Add #define's for the output enable registers (0x3004, 0x3005, 0x3006),
also known as SC_CTRL_04, SC_CTRL_05, SC_CTRL_04. Use those register
definitions instead of the raw values in the `common_regs` struct.
All values are based on the OV9281 datasheet v1.53 (january 2019).
Reviewed-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Signed-off-by: Richard Leitner <richard.leitner@linux.dev>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Add the new strobe duration and hardware strobe output enable to v4l
uAPI documentation. Additionally add labels for cross-referencing v4l
controls.
Signed-off-by: Richard Leitner <richard.leitner@linux.dev>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Add a control V4L2_CID_FLASH_STROBE_OE to en- or disable the
strobe output of v4l2 devices (most likely sensors).
Signed-off-by: Richard Leitner <richard.leitner@linux.dev>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Add a V4L2_CID_FLASH_DURATION control to set the duration of a
flash/strobe pulse. This controls the length of the flash/strobe pulse
output by device (typically a camera sensor) and connected to the flash
controller. This is different to the V4L2_CID_FLASH_TIMEOUT control,
which is implemented by the flash controller and defines a limit after
which the flash is "forcefully" turned off again.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Richard Leitner <richard.leitner@linux.dev>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
The imx219_update_pad_format() is short and called from a single place,
in imx219_set_pad_format(). Inline the code in the caller to keep all
format adjustments grouped in a single place and improve readability.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Dave Stevenson <dave.stevenson@raspberrypi.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Prepare for adding more media bus codes.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
ccs_propagate() no longer stores information in the driver's context
struct. The which parameter can thus be removed.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Compute scaling configuration from sub-device state instead of storing it
to the driver's device context struct.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Calculate binning configuration from sub-device state so the state related
configuration can be removed from the driver's device context struct.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Replace the use of mutex_{,un}lock() by guard() and scoped_guard() where
it makes sense (i.e. everywhere).
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Rely on sub-device state locking to serialise access to driver's data
structures. The driver-provided mutex is used as the state lock for all
driver sub-devices.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
As it's now easier to obtain the necessary information on crop and compose
rectangles after moving to sub-device state, remove the
ccs_get_crop_compose helper.
Also remove the comp arguments of the compose goodness calculators and
make related local variables and function arguments const where
applicable.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Digital gain wraps-around at the maximum of 16838 / 0x3fff.
Fix the maximum digital gain by setting it to 0x3fff.
Signed-off-by: Mehdi Djait <mehdi.djait@linux.intel.com>
Reviewed-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Fixes: 0827b58dab ("media: i2c: add ov01a10 image sensor driver")
Cc: stable@vger.kernel.org
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Fix indentation in ov01a10_get_pm_resources().
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Remove Hongju's e-mail address that no longer works.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
The struct dcmipp_bytecap_device embeds a cmier variable which is
actually only set once and then used in several places for interrupt
mask or interrupt enable/disable. Use the DCMIPP_CMIER_P0ALL
bitfield macro directly instead of using it through the cmier
variable.
There is also no need to mask again vcap->cmsr2 against enabled
interrupts bit since this has already been done when storing the
interrupt status into the cmsr2 variable.
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Since the byteproc is working at byte level, it is not possible to
perform decimation with formats other than 8 or 16 bit per pixel.
Ensure any other format won't be able to set a compose.
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
DCMIPP can handle Y8 to Y14 input formats however until now only
Y8 was handled. Add support for Y10-Y12-Y14 in all relevant
subdevs.
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Add entries within several subdevs in order to handle RGB888 format.
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Avoid possibility to perform compose on all frames which mbus code is
within the bayer range or jpeg format.
Fixes: 822c72eb15 ("media: stm32: dcmipp: add bayer 10~14 bits formats")
Cc: stable@vger.kernel.org
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Ensure that there are no pending interrupts after we have stopped the
pipeline. Indeed, it could happen that new interrupt has been generated
during the stop_streaming processing hence clear them in order to avoid
getting a new interrupt right from the start of a next start_streaming.
