GPU saves accumulated context runtime (in CS timestamp units) in PPHWSP
which will be useful for us in cases when we are not able to track context
busyness ourselves (like with GuC). Keep a copy of this in struct
intel_context from where it can be easily read even if the context is not
pinned.
v2:
(Chris)
* Do not store pphwsp address in intel_context.
* Log CS wrap-around.
* Simplify calculation by relying on integer wraparound.
v3:
* Include total/avg in traces and error state for debugging
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200216133620.394962-1-chris@chris-wilson.co.uk
The last in-kernel user of the old framebuffer driver is the
IM-PD1 module for the Integrator/AP. Let's implement support for
this remaining user so we can migrate the last user over to
DRM and delete the old FB driver.
On the Integrator/AP the IM-PD1 system controller will exist
alongside the common Integrator system controller so make
sure to do a special lookup for the IM-PD1 syscon and make it
take precedence if found.
Tested on the Integrator/AP with the IM-PD1 mounted.
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200213124833.35545-1-linus.walleij@linaro.org
Currently when we load the driver we set distrust_bios_wm=true, which
will cause active_pipe_changes to get flagged even when we're not
toggling any pipes on/off. The reason being that we want to fully
redistribute the dbuf among the active pipes and ignore whatever
state the firmware left behind.
Unfortunately when the code flags active_pipe_changes it doesn't
set state->modeset to true, which means the hardware dbuf state
won't actually get updated. Hence the hardware and software
states go out of sync, which can result in planes trying to use a
disabled dbuf slice. Suprisingly that only seems to corrupt the
display rather than making the whole display engine keel over.
Let's fix this for now by flagging state->modeset whenever
distrust_bios_wm is set.
Eventually we'll likely want to rip out all of this mess and
introduce proper statye tracking for dbuf. But that requires
more work. Toss in a FIXME to that effect.
Cc: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Fixes: ff2cd8635e ("drm/i915: Correctly map DBUF slices to pipes")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200213140412.32697-4-stanislav.lisovskiy@intel.com
TGL BIOS seems to enable both DBuf slices ocasionally, depending
how many displays are connected, while i915 according to BSpec
was powering on S1 DBuf slice, until a modeset was done.
This was causing a brief flash during the boot as we were
disabling slice, previously used by BIOS with that.
To prevent this, now we are ensuring tht we are enabling
_at least_ one slice, but if there are more, let's not
power them off.
Fixes: ff2cd8635e ("drm/i915: Correctly map DBUF slices to pipes")
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200213140412.32697-2-stanislav.lisovskiy@intel.com
Currently we assume any 18 byte descriptor to be a display descritor
if only the tag byte matches the expected value. But for detailed
timing descriptors that same byte is just the lower 8 bits of
hblank, and as such can match any display descriptor tag. To
properly validate that the 18 byte descriptor is in fact a
display descriptor we must also examine bytes 0-2 (just byte 1
should actually suffice but the spec does say that bytes 0 and
2 must also always be zero for display descriptors so we check
those too).
Unlike Allen's original proposed patch to just fix is_rb() we
roll this out across the board to fix everything.
Cc: Allen Chen <allen.chen@ite.com.tw>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200124200231.10517-2-ville.syrjala@linux.intel.com
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
CEA-861 says :
"d = offset for the byte following the reserved data block.
If no data is provided in the reserved data block, then d=4.
If no DTDs are provided, then d=0."
So let's not look for DTDs when d==0. In fact let's just make that
<4 since those values would just mean that he DTDs overlap the block
header. And let's also check that d isn't so big as to declare
the descriptors to live past the block end, although the code
does already survive that case as we'd just end up with a negative
number of descriptors and the loop would not do anything.
Cc: Allen Chen <allen.chen@ite.com.tw>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200124200231.10517-1-ville.syrjala@linux.intel.com
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Uma Shankar <uma.shankar@intel.com>
there was a type in the terminate command.
We should be calling psp_dtm_unload() instead of psp_hdcp_unload()
Fixes: 143f230533 ("drm/amdgpu: psp DTM init")
Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The driver currently uses runtime PM to perform some of the module
initialization and cleanup. This has three problems:
1) There is no Kconfig dependency on CONFIG_PM, so if runtime PM is
disabled, the driver will not work at all, since the module will
never be initialized.
2) The driver does not ensure that the device is suspended when
sun6i_dsi_probe() fails or when sun6i_dsi_remove() is called. It
simply disables runtime PM. From the docs of pm_runtime_disable():
The device can be either active or suspended after its runtime PM
has been disabled.
