Biju Das [Thu, 5 Sep 2024 11:18:26 +0000 (12:18 +0100)]
media: platform: rzg2l-cru: rzg2l-video: Set AXI burst max length
As per the hardware manual section 35.2.3.26 'AXI Master Transfer
Setting Register for CRU Image Data', it is mentioned that to improve
the transfer performance of CRU, it is recommended to use AXILEN value
'0xf' for AXI burst max length setting for image data.
Signed-off-by: Hien Huynh <hien.huynh.px@renesas.com> Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Link: https://lore.kernel.org/r/20240905111828.159670-1-biju.das.jz@bp.renesas.com Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Replace the ternary operator with an if. In this case the code is more
clear and also fixes the following cocci warnings:
drivers/staging/media/atomisp/pci/sh_css_frac.h:40:17-18: WARNING opportunity for max()
drivers/staging/media/atomisp/pci/sh_css_frac.h:50:17-18: WARNING opportunity for max()
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org> Acked-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Paul Elder [Mon, 9 Sep 2024 15:48:28 +0000 (17:48 +0200)]
media: platform: video-mux: Fix mutex locking
The current order of locking between the driver mutex and the v4l2 subdev
state lock causes a circular locking dependency when trying to set up a
link. Fix this.
Signed-off-by: Paul Elder <paul.elder@ideasonboard.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.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>
Lad Prabhakar [Mon, 7 Oct 2024 12:38:09 +0000 (13:38 +0100)]
v4l2-subdev: Return -EOPNOTSUPP for unsupported pad type in call_get_frame_desc()
The `get_frame_desc()` operation should always be called on a source pad,
which is indicated by the `MEDIA_PAD_FL_SOURCE` flag. This patch adds a
check in `call_get_frame_desc()` to ensure that the `MEDIA_PAD_FL_SOURCE`
flag is set for the pad before invoking `get_frame_desc()`. If the pad is
not a source pad, the function will return an `-EOPNOTSUPP` error,
signaling that the operation is not supported on non-source pads.
Colin Ian King [Wed, 2 Oct 2024 16:53:29 +0000 (17:53 +0100)]
media: i2c: ds90ub960: Fix missing return check on ub960_rxport_read call
The function ub960_rxport_read is being called and afterwards ret is
being checked for any failures, however ret is not being assigned to
the return of the function call. Fix this by assigning ret to the
return of the call which appears to be missing.
Fixes: afe267f2d368 ("media: i2c: add DS90UB960 driver") Signed-off-by: Colin Ian King <colin.i.king@gmail.com> Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Document the expected {enable,disable}_streams callback behaviour for
drivers that are stream-unaware i.e. don't specify the
V4L2_SUBDEV_CAP_STREAMS sub-device capability flag. In this specific case,
the mask argument can be ignored.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Sakari Ailus [Tue, 17 Sep 2024 12:14:44 +0000 (15:14 +0300)]
media: Documentation: Deprecate s_stream video op, update docs
The scope of the s_stream video operation is now fully supported by
{enable,disable}_streams. Explicitly document the s_stream() op as
deprecated and update the related documentation.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Biju Das [Mon, 26 Aug 2024 11:07:30 +0000 (12:07 +0100)]
media: platform: rzg2l-cru: rzg2l-video: Move request_irq() to probe()
Move request_irq() to probe(), in order to avoid requesting IRQ during
device start which happens frequently. As this function is in probe(), it
is better to replace it with its devm variant for managing the resource
efficiently.
While at it, drop the IRQF_SHARED flag as currently there is a single user
for this IRQ.
Tomi Valkeinen [Thu, 3 Oct 2024 10:31:10 +0000 (13:31 +0300)]
media: uapi: Add meta formats for PiSP FE config and stats
Add two meta formats for PiSP FE: V4L2_META_FMT_RPI_FE_CFG and
V4L2_META_FMT_RPI_FE_STATS. The former is used to provide configuration
for the FE and the latter is used to read the statistics from the FE.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Some R-Car SoCs are capable of capturing RAW10. Extend the format
enumeration, validation and setup to expose and configure for this
format if the particular device supports it.