Fixes: 28e0f37722 ("media: stm32-dcmipp: STM32 DCMIPP camera interface driver")
Cc: stable@vger.kernel.org
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
adv7604 and et8ek8 sensor drivers have mixed up logical and line level
for reset/powerdown signal. They call it a reset signal (it indeed
behaves like that), but drivers assert the reset to operate which is
clearly incorrect and relies on wrong ACTIVE_HIGH flag in the DTS.
People in discussions copy existing poor code and claim they can repeat
same mistake, so add a note to prevent that.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
When DCMIPP requires only a single clock (kclk), avoid relying on its
name to obtain it. The introduction of MP25 support added the mclk,
which necessitated naming the first clock kclk. However, this breaks
backward compatibility with existing MP13 device trees that do not
specify clock names.
Fixes: 686f27f7ea ("media: stm32: dcmipp: add core support for the stm32mp25")
Signed-off-by: Alain Volmat <alain.volmat@foss.st.com>
Cc: Stable@vger.kernel.org # 6.14.x: 7f487562af49 media: stm32: dcmipp: correct ret type in dcmipp_graph_notify_bound
Cc: Stable@vger.kernel.org # 6.14.x: c715dd62da30 media: stm32: dcmipp: add has_csi2 & needs_mclk in match data
Cc: Stable@vger.kernel.org # 6.14.x:
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Move Rui Miguel Silva to reviewer since he does not have sufficient time to
maintain i.MX7/8 media drivers. Frank Li has the hardware, complete
documentation, and has participated actively in patch review. Promote him
as the i.MX7/8 media drivers maintainer.
Reviewed-by: Rui Miguel Silva <rmfrfs@gmail.com>
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
We need to set correct mode for PLL to calculate correct frequency.
Signalling mode is known at this point, so use it for that.
Fixes: 47b6eaf36e ("media: ccs-pll: Differentiate between CSI-2 D-PHY and C-PHY")
Reviewed-by: Mehdi Djait <mehdi.djait@linux.intel.com>
Signed-off-by: David Heidelberg <david@ixit.cz>
[Sakari Ailus: Drop extra newline.]
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Rely on devm_v4l2_sensor_clk_get() to obtain the sensor's external clock.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Fix coding style issues and do other small cleanups in ccs_get_hwcfg().
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: David Heidelberg <david@ixit.cz>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Switch from s_stream to enable_streams and disable_streams callbacks.
Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Tarang Raval <tarang.raval@siliconsignals.io>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Switch to using the sub-device state lock and properly call
v4l2_subdev_init_finalize() / v4l2_subdev_cleanup() on probe() /
remove().
Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com>
Reviewed-by: Tarang Raval <tarang.raval@siliconsignals.io>
[Sakari Ailus: Remove now-redundant assignment of control handler lock.]
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Use the new common CCI register access helpers to replace the private
register access helpers in the ov5647 driver. This simplifies the driver
by reducing the amount of code.
Signed-off-by: Xiaolei Wang <xiaolei.wang@windriver.com>
Reviewed-by: Tarang Raval <tarang.raval@siliconsignals.io>
Reviewed-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Remove blank lines before closing braces to comply with
the Linux kernel coding style.
Issue identified by checkpatch.pl.
Signed-off-by: ArchitAnant <architanant5@gmail.com>
Link: https://patch.msgid.link/20260114094949.14825-3-architanant5@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Replace large if (memcmp(p+2, WMM_IE, 6)) inside the for (;;) loop,
which ends with a break, with an early continue.
This reduces deep nesting and is purely stylistic.
Signed-off-by: William Hansen-Baird <william.hansen.baird@gmail.com>
Link: https://patch.msgid.link/20260113211433.89512-2-william.hansen.baird@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Rewrite if-else construct with an early exit to reduce indentation,
and make the execution clearer.