And indeed, the device will likely still be active if sun6i_dsi_probe
fails. For example, if the panel driver is not yet loaded, we have
the following sequence:
sun6i_dsi_probe()
pm_runtime_enable()
mipi_dsi_host_register()
of_mipi_dsi_device_add(child)
...device_add()...
__device_attach()
pm_runtime_get_sync(dev->parent) -> Causes resume
bus_for_each_drv()
__device_attach_driver() -> No match for panel
pm_runtime_put(dev->parent) -> Async idle request
component_add()
__component_add()
try_to_bring_up_masters()
try_to_bring_up_master()
sun4i_drv_bind()
component_bind_all()
component_bind()
sun6i_dsi_bind() -> Fails with -EPROBE_DEFER
mipi_dsi_host_unregister()
pm_runtime_disable()
__pm_runtime_disable()
__pm_runtime_barrier() -> Idle request is still pending
cancel_work_sync() -> DSI host is *not* suspended!
Since the device is not suspended, the clock and regulator are never
disabled. The imbalance causes a WARN at devres free time.
3) The driver relies on being suspended when sun6i_dsi_encoder_enable()
is called. The resume callback has a comment that says:
Some part of it can only be done once we get a number of
lanes, see sun6i_dsi_inst_init
And then part of the resume callback only runs if dsi->device is not
NULL (that is, if sun6i_dsi_attach() has been called). However, as
the above call graph shows, the resume callback is guaranteed to be
called before sun6i_dsi_attach(); it is called before child devices
get their drivers attached.
Therefore, part of the controller initialization will only run if the
device is suspended between the calls to mipi_dsi_host_register() and
component_add() (which ends up calling sun6i_dsi_encoder_enable()).
Again, as shown by the above call graph, this is not the case. It
appears that the controller happens to work because it is still
initialized by the bootloader.
Because the connector is hardcoded to always be connected, the
device's runtime PM reference is not dropped until system suspend,
when sun4i_drv_drm_sys_suspend() ends up calling
sun6i_dsi_encoder_disable(). However, that is done as a system sleep
PM hook, and at that point the system PM core has already taken
another runtime PM reference, so sun6i_dsi_runtime_suspend() is
not called. Likewise, by the time the PM core releases its reference,
sun4i_drv_drm_sys_resume() has already re-enabled the encoder.
So after system suspend and resume, we have *still never called*
sun6i_dsi_inst_init(), and now that the rest of the display pipeline
has been reset, the DSI host is unable to communicate with the panel,
causing VBLANK timeouts.
Fix all of these issues by inlining the runtime PM hooks into the
encoder enable/disable functions, which are guaranteed to run after a
panel is attached. This allows sun6i_dsi_inst_init() to be called
unconditionally. Furthermore, this causes the hardware to be turned off
during system suspend and reinitialized on resume, which was not
happening before.
Fixes: 133add5b5a ("drm/sun4i: Add Allwinner A31 MIPI-DSI controller support")
Signed-off-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211072858.30784-4-samuel@sholland.org
Currently, the DSI host blocks binding the display pipeline until the
panel is available. This unnecessarily prevents other display outpus
from working, and adds logspam to dmesg when the panel driver is built
as a module (the component master is unsuccessfully brought up several
times during boot).
Flip the dependency, instead requiring the host to be bound before the
panel is attached. The panel driver provides no functionality outside of
the display pipeline anyway.
Since the panel is now probed after the DRM connector, we need a hotplug
event to turn on the connector after the panel is attached.
This has the added benefit of fixing panel module removal/insertion.
Previously, the panel would be turned off when its module was removed.
But because the connector state was hardcoded, nothing knew to turn the
panel back on when it was re-attached. Now, with hotplug events
available, the connector state will follow the panel module state, and
the panel will be re-enabled properly.
Fixes: 133add5b5a ("drm/sun4i: Add Allwinner A31 MIPI-DSI controller support")
Signed-off-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211072858.30784-3-samuel@sholland.org
The continued use of an ERR_PTR to signify "no panel" outside of
sun6i_dsi_attach is confusing because it is a double negative. Because
the connector always reports itself as connected, there is also the
possibility of sending an ERR_PTR to drm_panel_get_modes(), which would
crash.
Solve both of these by only storing the panel pointer if it is valid.