The VIN is usually capable of converting from most media bus formats to
a range of different pixel formats. However if the input media bus
format is a RAW10 format this is not possible so the corresponding pixel
output format is forced.
Niklas Söderlund [Wed, 2 Oct 2024 10:33:17 +0000 (12:33 +0200)]
media: rcar-isp: Add family compatible for R-Car Gen4 family
Add the Gen4 family compatible. This will be used instead of a SoC
specific compatible for the new Gen4 SoC V4M. Two Gen4 boards (V3U and
V4H) have already been added prior and their bindings need to be kept
for backward compatibility.
Niklas Söderlund [Wed, 2 Oct 2024 10:33:16 +0000 (12:33 +0200)]
dt-bindings: media: renesas,isp: Add Gen4 family fallback
The ISP Channel Selector IP is the same for all current Gen4 devices.
This was not known when adding support for V3U and V4H and a single SoC
specific compatible was used.
Before adding more SoC specific bindings for V4M add a family compatible
fallback for Gen4. That way the driver only needs to be updated once for
Gen4, and we still have the option to fix any problems in the driver if
any testable differences between the SoCs are found.
There are already DTS files using the V3U and V4H compatibles which
needs to be updated to not produce a warning for DTS checks. The driver
also needs to kept the compatible values to be backward compatible , but
for new Gen4 SoCs such as V4M we can avoid this.
Niklas Söderlund [Wed, 2 Oct 2024 09:23:31 +0000 (11:23 +0200)]
media: staging: max96712: Add support for MAX96724
The MAX96724 is almost identical to the MAX96712 and can be supported by
the same driver, add support for it.
For the staging driver which only supports patter generation the big
difference is that the datasheet (rev 4) for MAX96724 do not describe
the DEBUG_EXTRA register, which is at offset 0x0009 on MAX96712. It's
not clear if this register is removed or moved to a different offset.
What is known is writing to register 0x0009 have no effect on MAX96724.
This makes it impossible to increase the test pattern clock frequency
from 25 MHz to 75Mhz on MAX96724. To be able to get a stable test
pattern the DPLL frequency have to be increase instead to compensate for
this. The frequency selected is found by experimentation as the MAX96724
datasheet is much sparser then what's available for MAX96712.
Niklas Söderlund [Wed, 2 Oct 2024 09:23:30 +0000 (11:23 +0200)]
media: staging: max96712: Document the DEBUG_EXTRA register
The DEBUG_EXTRA register is not part of soon to be added MAX96724
device. To make it easier to understand the differences when reading the
code prepare for the addition by creating named defines for the
register.
Niklas Söderlund [Wed, 2 Oct 2024 09:23:29 +0000 (11:23 +0200)]
media: staging: max96712: Move link frequency setting to device struct
Prepare for supporting MAX96724 by moving the soon device specific link
frequency setting into information structure. This struct will be
extended to carry more differences between the two devices supported.
While at it remove trailing comma in device table, no entries will be
appended after the sentinel.
Niklas Söderlund [Wed, 2 Oct 2024 09:23:28 +0000 (11:23 +0200)]
media: staging: max96712: Remove device id check
This check is incorrect and checks the wrong register. Furthermore there
is no documented shared device id register for MAX96712. There might be
overlap with the soon to be added MAX96724 device which do document such
a register.
However as the check was merely a precaution and to check during
development that the driver could talk to the device there is no harm in
removing it all together. A correct and more sophisticated check can be
added later if there ever is a need to differentiate between different
versions of a device.
Niklas Söderlund [Wed, 2 Oct 2024 09:23:27 +0000 (11:23 +0200)]
dt-bindings: i2c: maxim,max96712: Add compatible for MAX96724
The MAX96712 and MAX96724 are both quad GMSL2 to CSI-2 deserializers and
are in parts similar, but not identical. The most obvious difference is
on the CSI-2 side where the MAX96712 have 4 PHYs and support D-PHY with
1x4, 2x2 and 4x2 lanes where the MAX96724 only have 2 PHYs and supports
D-PHY with 2x4 or 4x2 lanes.
