| Age | Commit message (Collapse) | Author |
|---|
|
Add CAMSS_8550 enum, sm8550 compatible and sm8550 camss drvier private
data, the private data just include some basic information now, later
changes will enumerate with csiphy, csid and vfe resources.
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
There is no CSID TPG on some SoCs, so the v4l2 ctrl in CSID driver
shouldn't be registered. Checking the supported TPG modes to indicate
if the TPG hardware exists or not and only registering v4l2 ctrl for
CSID only when the TPG hardware is present.
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Add a default case in vfe_src_pad_code to get rid of a compile
warning if a new hw enum is added.
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
The RUP registers and buf done irq are moved from the IFE to CSID register
block on recent CAMSS implementations. Add callbacks structure to wrapper
the location change with minimum logic disruption.
Co-developed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Some v4l2 buffer related logic functions can be moved to vfe core as
common code, then the vfe driver of different hw version can reuse them,
this also can avoid adding duplicate code for new version supporting.
Suggested-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
The get hw version and src pad code functions can be common code in csid
core file, then the csid driver of different hw version can reuse them,
rather than adding duplicate code in csid driver for each version.
Suggested-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
control regs
New versions of the CSIPHY locate the control registers at offset 0x1000
not offset 0x800.
Provide a variable to base an offset from for the purposes of redirecting
the base offset for the new PHY regs layout.
The existing setup bases from 0x800, the new from 0x1000 with some of the
'EXT' registers dropped but the lower-order lane config regs at offset 0x00
and up the same as before.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
csiphy struct
A .data field in the csiphy device structure allows us to extend out the
register layout of the three phase capable CSIPHY layer.
Move the existing lane configuration structure to an encapsulating
structure -> struct csiphy_device_regs which is derived from the .data
field populated at PHY init time, as opposed to calculated at lane
configuration.
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Add a nop init callback to CSIPHY devices, this callback is used to add
some HW register offset and register configuration for specific platform,
then different platform can reuse the same CSIPHY driver. Later changes
will enumerate with enabling code.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
The existing structure captures the configuration of CSIPHY lane registers.
Rename to struct csiphy_lane_regs to reflect.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Reviewed-by: Elliot Berman <quic_eberman@quicinc.com>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Adding a new CSIPHY init sequence using downstream as a reference prompted
me to look at why we are splitting up the init sequence into chunks.
Right now we declare CSI PHY init sequences as an array of five equally
sized writes with a hard-coded control loop to iterate through each of the
five indexes. One bug in this model is that if you don't have an even
number of writes, you can't init the PHY as you wish.
In downstream the original code has something of the character
phy_init_seq[MAX_LANES][MAX_PARAMS] which in upstream we have translated
into phy_init_seq[5][SOME_NUMBER_OF_EQUAL_WRITES];
What the code does is take a pointer to the first index of the
r = &phy_init_seq[0][0]; and then literally does write(r, value); r++;
The controlling loop that hard-codes '5' in-lieu of MAX_LANES does no
special sleep, fabric-coherence sync or even a printk() to justify its
existence. Our compilers are optimising all of this away anyway so lets
drop.
Reduce the array declaration down to one flat aggregate init and let the
code just step through. As a happy side-effect we can then also handle
odd-number writes as the number of elements in the init sequence will no
longer have to be evenly divisible.
Reviewed-by: Elliot Berman <quic_eberman@quicinc.com>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Replace space with tab in define indentations.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Depeng Shao <quic_depengs@quicinc.com>
Reviewed-by: Elliot Berman <quic_eberman@quicinc.com>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Update clock names to make them consistent with existing platform i.e
sc8280xp. Rename gcc_cam_hf_axi to gcc_axi_hf and add gcc_axi_sf.
