikula@intel.com

drm: add drm_core_check_all_features() to check for a mask of features

Add new drm_core_check_all_features() function to check for a mask of
features. All features in the mask are required.

Redefine existing drm_core_check_feature() in terms of this function,
using the drm_driver_feature enum for the parameter.

v3:
- add drm_core_check_all_features() (Thomas)

v2:
- fix kernel-doc (Ville)
- add an extra variable for clarity (Ville)

Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200123124801.14958-1-jani.nikula@intel.com

drm/panel: Fix boe,tv101wum-n53 htotal timing

The datasheet suggests 60 for tHFP, so let's adjust the number
accordingly.

This also makes the framerate be 60Hz as intended:
159916.0 * 1000 / ((1200 + 80 + 24 + 60)*(1920 + 20 + 4 + 10))
=> 60.00 Hz

Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Fixes: e6d020eeefcd ("drm/panel: support for boe, tv101wum-n53 wuxga dsi video mode panel")
Cc: Jitao Shi <jitao.shi@mediatek.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org> [added fixes tag]
Link: https://patchwork.freedesktop.org/patch/msgid/20200125050256.107404-1-drinkcat@chromium.org

drm: Avoid drm_global_mutex for simple inc/dec of dev->open_count

Since drm_global_mutex is a true global mutex across devices, we don't
want to acquire it unless absolutely necessary. For maintaining the
device local open_count, we can use atomic operations on the counter
itself, except when making the transition to/from 0. Here, we tackle the
easy portion of delaying acquiring the drm_global_mutex for the final
release by using atomic_dec_and_mutex_lock(), leaving the global
serialisation across the device opens.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Thomas Hellström (VMware) <thomas_os@shipmail.org>
Reviewed-by: Thomas Hellström <thellstrom@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200124130107.125404-1-chris@chris-wilson.co.uk

drm: Release filp before global lock

The file is not part of the global drm resource and can be released
prior to take the global mutex to drop the open_count (and potentially
close) the drm device. As the global mutex is indeed global, not only
within the device but across devices, a slow file release mechanism can
bottleneck the entire system.

However, inside drm_close_helper() there are a number of dev->driver
callbacks that take the drm_device as the first parameter... Worryingly
some of those callbacks may be (implicitly) depending on the global
mutex.

v2: Drop the debug message for the open-count, it's included with the
drm_file_free() debug message -- and for good measure make that up as
reading outside of the mutex.

v3: Separate the calling of the filp cleanup outside of
drm_global_mutex into a new drm_release_noglobal() hook, so that we can
phase the transition. drm/savage relies on the global mutex, and there
may be more, so be cautious.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Thomas Hellström (VMware) <thomas_os@shipmail.org>
Reviewed-by: Thomas Hellström (VMware) <thomas_os@shipmail.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200124125627.125042-1-chris@chris-wilson.co.uk

dt-bindings: restrict properties for sitronix,st7735r

David Lechner noticed (paraphrased):
- not all properties from panel-common are applicable.
- missing optional rotation and backlight properties

Fix this by listing all allowed properties,
and do not allow other properties.

Fixes: abdd9e3705c8 ("dt-bindings: display: sitronix,st7735r: Convert to DT schema")
Reported-by: David Lechner <david@lechnology.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Rob Herring <robh@kernel.org>
Cc: David Lechner <david@lechnology.com>
Cc: dri-devel@lists.freedesktop.org
Link: https://patchwork.freedesktop.org/patch/msgid/20200120190249.GA9619@ravnborg.org

drm/panel: simple: add display timings for logic technologies displays

Add display timings for the following 3 display panels manufactured by
Logic Technologies Limited:

- LT161010-2NHC e.g. as found in the Toradex Capacitive Touch Display
  7" Parallel [1]
- LT161010-2NHR e.g. as found in the Toradex Resistive Touch Display 7"
  Parallel [2]
- LT170410-2WHC e.g. as found in the Toradex Capacitive Touch Display
  10.1" LVDS [3]

Those panels may also be distributed by Endrich Bauelemente Vertriebs
GmbH [4].