The register layout overlap in part but there are differences and holes.
Most of the differences are related to the different number of CSI-2
PHYs, but there are other capability differences between the two.
Add a specific compatible for MAX96724 to the max96712 bindings. The
bindings do not yet support validating all DT properties to limit it the
each devices capabilities. However to allow for this in future a
specific compatible for the two different devices are needed.
Niklas Söderlund [Fri, 30 Aug 2024 20:31:04 +0000 (22:31 +0200)]
media: rcar-csi2: Add support for R-Car V4M
The V4M is the second Gen4 device that is enabled in the rcar-csi2
driver. There is much overlap with the already supported V4H device. The
registers that where new on Gen4 and where added with the V4H prefix are
retained and only new registers unique to the V4M are added with the new
V4M prefix. This follows the style for when V4H was added which had an
overlap with Gen3 registers.
The V4M CSI-2 receiver supports D-PHY mode only, either in 1-, 2- or
4-lane configuration. The datasheets do not document lane swapping and
is left out for now.
While the V4M only supports D-PHY the configuration for it is added in
such a way that it can be reused for V4H which supports both C-PHY and
D-PHY. No known SoC exists to test the D-PHY configuration on V4H so
it's not wired-up.
Niklas Söderlund [Fri, 30 Aug 2024 20:31:02 +0000 (22:31 +0200)]
media: rcar-csi2: Move PHTW write helpers
Prepare for V4M support by moving the PHTW write helpers to the generic
write helpers. This is needed as adding V4M support will involve
interact with the PHTW register from code that are logically grouped
with similar code in such a way that forward declarations of these
helpers would otherwise be needed.
Niklas Söderlund [Fri, 30 Aug 2024 20:31:01 +0000 (22:31 +0200)]
media: rcar-csi2: Add helper to lookup mbps settings
The structure mapping a configuration information to a particular mpbs
setting needs to be extended with more information to support future
SoCs. Before it is extended reduce code duplication by creating a helper
to lookup information from an array of mbps setting, the lookup code
has already been copied to two speared locations.
While at it rename the structure to make it clear it contains
information related to a mbps setting, not just a single register value.
Niklas Söderlund [Fri, 30 Aug 2024 20:31:00 +0000 (22:31 +0200)]
media: rcar-csi2: Abstract PHTW and PHYPLL register offsets
Most of the registers used on the R-Car V4M CSI-2 IP are shared with the
devices already supported by the rcar-csi2 driver. Two registers which
function and layout are the same are however found on different offsets.
Prepare for adding support for R-Car V4M by storing the offset to these
two registers offsets in the device information structured. This way the
code, which is shared between the devices, can be reused when V4M
support is added.
Niklas Söderlund [Fri, 30 Aug 2024 20:30:59 +0000 (22:30 +0200)]
media: rcar-csi2: Allow writing any code and data value to PHTW
The helper to write an array of code and data values to the PHY Test
Interface Write Register (PHTW) register uses the case where both code
and data are zero as an exit condition. This prevents writing data = 0
and code = 0 to the register.
Up until now this has been OK as no such combination where needed, and
it was a convenient exit condition. In future writing data = 0 and code
= 0 to the PHTW register will be needed.
Avoid using an exit condition when writing an array of PHTW values and
instead pass the length of the array to the helper. This allows any
combination of code and data to be written.
media: staging/intel-ipu3: css: Convert comma to semicolon
The return of function imgu_css_grid_end_calc is void.
To ensure code clarity and prevent potential errors, it's advisable
to employ the ';' as a statement separator, except when ',' are
intentionally used for specific purposes.
Benjamin Bara [Tue, 3 Sep 2024 14:54:34 +0000 (16:54 +0200)]
media: i2c: imx290: Check for availability in probe()
Currently, the V4L2 subdevice is also created when the device is not
available/connected. From userspace perspective, there is no visible
difference between a working and not-working subdevice (except when
trying it out).
This commit adds a simple preparation step, which includes an
availability check, before the subdev is initialized.
Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Benjamin Bara [Tue, 3 Sep 2024 14:54:33 +0000 (16:54 +0200)]
media: i2c: imx290: Avoid communication during probe()
As we don't know the mode during probe(), it doesn't make sense to
update the sensors' registers with assumptions. As imx290_set_ctrl(),
which is responsible for the happening communication, already ensures that
there is no communication with a suspended sensor, put the sensor to
suspend before calling it.
To clarify the dependency of the PM runtime to the link of the subdev
and the imx290 instance, put the block together.
Suggested-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Benjamin Bara [Tue, 3 Sep 2024 14:54:32 +0000 (16:54 +0200)]
media: i2c: imx290: Remove CHIP_ID reg definition
This register is not described in the public available imx290 datasheet.
Additionally, a read returns '0x07d0' for an imx327lqr and also for an
imx462, which means it cannot be used to distinguish between those two
imx290 derivatives.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
media: i2c: Fix typos in comments across various files
This commit corrects spelling errors in comments within
the media/i2c directory found by codespell to enhance clarity
and maintainability of the code.
This change does not affect the functionality.
Controls can be exposed to userspace via a v4l-subdevX device, and
userspace has to be able to subscribe to control events so that it is
notified when the control changes value.
Add missing HAS_EVENTS support: flag and .(un)subscribe_event().
Controls can be exposed to userspace via a v4l-subdevX device, and
userspace has to be able to subscribe to control events so that it is
notified when the control changes value.
Add missing HAS_EVENTS support: flag and .(un)subscribe_event().
Benjamin Gaignard [Mon, 26 Aug 2024 17:24:10 +0000 (17:24 +0000)]
media: verisilicon: Use V4L2_FMTDESC_FLAG_ENUM_ALL flag
By adding support for the V4L2_FMTDESC_FLAG_ENUM_ALL flag into the driver
we allow userspace applications to discover all possible
pixel formats of the hardware block. This way userspace can decide
which decoder to use given the supported pixel formats.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Benjamin Gaignard [Mon, 26 Aug 2024 17:24:10 +0000 (17:24 +0000)]
media: test-drivers: Use V4L2_FMTDESC_FLAG_ENUM_ALL flag
Since the V4L2_FMTDESC_FLAG_ENUM_ALL flag mostly targets stateless
decoder pixel-format enumeration, update visl test driver to use it.
When V4L2_FMTDESC_FLAG_ENUM_ALL flag is set let the driver returns
one more pixel format.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Benjamin Gaignard [Mon, 26 Aug 2024 17:24:09 +0000 (17:24 +0000)]
media: videodev2: Add flag to unconditionally enumerate pixel formats
When the index is ORed with V4L2_FMTDESC_FLAG_ENUM_ALL the
driver clears the flag and enumerate all the possible formats,
ignoring any limitations from the current configuration.
Drivers which do not support this flag yet always return an EINVAL.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
[hverkuil: improved doc when the new flag is not supported by the driver]
Vladimir Zapolskiy [Thu, 26 Sep 2024 21:19:57 +0000 (00:19 +0300)]
media: qcom: camss: move SM8250 regulators from CSID to CSIPHY subdevice
On Qualcomm SM8250 SoC there are two sets of regulators, and each of
both sets is specific to six CSIPHY IPs. At the moment there is no
proper split of two "combined" regulators with quite arbitrary selected
names in the driver or platform CAMSS device tree node, however for sake
of clarity and better hardware description it makes sense to move the
currently existing regulator resources from all CSID subdevices to all
CSIPHY subdevices.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Vladimir Zapolskiy [Thu, 26 Sep 2024 21:19:56 +0000 (00:19 +0300)]
media: qcom: camss: add management of supply regulators to CSIPHY
This change allows to properly assign and manage supply regulator
resources by CSIPHY subdevices of CAMSS, this is needed to fine tune
description of supply regulators on newer platforms, conversion of
old platforms to the new scheme is also anticipated.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org> Tested-by: Depeng Shao <quic_depengs@quicinc.com> # SM8550 Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Bryan O'Donoghue [Wed, 14 Aug 2024 16:14:34 +0000 (17:14 +0100)]
media: qcom: camss: Add hooks to get CSID wrapper resources
New SoCs have CSID devices inside of a shared "wrapper" i.e. a set of regs
which is responsible for manging the muxes of the CSID to various other
blocks throughout CAMSS.
