Laurent Pinchart [Mon, 15 Jan 2024 02:16:29 +0000 (04:16 +0200)]
media: nxp: imx8-isi: Mark all crossbar sink pads as MUST_CONNECT
All the sink pads of the crossbar switch require an active link if
they're part of the pipeline. Mark them with the
MEDIA_PAD_FL_MUST_CONNECT flag to fail pipeline validation if they're
not connected. This allows removing a manual check when translating
streams.
Cc: stable@vger.kernel.org # 6.1 Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Laurent Pinchart [Sun, 14 Jan 2024 23:04:52 +0000 (01:04 +0200)]
media: mc: Expand MUST_CONNECT flag to always require an enabled link
The MEDIA_PAD_FL_MUST_CONNECT flag indicates that the pad requires an
enabled link to stream, but only if it has any link at all. This makes
little sense, as if a pad is part of a pipeline, there are very few use
cases for an active link to be mandatory only if links exist at all. A
review of in-tree drivers confirms they all need an enabled link for
pads marked with the MEDIA_PAD_FL_MUST_CONNECT flag.
Expand the scope of the flag by rejecting pads that have no links at
all. This requires modifying the pipeline build code to add those pads
to the pipeline.
Cc: stable@vger.kernel.org # 6.1 Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Laurent Pinchart [Sun, 14 Jan 2024 22:30:02 +0000 (00:30 +0200)]
media: mc: Rename pad variable to clarify intent
The pad local variable in the media_pipeline_explore_next_link()
function is used to store the pad through which the entity has been
reached. Rename it to origin to reflect that and make the code easier to
read. This will be even more important in subsequent commits when
expanding the function with additional logic.
Cc: stable@vger.kernel.org # 6.1 Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Laurent Pinchart [Sun, 14 Jan 2024 22:30:02 +0000 (00:30 +0200)]
media: mc: Add num_links flag to media_pad
Maintain a counter of the links connected to a pad in the media_pad
structure. This helps checking if a pad is connected to anything, which
will be used in the pipeline building code.
Cc: stable@vger.kernel.org # 6.1 Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Marek Vasut [Fri, 1 Dec 2023 15:06:04 +0000 (16:06 +0100)]
media: nxp: imx8-isi: Check whether crossbar pad is non-NULL before access
When translating source to sink streams in the crossbar subdev, the
driver tries to locate the remote subdev connected to the sink pad. The
remote pad may be NULL, if userspace tries to enable a stream that ends
at an unconnected crossbar sink. When that occurs, the driver
dereferences the NULL pad, leading to a crash.
Prevent the crash by checking if the pad is NULL before using it, and
return an error if it is.
Laurent Pinchart [Sun, 14 Jan 2024 22:24:12 +0000 (00:24 +0200)]
media: mc: Fix flags handling when creating pad links
The media_create_pad_link() function doesn't correctly clear reject link
type flags, nor does it set the DATA_LINK flag. It only works because
the MEDIA_LNK_FL_DATA_LINK flag's value is 0.
Fix it by returning an error if any link type flag is set. This doesn't
introduce any regression, as nobody calls the media_create_pad_link()
function with link type flags (easily checked by grepping for the flag
in the source code, there are very few hits).
Set the MEDIA_LNK_FL_DATA_LINK explicitly, which is a no-op that the
compiler will optimize out, but is still useful to make the code more
explicit and easier to understand.
Cc: stable@vger.kernel.org # 6.1 Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Laurent Pinchart [Sun, 14 Jan 2024 13:55:40 +0000 (15:55 +0200)]
media: mc: Add local pad to pipeline regardless of the link state
When building pipelines by following links, the
media_pipeline_explore_next_link() function only traverses enabled
links. The remote pad of a disabled link is not added to the pipeline,
and neither is the local pad. While the former is correct as disabled
links should not be followed, not adding the local pad breaks processing
of the MEDIA_PAD_FL_MUST_CONNECT flag.
The MEDIA_PAD_FL_MUST_CONNECT flag is checked in the
__media_pipeline_start() function that iterates over all pads after
populating the pipeline. If the pad is not present, the check gets
skipped, rendering it useless.
Fix this by adding the local pad of all links regardless of their state,
only skipping the remote pad for disabled links.
