This series adds a deadline hint to fences, so realtime deadlines
such as vblank can be communicated to the fence signaller for power/
frequency management decisions.
This is partially inspired by a trick i915 does, but implemented
via dma-fence for a couple of reasons:
1) To continue to be able to use the atomic helpers
2) To support cases where display and gpu are different drivers
See https://patchwork.freedesktop.org/series/93035/
This does not yet add any UAPI, although this will be needed in
a number of cases:
1) Workloads "ping-ponging" between CPU and GPU, where we don't
want the GPU freq governor to interpret time stalled waiting
for GPU as "idle" time
2) Cases where the compositor is waiting for fences to be signaled
before issuing the atomic ioctl, for example to maintain 60fps
cursor updates even when the GPU is not able to maintain that
framerate.
Signed-off-by: Rob Clark <robdclark@chromium.org>
For an atomic commit updating a single CRTC (ie. a pageflip) calculate
the next vblank time, and inform the fence(s) of that deadline.
v2: Comment typo fix (danvet)
v3: If there are multiple CRTCs, consider the time of the soonest vblank
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Will be used in the next commit to set a deadline on fences that an
atomic update is waiting on.
v2: Calculate time at *start* of vblank period, not end
v3: Fix kbuild complaints
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Mario Kleiner <mario.kleiner.de@gmail.com>
As the finished fence is the one that is exposed to userspace, and
therefore the one that other operations, like atomic update, would
block on, we need to propagate the deadline from from the finished
fence to the actual hw fence.
v2: Split into drm_sched_fence_set_parent() (ckoenig)
v3: Ensure a thread calling drm_sched_fence_set_deadline_finished() sees
fence->parent set before drm_sched_fence_set_parent() does this
test_bit(DMA_FENCE_FLAG_HAS_DEADLINE_BIT).
Signed-off-by: Rob Clark <robdclark@chromium.org>
Acked-by: Luben Tuikov <luben.tuikov@amd.com>
It is preferred to use typed property access functions (i.e.
of_property_read_<type> functions) rather than low-level
of_get_property/of_find_property functions for reading properties. As
part of this, convert of_get_property/of_find_property calls to the
recently added of_property_present() helper when we just want to test
for presence of a property and nothing more.
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Reviewed-by: Liu Ying <victor.liu@nxp.com> # i.MX bridge
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://lore.kernel.org/r/20230310144705.1542207-1-robh@kernel.org
Signed-off-by: Rob Herring <robh@kernel.org>
The WA states that we need to alert the GSC FW before doing a GSC engine
reset and then wait for 200ms. The GuC owns engine reset, so on the i915
side we only need to apply this for full GT reset.
Given that we do full GT resets in the resume paths to cleanup the HW
state and that a long wait in those scenarios would not be acceptable,
a faster path has been introduced where, if the GSC is idle, we try first
to individually reset the GuC and all engines except the GSC and only fall
back to full reset if that fails.
Note: according to the WA specs, if the GSC is idle it should be possible
to only wait for the uC wakeup time (~15ms) instead of the whole 200ms.
However, the GSC FW team have mentioned that the wakeup time can change
based on other things going on in the HW and pcode, so a good security
margin would be required. Given that when the GSC is idle we already
skip the wait & reset entirely and that this reduced wait would still
likely be too long to use in resume paths, it's not worth adding support
for this reduced wait.
v2: add comment to explain why it is safe to skip the GSC reset (John)
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: John Harrison <john.c.harrison@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230323231857.2194435-2-daniele.ceraolospurio@intel.com
Commit 3db9d59055 ("drm/i915/gt: Reset twice") modified the code to
always hit the GDRST register twice when doing a reset, with the
reported aim to fix invalid post-reset engine state on some platforms
(Jasperlake being the only one actually mentioned).
This is a problem on MTL, due to the fact that we have to apply a time
consuming WA (coming in the next patch) every time we hit the GDRST
register in a way that can include the GSC engine. Even post MTL, the
expectation is that we'll have some work to do before and after hitting
the GDRST if the GSC is involved.
Since the issue requiring the double reset seems to be limited to older
platforms, instead of trying to handle the double-reset on MTL and
future platforms it is just easier to turn it off. The default on MTL is
also for GuC to own engine reset, with i915 only covering full-GT reset.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com>
Cc: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230323231857.2194435-1-daniele.ceraolospurio@intel.com
In the rare case where we do a full GT reset after starting the HuC
load and before it completes (which basically boils down to i915 hanging
during init), we need to cancel the delayed load fence, as it will be
re-initialized in the post-reset recovery.