[1] https://docs.toradex.com/104497-7-inch-parallel-capacitive-touch-display-800x480-datasheet.pdf
[2] https://docs.toradex.com/104498-7-inch-parallel-resistive-touch-display-800x480.pdf
[3] https://docs.toradex.com/105952-10-1-inch-lvds-capacitive-touch-display-1280x800-datasheet.pdf
[4] https://www.endrich.com/isi50_isi30_tft-displays/lt170410-1whc_isi30

Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Philippe Schenker <philippe.schenker@toradex.com>
Reviewed-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200120080100.170294-3-marcel@ziswiler.com

dt-bindings: panel-simple: add bindings for logic technologies displays

Add bindings for the following 3 to be added display panels manufactured
by Logic Technologies Limited:

- LT161010-2NHC e.g. as found in the Toradex Capacitive Touch Display
7" Parallel [1]
- LT161010-2NHR e.g. as found in the Toradex Resistive Touch Display 7"
Parallel [2]
- LT170410-2WHC e.g. as found in the Toradex Capacitive Touch Display
10.1" LVDS [3]

Those panels may also be distributed by Endrich Bauelemente Vertriebs
GmbH [4].

[1] https://docs.toradex.com/104497-7-inch-parallel-capacitive-touch-display-800x480-datasheet.pdf
[2] https://docs.toradex.com/104498-7-inch-parallel-resistive-touch-display-800x480.pdf
[3] https://docs.toradex.com/105952-10-1-inch-lvds-capacitive-touch-display-1280x800-datasheet.pdf
[4] https://www.endrich.com/isi50_isi30_tft-displays/lt170410-1whc_isi30

Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200120080100.170294-2-marcel@ziswiler.com

dt-bindings: add vendor prefix for logic technologies limited

Add vendor prefix for Logic Technologies Limited [1] which is a Chinese
display manufacturer e.g. distributed by German Endrich Bauelemente
Vertriebs GmbH [2].

[1] https://logictechno.com/contact-us/
[2] https://www.endrich.com/isi50_isi30_tft-displays/lt170410-1whc_isi30

Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Philippe Schenker <philippe.schenker@toradex.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200120080100.170294-1-marcel@ziswiler.com

dt-bindings: display: Convert a bunch of panels to DT schema

Convert all the 'simple' panels which match the constraints of the
common panel-simple.yaml schema. This conversion is based on how the
panels are documented. Some may turn out to be more complex once the
schema is applied to actual dts files.

Cc: Maxime Ripard <mripard@kernel.org>
Cc: Chen-Yu Tsai <wens@csie.org>
Cc: Thierry Reding <thierry.reding@gmail.com>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200119210907.24152-1-robh@kernel.org

Merge tag 'topic/drm-warn-2020-01-22' of git://anongit.freedesktop.org/drm/drm-intel into drm-misc-next

struct drm_device based drm_WARN* macros

Signed-off-by: Maxime Ripard <maxime@cerno.tech>
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/87eevrecnf.fsf@intel.com

drm/dp_mst: Mention max_payloads in proposed_vcpis/payloads docs

Mention that the size of these two structs is determined by
max_payloads. Suggested by Ville Syrjälä.

Signed-off-by: Lyude Paul <lyude@redhat.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200122194321.14953-2-lyude@redhat.com

drm/dp_mst: Fix clearing payload state on topology disable

The issues caused by:

commit 64e62bdf04ab ("drm/dp_mst: Remove VCPI while disabling topology
mgr")

Prompted me to take a closer look at how we clear the payload state in
general when disabling the topology, and it turns out there's actually
two subtle issues here.

