| Age | Commit message (Collapse) | Author |
|---|
|
IPU7 ISYS and PSYS auxiliary devices are released after
ipu7_bus_del_devices(), so driver can not reference the MMU devices
from ISYS and PSYS auxiliary devices, so move the MMUs cleanup before
releasing the auxiliary devices.
Fixes: b7fe4c0019b1 ("media: staging/ipu7: add Intel IPU7 PCI device driver")
Signed-off-by: Bingbu Cao <bingbu.cao@intel.com>
[Sakari Ailus: Drop extra newline.]
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Driver better not dev_set_name() to change the PCI device
name, so remove it.
Fixes: b7fe4c0019b1 ("media: staging/ipu7: add Intel IPU7 PCI device driver")
Signed-off-by: Bingbu Cao <bingbu.cao@intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
pci_enable_msi() is a deprecated API, thus switch to use modern
pci_alloc_irq_vectors().
Fixes: b7fe4c0019b1 ("media: staging/ipu7: add Intel IPU7 PCI device driver")
Signed-off-by: Bingbu Cao <bingbu.cao@intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Add support for negotiating the highest possible pixel mode (from
single, dual, quad) with the Cadence CSI2RX bridge. This is required to
drain the Cadence stream FIFOs without overflowing when the source is
operating at a high link-frequency [1].
Also, update the Kconfig as this introduces a hard build-time dependency
on the Cadence CSI2RX driver, even for a COMPILE_TEST.
[1] Section 12.6.1.4.8.14 CSI_RX_IF Programming Restrictions of AM62 TRM
Link: https://www.ti.com/lit/pdf/spruj16
Tested-by: Yemike Abhilash Chandra <y-abhilashchandra@ti.com> (on SK-AM68)
Signed-off-by: Jai Luthra <jai.luthra@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The output pixel interface is a parallel bus (32 bits), which
supports sending multiple pixels (1, 2 or 4) per clock cycle for
smaller pixel widths like RAW8-RAW16.
Dual-pixel and Quad-pixel modes can be a requirement if the export rate
of the Cadence IP in Single-pixel mode maxes out before the maximum
supported DPHY-RX frequency, which is the case with TI's integration of
this IP [1].
So, we export a function that lets the downstream hardware block request
a higher pixel-per-clock on a particular output pad.
We check if we can support the requested pixels per clock given the
known maximum for the currently configured format. If not, we set it
to the highest feasible value and return this value to the caller.
[1] Section 12.6.1.4.8.14 CSI_RX_IF Programming Restrictions of AM62 TRM
Link: https://www.ti.com/lit/pdf/spruj16
Tested-by: Yemike Abhilash Chandra <y-abhilashchandra@ti.com> (on SK-AM68)
Signed-off-by: Jai Luthra <jai.luthra@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Use v4l2_subdev_get_fwnode_pad_1_to_1() as the get_fwnode_pad operation.
Cadence CSI2RX maps port numbers and pad indices 1:1.
Reviewed-by: Changhuang Liang <changhuang.liang@starfivetech.com>
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
Tested-by: Yemike Abhilash Chandra <y-abhilashchandra@ti.com> (on SK-AM68)
Signed-off-by: Jai Luthra <jai.luthra@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
We don't use OF ports and remote-endpoints to connect the CSI2RX bridge
and this device in the device tree, thus it is wrong to use
v4l2_create_fwnode_links_to_pad() to create the media graph link between
the two.
It works out on accident, as neither the source nor the sink implement
the .get_fwnode_pad() callback, and the framework helper falls back on
using the first source and sink pads to create the link between them.
Instead, manually create the media link from the first source pad of the
bridge to the first sink pad of the J721E CSI2RX.
Fixes: b4a3d877dc92 ("media: ti: Add CSI2RX support for J721E")
Cc: stable@vger.kernel.org
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
Tested-by: Yemike Abhilash Chandra <y-abhilashchandra@ti.com> (on SK-AM68)
Signed-off-by: Jai Luthra <jai.luthra@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Simplify notifier registration logic. Instead of first getting the
device node, get the fwnode of the child directly.
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
Tested-by: Yemike Abhilash Chandra <y-abhilashchandra@ti.com> (on SK-AM68)
Signed-off-by: Jai Luthra <jai.luthra@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Ensure that we clean up the platform bus when we remove this driver.
This fixes a crash seen when reloading the module for the child device
with the parent not yet reloaded.
Fixes: b4a3d877dc92 ("media: ti: Add CSI2RX support for J721E")
Cc: stable@vger.kernel.org
Reviewed-by: Devarsh Thakkar <devarsht@ti.com>
Tested-by: Yemike Abhilash Chandra <y-abhilashchandra@ti.com> (on SK-AM68)
Signed-off-by: Jai Luthra <jai.luthra@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Remove the workaround that allowed calling v4l2_get_link_freq() on the
control handler.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Call v4l2_get_link_freq() on the source pad so we can remove the
workaround that allows calling it on a control handler.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Reviewed-by: Keke Li <keke.li@amlogic.com>
Tested-by: Keke Li <keke.li@amlogic.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Remove unnecessary semicolons.
Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
Reviewed-by: Kieran Bingham <kieran.bingham@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Make it easier to drop the legacy support from the driver later.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
It is more logical to call the PLL update in imx214_ctrls_init(). So
let's move it there.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Use the __free(fwnode_handle) hook to free the endpoint when the
function exits to simplify the error path.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The device can be accessed from struct imx214.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The parentheses are not needed to calculate bit_rate_mbps.
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: André Apitzsch <git@apitzsch.eu>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Add support for .get_frame_desc() to report CSI-2 virtual channel
and data type information. This allows CSI-2 receivers to properly
interpret the stream without inferring the data type from the pixel
format.
Signed-off-by: Julien Massot <julien.massot@collabora.com>
Reviewed-by: Benjamin Mugnier <benjamin.mugnier@foss.st.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The MT9M114 supports the different slew rates (0 to 7) on the output pads.
At the moment, this is hardcoded to 7 (the fastest rate).
The user might want to change this values due to EMC requirements.
Read the 'slew-rate' from the DT and configure the pad slew rates of
the output pads accordingly in mt9m114_initialize().
Remove the hardcoded slew rate setting from the mt9m114_init table.
Signed-off-by: Mathis Foerst <mathis.foerst@mt.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The current implementation does not apply changes to the crop-
configuration of the sensor immediately if the sensor is in
streaming state. The user has to stop and restart the stream for
the changes to be applied.
This can be undesirable e.g. in a calibration usecase where the user
wants to see the impact of his changes in a live video stream.
Under the condition that the width & height of the cropped image area
does not change, the changed cropping configuration can be applied to
the pixel-array immediately without disturbing the IFP.
Call mt9m114_configure_pa() in mt9m114_pa_set_selection() if the sensor is
in streaming state and the size of the cropping rectangle didn't change,
issue a CONFIG_CHANGE to apply the changes immediately.
Signed-off-by: Mathis Foerst <mathis.foerst@mt.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Don't populate the read-only arrays regs and values on the stack at run
time, instead make them static const.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Drop function pointer to configure CSI to avoid code duplication
by checking the presence of vc select register in rzg2l_cru_info.
After this change, limit the scope of the rzg2l_cru_csi2_setup()
to static.
Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com>
Tested-by: Tommaso Merciai <tommaso.merciai.xr@bp.renesas.com>
Reviewed-by: Tommaso Merciai <tommaso.merciai.xr@bp.renesas.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Added support for three additional test patterns in the
IMX334 driver: Black and Grey Bars, Black Color, and White Color.
Signed-off-by: Shravan Chippa <shravan.chippa@microchip.com>
[Sakari Ailus: Fix hexadecimal value.]
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
rj54n1_probe() won't clean all the allocated resources in fail
path, which may causes the memleaks. Add v4l2_ctrl_handler_free() to
prevent memleak.
Fixes: f187352dcd45 ("media: i2c: Copy rj54n1cb0c soc_camera sensor driver")
Signed-off-by: Zhang Shurong <zhang_shurong@foxmail.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
linux/version.h isn't needed by the driver, don't include it.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The atomisp gc0310 sensor driver has now been fully converted to
a standard v4l2 sensor driver. Move it to drivers/media/i2c/
to reflect this.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
MEDIA_CONTROLLER is selected for vim2m already since commit 016baa59bf9f
("media: Kconfig: Don't expose the Request API option"). Also remove the
related #ifdefs.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Assign the driver to Nikita as discussed on LMML.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The IPU6 ISYS driver supported metadata formats but was missing correct
embedded data type in the receiver configuration. Add it now.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The ipu7-isys driver uses v4l2_ctrl_subscribe_event() to handle control
event subscription on a subdev. While this works, it is the wrong API.
Use the subdev-specific v4l2_ctrl_subdev_subscribe_event() helper
instead.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The ipu6-isys driver uses v4l2_ctrl_subscribe_event() to handle control
event subscription on a subdev. While this works, it is the wrong API.
Use the subdev-specific v4l2_ctrl_subdev_subscribe_event() helper
instead.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The csi_unsubscribe_event() function simply calls
v4l2_event_unsubscribe(), forwarding its arguments. Replace it with the
v4l2_event_subdev_unsubscribe() helper that performs exactly the same
operation.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The isp_subdev_unsubscribe_event() function simply calls
v4l2_event_unsubscribe(), forwarding its arguments. Replace it with the
v4l2_event_subdev_unsubscribe() helper that performs exactly the same
operation.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The ccdc_unsubscribe_event() and omap3isp_stat_unsubscribe_event()
functions simply call v4l2_event_unsubscribe(), forwarding their
arguments. Replace them with the v4l2_event_subdev_unsubscribe() helper
that performs exactly the same operation.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The v4l2_subdev_stream_config structure holds configuration data for a
stream. It was meant to be used internally only, but already found its
way into the ds90ub913 driver. Now that the driver has been fixed, make
the structure private to v4l2-subdev.c to avoid using it by accident.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The IPU7 ISYS driver can't capture multiple streams on the same video
device. Disallow source multiplexing in the routes of the internal
subdev to reflect that limitation. As a result we can hardcode the
source stream to 0, simplifying the driver.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The v4l2_subdev_stream_config structure will be made private. Stop
accessing it directly, iterate over routes instead to initialize
formats.
