media: atomisp: Changes for libcamera support
Subsystems (4)
cover media: atomisp: Changes for libcamera support
Hi All,
Here is a series of cleanups (prep work) + some small(ish) changes to make the atomisp driver work with libcamera.
This has been tested with libcamera's simple pipelinehandler together with an ov2680 sensor.
Regards,
Hans
p.s.
Andy, I known I still need to do some small cleanups from my last pull-requests. I have this on my TODO list :)
Hans de Goede (9):
media: atomisp: Remove isp_subdev_propagate()
media: atomisp: Rename atomisp_set_crop_and_fmt()
media: atomisp: Remove custom VCM handling
media: atomisp: Remove ISP controls which get passed through to the
camera
media: atomisp: Stop setting sd->devnode for the ATOMISP_SUBDEV
v4l2-subdev
media: atomisp: Add DMABUF support
media: atomisp: Change ISP subdev name to "ATOM ISP"
media: atomisp: Make MC link from ISP to /dev/video# output node
immutable
media: atomisp: Implement link_setup op for ISP subdev MC entity
.../media/atomisp/include/linux/atomisp.h | 24 -- .../atomisp/include/linux/atomisp_platform.h | 5 +- .../staging/media/atomisp/pci/atomisp_cmd.c | 12 +- .../media/atomisp/pci/atomisp_internal.h | 4 - .../staging/media/atomisp/pci/atomisp_ioctl.c | 333 +----------------- .../media/atomisp/pci/atomisp_subdev.c | 95 +++-- .../staging/media/atomisp/pci/atomisp_v4l2.c | 18 +- 7 files changed, 72 insertions(+), 419 deletions(-)
Signature · 1 line
2.43.0
PATCH 1/9 media: atomisp: Remove isp_subdev_propagate()
isp_subdev_propagate() is a wrapper around atomisp_subdev_set_selection() which gets only used in a single place.
Scall atomisp_subdev_set_selection() directly in that single place instead.
--- .../media/atomisp/pci/atomisp_subdev.c | 38 ++++--------------- 1 file changed, 7 insertions(+), 31 deletions(-)
drivers/staging/media/atomisp/pci/atomisp_subdev.c +7 -31
@@ -287,35 +287,6 @@ static void isp_get_fmt_rect(struct v4l2_subdev *sd, } } -static void isp_subdev_propagate (struct v4l2_subdev *sd, -struct v4l2_subdev_state *sd_state, -u32 which,uint32_t pad,uint32_t target, -uint32_t flags) -{ -struct v4l2_mbus_framefmt *ffmt[ATOMISP_SUBDEV_PADS_NUM ]; -struct v4l2_rect *crop[ATOMISP_SUBDEV_PADS_NUM ], - *comp[ATOMISP_SUBDEV_PADS_NUM ]; - -if (flags &V4L2_SEL_FLAG_KEEP_CONFIG ) -return ; - -isp_get_fmt_rect (sd, sd_state, which, ffmt, crop, comp); - -switch (pad) { -case ATOMISP_SUBDEV_PAD_SINK : { -struct v4l2_rect r = {0 }; - -/* Only crop target supported on sink pad. */ - r.width = ffmt[pad]->width; - r.height = ffmt[pad]->height; - -atomisp_subdev_set_selection (sd, sd_state, which, pad, - target, flags, &r); -break ; - } - } -} -static int isp_subdev_get_selection (struct v4l2_subdev *sd,struct v4l2_subdev_state *sd_state,struct v4l2_subdev_selection *sel)
@@ -541,6 +512,7 @@ void atomisp_subdev_set_ffmt(struct v4l2_subdev *sd,case ATOMISP_SUBDEV_PAD_SINK : {const struct atomisp_in_fmt_conv *fc =atomisp_find_in_fmt_conv (ffmt->code); +struct v4l2_rect r = {};if (!fc) { fc = atomisp_in_fmt_conv;
@@ -551,8 +523,12 @@ void atomisp_subdev_set_ffmt(struct v4l2_subdev *sd, *__ffmt = *ffmt; -isp_subdev_propagate (sd, sd_state, which, pad, -V4L2_SEL_TGT_CROP ,0 ); +/* Propagate new ffmt to selection */ + r.width = ffmt->width; + r.height = ffmt->height; +/* Only crop target supported on sink pad. */ +atomisp_subdev_set_selection (sd, sd_state, which, pad, +V4L2_SEL_TGT_CROP ,0 , &r);if (which ==V4L2_SUBDEV_FORMAT_ACTIVE ) { atomisp_css_input_set_resolution(isp_sd,
Signature · 1 line
2.43.0
Re: [PATCH 1/9] media: atomisp: Remove isp_subdev_propagate()
Quoting Hans de Goede (2024-02-17 11:24:30)
Quoted 4 lines
isp_subdev_propagate()is a wrapper aroundatomisp_subdev_set_selection()which gets only used in a single place.Scall
atomisp_subdev_set_selection()directly in that single place instead.