Ricardo B. Marliere [Sat, 3 Feb 2024 15:31:27 +0000 (12:31 -0300)]
media: media-devnode: make media_bus_type const
Now that the driver core can properly handle constant struct bus_type,
move the media_bus_type variable to be a constant structure as well,
placing it into read-only memory which can not be modified at runtime.
Ricardo B. Marliere [Sat, 3 Feb 2024 15:31:26 +0000 (12:31 -0300)]
media: cec: make cec_bus_type const
Now that the driver core can properly handle constant struct bus_type,
move the cec_bus_type variable to be a constant structure as well,
placing it into read-only memory which can not be modified at runtime.
Rob Herring [Fri, 2 Feb 2024 22:23:25 +0000 (16:23 -0600)]
media: dt-bindings: techwell,tw9900: Fix port schema ref
The port@0 node doesn't define any extra properties in the port or endpoint
nodes, so the $ref should be to "/properties/port" instead as it restricts
extra properties.
Fixes: 0f82ffa9a295 ("media: dt-bindings: media: i2c: Add bindings for TW9900") Signed-off-by: Rob Herring <robh@kernel.org> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Hans Verkuil [Fri, 2 Feb 2024 08:25:44 +0000 (09:25 +0100)]
media: v4l2-common.h: kerneldoc: correctly format return values
Building the kerneldoc resulted in two errors:
Documentation/media/driver-api/v4l2-common:6: ./include/media/v4l2-common.h:566: ERROR: Unexpected indentation.
Documentation/media/driver-api/v4l2-common:6: ./include/media/v4l2-common.h:567: WARNING: Block quote ends without a blank line; unexpected unindent.
Format v4l2_link_freq_to_bitmap according to the kerneldoc standard.
The v4l2_fill_pixfmt_mp function also had incorrect return value
formatting, although that didn't report an error/warning, but it looked
ugly in the generated documentation. So fix that one as well.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Fixes: a68e88e2cf9e ("media: v4l: Add a helper for setting up link-frequencies control") Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Zhipeng Lu [Thu, 1 Feb 2024 12:48:44 +0000 (20:48 +0800)]
media: v4l2-mem2mem: fix a memleak in v4l2_m2m_register_entity
The entity->name (i.e. name) is allocated in v4l2_m2m_register_entity
but isn't freed in its following error-handling paths. This patch
adds such deallocation to prevent memleak of entity->name.
Fixes: be2fff656322 ("media: add helpers for memory-to-memory media controller") Signed-off-by: Zhipeng Lu <alexious@zju.edu.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Zhipeng Lu [Thu, 1 Feb 2024 12:47:53 +0000 (20:47 +0800)]
media: v4l2-tpg: fix some memleaks in tpg_alloc
In tpg_alloc, resources should be deallocated in each and every
error-handling paths, since they are allocated in for statements.
Otherwise there would be memleaks because tpg_free is called only when
tpg_alloc return 0.
Fixes: 63881df94d3e ("[media] vivid: add the Test Pattern Generator") Signed-off-by: Zhipeng Lu <alexious@zju.edu.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Ricardo Ribalda [Fri, 26 Jan 2024 23:16:08 +0000 (23:16 +0000)]
media: mediatek: vcodec: Fix kerneldoc
Those fields have been removed. They do not need to be documented.
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org> Reviewed-by: Randy Dunlap <rdunlap@infradead.org> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Ricardo Ribalda [Fri, 26 Jan 2024 23:16:07 +0000 (23:16 +0000)]
media: mediatek: jpeg: Fix kerneldoc
The field is gone, remove the documentation for it.
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org> Reviewed-by: Randy Dunlap <rdunlap@infradead.org> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Ricardo Ribalda [Fri, 26 Jan 2024 23:16:05 +0000 (23:16 +0000)]
media: i2c: css-quirk.h: Fix kerneldoc
Kerneldoc does not seem to understand that embed doc:
drivers/media/i2c/ccs/ccs-quirk.h:50: warning: Excess struct member 'write' description in 'ccs_quirk'
drivers/media/i2c/ccs/ccs-quirk.h:50: warning: Excess struct member 'reg' description in 'ccs_quirk'
drivers/media/i2c/ccs/ccs-quirk.h:50: warning: Excess struct member 'val' description in 'ccs_quirk'
Convert into a standard doc.