Not every SoC has this top-level muxing layer so make it optional depending
on whether its declared as a resource or not.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Vladimir Zapolskiy [Tue, 13 Aug 2024 21:03:42 +0000 (00:03 +0300)]
media: qcom: camss: fix error path on configuration of power domains
There is a chance to meet runtime issues during configuration of CAMSS
power domains, because on the error path dev_pm_domain_detach() is
unexpectedly called with NULL or error pointer.
One of the simplest ways to reproduce the problem is to probe CAMSS
driver before registration of CAMSS power domains, for instance if
a platform CAMCC driver is simply not built.
Warning backtrace example:
Unable to handle kernel NULL pointer dereference at virtual address 00000000000001a2
<snip>
pc : dev_pm_domain_detach+0x8/0x48
lr : camss_probe+0x374/0x9c0
The cause of the problem is that when using sysfs to dynamically register
an i2c device, there is no platform data, but the probe process of ts2020
needs to use platform data, resulting in a null pointer being accessed.
Solve this problem by adding checks to platform data.
Fixes: dc245a5f9b51 ("[media] ts2020: implement I2C client bindings") Cc: <stable@vger.kernel.org> Signed-off-by: Li Zetao <lizetao1@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Alexander Shiyan [Wed, 9 Oct 2024 06:05:44 +0000 (09:05 +0300)]
media: i2c: tc358743: Fix crash in the probe error path when using polling
If an error occurs in the probe() function, we should remove the polling
timer that was alarmed earlier, otherwise the timer is called with
arguments that are already freed, which results in a crash.
Fixes: 4e66a52a2e4c ("[media] tc358743: Add support for platforms without IRQ line") Signed-off-by: Alexander Shiyan <eagle.alexander923@gmail.com> Cc: stable@vger.kernel.org Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Arnd Bergmann [Fri, 4 Oct 2024 10:02:23 +0000 (10:02 +0000)]
media: cec: seco: add HAS_IOPORT dependency
This driver is now enabled for compile-testing on architectures
that may not have I/O port access:
drivers/media/cec/platform/seco/seco-cec.c: In function 'smb_word_op.constprop.isra':
include/asm-generic/io.h:542:14: error: call to '_inb' declared with attribute error: inb()) requires CONFIG_HAS_IOPORT
Add a Kconfig dependency again.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Rohan Barar [Thu, 3 Oct 2024 09:40:40 +0000 (19:40 +1000)]
media: cx231xx: Add support for Dexatek USB Video Grabber 1d19:6108
Add Dexatek Technology Ltd USB Video Grabber 1d19:6108 to the cx231xx
driver. This device is sold under the name "BAUHN DVD Maker (DK8723)" by
ALDI in Australia.
This device is similar to 1d19:6109, which is already included in cx231xx.
Both video and audio capture function correctly after installing the
patched cx231xx driver.
Patch Changelog
v1:
- Initial submission.
v2:
- Fix SoB + Improve subject.
v3:
- Rephrase message to not exceed 75 characters per line.
- Removed reference to external GitHub URL.
Signed-off-by: Rohan Barar <rohan.barar@gmail.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
media: cx231xx: Fix the S-Video capture on August VGB100
There are three separate issues preventing color capture with S-Video
on August VGB100 with the cx231xx driver (same vid/pid as OTG102):
1. `cx231xx_set_decoder_video_input` is called with a u32 passed
as its third argument, yet this functions expects a u8 instead.
Some information about the configuration of the video mux is lost
in the conversion, so that ch2 and ch3 do not get set by
`cx231xx_afe_set_input_mux` (expecting a u32 but being passed a u8).
2. The input pin for the chroma signal is not correctly specified
in cx231xx-cards.c, which can be verified by looking at the inf file
coming with the VGB100 and OTG102' drivers for Windows.