Fixes: 133add5b5a ("drm/sun4i: Add Allwinner A31 MIPI-DSI controller support")
Signed-off-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211072858.30784-2-samuel@sholland.org
drm/i915 fixes for v5.6-rc2
Most of these were aimed at a "next fixes" pull already during the merge
window, but there were issues with the baseline I used, which resulted
in a lot of issues in CI. I've regenerated this stuff piecemeal now,
adding gradually to it, and it seems healthy now.
Due to the issues this is much bigger than I'd like. But it was
obviously necessary to take the time to ensure it's not garbage...
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/878sl6yfrn.fsf@intel.com
Unlike DP 1.2 edid corruption test, DP 1.4 requires to calculate
real CRC value of the last edid data block, and write it back.
Current edid CRC calculates routine adds the last CRC byte,
and check if non-zero.
This behavior is not accurate; actually, we need to return
the actual CRC value when corruption is detected.
This commit changes this issue by returning the calculated CRC,
and initiate the required sequence.
Change since v7
- Fix for CI.CHECKPATCH
Change since v6
- Add return check
Change since v5
- Obtain real CRC value before dumping bad edid
Change since v4
- Fix for CI.CHECKPATCH
Change since v3
- Fix a minor typo.
Change since v2
- Rewrite checksum computation routine to avoid duplicated code.
- Rename to avoid confusion.
Change since v1
- Have separate routine for returning real CRC.
Signed-off-by: Jerry (Fangzhi) Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Rodrigo Siqueira <rodrigosiqueiramelo@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211160832.24259-1-Jerry.Zuo@amd.com
Move vga switcheroo and dsm handler register later in
i915_driver_register(), and unregister in i915_driver_unregister(). The
dsm handler unregister is a nop, and is only added for completeness.
My unsubstantiated suspicion is that the vga switcheroo state change
would not work as early as we register the hooks currently. In any case
exposing the interfaces to the world only after we've got everything set
up seems prudent.
Also replace the error handling in vga switcheroo register with a simple
error message. This is done at the same time due to lack of error
propagation from i915_driver_register().
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Acked-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211162802.16180-1-jani.nikula@intel.com
The component order between the two was swapped, resulting in incorrect
color when games with 565 visual hit the overlay path instead of GPU
composition.
Fixes: 25fdd5933e ("drm/msm: Add SDM845 DPU support")
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
For a given byte clock, if VCO recalc value is exactly same as
vco set rate value, vco_set_rate does not get called assuming
VCO is already set to required value. But Due to GDSC toggle,
VCO values are erased in the HW. To make sure VCO is programmed
correctly, we forcefully call set_rate from vco_prepare.
Signed-off-by: Harigovindan P <harigovi@codeaurora.org>
Reviewed-by: Jeffrey Hugo <jeffrey.l.hugo@gmail.com>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Save pll state before dsi host is powered off. Without this change
some register values gets resetted.
Signed-off-by: Harigovindan P <harigovi@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Since 'smu_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)' will always return
false considering the 'smu_system_features_control(smu, false)' disabled
all SMU features.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Originally, I did not expect having to rewind a context upon
timeslicing: the point was to replace the executing context with a
non-executing one! However, given a second context that depends on
requests from the first, we may have to split the requests along the
first context to execute the second, causing us to partially replay the
first context and so have to rewind its RING_TAIL.
References: 5ba32c7be8 ("drm/i915/execlists: Always force a context reload when rewinding RING_TAIL")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200213140150.3639027-1-chris@chris-wilson.co.uk
All non-legacy users of VBLANK functions in struct drm_driver have been
converted to use the respective interfaces in struct drm_crtc_funcs. The
remaining users of VBLANK callbacks in struct drm_driver are legacy drivers
with userspace modesetting.
All users of struct drm_driver.get_scanout_position() have been
converted to the respective CRTC helper function. Remove the callback
from struct drm_driver.
There are no users left of get_vblank_timestamp(), so the callback is
being removed. The other VBLANK callbacks are being moved to the legacy
section at the end of struct drm_driver.
Also removed is drm_calc_vbltimestamp_from_scanoutpos(). Callers of this
function have been converted to use the CRTC instead.
v4:
* more readable code for setting high_prec (Ville, Jani)
v2:
* merge with removal of struct drm_driver.get_scanout_position()
* remove drm_calc_vbltimestamp_from_scanoutpos()
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Tested-by: Yannick Fertré <yannick.fertre@st.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200123135943.24140-22-tzimmermann@suse.de