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org> Reviewed-by: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Ricardo Ribalda [Fri, 26 Jan 2024 23:16:01 +0000 (23:16 +0000)]
media: videodev2.h: Fix kerneldoc
Named nested unions need their prefix:
https://www.kernel.org/doc/html/latest/doc-guide/kernel-doc.html#nested-structs-unions
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org> Reviewed-by: Randy Dunlap <rdunlap@infradead.org> Tested-by: Randy Dunlap <rdunlap@infradead.org> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Nini Song [Thu, 25 Jan 2024 13:28:45 +0000 (21:28 +0800)]
media: cec: core: remove length check of Timer Status
The valid_la is used to check the length requirements,
including special cases of Timer Status. If the length is
shorter than 5, that means no Duration Available is returned,
the message will be forced to be invalid.
However, the description of Duration Available in the spec
is that this parameter may be returned when these cases, or
that it can be optionally return when these cases. The key
words in the spec description are flexible choices.
Remove the special length check of Timer Status to fit the
spec which is not compulsory about that.
Signed-off-by: Nini Song <nini.song@mediatek.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Hans Verkuil [Wed, 17 Jan 2024 14:52:04 +0000 (15:52 +0100)]
media: v4l2-ctrls: show all owned controls in log_status
VIDIOC_LOG_STATUS will log the controls owned by the driver. But the
code didn't take into account the case where a single driver creates
multiple control handlers. A good example is the vivid driver, but
others use it as well.
Modify v4l2_ctrl_handler_log_status() so that it really shows all
controls owned by this driver.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Hans Verkuil [Wed, 17 Jan 2024 11:44:03 +0000 (12:44 +0100)]
media: cx231xx: controls are from another device, mark this
The last argument of v4l2_ctrl_add_handler() indicates whether the controls
you add are from a control handler owned by another driver (true) or from
the same driver (false). In this case the last argument was incorrectly set
to false. The controls come from the cx25840 subdev.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Nikita Zhandarovich [Fri, 12 Jan 2024 13:42:26 +0000 (05:42 -0800)]
media: em28xx: annotate unchecked call to media_device_register()
Static analyzers generate alerts for an unchecked call to
`media_device_register()`. However, in this case, the device will work
reliably without the media controller API.
Add a comment above the call to prevent future unnecessary changes.
Gui-Dong Han [Fri, 22 Dec 2023 05:50:30 +0000 (13:50 +0800)]
media: xc4000: Fix atomicity violation in xc4000_get_frequency
In xc4000_get_frequency():
*freq = priv->freq_hz + priv->freq_offset;
The code accesses priv->freq_hz and priv->freq_offset without holding any
lock.
In xc4000_set_params():
// Code that updates priv->freq_hz and priv->freq_offset
...
xc4000_get_frequency() and xc4000_set_params() may execute concurrently,
risking inconsistent reads of priv->freq_hz and priv->freq_offset. Since
these related data may update during reading, it can result in incorrect
frequency calculation, leading to atomicity violations.
This possible bug is found by an experimental static analysis tool
developed by our team, BassCheck[1]. This tool analyzes the locking APIs
to extract function pairs that can be concurrently executed, and then
analyzes the instructions in the paired functions to identify possible
concurrency bugs including data races and atomicity violations. The above
possible bug is reported when our tool analyzes the source code of
Linux 6.2.
To address this issue, it is proposed to add a mutex lock pair in
xc4000_get_frequency() to ensure atomicity. With this patch applied, our
tool no longer reports the possible bug, with the kernel configuration
allyesconfig for x86_64. Due to the lack of associated hardware, we cannot
test the patch in runtime testing, and just verify it according to the
code logic.
[1] https://sites.google.com/view/basscheck/
Fixes: 4c07e32884ab ("[media] xc4000: Fix get_frequency()") Cc: stable@vger.kernel.org Reported-by: BassCheck <bass@buaa.edu.cn> Signed-off-by: Gui-Dong Han <2045gemini@gmail.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Duc-Long, Le [Sun, 15 Oct 2023 00:19:46 +0000 (20:19 -0400)]
media: tc358746: fix the pll calculating function
Following formula of Pll_clk in 5.2 section, 50th page of
TC358746AXBG/748XBG/748IXBG Functional Specification Rev 1.1 document.
The formula of fout is as below:
fout = refclk * mul / (prediv * postdiv)
Remove "p" to avoid using 2 times of prediv in pll calculating function.