Signed-off-by: Vikram Sharma <quic_vikramsa@quicinc.com>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media updates from Mauro Carvalho Chehab:
- Sensor driver fixes
- remove dead TI wl128x FM radio driver
- Add support for the imx462 sensor at the IMX290 binding
- V4L2 pixel data transmitter and receiver documentation improvements
- Add support for MIPI Discovery and Configuration for C-PHY line
orders
- imx8-isi fixes and improvements
- stm32: dcmipp: add core support for the stm32mp25
- qcom: camss: Add sc7280 support
- Various fixes and enhancements
* tag 'media/v6.14-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (152 commits)
media: nuvoton: Fix an error check in npcm_video_ece_init()
media: dvb-usb-v2: af9035: fix ISO C90 compilation error on af9035_i2c_master_xfer
media: platform: rzg2l-cru: rzg2l-video: Fix the comment in rzg2l_cru_start_streaming_vq()
media: fix secfeed undefined when filter alloc fail
media: dt-bindings: trivial white-space and example cleanup
MAINTAINERS: repair file entry in MEDIA DRIVERS FOR STM32 - CSI
media: solo6x10: Use const 'struct bin_attribute' callback
media: saa7164: Remove unused values
staging: media: imx: fix OF node leak in imx_media_add_of_subdevs()
media: platform: exynos4-is: Remove unused __is_get_frame_size
media: vidtv: Fix a null-ptr-deref in vidtv_mux_stop_thread
media: mmp: Bring back registration of the device
media: cec: include linux/debugfs.h and linux/seq_file.h where needed
Revert "media: qcom: camss: Restructure camss_link_entities"
media: venus: Remove unused hfi_core_ping()
media: dt-bindings: qcom-venus: Deprecate video-decoder and video-encoder where applicable
media: venus: Populate video encoder/decoder nodename entries
media: venus: Add support for static video encoder/decoder declarations
media: venus: match instance creation and destruction order
media: venus: destroy hfi session after m2m_ctx release
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks
Pull kthread updates from Frederic Weisbecker:
"Kthreads affinity follow either of 4 existing different patterns:
1) Per-CPU kthreads must stay affine to a single CPU and never
execute relevant code on any other CPU. This is currently handled
by smpboot code which takes care of CPU-hotplug operations.
Affinity here is a correctness constraint.
2) Some kthreads _have_ to be affine to a specific set of CPUs and
can't run anywhere else. The affinity is set through
kthread_bind_mask() and the subsystem takes care by itself to
handle CPU-hotplug operations. Affinity here is assumed to be a
correctness constraint.
3) Per-node kthreads _prefer_ to be affine to a specific NUMA node.
This is not a correctness constraint but merely a preference in
terms of memory locality. kswapd and kcompactd both fall into this
category. The affinity is set manually like for any other task and
CPU-hotplug is supposed to be handled by the relevant subsystem so
that the task is properly reaffined whenever a given CPU from the
node comes up. Also care should be taken so that the node affinity
doesn't cross isolated (nohz_full) cpumask boundaries.
4) Similar to the previous point except kthreads have a _preferred_
affinity different than a node. Both RCU boost kthreads and RCU
exp kworkers fall into this category as they refer to "RCU nodes"
from a distinctly distributed tree.
Currently the preferred affinity patterns (3 and 4) have at least 4
identified users, with more or less success when it comes to handle
CPU-hotplug operations and CPU isolation. Each of which do it in its
own ad-hoc way.
This is an infrastructure proposal to handle this with the following
API changes:
- kthread_create_on_node() automatically affines the created kthread
to its target node unless it has been set as per-cpu or bound with
kthread_bind[_mask]() before the first wake-up.
- kthread_affine_preferred() is a new function that can be called
right after kthread_create_on_node() to specify a preferred
affinity different than the specified node.
When the preferred affinity can't be applied because the possible
targets are offline or isolated (nohz_full), the kthread is affine to
the housekeeping CPUs (which means to all online CPUs most of the time
or only the non-nohz_full CPUs when nohz_full= is set).
kswapd, kcompactd, RCU boost kthreads and RCU exp kworkers have been
converted, along with a few old drivers.