Fixes: 27536e0327 ("drm/i915/huc: track delayed HuC load with a fence")
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Alan Previn <alan.previn.teres.alexis@intel.com>
Reviewed-by: Alan Previn <alan.previn.teres.alexis@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230313205556.1174503-1-daniele.ceraolospurio@intel.com
Essentially this is a revert of commit d9d54a530a ("drm/i915: Put
future HW and their uAPIs under STAGING & BROKEN").
We currently have no users for this config option. The last one was
removed in commit 8c26491f58 ("drm/i915: Kill the fake lmem
support"). Drop it altogether; it's easy enough to resurrect if need
arises.
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230327105330.312131-2-jani.nikula@intel.com
Although we now sanitycheck MMIO access during driver load to make sure
the MMIO BAR isn't returning all 0xFFFFFFFF, there have been a few cases
where (temporarily?) unreliable MMIO access has happened after GPU
resets or power events. We'll often notice this on our next GT register
access since forcewake handling will fail; let's change our handling
slightly so that when this happens we print a more meaningful message
clarifying that the problem is the MMIO access, not forcewake
specifically.
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230327195547.356584-3-andi.shyti@linux.intel.com
We occasionally see the PCI device in a non-accessible state at the
point the driver is loaded. When this happens, all BAR accesses will
read back as 0xFFFFFFFF. Rather than reading registers and
misinterpreting their (invalid) values, let's specifically check for
0xFFFFFFFF in a register that cannot have that value to see if the
device is accessible.
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230327195547.356584-2-andi.shyti@linux.intel.com
Add extras to support i.MX8M Plus. The main change is the removal of
HS/VS/DE signal inversion in the LCDIFv3-DSIM glue logic, otherwise
the implementation of this IP in i.MX8M Plus is very much compatible
with the i.MX8M Mini/Nano one.
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Acked-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <m.szyprowski@samsung.com>
Samsung MIPI DSIM master can also be found in i.MX8M Mini/Nano SoC.
Add compatible and associated driver_data for it.
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Acked-by: Robert Foss <robert.foss@linaro.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Tested-by: Adam Ford <aford173@gmail.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Samsung MIPI DSIM controller is common DSI IP that can be used in various
SoCs like Exynos, i.MX8M Mini/Nano.
In order to access this DSI controller between various platform SoCs,
the ideal way to incorporate this in the drm stack is via the drm bridge
driver.
We already have a consolidated code for supporting component and bridge
based DRM drivers, so keep the exynos component based code in existing
exynos_drm_dsi.c and move generic bridge code as part of samsung-dsim.c
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
IRQ handler for te-gpio seems to be common across DSIM host.
However, Exynos is handling this via CRTC drivers but there is no clear
evidence on how the same has been handled in i.MX8MM. Keeping the handler
as-it-is can be a viable option but adding DSIM bridge core in upcoming
patches is not possible to call Exynos CRTC handler as DSIM bridge has
to be common across DRM bridge core instead of platform specific DRM
drivers like Exynos here.
So, this patch handles the handler via platform host helper, so-that
handling platform specific hook across Exynos and generic can be
reasonable till it makes it generic across all platforms.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
DSI host registration, attach and detach operations are quite
different for the component and bridge-based DRM drivers.
Supporting generic bridge driver to use both component and bridge
based DRM drivers can be tricky and would require additional host
related operation hooks.
Add host operation hooks for registering and unregistering Exynos
and generic drivers, where Exynos hooks are used in existing Exynos
component based DRM drivers and generic hooks are used in i.MX8M
bridge based DRM drivers.
Add host attach and detach operation hooks for Exynos component
DRM drivers and those get invoked while DSI core host attach and
detach gets called.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Finding the right input bus format throughout the pipeline is hard
so add atomic_get_input_bus_fmts callback and initialize with the
proper input format from list of supported output formats.
This format can be used in pipeline for negotiating bus format between
the DSI-end of this bridge and the other component closer to pipeline
components.