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Jacopo Mondi <jacopo.mondi@ideasonboard.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Use string_choices API instead of manually determining the output using
ternary operator to improve code readability and consistency.
This also fixes cocci warning.
Signed-off-by: Nai-Chen Cheng <bleach1827@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
The platform device name is already set by V4L2 and MC frameworks for
media device bus_info and V4L2 querycap bus_info fields. Don't do it in
the driver.
Acked-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
|
|
Amlogic clock controller drivers are all doing the same thing, more or
less. Yet, over the years, tiny (and often pointless) differences have
emerged.
This makes reviews more difficult, allows some errors to slip through and
make it more difficult to exploit SoC commonalities, leading to code
duplication.
This change enforce, wherever possible, a consistent and predictable scheme
when it comes to code organisation and naming, The scheme chosen is what
was used the most already, to try and minimise the size of the ugly
resulting diff. Here are some of the rules applied:
- Aligning clock names, variable names and IDs.
- ID cannot change (used in DT)
- Variable names w/ SoC name prefixes
- Clock names w/o SoC name prefixes, except pclks for historic reasons
- Composite clock systematic naming : mux: X_sel, div:X_div, gate:X
- Parent table systematically named with the same name as the clock and
a '_parents' suffix
- Group various tables next to the related clock
- etc ...
Doing so removes what would otherwise show up as unrelated diff in
following changes. It will allow to introduce common definitions for
peripheral clocks, probe helpers, composite clocks, etc ... making further
review and maintenance easier.
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-1-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-2-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-3-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-4-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-5-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-6-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-7-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-8-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-9-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-10-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-11-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-12-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-13-e163c9a1fc21@baylibre.com
Link: https://lore.kernel.org/r/20250702-meson-clk-cleanup-24-v1-14-e163c9a1fc21@baylibre.com
Reviewed-by: Chuan Liu <chuan.liu@amlogic.com> # For c3 and s4
[jbrunet: squashed all naming alignment changes together]
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
|
|
This patch introduces dm-pcache, a new DM target that places a DAX-
capable persistent-memory device in front of any slower block device and
uses it as a high-throughput, low-latency cache.
Design highlights
-----------------
- DAX data path – data is copied directly between DRAM and the pmem
mapping, bypassing the block layer’s overhead.
- Segmented, crash-consistent layout
- all layout metadata are dual-replicated CRC-protected.
- atomic kset flushes; key replay on mount guarantees cache integrity
even after power loss.
- Striped multi-tree index
- Multi‑tree indexing for high parallelism.
- overlap-resolution logic ensures non-intersecting cached extents.
- Background services
- write-back worker flushes dirty keys in order, preserving backing-device
crash consistency. This is important for checkpoint in cloud storage.
- garbage collector reclaims clean segments when utilisation exceeds a
tunable threshold.
- Data integrity – optional CRC32 on cached payload; metadata always protected.
Comparison with existing block-level caches
---------------------------------------------------------------------------------------------------------------------------------
| Feature | pcache (this patch) | bcache | dm-writecache |
|----------------------------------|---------------------------------|------------------------------|---------------------------|
| pmem access method | DAX | bio (block I/O) | DAX |
| Write latency (4 K rand-write) | ~5 µs | ~20 µs | ~5 µs |
| Concurrency | multi subtree index | global index tree | single tree + wc_lock |
| IOPS (4K randwrite, 32 numjobs) | 2.1 M | 352 K | 283 K |
| Read-cache support | YES | YES | NO |
| Deployment | no re-format of backend | backend devices must be | no re-format of backend |
| | | reformatted | |
| Write-back ordering | log-structured; | no ordering guarantee | no ordering guarantee |
| | preserves app-IO-order | | |
| Data integrity checks | metadata + data CRC(optional) | metadata CRC only | none |
---------------------------------------------------------------------------------------------------------------------------------
Signed-off-by: Dongsheng Yang <dongsheng.yang@linux.dev>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
|
|
dma_free_coherent() must only be called if the corresponding
dma_alloc_coherent() call has succeeded. Calling it when the allocation fails
leads to undefined behavior.
Delete the wrong call.
[ bp: Massage commit message. ]
Fixes: 71bcada88b0f3 ("edac: altera: Add Altera SDRAM EDAC support")
Signed-off-by: Salah Triki <salah.triki@gmail.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Acked-by: Dinh Nguyen <dinguyen@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/aIrfzzqh4IzYtDVC@pc
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-10-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-9-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-8-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-7-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-6-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-5-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-4-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
|
The drivers do not require their own error messages for error
-ENOMEM, memory allocation failures. So remove the dev_err()
messages from the probe().
Signed-off-by: Dixit Parmar <dixitparmar19@gmail.com>
Link: https://patch.msgid.link/20250822-enomam_logs-v1-3-db87f2974552@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
|