s/Scall/Call/
Quoted 72 lines
Signed-off-by Hans de Goede <[email protected]>--- .../media/atomisp/pci/atomisp_subdev.c | 38 ++++--------------- 1 file changed, 7 insertions(+), 31 deletions(-)drivers/staging/media/atomisp/pci/atomisp_subdev.c +7
-31@@ -287,35 +287,6 @@ static void isp_get_fmt_rect(struct v4l2_subdev *sd, } } -static void isp_subdev_propagate (struct v4l2_subdev *sd, -struct v4l2_subdev_state *sd_state, -u32 which,uint32_t pad,uint32_t target, -uint32_t flags) -{ -struct v4l2_mbus_framefmt *ffmt[ATOMISP_SUBDEV_PADS_NUM ]; -struct v4l2_rect *crop[ATOMISP_SUBDEV_PADS_NUM ], - *comp[ATOMISP_SUBDEV_PADS_NUM ]; - -if (flags &V4L2_SEL_FLAG_KEEP_CONFIG ) -return ; - -isp_get_fmt_rect (sd, sd_state, which, ffmt, crop, comp); - -switch (pad) { -case ATOMISP_SUBDEV_PAD_SINK : { -struct v4l2_rect r = {0 }; - -/* Only crop target supported on sink pad. */ - r.width = ffmt[pad]->width; - r.height = ffmt[pad]->height; - -atomisp_subdev_set_selection (sd, sd_state, which, pad, - target, flags, &r); -break ; - } - } -} -static int isp_subdev_get_selection (struct v4l2_subdev *sd,struct v4l2_subdev_state *sd_state,struct v4l2_subdev_selection *sel)@@ -541,6 +512,7 @@ void atomisp_subdev_set_ffmt(struct v4l2_subdev *sd,case ATOMISP_SUBDEV_PAD_SINK : {const struct atomisp_in_fmt_conv *fc =atomisp_find_in_fmt_conv (ffmt->code); +struct v4l2_rect r = {};if (!fc) { fc = atomisp_in_fmt_conv;@@ -551,8 +523,12 @@ void atomisp_subdev_set_ffmt(struct v4l2_subdev *sd, *__ffmt = *ffmt; -isp_subdev_propagate (sd, sd_state, which, pad, -V4L2_SEL_TGT_CROP ,0 ); +/* Propagate new ffmt to selection */ + r.width = ffmt->width; + r.height = ffmt->height; +/* Only crop target supported on sink pad. */ +atomisp_subdev_set_selection (sd, sd_state, which, pad, +V4L2_SEL_TGT_CROP ,0 , &r);if (which ==V4L2_SUBDEV_FORMAT_ACTIVE ) { atomisp_css_input_set_resolution(isp_sd,Signature · 1 line
2.43.0
PATCH 2/9 media: atomisp: Rename atomisp_set_crop_and_fmt()
In reply to Hans de GoedeRename atomisp_set_crop_and_fmt() to atomisp_set_sensor_crop_and_fmt() to make clear that it operates on the sensor subdev.
--- drivers/staging/media/atomisp/pci/atomisp_cmd.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/staging/media/atomisp/pci/atomisp_cmd.c +6 -6
@@ -3721,9 +3721,9 @@ void atomisp_get_padding(struct atomisp_device *isp, u32 width, u32 height, *padding_h =max_t (u32, *padding_h, min_pad_h); } -static int atomisp_set_crop_and_fmt (struct atomisp_device *isp, -struct v4l2_mbus_framefmt *ffmt, -int which) +static int atomisp_set_sensor_crop_and_fmt (struct atomisp_device *isp, +struct v4l2_mbus_framefmt *ffmt, +int which) {struct atomisp_input_subdev *input = &isp->inputs[isp->asd.input_curr];struct v4l2_subdev_selection sel = {
@@ -3817,7 +3817,7 @@ int atomisp_try_fmt(struct atomisp_device *isp, struct v4l2_pix_format *f,dev_dbg (isp->dev,"try_mbus_fmt: try %ux%u\n" , ffmt.width, ffmt.height); - ret =atomisp_set_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY ); + ret =atomisp_set_sensor_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY );if (ret)return ret;
@@ -4263,7 +4263,7 @@ static int atomisp_set_fmt_to_snr(struct video_device *vdev, const struct v4l2_p/* Disable dvs if resolution can't be supported by sensor */ if (asd->params.video_dis_en && asd->run_mode->val ==ATOMISP_RUN_MODE_VIDEO ) { - ret =atomisp_set_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY ); + ret =atomisp_set_sensor_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY );if (ret)return ret;
@@ -4281,7 +4281,7 @@ static int atomisp_set_fmt_to_snr(struct video_device *vdev, const struct v4l2_p } } - ret =atomisp_set_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_ACTIVE ); + ret =atomisp_set_sensor_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_ACTIVE );if (ret)return ret;
Signature · 1 line
2.43.0
Re: [PATCH 2/9] media: atomisp: Rename atomisp_set_crop_and_fmt()
Quoting Hans de Goede (2024-02-17 11:24:31)
Quoted 13 lines
Rename
atomisp_set_crop_and_fmt()toatomisp_set_sensor_crop_and_fmt()to make clear that it operates on the sensor subdev.Signed-off-by Hans de Goede <[email protected]>--- drivers/staging/media/atomisp/pci/atomisp_cmd.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)drivers/staging/media/atomisp/pci/atomisp_cmd.c +0
-0
drivers/staging/media/atomisp/pci/atomisp_cmd.c void atomisp_get_padding(struct atomisp_device *isp, u32 width, u32 height, · line 3726
Quoted 9 lines
drivers/staging/media/atomisp/pci/atomisp_cmd.c +3
-3@@ -3721,9 +3721,9 @@ void atomisp_get_padding(struct atomisp_device *isp, u32 width, u32 height, *padding_h =max_t (u32, *padding_h, min_pad_h); } -static int atomisp_set_crop_and_fmt (struct atomisp_device *isp, -struct v4l2_mbus_framefmt *ffmt, -int which) +static int atomisp_set_sensor_crop_and_fmt (struct atomisp_device *isp, +struct v4l2_mbus_framefmt *ffmt, +int which)
I went to look this up and I only see atomisp_set_crop() so I guess I've missed some patches.