The first is that we're not grabbing &mgr.payload_lock when clearing the
payloads in drm_dp_mst_topology_mgr_set_mst(). Seeing as the canonical
lock order is &mgr.payload_lock -> &mgr.lock (because we always want
&mgr.lock to be the inner-most lock so topology validation always
works), this makes perfect sense. It also means that -technically- there
could be racing between someone calling
drm_dp_mst_topology_mgr_set_mst() to disable the topology, along with a
modeset occurring that's modifying the payload state at the same time.

The second is the more obvious issue that Wayne Lin discovered, that
we're not clearing proposed_payloads when disabling the topology.

I actually can't see any obvious places where the racing caused by the
first issue would break something, and it could be that some of our
higher-level locks already prevent this by happenstance, but better safe
then sorry. So, let's make it so that drm_dp_mst_topology_mgr_set_mst()
first grabs &mgr.payload_lock followed by &mgr.lock so that we never
race when modifying the payload state. Then, we also clear
proposed_payloads to fix the original issue of enabling a new topology
with a dirty payload state. This doesn't clear any of the drm_dp_vcpi
structures, but those are getting destroyed along with the ports anyway.

Changes since v1:
* Use sizeof(mgr->payloads[0])/sizeof(mgr->proposed_vcpis[0]) instead -
  vsyrjala

Cc: Sean Paul <sean@poorly.run>
Cc: Wayne Lin <Wayne.Lin@amd.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: stable@vger.kernel.org # v4.4+
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200122194321.14953-1-lyude@redhat.com

drm/dp_mst: Fix indenting in drm_dp_mst_topology_mgr_set_mst()

This has always bugged me but somehow I've never remembered to actually
fix it. So let's do that.

Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200117224749.128994-1-lyude@redhat.com

drm/bridge: Add I2C based driver for ps8640 bridge

This patch adds drm_bridge driver for parade DSI to eDP bridge chip.

Signed-off-by: Jitao Shi <jitao.shi@mediatek.com>
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
[uli: followed API changes, removed FW update feature]
Signed-off-by: Ulrich Hecht <uli@fpond.eu>
Signed-off-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Tested-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Ezequiel Garcia <ezequiel@collabora.com>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191230090419.137141-3-enric.balletbo@collabora.com

Documentation: bridge: Add documentation for ps8640 DT properties

Add documentation for DT properties supported by
ps8640 DSI-eDP converter.

Signed-off-by: Jitao Shi <jitao.shi@mediatek.com>
Acked-by: Rob Herring <robh@kernel.org>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Ulrich Hecht <uli@fpond.eu>
Signed-off-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191230090419.137141-2-enric.balletbo@collabora.com

drm/print: introduce new struct drm_device based WARN* macros

Add new struct drm_device based WARN* macros. These are modeled after
the core kernel device based WARN* macros. These would be preferred
over the regular WARN* macros, where possible.

These macros include device information in the backtrace, so we know
what device the warnings originate from.

Knowing the device specific information in the backtrace would be
helpful in development all around.

Signed-off-by: Pankaj Bharadiya <pankaj.laxminarayan.bharadiya@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Maxime Ripard <mripard@kernel.org>
Acked-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Sean Paul <sean@poorly.run>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115034455.17658-2-pankaj.laxminarayan.bharadiya@intel.com

drm/ttm: flush the fence on the bo after we individualize the reservation object

As we move the ttm_bo_individualize_resv() upwards, we need flush the
copied fence too. Otherwise the driver keeps waiting for fence.

run&Kill kfdtest, then perf top.

  25.53%  [ttm]                     [k] ttm_bo_delayed_delete
  24.29%  [kernel]                  [k] dma_resv_test_signaled_rcu
  19.72%  [kernel]                  [k] ww_mutex_lock

Fix: 378e2d5b("drm/ttm: fix ttm_bo_cleanup_refs_or_queue once more")
Signed-off-by: xinhui pan <xinhui.pan@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/series/72339/
Signed-off-by: Christian König <christian.koenig@amd.com>

drm/drm_dp_mst:remove set but not used variable 'origlen'

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/gpu/drm/drm_dp_mst_topology.c:3693:16: warning: variable
‘origlen’ set but not used [-Wunused-but-set-variable]
  int replylen, origlen, curreply;

It looks like never use variable origlen after assign value to it.