Summary of the changes:
- Consolidate a bunch of ad-hoc implementations of
kthread_run_on_cpu()
- Introduce task_cpu_fallback_mask() that defines the default last
resort affinity of a task to become nohz_full aware
- Add some correctness check to ensure kthread_bind() is always
called before the first kthread wake up.
- Default affine kthread to its preferred node.
- Convert kswapd / kcompactd and remove their halfway working ad-hoc
affinity implementation
- Implement kthreads preferred affinity
- Unify kthread worker and kthread API's style
- Convert RCU kthreads to the new API and remove the ad-hoc affinity
implementation"
* tag 'kthread-for-6.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks:
kthread: modify kernel-doc function name to match code
rcu: Use kthread preferred affinity for RCU exp kworkers
treewide: Introduce kthread_run_worker[_on_cpu]()
kthread: Unify kthread_create_on_cpu() and kthread_create_worker_on_cpu() automatic format
rcu: Use kthread preferred affinity for RCU boost
kthread: Implement preferred affinity
mm: Create/affine kswapd to its preferred node
mm: Create/affine kcompactd to its preferred node
kthread: Default affine kthread to its preferred NUMA node
kthread: Make sure kthread hasn't started while binding it
sched,arm64: Handle CPU isolation on last resort fallback rq selection
arm64: Exclude nohz_full CPUs from 32bits el0 support
lib: test_objpool: Use kthread_run_on_cpu()
kallsyms: Use kthread_run_on_cpu()
soc/qman: test: Use kthread_run_on_cpu()
arm/bL_switcher: Use kthread_run_on_cpu()
|
|
kthread_create() creates a kthread without running it yet. kthread_run()
creates a kthread and runs it.
On the other hand, kthread_create_worker() creates a kthread worker and
runs it.
This difference in behaviours is confusing. Also there is no way to
create a kthread worker and affine it using kthread_bind_mask() or
kthread_affine_preferred() before starting it.
Consolidate the behaviours and introduce kthread_run_worker[_on_cpu]()
that behaves just like kthread_run(). kthread_create_worker[_on_cpu]()
will now only create a kthread worker without starting it.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
|
|
When function of_find_device_by_node() fails, it returns NULL instead of
an error code. So the corresponding error check logic should be modified
to check whether the return value is NULL and set the error code to be
returned as -ENODEV.
Fixes: 46c15a4ff1f4 ("media: nuvoton: Add driver for NPCM video capture and encoding engine")
Cc: stable@vger.kernel.org
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20241015014053.669-1-thunder.leizhen@huawei.com
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
rzg2l_cru_start_streaming_vq()
Replace "buffer." -> "buffer", for consistency with rest of the
comment blocks in rzg2l_cru_start_streaming_vq().
Reported-by: Pavel Machek <pavel@denx.de>
Closes: https://lore.kernel.org/all/ZtWNFuw70nkB37EK@duo.ucw.cz/
Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com>
Link: https://lore.kernel.org/r/20240905112508.160560-1-biju.das.jz@bp.renesas.com
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The last use of __is_get_frame_size() was removed in 2013 by
commit 5cfaad64d88a ("[media] exynos4-is: Fix format propagation on
FIMC-IS-ISP subdev")
Remove it.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
In commit 4af65141e38e ("media: marvell: cafe: Register V4L2 device
earlier"), a call to v4l2_device_register() was moved away from
mccic_register() into its caller, marvell/cafe's cafe_pci_probe().
This is not the only caller though -- there's also marvell/mmp.
Add v4l2_device_register() into mmpcam_probe() to unbreak the MMP camera
driver, in a fashion analogous to what's been done to the Cafe driver.
Same for the teardown path.
Fixes: 4af65141e38e ("media: marvell: cafe: Register V4L2 device earlier")
Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Cc: stable@vger.kernel.org # v6.6+
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
This reverts commit cc1ecabe67d92a2da0b0402f715598e8dbdc3b9e.
