Included necessary input system header files (local, private, public
scopes) for both types of atomisp. There is also a conflict of
declaration issue so the unused input_system_state_s is removed.

Link: https://lore.kernel.org/r/20230713100231.308923-4-hpa@redhat.comSigned-off-by: Kate Hsuan <hpa@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The original mipi_format_t represents the MIPI format settings for both
atomisp with the same name and the enum member was controlled by #if
defined(ISP2401). However, if we want to make this driver generic, this
enum should be individually defined for them.

Since the format type is redefined, the format setting function
ia_css_isys_set_fmt_type() should be reworked for ISP2400 and ISP2401
called ia_css_isys_2400_set_fmt_type() and ia_css_isys_2401_set_fmt_type()
.

Additionally, the function called is_mipi_format_yuv420() isn't used by any
functions so it can be removed.

Link: https://lore.kernel.org/r/20230713100231.308923-3-hpa@redhat.comSigned-off-by: Kate Hsuan <hpa@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

This portion of debug code was based on a structure for both ISP2400 and
2401 with the same name and controlled by #if defined(ISP2401). This
structure was only used to dump the status of atomisp. So, to make the
driver generic, this part of debug codes can be removed.

Link: https://lore.kernel.org/r/20230713100231.308923-2-hpa@redhat.comSigned-off-by: Kate Hsuan <hpa@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The v4l2_ctrl_s_ctrl(asd->run_mode) call in atomisp_subdev_init_struct()
gets immediately overridden by a second call directly after
atomisp_subdev_init_struct() is called.

And the second call in atomisp_open() also is not helpful.
ATOMISP_RUN_MODE_PREVIEW is the default and if changed controls
are supposed to stay changed over an open/close of the /dev/video#
node. So drop both calls.

Link: https://lore.kernel.org/r/20230619105212.303653-7-hdegoede@redhat.comSigned-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Remove the support to update a V4L2_CID_RUN_MODE run-mode control
on sensors when changing the atomisp run-mode or directly by calling
the custom ATOMISP_IOC_S_SENSOR_RUNMODE IOCTL.

No sensor drivers implement this and having custom controls / IOCTLs
is undesirable.

Even if there was such a control on sensors then userspace should directly
talk to the sensor v4l2-subdev, rather than relying on a custom IOCTLs
on the output /dev/video# node to pass this through to the senor.

Link: https://lore.kernel.org/r/20230619105212.303653-6-hdegoede@redhat.comSigned-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Add some higher resolutions to the fixed list of resolutions which
atomisp_enum_framesizes() uses on sensors which can do cropping and can
thus make any resolution that will fit.

This is useful for higher resolution sensors like the 2560x1920 ov5693
sensor.

Note the highest resolutions added here are 1920x<height> because
the atomisp firmware does not support widths > 1920 with the default
asd->run_mode->val == ATOMISP_RUN_MODE_PREVIEW setting.

Link: https://lore.kernel.org/r/20230619105212.303653-5-hdegoede@redhat.comSigned-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Change atomisp_enum_framesizes() cut off for too small resolutions
from 2/3th to 5/8th this results in more resolutions being available
with some sensors.

E.g. this allows using 800x600 with a 1280x960 sensor.

Link: https://lore.kernel.org/r/20230619105212.303653-4-hdegoede@redhat.comSigned-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

The pipeline firmware-binaries used in previed mode have
ia_css_binary_xinfo.output.max_width set to 1920.

This causes ia_css_binary_find() to fail when trying to set a higher
resolution resulting in the dump_stack() call in ia_css_binary_find()
triggering and resulting in the try_fmt() or set_fmt() IOCTL failing.

Fix this by clamping the width to max 1920 when in preview mode.

Link: https://lore.kernel.org/r/20230619105212.303653-3-hdegoede@redhat.comSigned-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Fix missing v4l2_fh_release() in atomisp_open()'s
"if (pipe->users)" error exit path.

Closes: https://lore.kernel.org/r/202306180511.XWN0Hr7F-lkp@intel.com/
Link: https://lore.kernel.org/r/20230619105212.303653-1-hdegoede@redhat.comReported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Since some parts of the data structure elements are determined in compile
time, the configuration data structure should be compiled for both two
ISP models. In order to set the configuration for both ISP models in
runtime, The sh_css_sp_group is unified to one data structure for the
configuration to ensure the data structure can be used for both ISP2400
and 2401 in runtime. Also, the unused codes for debug purpose are removed.