Does the atomisp use the usual v4l2-subdev drivers for sensors? or are they separate for now ?
I'm curious about how it's handling the binning in that function, but that's not specific to this patch which looks reasonable so far so:
Quoted 33 lines
drivers/staging/media/atomisp/pci/atomisp_cmd.c +3
-3{struct atomisp_input_subdev *input = &isp->inputs[isp->asd.input_curr];struct v4l2_subdev_selection sel = {@@ -3817,7 +3817,7 @@ int atomisp_try_fmt(struct atomisp_device *isp, struct v4l2_pix_format *f,dev_dbg (isp->dev,"try_mbus_fmt: try %ux%u\n" , ffmt.width, ffmt.height); - ret =atomisp_set_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY ); + ret =atomisp_set_sensor_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY );if (ret)return ret;@@ -4263,7 +4263,7 @@ static int atomisp_set_fmt_to_snr(struct video_device *vdev, const struct v4l2_p/* Disable dvs if resolution can't be supported by sensor */ if (asd->params.video_dis_en && asd->run_mode->val ==ATOMISP_RUN_MODE_VIDEO ) { - ret =atomisp_set_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY ); + ret =atomisp_set_sensor_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_TRY );if (ret)return ret;@@ -4281,7 +4281,7 @@ static int atomisp_set_fmt_to_snr(struct video_device *vdev, const struct v4l2_p } } - ret =atomisp_set_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_ACTIVE ); + ret =atomisp_set_sensor_crop_and_fmt (isp, &ffmt,V4L2_SUBDEV_FORMAT_ACTIVE );if (ret)return ret;Signature · 1 line
2.43.0
PATCH 3/9 media: atomisp: Remove custom VCM handling
In reply to Hans de GoedeRemove the custom VCM handling, instead the VCM should be controlled through its own v4l2-subdev and the new ipu-bridge.c code already supports instantiating an i2c_client for this and setting up the necessary endpoints in the fwnode graph.
--- .../media/atomisp/include/linux/atomisp.h | 3 - .../atomisp/include/linux/atomisp_platform.h | 5 +- .../media/atomisp/pci/atomisp_internal.h | 4 - .../staging/media/atomisp/pci/atomisp_ioctl.c | 86 +------------------ .../staging/media/atomisp/pci/atomisp_v4l2.c | 15 ---- 5 files changed, 3 insertions(+), 110 deletions(-)
drivers/staging/media/atomisp/include/linux/atomisp.h +0 -3
@@ -914,9 +914,6 @@ enum atomisp_burst_capture_options {/* VCM step time */ #define V4L2_CID_VCM_TIMING (V4L2_CID_CAMERA_LASTP1 + 12) -/* Query Focus Status */ -#define V4L2_CID_FOCUS_STATUS (V4L2_CID_CAMERA_LASTP1 + 14) -/* number of frames to skip at stream start */ #define V4L2_CID_G_SKIP_FRAMES (V4L2_CID_CAMERA_LASTP1 + 17)
drivers/staging/media/atomisp/include/linux/atomisp_platform.h +2 -3
@@ -111,9 +111,8 @@ enum atomisp_input_format {enum intel_v4l2_subdev_type {RAW_CAMERA =1 , -CAMERA_MOTOR =2 , -LED_FLASH =3 , -TEST_PATTERN =4 , +LED_FLASH =2 , +TEST_PATTERN =3 , };struct intel_v4l2_subdev_id {
drivers/staging/media/atomisp/pci/atomisp_internal.h +0 -4
@@ -134,9 +134,6 @@ struct atomisp_input_subdev {struct v4l2_rect active_rect;/* Sensor state for which == V4L2_SUBDEV_FORMAT_TRY calls */ struct v4l2_subdev_state *try_sd_state; - -struct v4l2_subdev *motor; -/* * To show this resource is used by * which stream, in ISP multiple stream mode
@@ -210,7 +207,6 @@ struct atomisp_device {unsigned int input_cnt;struct atomisp_input_subdev inputs[ATOM_ISP_MAX_INPUTS ];struct v4l2_subdev *flash; -struct v4l2_subdev *motor;struct atomisp_regs saved_regs;struct atomisp_css_env css_env;
drivers/staging/media/atomisp/pci/atomisp_ioctl.c +1 -85
149 lines not on this page · raw message
drivers/staging/media/atomisp/pci/atomisp_v4l2.c +0 -15