This commit has a basic flaw in that it relies on camss->res->csid_num as a
control to index the array camss->vfe[i].
Testing on a platform where csid_num > vfe_num showed this bug up.
camss->vfe should only be indexed by camss->res->vfe_num. Since this commit
is meant to make the code be more readable reverting will simply restore
the previous correct bounds checking.
We can make another pass at making camss_link_entities look prettier but,
for now we should zap the bug introduced.
Fixes: cc1ecabe67d9 ("media: qcom: camss: Restructure camss_link_entities")
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
hfi_core_ping() was added by 2017's
commit 09c2845e8fe4 ("[media] media: venus: hfi: add Host Firmware
Interface (HFI)")
but has remained unused.
Remove it.
It was the only caller of the ->core_ping member of hfi_ops,
so remove it, and the venus_core_ping that it pointed to.
Note I've left pky_sys_ping which seems to be the lowest level
definition of the command.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.k.varbanov@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Populate encoder and decoder node-name entries for the upstream parts. Once
done the compat="video-encoder" and compat="video-decoder" in the dtsi can
be dropped though the venus driver will continue to favour DT declared
video-encoder/video-decoder declarations over static declarations for
compatibility.
Tested-by: Renjiang Han <quic_renjiang@quicinc.com>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.k.varbanov@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Add resource structure data and probe() logic to support static
declarations of encoder and decoder.
Right now we rely on video encoder/decoder selection happening in the dtb
but, this goes against the remit of device tree which is supposed to
describe hardware, not select functional logic in Linux drivers.
Provide two strings in the venus resource structure enc_nodename and
dec_nodename.
When set the venus driver will create an OF entry in-memory consistent
with:
dec_nodename {
compat = "video-decoder";
};
and/or
enc_nodename {
compat = "video-encoder";
};
This will allow us to reuse the existing driver scheme of relying on compat
names maintaining compatibility with old dtb files.
dec_nodename can be "video-decoder" or "video0"
enc_nodename can be "video-encoder" or "video1"
This change relies on of_changeset() API as a result select OF_DYNAMIC will
be added to venus/Kconfig
Tested-by: Renjiang Han <quic_renjiang@quicinc.com>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.k.varbanov@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Match order of instances' initialization in venc_open()/vdec_close()
and order of destruction in venus_common_close().
Suggested-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Signed-off-by: Stanimir Varbanov <stanimir.k.varbanov@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
This partially reverts commit that made hfi_session_destroy()
the first step of vdec/venc close(). The reason being is a
regression report when, supposedly, encode/decoder is closed
with still active streaming (no ->stop_streaming() call before
close()) and pending pkts, so isr_thread cannot find instance
and fails to process those pending pkts. This was the idea
behind the original patch - make it impossible to use instance
under destruction, because this is racy, but apparently there
are uses cases that depend on that unsafe pattern. Return to
the old (unsafe) behaviour for the time being (until a better
fix is found).
Fixes: 45b1a1b348ec ("media: venus: sync with threaded IRQ during inst destruction")
Cc: stable@vger.kernel.org
Reported-by: Nathan Hebert <nhebert@google.com>
Signed-off-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Signed-off-by: Stanimir Varbanov <stanimir.k.varbanov@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Add ISI support for i.MX8ULP.
Signed-off-by: Guoniu.zhou <guoniu.zhou@nxp.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://lore.kernel.org/r/20241012084732.1036652-2-guoniu.zhou@oss.nxp.com
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
Running the v4l2-compliance (1.27.0-5208, SHA: af114250d48d) on the m2m
device fails on the MMAP streaming tests, with the following messages:
fail: v4l2-test-buffers.cpp(240): g_field() == V4L2_FIELD_ANY
fail: v4l2-test-buffers.cpp(1508): buf.qbuf(node)
Apparently, the driver does not properly set the field member of
vb2_v4l2_buffer struct, returning the default V4L2_FIELD_ANY value which
is against the guidelines.