The mistake in the cx231xx driver likely stems from a wrong comment
in the same file, suggesting VIN1_2 for chroma,
while VIN3_2 is actually used.
3. Even after fixing the two issues above, the captured stream remains
essentially B&W (although acquiring some pale green shades, suggesting
we're moving in the right direction).
After tests with somewhat random changes, it was found that removing
`CX25840_SVIDEO_ON` from the vmux configuration in cx231xx-cards.c
results in a captured stream with the expected colors.
Signed-off-by: Fabio Luongo <f.langufo.l@gmail.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
media: wl128x: Fix atomicity violation in fmc_send_cmd()
Atomicity violation occurs when the fmc_send_cmd() function is executed
simultaneously with the modification of the fmdev->resp_skb value.
Consider a scenario where, after passing the validity check within the
function, a non-null fmdev->resp_skb variable is assigned a null value.
This results in an invalid fmdev->resp_skb variable passing the validity
check. As seen in the later part of the function, skb = fmdev->resp_skb;
when the invalid fmdev->resp_skb passes the check, a null pointer
dereference error may occur at line 478, evt_hdr = (void *)skb->data;
To address this issue, it is recommended to include the validity check of
fmdev->resp_skb within the locked section of the function. This
modification ensures that the value of fmdev->resp_skb does not change
during the validation process, thereby maintaining its validity.
This possible bug is found by an experimental static analysis tool
developed by our team. 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.
Fixes: e8454ff7b9a4 ("[media] drivers:media:radio: wl128x: FM Driver Common sources") Cc: stable@vger.kernel.org Signed-off-by: Qiu-ji Chen <chenqiuji666@gmail.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
media: Switch back to struct platform_driver::remove()
After commit 0edb555a65d1 ("platform: Make platform_driver::remove()
return void") .remove() is (again) the right callback to implement for
platform drivers.
Convert all platform drivers below drivers/media to use .remove(), with
the eventual goal to drop struct platform_driver::remove_new(). As
.remove() and .remove_new() have the same prototypes, conversion is done
by just changing the structure member name in the driver initializer.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
media: dvb: Fix typos in comments across various files
This patch corrects spelling errors in comments within
the media/dvb-frontends directory found by codespell to enhance
clarity and maintainability of the code.
This change does not affect the functionality.
Signed-off-by: Yu Jiaoliang <yujiaoliang@vivo.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
media:tuners: Fix typos in comments across various files
This commit corrects spelling errors in comments within
the media/tuners directory found by codespell to enhance
clarity and maintainability of the code.
This change does not affect the functionality.
Ming Qian [Fri, 13 Sep 2024 06:22:54 +0000 (15:22 +0900)]
media: imx-jpeg: Ensure power suppliers be suspended before detach them
The power suppliers are always requested to suspend asynchronously,
dev_pm_domain_detach() requires the caller to ensure proper
synchronization of this function with power management callbacks.
otherwise the detach may led to kernel panic, like below:
Ming Qian [Fri, 13 Sep 2024 06:21:45 +0000 (15:21 +0900)]
media: amphion: Set video drvdata before register video device
The video drvdata should be set before the video device is registered,
otherwise video_drvdata() may return NULL in the open() file ops, and led
to oops.
Ming Qian [Fri, 13 Sep 2024 06:21:44 +0000 (15:21 +0900)]
media: imx-jpeg: Set video drvdata before register video device
The video drvdata should be set before the video device is registered,
otherwise video_drvdata() may return NULL in the open() file ops, and led
to oops.
Fixes: 2db16c6ed72c ("media: imx-jpeg: Add V4L2 driver for i.MX8 JPEG Encoder/Decoder") Cc: <stable@vger.kernel.org> Signed-off-by: Ming Qian <ming.qian@nxp.com> Reviewed-by: TaoJiang <tao.jiang_2@nxp.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
media: mtk-jpeg: Fix null-ptr-deref during unload module
The workqueue should be destroyed in mtk_jpeg_core.c since commit 09aea13ecf6f ("media: mtk-jpeg: refactor some variables"), otherwise
the below calltrace can be easily triggered.