Signed-off-by: William Hansen-Baird <william.hansen.baird@gmail.com>
Link: https://patch.msgid.link/20260113211433.89512-1-william.hansen.baird@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
renamed variables supportRate and supportRateNum to support_rate and
support_rate_num to adhere to the linux kernel coding style which mandates
snake_case for variable names.
This fixes the following checkpatch.pl warning:
CHECK: Avoid CamelCase: <supportRate>
CHECK: Avoid CamelCase: <supportRateNum>
Signed-off-by: Matous Jarolim <matous.jarolim.lkml@gmail.com>
Link: https://patch.msgid.link/20260113170346.557634-1-matous.jarolim.lkml@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix checkpatch.pl warnings regarding block comment alignment.
The warnings were: "Block comments should align the * on each line".
This patch aligns the asterisks in the block comments so they adhere to
the kernel coding style.
Signed-off-by: Mahad Ibrahim <mahad.ibrahim.dev@gmail.com>
Link: https://patch.msgid.link/20260112174227.14922-5-mahad.ibrahim.dev@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix checkpatch.pl checks in HalHWImg8723B_MAC.c regarding:
-Spaces preferred around that operator
Adhere to kernel coding style by adding spaces around operators.
Signed-off-by: Mahad Ibrahim <mahad.ibrahim.dev@gmail.com>
Link: https://patch.msgid.link/20260112174227.14922-3-mahad.ibrahim.dev@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix checkpatch warnings regarding block comment alignment.
The warnings were: "Block comments should align the * on each line".
This patch aligns the asterisks in the block comments to match the
kernel coding style.
Signed-off-by: Mahad Ibrahim <mahad.ibrahim.dev@gmail.com>
Link: https://patch.msgid.link/20260112174227.14922-2-mahad.ibrahim.dev@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This continues the effort to refactor workqueue APIs, which began with
the introduction of new workqueues and a new alloc_workqueue flag in:
commit 128ea9f6cc ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566 ("workqueue: Add new WQ_PERCPU flag")
The refactoring is going to alter the default behavior of
alloc_workqueue() to be unbound by default.
With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND),
any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND
must now use WQ_PERCPU. For more details see the Link tag below.
In order to keep alloc_workqueue() behavior identical, explicitly request
WQ_PERCPU.
Link: https://lore.kernel.org/all/20250221112003.1dSuoGyc@linutronix.de/
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Link: https://patch.msgid.link/20260113135737.190636-1-marco.crivellari@suse.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes the checkpatch.pl warning:
"WARNING: else is not generally useful after a break or return"
In sdio_ops_linux.c, the else blocks after break statements are
redundant. Removing them reduces indentation level and improves
code readability.
Signed-off-by: Vo Thanh Cong <thanhcongvo079@gmail.com>
Link: https://patch.msgid.link/20260113103738.207570-1-thanhcongvo079@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This contains a few small fixes for the current cycle.