Fixes: ad7f8a1f9ced ("drm/helper: add Displayport multi-stream helper (v0.6)")
Signed-off-by: Bo YU <tsu.yubo@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20200118080628.mxcx7bfwdas5m7un@kaowomen.cn

drm/panfrost: Prefix interrupt handlers' names

Currently, the interrupt lines requested by Panfrost
use unmeaningful names, which adds some obscurity
to interrupt introspection (i.e. any tool based
on procfs' interrupts file).

In order to improve this, prefix each requested
interrupt with the module name: panfrost-{gpu,job,mmu}.

Signed-off-by: Ezequiel Garcia <ezequiel@collabora.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20191214045952.9452-1-ezequiel@collabora.com

drm/panfrost: Remove core stack power management

Explicit management of the GPU's core stacks is only necessary in the
case of a broken integration with the PDC. Since there are no known
platforms which have such a broken integration let's remove the explicit
control from the driver since this apparently causes problems on other
platforms and will have a small performance penality.

The out of tree mali_kbase driver contains this text regarding
controlling the core stack (CONFIGMALI_CORESTACK):

  Enabling this feature on supported GPUs will let the driver powering
  on/off the GPU core stack independently without involving the Power
  Domain Controller. This should only be enabled on platforms which
  integration of the PDC to the Mali GPU is known to be problematic.
  This feature is currently only supported on t-Six and t-HEx GPUs.

  If unsure, say N.

Signed-off-by: Steven Price <steven.price@arm.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Nicolas Boichat <drinkcat@chromium.org>
Tested-by: Nicolas Boichat <drinkcat@chromium.org>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200109133104.11661-1-steven.price@arm.com

drm/syncobj: Add documentation for timeline syncobj

We've added a set of new APIs to manipulate syncobjs holding timelines
of dma_fence. This adds a bit of documentation about how this works.

v2: Small language nits (Lionel)

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/348578/
Cc: Christian Koenig <Christian.Koenig@amd.com>
Cc: Jason Ekstrand <jason@jlekstrand.net>
Cc: David(ChunMing) Zhou <David1.Zhou@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>

drm: tiny: st7735r: Add support for Okaya RH128128T

Add support for the Okaya RH128128T display to the st7735r driver on
DT-enabled systems.

The RH128128T is a 128x128 1.44" TFT display driven by a Sitronix
ST7715R TFT Controller/Driver.  The latter is very similar to the
ST7735R, and can be handled by the existing st7735r driver.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115124548.3951-6-geert+renesas@glider.be

drm: tiny: st7735r: Prepare for adding support for more displays

Currently the st7735r driver supports only a single display panel.
Prepare for adding support for other display panels by factoring out the
display-specific parameters in struct st7735r_cfg.

For now, the following parameters can be configured:
  - Display resolution,
  - Horizontal/vertical display offsets,
  - Read-write versus read-only controllers,
  - RGB versus BGR color component ordering.

Rename jd_t18003_t01_pipe_enable() and jd_t18003_t01_pipe_funcs() to
st7735r_pipe_enable() resp. st7735r_pipe_funcs(), as they are not really
specific to the Jianda JD-T18003-T01 display anymore.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115124548.3951-5-geert+renesas@glider.be

drm/mipi_dbi: Add support for display offsets

If the resolution of the TFT display is smaller than the maximum
resolution supported by the display controller, the display may be
connected to the driver output arrays with a horizontal and/or vertical
offset, leading to a shifted image.

Add support for specifying these offsets.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Noralf Trønnes <noralf@tronnes.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115124548.3951-4-geert+renesas@glider.be

dt-bindings: display: sitronix,st7735r: Add Okaya RH128128T

Document support for the Okaya RH128128T display, which is a 128x128
1.44" TFT display driven by a Sitronix ST7715R TFT Controller/Driver.
It can be found on e.g. the Renesas YRSK-LCD-PMOD extension board, which
comes with various Renesas development kits (e.g. Renesas Starter Kit+
for RZ/A1H[1]).