Fixes: cf21f328fcaf ("media: nxp: Add i.MX8 ISI driver")
Signed-off-by: Laurentiu Palcu <laurentiu.palcu@oss.nxp.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://lore.kernel.org/r/20240924103304.124085-1-laurentiu.palcu@oss.nxp.com
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fixes from Mauro Carvalho Chehab:
- fix a clang build issue with mediatec vcodec
- add missing variable initialization to dib3000mb write function
* tag 'media/v6.13-3' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
media: mediatek: vcodec: mark vdec_vp9_slice_map_counts_eob_coef noinline
media: dvb-frontends: dib3000mb: fix uninit-value in dib3000_write_reg
|
|
Add check for the return value of clk_enable() to gurantee the success.
Fixes: babde1c243b2 ("[media] V4L: Add driver for S3C24XX/S3C64XX SoC series camera interface")
Signed-off-by: Jiasheng Jiang <jiashengjiangcool@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add check for the return value of clk_enable() to gurantee the success.
Fixes: b5f1220d587d ("[media] v4l: Add v4l2 subdev driver for S5P/EXYNOS4 MIPI-CSI receivers")
Signed-off-by: Jiasheng Jiang <jiashengjiangcool@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The Unicam hardware has been observed to cause a buffer overrun when
using the dummy buffer as a circular buffer. The conditions that cause
the overrun are not fully known, but it seems to occur when the memory
bus is heavily loaded.
To avoid the overrun, program the hardware with a buffer size of 0 when
using the dummy buffer. This will cause overrun into the allocated dummy
buffer, but avoid out of bounds writes.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The imx219/imx708 sensors frequently generate a single corrupt frame
(image or embedded data) when the sensor first starts. This can either
be a missing line, or invalid samples within the line. This only occurrs
using the upstream Unicam kernel driver.
Disabling trigger mode elimiates this corruption. Since trigger mode is
a legacy feature copied from the firmware driver and not expected to be
needed, remove it. Tested on the Raspberry Pi cameras and shows no ill
effects.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
When matching formats via try_fmt/set_fmt ioctls, test for the unpacked
formats as well as packed formats. This allows userland clients setup
unpacking to 16-bits from the 10/12/14-packed CSI2 formats.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Ensure that the frame sequence counter is incremented only if a previous
frame start interrupt has occurred, or a frame start + frame end has
occurred simultaneously.
This corresponds the sequence number with the actual number of frames
produced by the sensor, not the number of frame buffers dequeued back
to userland.
Signed-off-by: Naushir Patuck <naush@raspberrypi.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The latest datasheet (Rev.0.70) updates the D-PHY start-up sequence for
D-PHY operation. Unfortunately the datasheet do not add any additional
documentation on the magic values.
This have been tested together with the MAX96724 available on the single
board test platform and it works as expected.
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Later versions of the V4H datasheet adds documentation for the line
order register needed to support all possible configurations. Extend the
driver to take the line order for each data line into account when
configuring the device.
Unfortunately not all registers initially thought to be involved in line
order configuration where directly related. One magic value is still in
the driver and left as-is, but it is not related to line order as that
procedure have now been documented.
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Add check for the return value of clk_enable() to guarantee the success.
Fixes: 81a409bfd551 ("media: marvell-ccic: provide a clock for the sensor")
Signed-off-by: Jiasheng Jiang <jiashengjiangcool@gmail.com>
[Sakari Ailus: Fix spelling in commit message.]
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The company name has update to Rockchip Electronics Co., Ltd.
since 2021.