Signed-off-by: Duc-Long, Le <duclong.linux@gmail.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Jason Chen [Wed, 24 Jan 2024 14:43:01 +0000 (15:43 +0100)]
media: ov08x40: Reduce start streaming time
Because video duration involves calculating the streaming time, and i2c
communication incurs too many XTALK register settings every 4 bytes with
i2c START and STOP.
So we have opted switch to the i2c burst method.
This method involves writing the XTALK registers in the order of
the register block.
The start streaming time can be reduced from around 400ms to 150ms
[Sakari Ailus: Drop unneeded dev_dbg().]
Signed-off-by: Jason Chen <jason.z.chen@intel.com> Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Jason Chen [Mon, 22 Jan 2024 02:54:34 +0000 (03:54 +0100)]
media: ov08x40: Avoid sensor probing in D0 state
When the system enters the D0 state and attempt to probe the device,
another component, such as LED, will also be pulled high due to the
hardware design. It's advisable to keep the device being probed in
a different D state.
Signed-off-by: Jason Chen <jason.z.chen@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Bo Liu [Wed, 17 Jan 2024 03:09:10 +0000 (04:09 +0100)]
media: i2c: isl7998x: convert to use maple tree register cache
The maple tree register cache is based on a much more modern data structure
than the rbtree cache and makes optimisation choices which are probably
more appropriate for modern systems than those made by the rbtree cache.
Signed-off-by: Bo Liu <liubo03@inspur.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Bo Liu [Wed, 17 Jan 2024 03:10:24 +0000 (04:10 +0100)]
media: i2c: max2175: convert to use maple tree register cache
The maple tree register cache is based on a much more modern data structure
than the rbtree cache and makes optimisation choices which are probably
more appropriate for modern systems than those made by the rbtree cache.
Signed-off-by: Bo Liu <liubo03@inspur.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Bo Liu [Wed, 17 Jan 2024 03:11:34 +0000 (04:11 +0100)]
media: i2c: tvp5150: convert to use maple tree register cache
The maple tree register cache is based on a much more modern data structure
than the rbtree cache and makes optimisation choices which are probably
more appropriate for modern systems than those made by the rbtree cache.
Signed-off-by: Bo Liu <liubo03@inspur.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Bo Liu [Wed, 17 Jan 2024 03:11:05 +0000 (04:11 +0100)]
media: i2c: mt9v032: convert to use maple tree register cache
The maple tree register cache is based on a much more modern data structure
than the rbtree cache and makes optimisation choices which are probably
more appropriate for modern systems than those made by the rbtree cache.
Signed-off-by: Bo Liu <liubo03@inspur.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Bo Liu [Wed, 17 Jan 2024 03:08:25 +0000 (04:08 +0100)]
media: i2c: imx274: convert to use maple tree register cache
The maple tree register cache is based on a much more modern data structure
than the rbtree cache and makes optimisation choices which are probably
more appropriate for modern systems than those made by the rbtree cache.
Signed-off-by: Bo Liu <liubo03@inspur.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Bo Liu [Wed, 17 Jan 2024 03:07:30 +0000 (04:07 +0100)]
media: i2c: imx214: convert to use maple tree register cache
The maple tree register cache is based on a much more modern data structure
than the rbtree cache and makes optimisation choices which are probably
more appropriate for modern systems than those made by the rbtree cache.
Signed-off-by: Bo Liu <liubo03@inspur.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Alexander Stein [Wed, 17 Jan 2024 07:39:36 +0000 (08:39 +0100)]
media: i2c: imx415: Add more clock configurations
Complete the list from "INCK Setting" section in IMX415-AAQR-C
(Rev. E19504, 2019/05/21). For consistency suffix all lane rate values by
UL, which is needed for 2376000000 anyway.
Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Julien Massot [Fri, 12 Jan 2024 09:52:28 +0000 (10:52 +0100)]
media: i2c: st-vgxy61: Convert to CCI register access helpers
Use the new common CCI register access helpers to replace the private
register access helpers in the st-vgxy61 driver. This simplifies the
driver by reducing the amount of code.
st-vgxy61 devices use little endianness arrangement, therefore
the driver uses the CCI_REGx_LE registers definition.
Drop duplicated UAPI specific portions of the CCS (kernel) documentation
and fix a spelling error in UAPI documentation previously fixed in driver
documentation.