List of Pixel formats are taken from,
AN13573 i.MX 8/RT MIPI DSI/CSI-2, Rev. 0, 21 March 2022
3.7.4 Pixel formats
Table 14. DSI pixel packing formats
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
LCDIF-DSIM glue logic inverts the HS/VS/DE signals and expecting
the i.MX8M Mini/Nano DSI host to add additional Data Enable signal
active low (DE_LOW). This makes the valid data transfer on each
horizontal line.
So, add additional bus flags DE_LOW setting via input_bus_flags
for i.MX8M Mini/Nano platforms.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Suggested-by: Marek Vasut <marex@denx.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Look like an explicit fixing up of mode_flags is required for DSIM IP
present in i.MX8M Mini/Nano SoCs.
At least the LCDIF + DSIM needs active low sync polarities in order
to correlate the correct sync flags of the surrounding components in
the chain to make sure the whole pipeline can work properly.
On the other hand the i.MX 8M Mini Applications Processor Reference Manual,
Rev. 3, 11/2020 says.
"13.6.3.5.2 RGB interface
Vsync, Hsync, and VDEN are active high signals."
i.MX 8M Mini Applications Processor Reference Manual Rev. 3, 11/2020
3.6.3.5.2 RGB interface
i.MX 8M Nano Applications Processor Reference Manual Rev. 2, 07/2022
13.6.2.7.2 RGB interface
both claim "Vsync, Hsync, and VDEN are active high signals.", the
LCDIF must generate inverted HS/VS/DE signals, i.e. active LOW.
No clear evidence about whether it can be documentation issues or
something, so added proper comments on the code.
Comments are suggested by Marek Vasut.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Host transfer() in the DSI master will invoke only when the DSI commands
are sent from DSI devices like DSI Panel or DSI bridges and this host
the transfer wouldn't invoke for I2C-based-DSI bridge drivers.
Handling DSI host initialization in transfer calls misses the controller
setup for I2C configured DSI bridges.
This patch updates the DSI host initialization by calling host to init
from bridge pre_enable as the bridge pre_enable API is invoked by core
as it is common across all classes of DSI device drivers.
The host init during pre_enable is conditional and not invoked for Exynos
as existing downstream drm panels and bridges in Exynos are expecting
the host initialization during DSI transfer.
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Samsung MIPI DSIM controller is common DSI IP that can be used
in various SoCs like Exynos, i.MX8M Mini/Nano/Plus.
Add hw_type enum via platform_data so that accessing the different
controller data between various platforms becomes easy and meaningful.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Suggested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Look like PLL PMS_P offset value varies between platforms that have
Samsung DSIM IP.
However, there is no clear evidence for it as both Exynos and i.MX
8M Mini Application Processor Reference Manual is still referring
the PMS_P offset as 13.
The offset 13 is not working for i.MX8M Mini SoCs but the downstream
NXP sec-dsim.c driver is using offset 14 for i.MX8M Mini SoC platforms
[1] [2].
PMS_P value set in sec_mipi_dsim_check_pll_out using PLLCTRL_SET_P()
with offset 13 and then an additional offset of one bit added in
sec_mipi_dsim_config_pll via PLLCTRL_SET_PMS().
Not sure whether it is reference manual documentation or something
else but this patch trusts the downstream code and handle PLL_P offset
via platform driver data so-that imx8mm driver data shall use
pll_p_offset to 14.
Similar to Mini the i.MX8M Nano/Plus also has P=14, unlike Exynos.
[1] https://source.codeaurora.org/external/imx/linux-imx/tree/drivers/gpu/drm/bridge/sec-dsim.c?h=imx_5.4.47_2.2.0#n210
[2] https://source.codeaurora.org/external/imx/linux-imx/tree/drivers/gpu/drm/bridge/sec-dsim.c?h=imx_5.4.47_2.2.0#n211
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
The same Samsung MIPI DSIM master can also be used in NXP's
i.MX8M Mini/Nano/Plus SoC.
In i.MX8M Mini/Nano/Plus SoC the DSI Phy requires a MIPI DPHY
bit to reset in order to activate the PHY and that can be done
via upstream i.MX8M blk-ctrl driver.
So, mark the phy get as optional.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
In general, for MIPI DSI there are three ways to represent the
pipeline for an upstream bridge to find the connected downstream
panel or bridge.
1. Child panel or bridge as a conventional device tree child node.
2. Child panel or bridge as an OF-graph port node.
3. Child panel or bridge as an OF-graph ports node.
There are three different downstream panels or bridges that are
possible to connect an upstream DSI host bridge - DSI Panel,
DSI Bridge, and I2C-Configured DSI bridge.