Li Zetao [Tue, 10 Sep 2024 15:48:03 +0000 (23:48 +0800)]
media: siano: remove redundant null pointer checks in cec_devnode_init()
Since the debugfs_create_dir() never returns a null pointer, checking
the return value for a null pointer is redundant, Remove this check
since debugfs_create_file can handle IS_ERR pointers.
Signed-off-by: Li Zetao <lizetao1@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Benjamin Gaignard [Tue, 10 Sep 2024 14:10:09 +0000 (14:10 +0000)]
media: verisilicon: av1: Fix reference video buffer pointer assignment
Always get new destination buffer for reference frame because nothing
garantees the one set previously is still valid or unused.
Fixes this chromium test suite:
https://chromium.googlesource.com/chromium/src/media/+/refs/heads/main/test/data/test-25fps.av1.ivf
Fixes: 727a400686a2 ("media: verisilicon: Add Rockchip AV1 decoder") Cc: <stable@vger.kernel.org> Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com> Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
[hverkuil: fix typo and add link to chromium test suite]
Chen Ni [Thu, 5 Sep 2024 02:25:32 +0000 (10:25 +0800)]
media: vde: Convert comma to semicolon
Replace comma between expressions with semicolons.
Using a ',' in place of a ';' can have unintended side effects.
Although that is not the case here, it is seems best to use ';'
unless ',' is intended.
Found by inspection.
No functional change intended.
Compile tested only.
Signed-off-by: Chen Ni <nichen@iscas.ac.cn> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Ming Qian [Wed, 28 Aug 2024 13:01:24 +0000 (16:01 +0300)]
media: amphion: Guard memory allocation to catch failures
The firmware will ask the driver for memory allocation, but it will not
check the completeness of the task. Therefore, the vpu will crash until
reboot. This code will guard this bug and make the driver fail gracefully
when memory allocation cannot be completed.
Signed-off-by: Ming Qian <ming.qian@nxp.com> Reviewed-by: Zhou Peng <eagle.zhou@nxp.com> Signed-off-by: Mihai Despotovici <mihai.despotovici@nxp.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
John Keeping [Mon, 12 Aug 2024 14:35:55 +0000 (15:35 +0100)]
media: platform: rga: fix 32-bit DMA limitation
The destination buffer flags are assigned twice but source is not set in
what looks like a copy+paste mistake. Assign the source queue flags so
the 32-bit DMA limitation is handled consistently.
Fixes: ec9ef8dda2a2 ("media: rockchip: rga: set dma mask to 32 bits") Cc: <stable@vger.kernel.org> Signed-off-by: John Keeping <jkeeping@inmusicbrands.com> Reviewed-by: Michael Tretter <m.tretter@pengutronix.de> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Fabio Estevam [Fri, 9 Aug 2024 01:20:13 +0000 (22:20 -0300)]
media: imx-jpeg: Switch to RUNTIME/SYSTEM_SLEEP_PM_OPS()
Replace SET_RUNTIME_PM_OPS()/SET SYSTEM_SLEEP_PM_OPS() with their modern
RUNTIME_PM_OPS() and SYSTEM_SLEEP_PM_OPS() alternatives.
The combined usage of pm_ptr() and RUNTIME_PM_OPS/SYSTEM_SLEEP_PM_OPS()
allows the compiler to evaluate if the runtime suspend/resume() functions
are used at build time or are simply dead code.
This allows removing the CONFIG_PM ifdefery from the runtime
suspend/resume() functions.
Signed-off-by: Fabio Estevam <festevam@denx.de> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Hans Verkuil [Mon, 2 Sep 2024 14:04:55 +0000 (16:04 +0200)]
media: vb2: use lock if wait_prepare/finish are NULL
If the wait_prepare or wait_finish callback is set, then call it.
If it is NULL and the queue lock pointer is not NULL, then just
unlock/lock that mutex.
This allows simplifying drivers by dropping the wait_prepare and
wait_finish ops (and eventually the vb2_ops_wait_prepare/finish helpers).