- dt-bindings: interconnect: qcom,sa8775p-rpmh: Fix incorrectly added reg and clocks
- MAINTAINERS: Add interconnect-clk.h to interconnect API entry
- interconnect: debugfs: initialize src_node and dst_node to empty strings
Signed-off-by: Georgi Djakov <djakov@kernel.org>
-----BEGIN PGP SIGNATURE-----
iQIzBAABCgAdFiEExQ4tCsdmOu34XfMFgNDMCsbYFmMFAmlo5hwACgkQgNDMCsbY
FmODQw/9G/9SaSnvbeRo6JoU1HaKJOnrUYhRl8/Hp7HQy8ZKxCbi3zQ1P0THPmrc
ljlKZzFOaOV5v9fppc7hFduRuK+16a9ovzEqivSphWotZawmM9zW2U2QZnJZ57Sf
LZM8bqTtb3NDcFSjQ1nmtga0p29mqBR1n0KIjD9V+RDCSKhfj55fYw0Qsbk7IG90
SCLMGZvPwhM2/p8KqarhF860PMbpzYrP6XfSOUWEK4LuHvVNGlC39Nyotqz8mCv8
wTg6UyQzplB2nfpbEpM3qMpwvIa/bc0xo6W4laDsR1yssVQTIa30NdL2D0mtOixQ
DFaQf58QpIIg0H7a2RXPMj3GT/CXwxR7mNXlEazW+qMYAmStrcdT79MHO09d3Fwn
bVHa0W0OtMRqzFa3TCZ3P0oOmRgO4Km6t4KBt9WHzgZaLHrgHBzRR9oVxHE5Ipq+
lEznzV1lD6o1qOC3w8WCpgiWO9DIHZ/jj5EyT/26cHmzNLIqNP+nljX3FSjcMUgT
Kr7ddQrxLjBqZbctrDqndSaundzztCYOpycXn6jyA3bZkBD0bY/fhKqraNte5Hbr
bBQc/LYKERnH4XaP3H3NxDF4DcC1vOfC/VBB9qLTYIVqLYIvo4/Az41BiI0Eghnv
BK3/3t2G/OaXn/nU/LhKbvOq7aryJ8FAZVTbGEHxUT2u56Qc/2Y=
=d8IB
-----END PGP SIGNATURE-----
Merge tag 'icc-6.19-rc6' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/djakov/icc into char-misc-linus
interconnect fixes for v6.19-rc
This contains a few small fixes for the current cycle.
- dt-bindings: interconnect: qcom,sa8775p-rpmh: Fix incorrectly added reg and clocks
- MAINTAINERS: Add interconnect-clk.h to interconnect API entry
- interconnect: debugfs: initialize src_node and dst_node to empty strings
Signed-off-by: Georgi Djakov <djakov@kernel.org>
* tag 'icc-6.19-rc6' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/djakov/icc:
interconnect: debugfs: initialize src_node and dst_node to empty strings
MAINTAINERS: Add interconnect-clk.h to interconnect API entry
dt-bindings: interconnect: qcom,sa8775p-rpmh: Fix incorrectly added reg and clocks
The WCH CH334/CH335[0] are USB2.0 protocol compliant 4-port USB HUB
controller chips, supporting USB2.0 high-speed and full-speed for
upstream ports, and USB2.0 high-speed 480Mbps, full-speed 12Mbps and
low-speed 1.5Mbps for downstream ports, supporting not only low-cost STT
mode (single TT schedules 4 downstream ports in time share), but also
supports high performance MTT mode (4 TTs each corresponding to 1 port,
concurrent processing).
[0]: https://www.wch-ic.com/downloads/CH334DS1_PDF.html
Signed-off-by: Chaoyi Chen <chaoyi.chen@rock-chips.com>
Link: https://patch.msgid.link/20260113095827.115-3-kernel@airkyi.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The WCH CH334/CH335[0] are USB2.0 protocol compliant 4-port USB HUB
controller chips, supporting USB2.0 high-speed and full-speed for
upstream ports, and USB2.0 high-speed 480Mbps, full-speed 12Mbps and
low-speed 1.5Mbps for downstream ports, supporting not only low-cost STT
mode (single TT schedules 4 downstream ports in time share), but also
supports high performance MTT mode (4 TTs each corresponding to 1 port,
concurrent processing).
Add a device tree binding for it.
[0]: https://www.wch-ic.com/downloads/CH334DS1_PDF.html
Signed-off-by: Chaoyi Chen <chaoyi.chen@rock-chips.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20260113095827.115-2-kernel@airkyi.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use dwc pointer in dwc3_readl() dwc3_writel() instead of passing
the dwc->regs. This would help us access the dwc structure and
log the base address in traces. There's no functional changes in
this patch, just refactoring existing APIs.
Signed-off-by: Prashanth K <prashanth.k@oss.qualcomm.com>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://patch.msgid.link/20260114100748.2950103-3-prashanth.k@oss.qualcomm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove dep->regs from struct dwc3_ep and reuse dwc->regs instead.