Form the aspect of ISP setting, the length and the content of the
configuration structure for ISP2400 and ISP2401 are different. So, we
need to pick up the necessary part of sp_group configuration for both
models and then copy those parts into a temporary buffer. This buffer
is finally written to the ISP with the corresponding length.

Link: https://lore.kernel.org/r/20230619052453.48598-4-hpa@redhat.comSigned-off-by: Kate Hsuan <hpa@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Since the debug codes around #ifdef HAS_WATCHDOG_SP_THREAD_DEBUG are
no longer used, the debug implementation of this can be removed.

Link: https://lore.kernel.org/r/20230619052453.48598-3-hpa@redhat.comSigned-off-by: Kate Hsuan <hpa@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

This portion of the codes was not referenced so it can be completely
removed.

Link: https://lore.kernel.org/r/20230619052453.48598-2-hpa@redhat.comSigned-off-by: Kate Hsuan <hpa@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>

Add new compatible for Amlogic's Meson-S4 IR Controller.
Meson IR Controller supports hardware decoder in Meson-S4 and later SoC.
Signed-off-by: Zelong Dong <zelong.dong@amlogic.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Meson IR Controller supports hardware decoder in Meson-S4 and later
SoC. So far, protocol NEC could be decoded by hardware decoder.
Signed-off-by: Zelong Dong <zelong.dong@amlogic.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The ir-rx51 is a pwm-based TX driver specific to the N900. This can be
handled entirely by the generic pwm-ir-tx driver, and in fact the
pwm-ir-tx driver has been compatible with ir-rx51 from the start.

Note that the suspend code in the ir-rx51 driver is unnecessary, since
during transmit, the process is not in interruptable sleep. The process
is not put to sleep until the transmit completes.
Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Tested-by: Sicelo A. Mhlongo <absicsz@gmail.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The generic pwm-ir-tx driver works for the Nokia n900, so nokia,n900-ir
can be removed.
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Acked-by: Lee Jones <lee@kernel.org>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Add driver for Video Capture/Differentiation Engine (VCD) and Encoding
Compression Engine (ECE) present on Nuvoton NPCM SoCs. As described in
the datasheet NPCM750D_DS_Rev_1.0, the VCD can capture frames from
digital video input and compare two frames in memory, and then the ECE
can compress the frame data into HEXTILE format. This driver implements
V4L2 interfaces and provides user controls to support KVM feature, also
tested with VNC Viewer ver.6.22.826 and openbmc/obmc-ikvm.
Signed-off-by: Marvin Lin <milkfafa@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Create controls for Nuvoton NPCM video driver to support setting
capture mode of Video Capture/Differentiation (VCD) engine and getting
the count of HEXTILE rectangles that is compressed by Encoding
Compression Engine (ECE).
Signed-off-by: Marvin Lin <milkfafa@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Add a control base for Nuvoton NPCM driver controls, and reserve 16
controls.
Signed-off-by: Marvin Lin <milkfafa@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Add HEXTILE compressed format which is defined in Remote Framebuffer
Protocol (RFC 6143, chapter 7.7.4 Hextile Encoding) and is used by
Encoding Compression Engine (ECE) present on Nuvoton NPCM SoCs.
Signed-off-by: Marvin Lin <milkfafa@gmail.com>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Add dt-bindings document for Graphics Core Information (GFXI) node. It
is used by NPCM video driver to retrieve Graphics core information.
Signed-off-by: Marvin Lin <milkfafa@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Add dt-bindings document for Video Capture/Differentiation Engine
(VCD) and Encoding Compression Engine (ECE) present on Nuvoton NPCM
SoCs.
Signed-off-by: Marvin Lin <milkfafa@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The last driver that still used this old framework has been converted
to the videobuf2 framework. So it is now time to delete the old videobuf
code.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>

This source still included the media/videobuf-vmalloc.h
header, even though it is no longer used.