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Hans Verkuil [Mon, 2 Sep 2024 14:04:54 +0000 (16:04 +0200)]
media: vb2: vb2_core_queue_init(): sanity check lock and wait_prepare/finish
Add two new checks:
1) wait_prepare and wait_finish callbacks are either both present or
both unset, you can't mix.
2) if lock == NULL, then wait_prepare (and due to check 1 also
wait_finish) must be present.
These checks should prevent the case where lock == NULL, but there
is no way to release/reacquire whatever lock is used when waiting
for a buffer to arrive in VIDIOC_DQBUF.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Hans Verkuil [Mon, 2 Sep 2024 14:04:53 +0000 (16:04 +0200)]
media: video-i2c: set lock before calling vb2_queue_init()
The vb2_queue_init() will expect the vb2_queue lock pointer to be set in
the future. So for those drivers that set the lock later, move it up to
before the vb2_queue_init() call.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Acked-by: Matt Ranostay <matt@ranostay.sg> Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Hans Verkuil [Mon, 2 Sep 2024 14:04:52 +0000 (16:04 +0200)]
media: rcar_drif.c: set lock before calling vb2_queue_init()
The vb2_queue_init() will expect the vb2_queue lock pointer to be set in
the future. So for those drivers that set the lock later, move it up to
before the vb2_queue_init() call.
Hans Verkuil [Mon, 2 Sep 2024 14:04:51 +0000 (16:04 +0200)]
media: airspy: set lock before calling vb2_queue_init()
The vb2_queue_init() will expect the vb2_queue lock pointer to be set in
the future. So for those drivers that set the lock later, move it up to
before the vb2_queue_init() call.
Hans Verkuil [Mon, 2 Sep 2024 14:04:50 +0000 (16:04 +0200)]
media: hackrf: set lock before calling vb2_queue_init()
The vb2_queue_init() will expect the vb2_queue lock pointer to be set in
the future. So for those drivers that set the lock later, move it up to
before the vb2_queue_init() call.
Hans Verkuil [Mon, 2 Sep 2024 14:04:49 +0000 (16:04 +0200)]
media: msi2500: set lock before calling vb2_queue_init()
The vb2_queue_init() will expect the vb2_queue lock pointer to be set in
the future. So for those drivers that set the lock later, move it up to
before the vb2_queue_init() call.
Hans Verkuil [Mon, 2 Sep 2024 14:04:48 +0000 (16:04 +0200)]
media: pwc: set lock before calling vb2_queue_init()
The vb2_queue_init() will expect the vb2_queue lock pointer to be set in
the future. So for those drivers that set the lock later, move it up to
before the vb2_queue_init() call.
Hans Verkuil [Mon, 2 Sep 2024 08:31:24 +0000 (10:31 +0200)]
media: venus: add missing wait_prepare/finish ops
Without these ops the v4l2-compliance blocking wait test will fail.
These ops are required to ensure that when VIDIOC_DQBUF has to
wait for buffers to arrive, the queue lock is correctly released
and retaken. Otherwise the wait for a buffer would block all other
queue ioctls.
Without these ops the v4l2-compliance blocking wait test will fail.
These ops are required to ensure that when VIDIOC_DQBUF has to
wait for buffers to arrive, the queue lock is correctly released
and retaken. Otherwise the wait for a buffer would block all other
queue ioctls.
Without these ops the v4l2-compliance blocking wait test will fail.
These ops are required to ensure that when VIDIOC_DQBUF has to
wait for buffers to arrive, the queue lock is correctly released
and retaken. Otherwise the wait for a buffer would block all other
queue ioctls.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
media: uvcvideo: Avoid race condition during unregister
The control events are handled asynchronously by the driver. Once the
control event are handled, the urb is re-submitted.
If we simply kill the urb, there is a chance that a control event is
waiting to be processed, which will re-submit the urb after the device is
disconnected.
Fix this by calling uvc_status_suspend(), which flushes the async
controls and kills the URB in a race-free manner.
projects / users / jedix / linux-maple.git / log