ST7715R and ST7735R are very similar.  Their major difference is that
the former is restricted to displays of up to 132x132 pixels, while the
latter supports displays up to 132x162 pixels.

[1] https://renesasrulz.com/the_vault/f/archive-forum/4981/upgrading-to-the-renesas-rz-a1h

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115124548.3951-3-geert+renesas@glider.be

dt-bindings: display: sitronix,st7735r: Convert to DT schema

Convert the DT binding documentation for Sitronix ST7735R displays to DT
schema.

Add a reference to the Adafruit 1.8" LCD while at it.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200115124548.3951-2-geert+renesas@glider.be

drm: fix parameters documentation style in drm_dma

Remove old documentation style and use new one to avoid warnings when
compiling with W=1

Signed-off-by: Benjamin Gaignard <benjamin.gaignard@st.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200114160135.14990-1-benjamin.gaignard@st.com

drm/mst: Some style improvements in drm_dp_mst_topology_mgr_set_mst()

Removing this lose code block and removing unnecessary bracket.

Cc: Lyude Paul <lyude@redhat.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Lyude Paul <lyude@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200117015837.402239-2-jose.souza@intel.com

Revert "drm/dp_mst: Remove VCPI while disabling topology mgr"

This reverts commit 64e62bdf04ab8529f45ed0a85122c703035dec3a.

This commit ends up causing some lockdep splats due to trying to grab the
payload lock while holding the mgr's lock:

[   54.010099]
[   54.011765] ======================================================
[   54.018670] WARNING: possible circular locking dependency detected
[   54.025577] 5.5.0-rc6-02274-g77381c23ee63 #47 Not tainted
[   54.031610] ------------------------------------------------------
[   54.038516] kworker/1:6/1040 is trying to acquire lock:
[   54.044354] ffff888272af3228 (&mgr->payload_lock){+.+.}, at:
drm_dp_mst_topology_mgr_set_mst+0x218/0x2e4
[   54.054957]
[   54.054957] but task is already holding lock:
[   54.061473] ffff888272af3060 (&mgr->lock){+.+.}, at:
drm_dp_mst_topology_mgr_set_mst+0x3c/0x2e4
[   54.071193]
[   54.071193] which lock already depends on the new lock.
[   54.071193]
[   54.080334]
[   54.080334] the existing dependency chain (in reverse order) is:
[   54.088697]
[   54.088697] -> #1 (&mgr->lock){+.+.}:
[   54.094440]        __mutex_lock+0xc3/0x498
[   54.099015]        drm_dp_mst_topology_get_port_validated+0x25/0x80
[   54.106018]        drm_dp_update_payload_part1+0xa2/0x2e2
[   54.112051]        intel_mst_pre_enable_dp+0x144/0x18f
[   54.117791]        intel_encoders_pre_enable+0x63/0x70
[   54.123532]        hsw_crtc_enable+0xa1/0x722
[   54.128396]        intel_update_crtc+0x50/0x194
[   54.133455]        skl_commit_modeset_enables+0x40c/0x540
[   54.139485]        intel_atomic_commit_tail+0x5f7/0x130d
[   54.145418]        intel_atomic_commit+0x2c8/0x2d8
[   54.150770]        drm_atomic_helper_set_config+0x5a/0x70
[   54.156801]        drm_mode_setcrtc+0x2ab/0x833
[   54.161862]        drm_ioctl+0x2e5/0x424
[   54.166242]        vfs_ioctl+0x21/0x2f
[   54.170426]        do_vfs_ioctl+0x5fb/0x61e
[   54.175096]        ksys_ioctl+0x55/0x75
[   54.179377]        __x64_sys_ioctl+0x1a/0x1e
[   54.184146]        do_syscall_64+0x5c/0x6d
[   54.188721]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
[   54.194946]
[   54.194946] -> #0 (&mgr->payload_lock){+.+.}:
[   54.201463]
[   54.201463] other info that might help us debug this:
[   54.201463]
[   54.210410]  Possible unsafe locking scenario:
[   54.210410]
[   54.217025]        CPU0                    CPU1
[   54.222082]        ----                    ----
[   54.227138]   lock(&mgr->lock);
[   54.230643]                                lock(&mgr->payload_lock);
[   54.237742]                                lock(&mgr->lock);
[   54.244062]   lock(&mgr->payload_lock);
[   54.248346]
[   54.248346]  *** DEADLOCK ***
[   54.248346]
[   54.254959] 7 locks held by kworker/1:6/1040:
[   54.259822]  #0: ffff888275c4f528 ((wq_completion)events){+.+.},
at: worker_thread+0x455/0x6e2
[   54.269451]  #1: ffffc9000119beb0
((work_completion)(&(&dev_priv->hotplug.hotplug_work)->work)){+.+.},
at: worker_thread+0x455/0x6e2
[   54.282768]  #2: ffff888272a403f0 (&dev->mode_config.mutex){+.+.},
at: i915_hotplug_work_func+0x4b/0x2be
[   54.293368]  #3: ffffffff824fc6c0 (drm_connector_list_iter){.+.+},
at: i915_hotplug_work_func+0x17e/0x2be
[   54.304061]  #4: ffffc9000119bc58 (crtc_ww_class_acquire){+.+.},
at: drm_helper_probe_detect_ctx+0x40/0xfd
[   54.314855]  #5: ffff888272a40470 (crtc_ww_class_mutex){+.+.}, at:
drm_modeset_lock+0x74/0xe2
[   54.324385]  #6: ffff888272af3060 (&mgr->lock){+.+.}, at:
drm_dp_mst_topology_mgr_set_mst+0x3c/0x2e4
[   54.334597]
[   54.334597] stack backtrace:
[   54.339464] CPU: 1 PID: 1040 Comm: kworker/1:6 Not tainted
5.5.0-rc6-02274-g77381c23ee63 #47
[   54.348893] Hardware name: Google Fizz/Fizz, BIOS
Google_Fizz.10139.39.0 01/04/2018
[   54.357451] Workqueue: events i915_hotplug_work_func
[   54.362995] Call Trace:
[   54.365724]  dump_stack+0x71/0x9c
[   54.369427]  check_noncircular+0x91/0xbc
[   54.373809]  ? __lock_acquire+0xc9e/0xf66
[   54.378286]  ? __lock_acquire+0xc9e/0xf66
[   54.382763]  ? lock_acquire+0x175/0x1ac
[   54.387048]  ? drm_dp_mst_topology_mgr_set_mst+0x218/0x2e4
[   54.393177]  ? __mutex_lock+0xc3/0x498
[   54.397362]  ? drm_dp_mst_topology_mgr_set_mst+0x218/0x2e4
[   54.403492]  ? drm_dp_mst_topology_mgr_set_mst+0x218/0x2e4
[   54.409620]  ? drm_dp_dpcd_access+0xd9/0x101
[   54.414390]  ? drm_dp_mst_topology_mgr_set_mst+0x218/0x2e4
[   54.420517]  ? drm_dp_mst_topology_mgr_set_mst+0x218/0x2e4
[   54.426645]  ? intel_digital_port_connected+0x34d/0x35c
[   54.432482]  ? intel_dp_detect+0x227/0x44e
[   54.437056]  ? ww_mutex_lock+0x49/0x9a
[   54.441242]  ? drm_helper_probe_detect_ctx+0x75/0xfd
[   54.446789]  ? intel_encoder_hotplug+0x4b/0x97
[   54.451752]  ? intel_ddi_hotplug+0x61/0x2e0
[   54.456423]  ? mark_held_locks+0x53/0x68
[   54.460803]  ? _raw_spin_unlock_irqrestore+0x3a/0x51
[   54.466347]  ? lockdep_hardirqs_on+0x187/0x1a4
[   54.471310]  ? drm_connector_list_iter_next+0x89/0x9a
[   54.476953]  ? i915_hotplug_work_func+0x206/0x2be
[   54.482208]  ? worker_thread+0x4d5/0x6e2
[   54.486587]  ? worker_thread+0x455/0x6e2
[   54.490966]  ? queue_work_on+0x64/0x64
[   54.495151]  ? kthread+0x1e9/0x1f1
[   54.498946]  ? queue_work_on+0x64/0x64
[   54.503130]  ? kthread_unpark+0x5e/0x5e
[   54.507413]  ? ret_from_fork+0x3a/0x50