Also add references both ways.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:38 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: add support for parallel pipe to improve FPS
In some chips, MDP3 has the ability to utilize two pipelines to
parallelly process a single frame.
To enable this feature, multiple CMDQ clients and packets need to
be configured at the same time.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Extend the component settings used in MT8195 MDP3.
Additionally, it is crucial to read all component settings in
a specific manner to ensure that shared memory data structure lengths
are aligned across different platforms.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
The configuration of the MT8195 components in the shared memory
is defined in the header file "mdp_sm_mt8195.h".
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Add MT8195 MDP3 basic configuration in file "mdp_cfg_data.c"
and corresponding driver data.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:34 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: add support for blending multiple components
Certain chips can combine several components to form complex virtual
units with advanced functions.
These components require simultaneous configuration of their MODs and
clocks.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:33 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: extend GCE event waiting in RDMA and WROT
Support for multiple RDMA/WROT waits for GCE events.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
The amount of MDP3 driver probes is determined by the registered
clocks of MMSYS.
Since MT8195 MDP3 utilizes VPPSYS0 and VPPSYS1, it's necessary to
prevent multiple driver registrations.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:31 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: add checks for dummy components
Some components act as bridges only and do not require full configuration.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:30 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: introduce more MDP3 components
Add configuration of more components in MT8195 MDP3.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:29 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: introduce more pipelines from MT8195
Increasing the number of sets built by MMSYS and MUTEX in MT8195
will enable the creation of more pipelines in MDP3.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:28 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: add support second sets of MUTEX
After setting up the second set of MMSYS (VPPSYS1), it is necessary
to have a corresponding second set of MUTEX (MUTEX2) to assist in
handling SOF/EOF.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 20 Dec 2023 10:18:27 +0000 (11:18 +0100)]
media: platform: mtk-mdp3: add support second sets of MMSYS
The MT8195 chipset features two MMSYS subsets: VPPSYS0 and VPPSYS1.
These subsets coordinate and control the clock, power, and
register settings required for the components of MDP3.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Sun, 31 Dec 2023 10:00:21 +0000 (11:00 +0100)]
media: atomisp: Update TODO
Remove the TODO items for using the main (drivers/media/i2c) ov2680 and
ov5693 drivers and removing the atomisp specific ones, this has been done.
Remove the TODO item for gracefully handling missing firmware the
"media: atomisp: Bind and do power-management without firmware" changes
have fixed this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Sat, 30 Dec 2023 14:37:54 +0000 (15:37 +0100)]
media: atomisp: Remove unnecessary msleep(10) from atomisp_mrfld_power() error path
Remove unnecessary msleep(10) from atomisp_mrfld_power() error-exit path,
the success exit from atomisp_mrfld_power() happens if a test succeeds
inside the do { } while loop above the msleep().
The error-exit path with the removed msleep is only hit it the power-on is
not reflected in the iUNIT ISPSSPM0 status bits after a timeout of 50 ms.
Sleeping an extra 10 ms in the timeout path makes little sense.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Sat, 30 Dec 2023 14:31:41 +0000 (15:31 +0100)]
media: atomisp: Bind and do power-management without firmware
The ISP needs to be turned off in a special manner for the SoC
to be able to reach S0i3 during suspend.
When the firmware is missing still bind to the PCI device and
run in pm-only mode which takes care of turning the ISP off
(including turning it off again if the firmware has turned it
on during resume).
In this new pm-only mode the atomisp driver works exactly the same
as the non-staging, pm-only drivers/platform/x86/intel/atomisp2/pm.c
driver.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Fri, 29 Dec 2023 21:04:43 +0000 (22:04 +0100)]
media: atomisp: Replace atomisp_drvfs attr with using driver.dev_groups attr
sysfs attributes preferably should not be manually be registered but
instead the driver.groups / driver.dev_groups driver struct members
should be used to have the driver core handle this in a race free
manner.
Using driver.groups would be the most direct replacement for
driver_[add|remove]_file, but some of the attributes actually need access
to the struct atomisp_device (*), so as part of modernizing this part of
the atomisp driver this change also makes the sysfs attribute device
attributes instead of driver attributes.