An example of the downstream panel represented as a child node,
&dsi {
compatible = "samsung,exynos5433-mipi-dsi";
ports {
port@0 {
reg = <0>;
dsi_to_mic: endpoint {
remote-endpoint = <&mic_to_dsi>;
};
};
};
panel@0 {
reg = <0>;
};
};
An example of the downstream bridge represented as a port node,
&i2c4 {
bridge@2c {
compatible = "ti,sn65dsi84";
ports {
port@0 {
reg = <0>;
bridge_in_dsi: endpoint {
remote-endpoint = <&dsi_out_bridge>;
data-lanes = <1 2>;
};
};
port@2 {
reg = <2>;
bridge_out_panel: endpoint {
remote-endpoint = <&panel_out_bridge>;
};
};
};
};
};
&dsi {
compatible = "fsl,imx8mm-mipi-dsim";
port {
dsi_in_lcdif: endpoint@0 {
reg = <0>;
remote-endpoint = <&lcdif_out_dsi>;
};
dsi_out_bridge: endpoint@1 {
reg = <1>;
remote-endpoint = <&bridge_in_dsi>;
};
};
};
An example of the downstream bridge represented as a ports node,
&dsi {
compatible = "fsl,imx8mm-mipi-dsim";
ports {
port@0 {
reg = <0>;
dsi_in_lcdif: endpoint@0 {
reg = <0>;
remote-endpoint = <&lcdif_out_dsi>;
};
};
port@1 {
reg = <1>;
dsi_out_bridge: endpoint {
remote-endpoint = <&bridge_in_dsi>;
};
};
};
In, summary it is possible to represent all three downstream slaves
devices using OF-graph port or ports node however only DSI Panel and
DSI Bridge are possible but not possible to represent I2C-Configured
DSI bridge child nodes since I2C-Configure bridges are child of I2C
node, not upstream DSI host bridge and it is must represent them
endpoint port linking.
This indeed means, the OF-graph port or ports representation is
mandatory for I2C-Configured DSI bridges.
This patch tries to add an OF-graph port or ports representation
detection code on top of existing child node detection.
It is possible to replace the entire detection code using existing
drm_of helper drm_of_find_panel_or_bridge but it will break the
Exynos DSI since the pipeline doesn't support OF-graph port or ports
node.
Overall, this patch has a combination of child and OF-graph pipeline
detections in order to support the backward compatibility of Exynos
DSI child node and i.MX8M Mini/Nano/Plus OF-graph port or ports
node pipelines.
This is the first common DSI host bridge driver that needs to support
all possible downstream connection pipeline combinations.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Exynos DSI already converted into a bridge driver, so bridge
detach will suppose happened during bridge chain removal done
by the bridge core.
Drop the explicit call chain to detach the bridge.
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Add driver data of mt8188 vdosys0 to mediatek-drm and the sub driver.
Signed-off-by: amy zhang <Amy.Zhang@mediatek.com>
Signed-off-by: Nathan Lu <nathan.lu@mediatek.com>
Signed-off-by: Jason-JH.Lin <jason-jh.lin@mediatek.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com>
Link: https://patchwork.kernel.org/project/linux-mediatek/patch/20230324100553.13719-1-jason-jh.lin@mediatek.com/
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
This panel is found in Lenovo Flex 5G laptop, so add the entry for it
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Acked-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20230324165909.131831-1-vkoul@kernel.org
A kzalloc()+memcpy() can be optimized in a single kmemdup().
This saves a few cycles because some memory doesn't need to be zeroed.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Variable dppclk_delay_subtotal is not effectively used, so delete it.
drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn314/display_rq_dlg_calc_314.c:1004:15: warning: variable 'dppclk_delay_subtotal' set but not used.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Link: https://bugzilla.openanolis.cn/show_bug.cgi?id=4584
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
Skip TMR unload for Navi12 and CHIP_SIENNA_CICHLID SRIOV as TMR is
not loaded at all
Signed-off-by: Tong Liu01 <Tong.Liu01@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Compiling AMD GPU drivers displays two warnings:
drivers/gpu/drm/scheduler/sched_main.c:738: warning: Function parameter or member 'file' not described in 'drm_sched_job_add_syncobj_dependency'
drivers/gpu/drm/scheduler/sched_main.c:738: warning: Excess function
parameter 'file_private' description in
'drm_sched_job_add_syncobj_dependency'
Get rid of them by renaming the variable name on the function description
Signed-off-by: Caio Novais <caionovais@usp.br>
Link: https://lore.kernel.org/r/20230325131532.6356-1-caionovais@usp.br
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Wrong register address is used to read the SAG block time. Fix
the register address according to the bspec.