@@ -823,13 +823,6 @@ static int atomisp_subdev_probe(struct atomisp_device *isp) isp->sensor_lanes[mipi_port] = subdevs->lanes; isp->sensor_subdevs[subdevs->port] = subdevs->subdev;break ; -case CAMERA_MOTOR : -if (isp->motor) { -dev_warn (isp->dev,"too many atomisp motors\n" ); -continue ; - } - isp->motor = subdevs->subdev; -break ;case LED_FLASH :if (isp->flash) {dev_warn (isp->dev,"too many atomisp flash devices\n" );
@@ -1066,14 +1059,6 @@ int atomisp_register_device_nodes(struct atomisp_device *isp)atomisp_init_sensor (input); -/* -* HACK: Currently VCM belongs to primary sensor only, but correct -* approach must be to acquire from platform code which sensor -* owns it. -*/ -if (i ==ATOMISP_CAMERA_PORT_PRIMARY ) - input->motor = isp->motor; - err = media_create_pad_link(&input->camera->entity, 0, &isp->csi2_port[i].subdev.entity, CSI2_PAD_SINK,
Signature · 1 line
2.43.0
Re: [PATCH 3/9] media: atomisp: Remove custom VCM handling
Quoting Hans de Goede (2024-02-17 11:24:32)
Quoted 4 lines
Remove the custom VCM handling, instead the VCM should be controlled through its own v4l2-subdev and the new
ipu-bridge.ccode already supports instantiating ani2c_clientfor this and setting up the necessary endpoints in the fwnode graph.
Agreed, this sounds like a good move forwards to align with other ISPs and existing VCM driver frameworks. I must get back to my rework of the VCM drivers I started though!
Quoted 260 lines
Signed-off-by Hans de Goede <[email protected]>--- .../media/atomisp/include/linux/atomisp.h | 3 - .../atomisp/include/linux/atomisp_platform.h | 5 +- .../media/atomisp/pci/atomisp_internal.h | 4 - .../staging/media/atomisp/pci/atomisp_ioctl.c | 86 +------------------ .../staging/media/atomisp/pci/atomisp_v4l2.c | 15 ---- 5 files changed, 3 insertions(+), 110 deletions(-)drivers/staging/media/atomisp/include/linux/atomisp.h +0
-3@@ -914,9 +914,6 @@ enum atomisp_burst_capture_options {/* VCM step time */ #define V4L2_CID_VCM_TIMING (V4L2_CID_CAMERA_LASTP1 + 12) -/* Query Focus Status */ -#define V4L2_CID_FOCUS_STATUS (V4L2_CID_CAMERA_LASTP1 + 14) -/* number of frames to skip at stream start */ #define V4L2_CID_G_SKIP_FRAMES (V4L2_CID_CAMERA_LASTP1 + 17)drivers/staging/media/atomisp/include/linux/atomisp_platform.h +2
-3@@ -111,9 +111,8 @@ enum atomisp_input_format {enum intel_v4l2_subdev_type {RAW_CAMERA =1 , -CAMERA_MOTOR =2 , -LED_FLASH =3 , -TEST_PATTERN =4 , +LED_FLASH =2 , +TEST_PATTERN =3 , };struct intel_v4l2_subdev_id {drivers/staging/media/atomisp/pci/atomisp_internal.h +0
-4@@ -134,9 +134,6 @@ struct atomisp_input_subdev {struct v4l2_rect active_rect;/* Sensor state for which == V4L2_SUBDEV_FORMAT_TRY calls */ struct v4l2_subdev_state *try_sd_state; - -struct v4l2_subdev *motor; -/* * To show this resource is used by * which stream, in ISP multiple stream mode @@ -210,7 +207,6 @@ struct atomisp_device {unsigned int input_cnt;struct atomisp_input_subdev inputs[ATOM_ISP_MAX_INPUTS ];struct v4l2_subdev *flash; -struct v4l2_subdev *motor;struct atomisp_regs saved_regs;struct atomisp_css_env css_env;drivers/staging/media/atomisp/pci/atomisp_ioctl.c +1
-85@@ -557,7 +557,6 @@ static int atomisp_enum_input(struct file *file, void *fh,struct video_device *vdev =video_devdata (file);struct atomisp_device *isp =video_get_drvdata (vdev);int index = input->index; -struct v4l2_subdev *motor;if (index >= isp->input_cnt)return -EINVAL ;@@ -569,28 +568,6 @@ static int atomisp_enum_input(struct file *file, void *fh,strscpy (input->name, isp->inputs[index].camera->name,sizeof (input->name)); -/* -* HACK: append actuator's name to sensor's -* As currently userspace can't talk directly to subdev nodes, this -* ioctl is the only way to enum inputs + possible external actuators -* for 3A tuning purpose. -*/ -if (!IS_ISP2401 ) - motor = isp->inputs[index].motor; -else - motor = isp->motor; - -if (motor &&strlen (motor->name) >0 ) { -const int cur_len =strlen (input->name); -const int max_size =sizeof (input->name) - cur_len -1 ; - -if (max_size >1 ) { - input->name[cur_len] ='+' ; -strscpy (&input->name[cur_len +1 ], - motor->name, max_size); - } - } - input->type =V4L2_INPUT_TYPE_CAMERA ; input->index = index; input->reserved[0 ] = isp->inputs[index].type;@@ -629,7 +606,6 @@ static int atomisp_s_input(struct file *file, void *fh, unsigned int input)struct atomisp_video_pipe *pipe =atomisp_to_video_pipe (vdev);struct atomisp_sub_device *asd = pipe->asd;struct v4l2_subdev *camera =NULL ; -struct v4l2_subdev *motor;int ret; ret =atomisp_pipe_check (pipe, true);@@ -666,17 +642,6 @@ static int atomisp_s_input(struct file *file, void *fh, unsigned int input)return ret; } -if (!IS_ISP2401 ) { - motor = isp->inputs[input].motor; - }else { - motor = isp->motor; -if (motor) - ret =v4l2_subdev_call (motor, core, s_power,1 ); - } - -if (motor) - ret =v4l2_subdev_call (motor, core, init,1 ); - asd->input_curr = input;/* mark this camera is used by the current stream */ isp->inputs[input].asd = asd;@@ -1433,26 +1398,8 @@ static int atomisp_s_ctrl(struct file *file, void *fh,static int atomisp_queryctl (struct file *file,void *fh,struct v4l2_queryctrl *qc) { -int i, ret = -EINVAL ;struct video_device *vdev =video_devdata (file); -struct atomisp_sub_device *asd =atomisp_to_video_pipe (vdev)->asd; -struct atomisp_device *isp =video_get_drvdata (vdev); - -switch (qc->id) { -case V4L2_CID_FOCUS_ABSOLUTE : -case V4L2_CID_FOCUS_RELATIVE : -case V4L2_CID_FOCUS_STATUS : -if (!IS_ISP2401 ) { -return v4l2_queryctrl (isp->inputs[asd->input_curr].camera-> - ctrl_handler, qc); - } -/* ISP2401 */ -if (isp->motor) -return v4l2_queryctrl (isp->motor->ctrl_handler, qc); -else -return v4l2_queryctrl (isp->inputs[asd->input_curr]. - camera->ctrl_handler, qc); - } +int i, ret = -EINVAL ;if (qc->id &V4L2_CTRL_FLAG_NEXT_CTRL )return ret;@@ -1478,16 +1425,10 @@ static int atomisp_camera_g_ext_ctrls(struct file *file, void *fh,struct video_device *vdev =video_devdata (file);struct atomisp_sub_device *asd =atomisp_to_video_pipe (vdev)->asd;struct atomisp_device *isp =video_get_drvdata (vdev); -struct v4l2_subdev *motor;struct v4l2_control ctrl;int i;int ret =0 ; -if (!IS_ISP2401 ) - motor = isp->inputs[asd->input_curr].motor; -else - motor = isp->motor; -for (i =0 ; i < c->count; i++) { ctrl.id = c->controls[i].id; ctrl.value = c->controls[i].value;@@ -1509,13 +1450,6 @@ static int atomisp_camera_g_ext_ctrls(struct file *file, void *fh,v4l2_g_ctrl (isp->inputs[asd->input_curr].camera-> ctrl_handler, &ctrl);break ; -case V4L2_CID_FOCUS_ABSOLUTE : -case V4L2_CID_FOCUS_RELATIVE : -case V4L2_CID_FOCUS_STATUS : -case V4L2_CID_FOCUS_AUTO : -if (motor) - ret =v4l2_g_ctrl (motor->ctrl_handler, &ctrl); -break ;case V4L2_CID_FLASH_STATUS :case V4L2_CID_FLASH_INTENSITY :case V4L2_CID_FLASH_TORCH_INTENSITY :@@ -1584,16 +1518,10 @@ static int atomisp_camera_s_ext_ctrls(struct file *file, void *fh,struct video_device *vdev =video_devdata (file);struct atomisp_sub_device *asd =atomisp_to_video_pipe (vdev)->asd;struct atomisp_device *isp =video_get_drvdata (vdev); -struct v4l2_subdev *motor;struct v4l2_control ctrl;int i;int ret =0 ; -if (!IS_ISP2401 ) - motor = isp->inputs[asd->input_curr].motor; -else - motor = isp->motor; -for (i =0 ; i < c->count; i++) {struct v4l2_ctrl *ctr;@@ -1616,18 +1544,6 @@ static int atomisp_camera_s_ext_ctrls(struct file *file, void *fh, isp->inputs[asd->input_curr].camera-> ctrl_handler, &ctrl);break ; -case V4L2_CID_FOCUS_ABSOLUTE : -case V4L2_CID_FOCUS_RELATIVE : -case V4L2_CID_FOCUS_STATUS : -case V4L2_CID_FOCUS_AUTO : -if (motor) - ret =v4l2_s_ctrl (NULL , motor->ctrl_handler, - &ctrl); -else - ret =v4l2_s_ctrl (NULL , - isp->inputs[asd->input_curr]. - camera->ctrl_handler, &ctrl); -break ;case V4L2_CID_FLASH_STATUS :case V4L2_CID_FLASH_INTENSITY :case V4L2_CID_FLASH_TORCH_INTENSITY :drivers/staging/media/atomisp/pci/atomisp_v4l2.c +0