Thus eliminating redundant iomem addresses and making register
access more consistent across the driver.
Signed-off-by: Prashanth K <prashanth.k@oss.qualcomm.com>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://patch.msgid.link/20260114100748.2950103-2-prashanth.k@oss.qualcomm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Rename the driver data pointer for consistency with the rest of the
driver and drop a redundant cast.
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Link: https://lore.kernel.org/r/20251210040258.60106-1-johan@kernel.org
Signed-off-by: Andi Shyti <andi.shyti@kernel.org>
The Command Queue Virtualization (CMDQV) hardware is part of the
SMMUv3 implementation on NVIDIA Tegra SoCs. It assists in
virtualizing the command queue for the SMMU.
Add a new device tree binding document for nvidia,tegra264-cmdqv.
Also update the arm,smmu-v3 binding to include an optional nvidia,cmdqv
property. This property is a phandle to the CMDQV device node, allowing
the SMMU driver to associate with its corresponding CMDQV instance.
Restrict this property usage to Nvidia Tegra264 only.
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Acked-by: Nicolin Chen <nicolinc@nvidia.com>
Signed-off-by: Ashish Mhetre <amhetre@nvidia.com>
Reviewed-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
When a filesystem is frozen, quotactl_block() enters a retry loop
waiting for the filesystem to thaw. It acquires s_umount, checks the
freeze state, drops s_umount and uses sb_start_write() - sb_end_write()
pair to wait for the unfreeze.
However, this retry loop can trigger a livelock issue, specifically on
kernels with preemption disabled.
The mechanism is as follows:
1. freeze_super() sets SB_FREEZE_WRITE and calls sb_wait_write().
2. sb_wait_write() calls percpu_down_write(), which initiates
synchronize_rcu().
3. Simultaneously, quotactl_block() spins in its retry loop, immediately
executing the sb_start_write() - sb_end_write() pair.
4. Because the kernel is non-preemptible and the loop contains no
scheduling points, quotactl_block() never yields the CPU. This
prevents that CPU from reaching an RCU quiescent state.
5. synchronize_rcu() in the freezer thread waits indefinitely for the
quotactl_block() CPU to report a quiescent state.
6. quotactl_block() spins indefinitely waiting for the freezer to
advance, which it cannot do as it is blocked on the RCU sync.
This results in a hang of the freezer process and 100% CPU usage by the
quota process.
While this can occur intermittently on multi-core systems, it is
reliably reproducing on a node with the following script, running both
the freezer and the quota toggle on the same CPU:
# mkfs.ext4 -O quota /dev/sda 2g && mkdir a_mount
# mount /dev/sda -o quota,usrquota,grpquota a_mount
# taskset -c 3 bash -c "while true; do xfs_freeze -f a_mount; \
xfs_freeze -u a_mount; done" &
# taskset -c 3 bash -c "while true; do quotaon a_mount; \
quotaoff a_mount; done" &
Adding cond_resched() to the retry loop fixes the issue. It acts as an
RCU quiescent state, allowing synchronize_rcu() in percpu_down_write()
to complete.
Fixes: 576215cffd ("fs: Drop wait_unfrozen wait queue")
Signed-off-by: Abhishek Bapat <abhishekbapat@google.com>
Link: https://patch.msgid.link/20260115213103.1089129-1-abhishekbapat@google.com
Signed-off-by: Jan Kara <jack@suse.cz>
Accesses to external memory are routed through the data backbone (DBB)
on Tegra264. A separate clock feeds this path and needs to be enabled
whenever an IP block makes an access to external memory. The external
memory controller driver is the best place to control this clock since
it knows how many devices are actively accessing memory.
Document the presence of this clock on Tegra264 only.
Acked-by: Rob Herring (Arm) <robh@kernel.org>
Signed-off-by: Thierry Reding <treding@nvidia.com>