The proper fix for this is probably cleanup the VCPI allocations when we're
enabling the topology, or on the first payload allocation. For now though,
let's just revert.

Signed-off-by: Lyude Paul <lyude@redhat.com>
Fixes: 64e62bdf04ab ("drm/dp_mst: Remove VCPI while disabling topology mgr")
Cc: Sean Paul <sean@poorly.run>
Cc: Wayne Lin <Wayne.Lin@amd.com>
Reviewed-by: Sean Paul <sean@poorly.run>
Link: https://patchwork.freedesktop.org/patch/msgid/20200117205149.97262-1-lyude@redhat.com

drm/panel: Add Feixin K101 IM2BA02 panel

Feixin K101 IM2BA02 is a 800x1280 4-lane MIPI-DSI LCD panel.

Add a panel driver for it.

Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116033636.512461-4-icenowy@aosc.io

dt-bindings: panel: add Feixin K101 IM2BA02 MIPI-DSI panel

Feixin K101 IM2BA02 is a 10.1" 800x1280 4-lane MIPI-DSI panel.

Add device tree binding for it.

Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116033636.512461-3-icenowy@aosc.io

dt-bindings: vendor-prefix: add Shenzhen Feixin Photoelectics Co., Ltd

Shenzhen Feixin Photoelectics Co., Ltd is a company to provide LCD
modules.

Add its vendor prefix.

Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116033636.512461-2-icenowy@aosc.io

drm/panel: support for auo, b101uan08.3 wuxga dsi video mode panel

Auo,auo,b101uan08.3's connector is same as boe,tv101wum-nl6.
The most codes can be reuse.
So auo,b101uan08.3 and boe,tv101wum-nl6 use one driver file.
Add the different parts in driver data.

Signed-off-by: Jitao Shi <jitao.shi@mediatek.com>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116021511.22675-6-jitao.shi@mediatek.com

drm/panel: support for boe, tv101wum-n53 wuxga dsi video mode panel

Boe,tv101wum-n53's connector is same as boe,tv101wum-nl6.
The most codes can be reuse.
So boe,tv101wum-n53 and boe,tv101wum-nl6 use one driver file.
Add the different parts in driver data.

Signed-off-by: Jitao Shi <jitao.shi@mediatek.com>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116021511.22675-5-jitao.shi@mediatek.com

drm/panel: support for auo, kd101n80-45na wuxga dsi video mode panel

Auo,kd101n80-45na's connector is same as boe,tv101wum-nl6.
The most codes can be reuse.
So auo,kd101n80-45na and boe,tv101wum-nl6 use one driver file.
Add the different parts in driver data.

Signed-off-by: Jitao Shi <jitao.shi@mediatek.com>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116021511.22675-4-jitao.shi@mediatek.com

drm/panel: support for boe tv101wum-nl6 wuxga dsi video mode panel

Add driver for BOE tv101wum-nl6 panel is a 10.1" 1200x1920 panel.

Signed-off-by: Jitao Shi <jitao.shi@mediatek.com>
Reviewed-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200116021511.22675-3-jitao.shi@mediatek.com