Since the old videobuf framework is about to be removed,
delete this include first.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>

These are never implemented, remove them.
Signed-off-by: Yue Haibing <yuehaibing@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Prefer struct_size() over open-coded versions of idiom:

sizeof(struct-with-flex-array) + sizeof(typeof-flex-array-elements) * count

where count is the max number of items the flexible array is supposed to
contain.
Signed-off-by: Yu Liao <liaoyu15@huawei.com>
Reviewed-by: Tomasz Figa <tfiga@chromium.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

There are a couple of spelling mistakes in pr_err error messages. Fix them.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Devicetree for the FIMC IS camera included duplicated PMU node as its
child like:

  soc@0 {
    system-controller@10020000 { ... }; // Real PMU

    camera@11800000 {
      fimc-is@12000000 {
        // FIMC IS camera node
        pmu@10020000 {
          reg = <0x10020000 0x3000>; // Fake PMU node
        };
      };
    };
  };

This is not a correct representation of the hardware.  Mapping the PMU
(Power Management Unit) IO memory should be via syscon-like phandle
(samsung,pmu-syscon, already used for other drivers), not by duplicating
"pmu" Devicetree node inside the FIMC IS.  Backward compatibility is
preserved.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The camera node's ranges property and unit addresses of its children
were not correct.  If camera is @11800000, then its fimc child is @0.
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The FIMC IS camera must access the PMU (Power Management Unit) IO memory
to control camera power.  This was achieved by duplicating the PMU node
as its child like:

  soc@0 {
    system-controller@10020000 { ... }; // Real PMU

    camera@11800000 {
      fimc-is@12000000 {
        // FIMC IS camera node
        pmu@10020000 {
          reg = <0x10020000 0x3000>; // Fake PMU node
        };
      };
    };
  };

This is not a correct representation of the hardware.  Mapping the PMU
(Power Management Unit) IO memory should be via syscon-like phandle
(samsung,pmu-syscon, already used for other drivers), not by duplicating
"pmu" Devicetree node inside the FIMC IS.
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The driver depends on CONFIG_OF, it is not necessary to use
of_match_ptr() here.
Signed-off-by: Ruan Jinjie <ruanjinjie@huawei.com>
Acked-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Acked-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Use the helper function devm_kmemdup() rather than duplicating its
implementation, which helps to enhance code readability.
Signed-off-by: Li Zetao <lizetao1@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Instead of creating a huge switch statement (that thoroughly confuses
smatch), create a table of coefficients and use that.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

It is possible that dma_request_chan will return EPROBE_DEFER,
which means that dma->xdev->dev is not ready yet. In this case,
dev_err(dma->xdev->dev), there will be no output. This patch
fixes the bug.
Signed-off-by: Wang Ming <machel@vivo.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Don't populate the arrays on the stack, instead make them static const.
Also add spaces between values to clean up checkpatch style warnings.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

The module loads firmware so add a MODULE_FIRMWARE macro to provide that
information via modinfo.
Signed-off-by: Juerg Haefliger <juerg.haefliger@canonical.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

There may be some a race condition between timer function
bttv_irq_timeout and bttv_remove. The timer is setup in
probe and there is no timer_delete operation in remove
function. When it hit kfree btv, the function might still be
invoked, which will cause use after free bug.

This bug is found by static analysis, it may be false positive.

Fix it by adding del_timer_sync invoking to the remove function.

cpu0                cpu1
                  bttv_probe
                    ->timer_setup
                      ->bttv_set_dma
                        ->mod_timer;
bttv_remove
  ->kfree(btv);
                  ->bttv_irq_timeout
                    ->USE btv

Fixes: 162e6376 ("media: pci: Convert timers to use timer_setup()")
Signed-off-by: Zheng Wang <zyytlz.wz@163.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>

Stopping streaming on a camera pipeline at system suspend time, and
restarting it at system resume time, requires coordinated action between
the bridge driver and the camera sensor driver. This is handled by the
bridge driver calling the sensor's .s_stream() handler at system suspend
and resume time. There is thus no need for the sensor to independently
implement system sleep PM operations. Drop them.

The streaming field of the driver's private structure is now unused,
drop it as well.
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@xs4all.nl>

Stopping streaming on a camera pipeline at system suspend time, and
restarting it at system resume time, requires coordinated action between
the bridge driver and the camera sensor driver. This is handled by the
bridge driver calling the sensor's .s_stream() handler at system suspend
and resume time. There is thus no need for the sensor to independently
implement system sleep PM operations. Drop them.

The streaming field of the driver's private structure is now unused,
drop it as well.
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@xs4all.nl>

Stopping streaming on a camera pipeline at system suspend time, and
restarting it at system resume time, requires coordinated action between
the bridge driver and the camera sensor driver. This is handled by the
bridge driver calling the sensor's .s_stream() handler at system suspend
and resume time. There is thus no need for the sensor to independently
implement system sleep PM operations. Drop them.

The streaming field of the driver's private structure is now unused,
drop it as well.
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@xs4all.nl>