-15@@ -823,13 +823,6 @@ static int atomisp_subdev_probe(struct atomisp_device *isp) isp->sensor_lanes[mipi_port] = subdevs->lanes; isp->sensor_subdevs[subdevs->port] = subdevs->subdev;break ; -case CAMERA_MOTOR : -if (isp->motor) { -dev_warn (isp->dev,"too many atomisp motors\n" ); -continue ; - } - isp->motor = subdevs->subdev; -break ;case LED_FLASH :if (isp->flash) {dev_warn (isp->dev,"too many atomisp flash devices\n" );@@ -1066,14 +1059,6 @@ int atomisp_register_device_nodes(struct atomisp_device *isp)atomisp_init_sensor (input); -/* -* HACK: Currently VCM belongs to primary sensor only, but correct -* approach must be to acquire from platform code which sensor -* owns it. -*/ -if (i ==ATOMISP_CAMERA_PORT_PRIMARY ) - input->motor = isp->motor; - err = media_create_pad_link(&input->camera->entity, 0, &isp->csi2_port[i].subdev.entity, CSI2_PAD_SINK,Signature · 1 line
2.43.0
PATCH 4/9 media: atomisp: Remove ISP controls which get passed through to the camera
In reply to Hans de GoedeDrop all ISP controls and ioctls which just get passed through to the camera subdev. Instead these calls should be done directly at the sensor subdev.
--- .../media/atomisp/include/linux/atomisp.h | 21 -- .../staging/media/atomisp/pci/atomisp_ioctl.c | 248 ------------------ 2 files changed, 269 deletions(-)
drivers/staging/media/atomisp/include/linux/atomisp.h +0 -21
@@ -837,9 +837,6 @@ enum atomisp_burst_capture_options {#define ATOMISP_IOC_S_PARAMETERS \ _IOW('v', BASE_VIDIOC_PRIVATE + 32, struct atomisp_parameters) -#define ATOMISP_IOC_EXT_ISP_CTRL \ - _IOWR('v', BASE_VIDIOC_PRIVATE + 35, struct atomisp_ext_isp_ctrl) -#define ATOMISP_IOC_EXP_ID_UNLOCK \ _IOW('v', BASE_VIDIOC_PRIVATE + 36, int)
@@ -909,19 +906,6 @@ enum atomisp_burst_capture_options {/* Set the flash mode (see enum atomisp_flash_mode) */ #define V4L2_CID_FLASH_MODE (V4L2_CID_CAMERA_LASTP1 + 10) -/* VCM slew control */ -#define V4L2_CID_VCM_SLEW (V4L2_CID_CAMERA_LASTP1 + 11) -/* VCM step time */ -#define V4L2_CID_VCM_TIMING (V4L2_CID_CAMERA_LASTP1 + 12) - -/* number of frames to skip at stream start */ -#define V4L2_CID_G_SKIP_FRAMES (V4L2_CID_CAMERA_LASTP1 + 17) - -/* Query sensor's 2A status */ -#define V4L2_CID_2A_STATUS (V4L2_CID_CAMERA_LASTP1 + 18) -#define V4L2_2A_STATUS_AE_READY BIT(0) -#define V4L2_2A_STATUS_AWB_READY BIT(1) -#define V4L2_CID_RUN_MODE (V4L2_CID_CAMERA_LASTP1 + 20)#define ATOMISP_RUN_MODE_VIDEO 1#define ATOMISP_RUN_MODE_STILL_CAPTURE 2
@@ -952,11 +936,6 @@ enum atomisp_burst_capture_options {/* Disable digital zoom */ #define V4L2_CID_DISABLE_DZ (V4L2_CID_CAMERA_LASTP1 + 32) -#define V4L2_CID_TEST_PATTERN_COLOR_R (V4L2_CID_CAMERA_LASTP1 + 33) -#define V4L2_CID_TEST_PATTERN_COLOR_GR (V4L2_CID_CAMERA_LASTP1 + 34) -#define V4L2_CID_TEST_PATTERN_COLOR_GB (V4L2_CID_CAMERA_LASTP1 + 35) -#define V4L2_CID_TEST_PATTERN_COLOR_B (V4L2_CID_CAMERA_LASTP1 + 36) -#define V4L2_CID_ATOMISP_SELECT_ISP_VERSION (V4L2_CID_CAMERA_LASTP1 + 38)#define V4L2_BUF_FLAG_BUFFER_INVALID 0x0400
drivers/staging/media/atomisp/pci/atomisp_ioctl.c +0 -248
332 lines not on this page · raw message
Signature · 1 line
2.43.0
Re: [PATCH 4/9] media: atomisp: Remove ISP controls which get passed through to the camera
Quoting Hans de Goede (2024-02-17 11:24:33)
Quoted 295 lines
Drop all ISP controls and ioctls which just get passed through to the camera subdev. Instead these calls should be done directly at the sensor subdev.