Bspec: 64608
Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com>
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230323114426.41136-3-vinod.govindapillai@intel.com
Wrong offsets are calculated to read QGV point registers. Fix it
to read from the correct registers.
Bspec: 64602
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com>
Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230323114426.41136-2-vinod.govindapillai@intel.com
If we fail to adjust the GuC run-control on opening the perf stream,
make sure we unwind the wakeref just taken.
v2: Retain old goto label names (Ashutosh)
v3: Drop bitfield boolean
Fixes: 01e7427467 ("drm/i915/guc: Support OA when Wa_16011777198 is enabled")
Signed-off-by: Chris Wilson <chris.p.wilson@linux.intel.com>
Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230323225901.3743681-2-umesh.nerlige.ramappa@intel.com
(cherry picked from commit 2810ac6c75)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Currently i915_gem_object_is_framebuffer() doesn't treat the
BO containing the framebuffer's DPT as a framebuffer itself.
This means eg. that the shrinker can evict the DPT BO while
leaving the actual FB BO bound, when the DPT is allocated
from regular shmem.
That causes an immediate oops during hibernate as we
try to rewrite the PTEs inside the already evicted
DPT obj.
TODO: presumably this might also be the reason for the
DPT related display faults under heavy memory pressure,
but I'm still not sure how that would happen as the object
should be pinned by intel_dpt_pin() while in active use by
the display engine...
Cc: stable@vger.kernel.org
Cc: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Imre Deak <imre.deak@intel.com>
Fixes: 0dc987b699 ("drm/i915/display: Add smem fallback allocation for dpt")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320090522.9909-2-ville.syrjala@linux.intel.com
Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
(cherry picked from commit 779cb5ba64)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
i915_gem_object_create_lmem_from_data() lacks the flush of the data
written to lmem to ensure the object is marked as dirty and the writes
flushed to the backing store. Once created, we can immediately release
the obj->mm.mapping caching of the vmap.
Fixes: 7acbbc7cf4 ("drm/i915/guc: put all guc objects in lmem when available")
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: Chris Wilson <chris.p.wilson@linux.intel.com>
Cc: <stable@vger.kernel.org> # v5.16+
Signed-off-by: Nirmoy Das <nirmoy.das@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Reviewed-by: Nirmoy Das <nirmoy.das@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230316165918.13074-1-nirmoy.das@intel.com
(cherry picked from commit e2ee10474c)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
The commit renaming icl_tc_phy_is_in_safe_mode() to
icl_tc_phy_take_ownership() didn't flip the function's return value
accordingly, fix this up.
This didn't cause an actual problem besides state check errors, since
the function is only used during HW readout.
Cc: José Roberto de Souza <jose.souza@intel.com>
Fixes: f53979d68a ("drm/i915/display/tc: Rename safe_mode functions ownership")
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230316131724.359612-4-imre.deak@intel.com
(cherry picked from commit f2c7959dda)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Keeping DC states enabled is incompatible with the _noarm()/_arm()
split we use for writing pipe/plane registers. When DC5 and PSR
are enabled, all pipe/plane registers effectively become self-arming
on account of DC5 exit arming the update, and PSR exit latching it.
What probably saves us most of the time is that (with PIPE_MISC[21]=0)
all pipe register writes themselves trigger PSR exit, and then
we don't re-enter PSR until the idle frame count has elapsed.
So it may be that the PSR exit happens already before we've
updated the state too much.
Also the PSR1 panel (at least on this KBL) seems to discard the first
frame we trasmit, presumably still scanning out from its internal
framebuffer at that point. So only the second frame we transmit is
actually visible. But I suppose that could also be panel specific
behaviour. I haven't checked out how other PSR panels behave, nor
did I bother to check what the eDP spec has to say about this.
And since this really is all about DC states, let's switch from
the MODESET domain to the DC_OFF domain. Functionally they are
100% identical. We should probably remove the MODESET domain...