*) Before this change accessing these attributes without the driver having
bound would result in a NULL pointer deref, this commit fixes this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
-Currently the driver uses pm_runtime_put_noidle() and relies on
userspace to open + close the /dev/video# node at least once to
actually turn the ISP off. Replace the pm_runtime_put_noidle()
with pm_runtime_put_sync() to make sure that the device is turned
off without relying on userspace for this.
This also ensures that atomisp_css_init() is run (by atomisp_power_on())
if the first userspace process opening /dev/video# wants to do more then
just query the v4l2-caps.
As part of this change move the pm setup code in probe() to just before
calling v4l2_async_nf_register() which registers the /dev/* nodes, so
that the device is left on for the entirety of the probe() function.
-Remove the turning off of the atomisp from the exit-error path,
the PCI subsystem and subsequent probe() attempts expect the device
to be in the on state when probe() fails.
This also fixes the atomisp driver causing the system to hang / freeze
when its firmware is missing. This freeze is caused by an unidentified
bug in the power-off on exit-error code which is now removed.
-Make sure that remove() properly powers on the device by replacing
pm_runtime_get_noresume() with pm_runtime_get_sync. The PCI subsystem
and subsequent probe() attempts expect the device to be in the on state
after unbinding the driver.
-Note this also swaps the order of put()/allow() and forbid()/get()
so that the sync versions actually work by calling allow() before put()
and forbid() after get()
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Wed, 27 Dec 2023 21:47:45 +0000 (22:47 +0100)]
media: atomisp: Fix atomisp_pci_remove()
Fix atomisp_pci_remove():
-Remove uninformative "Removing atomisp driver" log message
-Add missing devm_free_irq(), atomisp_uninitialize_modules() and
pci_free_irq_vectors() calls
-Move atomisp_msi_irq_uninit() down so that the remove() order is
an exact mirror of the probe() order
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Wed, 27 Dec 2023 20:45:48 +0000 (21:45 +0100)]
media: atomisp: Fix probe error-exit path
Fix probe error-exit path:
-Add a missing ia_css_unload_firmware() call when v4l2_async_nf_register()
fails
-Add a missing pm_runtime_forbid() call to undo the pm_runtime_allow() call
-Remove the unnecessary pcim_iounmap_regions() those are devm managed
so they will get cleaned up automatically on an error exit from probe()
-Rename the labels to avoid having multiple labels pointing to the same
place in the error-exit path
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans de Goede [Tue, 26 Dec 2023 15:32:10 +0000 (16:32 +0100)]
media: atomisp: Drop is_valid_device() function
Now that a single build supports both the ISP 2400 and the ISP 2401
this function is no longer necessary. The main probe() already
contains a similar switch (id->device & ATOMISP_PCI_DEVICE_SOC_MASK)
checking for a known device_id.
Move the revision check into the main probe() and drop
the is_valid_device() function.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
There is a bunch of dead code left from the deferred firmware loading
support which was removed in commit 8972ed6ea7a0 ("media: atomisp: Remove
deferred firmware loading support").
Remove this dead code:
- Remove the skip_fwload module parameter
- Remove the now always NULL fw argument from ia_css_init()
- Remove the fw_explicitly_loaded boolean from sh_css.c
(this was always true now)
- Adjust some function kernel-doc comments to match
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Dipendra Khadka [Sat, 23 Dec 2023 07:22:45 +0000 (08:22 +0100)]
media: atomisp: Fix spelling mistakes in rmgr_vbuf.c
codespell reported following spelling mistake
in rmgr_vbuf.cas below:
'''
./runtime/rmgr/src/rmgr_vbuf.c:201: succes ==> success
./runtime/rmgr/src/rmgr_vbuf.c:211: succes ==> success
./runtime/rmgr/src/rmgr_vbuf.c:215: succes ==> success
'''
This patch fixes these spelling mistakes.
It is good to use variable name that gives
proper meaning and spelling error free.
Dipendra Khadka [Sat, 23 Dec 2023 06:41:52 +0000 (07:41 +0100)]
media: atomisp: Fix spelling mistakes in sh_css_mipi.c
codespell reported following spelling mistake
in sh_css_mipi.c as below:
'''
./sh_css_mipi.c:177: separatelly ==> separately
./sh_css_mipi.c:540: ofset ==> offset, of set
'''
This patch fixes these spelling mistakes by changing
"separatelly" to "separately" and "ofset" to "offset".
projects / linux.git / log