And change Co.Ltd to Co., Ltd. to fix mail server warning:
DBL_SPAM(6.50)[co.ltd:url];
Signed-off-by: Andy Yan <andy.yan@rock-chips.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Heiko Stuebner <heiko@sntech.de>
Link: https://lore.kernel.org/r/20241216100444.3726048-1-andyshrk@163.com
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
|
|
Store the chroma and motion vector offsets for each frame so that they
can be used later when the resolution changes.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The Hantro decoder non post-processed pixel-format steps are different
from the post-processed ones. Fix the steps according to the hardware
limitations. Since reference frame pixel-format issues have been fixed,
it is possible to use V4L2_PIX_FMT_NV15_4L4 rather V4L2_PIX_FMT_P010_4L4
for 10bit streams.
Fluster VP9 score goes up to 207/305.
HEVC score is still 141/147.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
The Hantro decoder always produces tiled pixel-formats, but when the
post-processor is used, the destination pixel-format is a non-tiled
pixel-format. This causes an incorrect computation of the reference
frame size and offsets. Get and save the correct tiled pixel-format for
8 and 10 bit streams to solve these computation issues.
Fluster VP9 score increase to 166/305 (vs 145/305).
HEVC score is still 141/147.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
With KASAN enabled, clang fails to optimize the inline version of
vdec_vp9_slice_map_counts_eob_coef() properly, leading to kilobytes
of temporary values spilled to the stack:
drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_vp9_req_lat_if.c:1526:12: error: stack frame size (2160) exceeds limit (2048) in 'vdec_vp9_slice_update_prob' [-Werror,-Wframe-larger-than]
This seems to affect all versions of clang including the latest (clang-20),
but the degree of stack overhead is different per release.
Marking the function as noinline_for_stack is harmless here and avoids
the problem completely.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Sebastian Fricke <sebastian.fricke@collabora.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
|
|
Use the cmdq_pkt_create() and cmdq_pkt_destroy() common functions
instead of implementing specific mdp3 versions.
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Reviewed-by: Matthias Brugger <matthias.bgg@gmail.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+huawei@kernel.org>
|
|
In order to have fine-grained control, use cmdq_pkt_eoc() and
cmdq_pkt_jump_rel() to replace cmdq_pkt_finalize().
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Reviewed-by: Matthias Brugger <matthias.bgg@gmail.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+huawei@kernel.org>
|
|
Just like was done with MM_REG_WRITE, remove the mask from the
MM_REG_POLL macro, leaving MM_REG_POLL_MASK to be used when a mask is
required, and update the call sites accordingly. In this case, all calls
require a mask, so MM_REG_POLL remains unused, but at least this makes
the MM_REG_POLL macros consistent with the MM_REG_WRITE ones.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.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+huawei@kernel.org>
|
|
There are two macros to issue a cmdq write: MM_REG_WRITE_MASK and
MM_REG_WRITE, but confusingly, both of them take a mask parameter. The
difference is that MM_REG_WRITE additionally checks whether the mask
passed in contains the register mask, in which case, the 0xffffffff mask
is passed to cmdq_pkt_write_mask(), effectively disregarding the mask
and calling cmdq_pkt_write() as an optimization.
Move that optimization to the MM_REG_WRITE_MASK macro and make
MM_REG_WRITE the variant that doesn't take a mask, directly calling to
cmdq_pkt_write().
Change the call sites to MM_REG_WRITE whenever a mask wasn't necessary
(ie 0xffffffff or a <register>_MASK was passed as mask) and in other
cases to MM_REG_WRITE_MASK.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.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+huawei@kernel.org>
|
|
A few macros declare variadic arguments even though the underlying
functions don't support them. Remove them.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.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+huawei@kernel.org>
|
|
cmdq_pkt_write_mask() boils down to a cmdq_pkt_write() when the mask is
0xFFFFFFFF. Call cmdq_pkt_write() directly in those cases to simplify
the code.
Suggested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.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+huawei@kernel.org>
|
|
The stm32mp25 supports both parallel & csi inputs.
An additional clock control is necessary.
Skeleton of the subdev structures for the stm32mp25 is added,
identical for the time being to the stm32mp13 however more subdeves
will be added in further commits.
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
|