And for good measure let's toss in an assert to the place where
we do the _noarm() register writes to make sure DC states are
in fact off.
v2: Just use intel_display_power_is_enabled() (Imre)
Cc: <stable@vger.kernel.org> #v5.17+
Cc: Manasi Navare <navaremanasi@google.com>
Cc: Drew Davenport <ddavenport@chromium.org>
Cc: Jouni Högander <jouni.hogander@intel.com>
Reviewed-by: Imre Deak <imre.deak@intel.com>
Fixes: d13dde4495 ("drm/i915: Split pipe+output CSC programming to noarm+arm pair")
Fixes: f8a005eb89 ("drm/i915: Optimize icl+ universal plane programming")
Fixes: 890b6ec4a5 ("drm/i915: Split skl+ plane update into noarm+arm pair")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320183532.17727-1-ville.syrjala@linux.intel.com
(cherry picked from commit 41b4c7fe72)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Unlike SKL/GLK the ICL CSC unit suffers from a new issue where
CSC_MODE arming is sticky. That is, once armed it remains armed
causing the CSC coeff/offset registers to become effectively
self-arming.
CSC coeff/offset registers writes no longer disarm the CSC,
but fortunately register read still do. So we can use that
to disarm the CSC unit once the registers for the current
frame have been latched. This avoid s the self-arming behaviour
from persisting into the next frame's .color_commit_noarm()
call.
Cc: <stable@vger.kernel.org> #v5.19+
Cc: Manasi Navare <navaremanasi@google.com>
Cc: Drew Davenport <ddavenport@chromium.org>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Jouni Högander <jouni.hogander@intel.com>
Fixes: d13dde4495 ("drm/i915: Split pipe+output CSC programming to noarm+arm pair")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320095438.17328-5-ville.syrjala@linux.intel.com
Reviewed-by: Imre Deak <imre.deak@intel.com>
(cherry picked from commit 92736f1b45)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
We're going to need stuff after the color management
register latching has happened. Add a corresponding hook.
Cc: <stable@vger.kernel.org> #v5.19+
Cc: Manasi Navare <navaremanasi@google.com>
Cc: Drew Davenport <ddavenport@chromium.org>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Jouni Högander <jouni.hogander@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320095438.17328-4-ville.syrjala@linux.intel.com
Reviewed-by: Imre Deak <imre.deak@intel.com>
(cherry picked from commit 3962ca4e08)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
SKL/GLK CSC unit suffers from a nasty issue where a CSC
coeff/offset register read or write between DC5 exit and
PSR exit will undo the CSC arming performed by DMC, and
then during PSR exit the hardware will latch zeroes into
the active CSC registers. This causes any plane going
through the CSC to output all black.
We can sidestep the issue by making sure the PSR exit has
already actually happened before we touch the CSC coeff/offset
registers. Easiest way to guarantee that is to just move the
CSC programming back into the .color_commir_arm() as we force
a PSR exit (and crucially wait for it to actually happen)
prior to touching the arming registers.
When PSR (and thus also DC states) are disabled we don't
have anything to worry about, so we can keep using the
more optional _noarm() hook for writing the CSC registers.
Cc: <stable@vger.kernel.org> #v5.19+
Cc: Manasi Navare <navaremanasi@google.com>
Cc: Drew Davenport <ddavenport@chromium.org>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Jouni Högander <jouni.hogander@intel.com>
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/8283
Fixes: d13dde4495 ("drm/i915: Split pipe+output CSC programming to noarm+arm pair")
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320095438.17328-3-ville.syrjala@linux.intel.com
Reviewed-by: Imre Deak <imre.deak@intel.com>
(cherry picked from commit 80a892a4c2)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
We're going to want different behavior for skl/glk vs. icl
in .color_commit_noarm(), so split the hook into two. Arguably
we already had slightly different behaviour since
csc_enable/gamma_enable are never set on icl+, so the old
code was perhaps a bit confusing as well.
Cc: <stable@vger.kernel.org> #v5.19+
Cc: Manasi Navare <navaremanasi@google.com>
Cc: Drew Davenport <ddavenport@chromium.org>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Jouni Högander <jouni.hogander@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320095438.17328-2-ville.syrjala@linux.intel.com
Reviewed-by: Imre Deak <imre.deak@intel.com>
(cherry picked from commit f161eb01f5)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>