Signed-off-by Hans de Goede <[email protected]>--- .../media/atomisp/include/linux/atomisp.h | 21 -- .../staging/media/atomisp/pci/atomisp_ioctl.c | 248 ------------------ 2 files changed, 269 deletions(-)drivers/staging/media/atomisp/include/linux/atomisp.h +0
-21@@ -837,9 +837,6 @@ enum atomisp_burst_capture_options {#define ATOMISP_IOC_S_PARAMETERS \ _IOW('v', BASE_VIDIOC_PRIVATE + 32, struct atomisp_parameters) -#define ATOMISP_IOC_EXT_ISP_CTRL \ - _IOWR('v', BASE_VIDIOC_PRIVATE + 35, struct atomisp_ext_isp_ctrl) -#define ATOMISP_IOC_EXP_ID_UNLOCK \ _IOW('v', BASE_VIDIOC_PRIVATE + 36, int)@@ -909,19 +906,6 @@ enum atomisp_burst_capture_options {/* Set the flash mode (see enum atomisp_flash_mode) */ #define V4L2_CID_FLASH_MODE (V4L2_CID_CAMERA_LASTP1 + 10) -/* VCM slew control */ -#define V4L2_CID_VCM_SLEW (V4L2_CID_CAMERA_LASTP1 + 11) -/* VCM step time */ -#define V4L2_CID_VCM_TIMING (V4L2_CID_CAMERA_LASTP1 + 12) - -/* number of frames to skip at stream start */ -#define V4L2_CID_G_SKIP_FRAMES (V4L2_CID_CAMERA_LASTP1 + 17) - -/* Query sensor's 2A status */ -#define V4L2_CID_2A_STATUS (V4L2_CID_CAMERA_LASTP1 + 18) -#define V4L2_2A_STATUS_AE_READY BIT(0) -#define V4L2_2A_STATUS_AWB_READY BIT(1) -#define V4L2_CID_RUN_MODE (V4L2_CID_CAMERA_LASTP1 + 20)#define ATOMISP_RUN_MODE_VIDEO 1#define ATOMISP_RUN_MODE_STILL_CAPTURE 2@@ -952,11 +936,6 @@ enum atomisp_burst_capture_options {/* Disable digital zoom */ #define V4L2_CID_DISABLE_DZ (V4L2_CID_CAMERA_LASTP1 + 32) -#define V4L2_CID_TEST_PATTERN_COLOR_R (V4L2_CID_CAMERA_LASTP1 + 33) -#define V4L2_CID_TEST_PATTERN_COLOR_GR (V4L2_CID_CAMERA_LASTP1 + 34) -#define V4L2_CID_TEST_PATTERN_COLOR_GB (V4L2_CID_CAMERA_LASTP1 + 35) -#define V4L2_CID_TEST_PATTERN_COLOR_B (V4L2_CID_CAMERA_LASTP1 + 36) -#define V4L2_CID_ATOMISP_SELECT_ISP_VERSION (V4L2_CID_CAMERA_LASTP1 + 38)#define V4L2_BUF_FLAG_BUFFER_INVALID 0x0400drivers/staging/media/atomisp/pci/atomisp_ioctl.c +0
-189@@ -82,15 +82,6 @@ static struct v4l2_queryctrl ci_v4l2_controls[] = { .step = 1, .default_value = 0x00, }, - { - .id = V4L2_CID_POWER_LINE_FREQUENCY, - .type = V4L2_CTRL_TYPE_MENU, - .name = "Light frequency filter", - .minimum = 1, - .maximum = 2, - .step = 1, - .default_value = 1, - }, { .id = V4L2_CID_COLORFX, .type = V4L2_CTRL_TYPE_INTEGER,@@ -100,15 +91,6 @@ static struct v4l2_queryctrl ci_v4l2_controls[] = { .step = 1, .default_value = 0, }, - { - .id = V4L2_CID_COLORFX_CBCR, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Image Color Effect CbCr", - .minimum = 0, - .maximum = 0xffff, - .step = 1, - .default_value = 0, - }, { .id = V4L2_CID_ATOMISP_BAD_PIXEL_DETECTION, .type = V4L2_CTRL_TYPE_INTEGER,@@ -172,142 +154,6 @@ static struct v4l2_queryctrl ci_v4l2_controls[] = { .step = 1, .default_value = 1, }, - { - .id = V4L2_CID_2A_STATUS, - .type = V4L2_CTRL_TYPE_BITMASK, - .name = "AE and AWB status", - .minimum = 0, - .maximum = V4L2_2A_STATUS_AE_READY | V4L2_2A_STATUS_AWB_READY, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_EXPOSURE, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "exposure", - .minimum = -4, - .maximum = 4, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_EXPOSURE_ZONE_NUM, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "one-time exposure zone number", - .minimum = 0x0, - .maximum = 0xffff, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_EXPOSURE_AUTO_PRIORITY, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Exposure auto priority", - .minimum = V4L2_EXPOSURE_AUTO, - .maximum = V4L2_EXPOSURE_APERTURE_PRIORITY, - .step = 1, - .default_value = V4L2_EXPOSURE_AUTO, - }, - { - .id = V4L2_CID_SCENE_MODE, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "scene mode", - .minimum = 0, - .maximum = 13, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_ISO_SENSITIVITY, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "iso", - .minimum = -4, - .maximum = 4, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_ISO_SENSITIVITY_AUTO, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "iso mode", - .minimum = V4L2_ISO_SENSITIVITY_MANUAL, - .maximum = V4L2_ISO_SENSITIVITY_AUTO, - .step = 1, - .default_value = V4L2_ISO_SENSITIVITY_AUTO, - }, - { - .id = V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "white balance", - .minimum = 0, - .maximum = 9, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_EXPOSURE_METERING, - .type = V4L2_CTRL_TYPE_MENU, - .name = "metering", - .minimum = 0, - .maximum = 3, - .step = 1, - .default_value = 1, - }, - { - .id = V4L2_CID_3A_LOCK, - .type = V4L2_CTRL_TYPE_BITMASK, - .name = "3a lock", - .minimum = 0, - .maximum =V4L2_LOCK_EXPOSURE |V4L2_LOCK_WHITE_BALANCE - | V4L2_LOCK_FOCUS, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_TEST_PATTERN, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Test Pattern", - .minimum = 0, - .maximum = 0xffff, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_TEST_PATTERN_COLOR_R, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Test Pattern Solid Color R", - .minimum = INT_MIN, - .maximum = INT_MAX, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_TEST_PATTERN_COLOR_GR, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Test Pattern Solid Color GR", - .minimum = INT_MIN, - .maximum = INT_MAX, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_TEST_PATTERN_COLOR_GB, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Test Pattern Solid Color GB", - .minimum = INT_MIN, - .maximum = INT_MAX, - .step = 1, - .default_value = 0, - }, - { - .id = V4L2_CID_TEST_PATTERN_COLOR_B, - .type = V4L2_CTRL_TYPE_INTEGER, - .name = "Test Pattern Solid Color B", - .minimum = INT_MIN, - .maximum = INT_MAX, - .step = 1, - .default_value = 0, - }, };static const u32 ctrls_num =ARRAY_SIZE (ci_v4l2_controls);@@ -1248,7 +1094,6 @@ static int atomisp_g_ctrl(struct file *file, void *fh, {struct video_device *vdev =video_devdata (file);struct atomisp_sub_device *asd =atomisp_to_video_pipe (vdev)->asd; -struct atomisp_device *isp =video_get_drvdata (vdev);int i, ret = -EINVAL ;for (i =0 ; i < ctrls_num; i++) {@@ -1262,27 +1107,6 @@ static int atomisp_g_ctrl(struct file *file, void *fh,return ret;switch (control->id) { -case V4L2_CID_IRIS_ABSOLUTE : -case V4L2_CID_EXPOSURE_ABSOLUTE : -case V4L2_CID_2A_STATUS : -case V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE : -case V4L2_CID_EXPOSURE : -case V4L2_CID_EXPOSURE_AUTO : -case V4L2_CID_SCENE_MODE : -case V4L2_CID_ISO_SENSITIVITY : -case V4L2_CID_ISO_SENSITIVITY_AUTO : -case V4L2_CID_CONTRAST : -case V4L2_CID_SATURATION : -case V4L2_CID_SHARPNESS : -case V4L2_CID_3A_LOCK : -case V4L2_CID_EXPOSURE_ZONE_NUM : -case V4L2_CID_TEST_PATTERN : -case V4L2_CID_TEST_PATTERN_COLOR_R : -case V4L2_CID_TEST_PATTERN_COLOR_GR : -case V4L2_CID_TEST_PATTERN_COLOR_GB : -case V4L2_CID_TEST_PATTERN_COLOR_B : -return v4l2_g_ctrl (isp->inputs[asd->input_curr].camera-> - ctrl_handler, control);case V4L2_CID_COLORFX : ret =atomisp_color_effect (asd,0 , &control->value);break ;@@ -1322,7 +1146,6 @@ static int atomisp_s_ctrl(struct file *file, void *fh, {struct video_device *vdev =video_devdata (file);struct atomisp_sub_device *asd =atomisp_to_video_pipe (vdev)->asd; -struct atomisp_device *isp =video_get_drvdata (vdev);int i, ret = -EINVAL ;for (i =0 ; i < ctrls_num; i++) {@@ -1336,28 +1159,6 @@ static int atomisp_s_ctrl(struct file *file, void *fh,return ret;switch (control->id) { -case V4L2_CID_AUTO_N_PRESET_WHITE_BALANCE : -case V4L2_CID_EXPOSURE : -case V4L2_CID_EXPOSURE_AUTO : -case V4L2_CID_EXPOSURE_AUTO_PRIORITY : -case V4L2_CID_SCENE_MODE : -case V4L2_CID_ISO_SENSITIVITY : -case V4L2_CID_ISO_SENSITIVITY_AUTO : -case V4L2_CID_POWER_LINE_FREQUENCY : -case V4L2_CID_EXPOSURE_METERING : -case V4L2_CID_CONTRAST : -case V4L2_CID_SATURATION : -case V4L2_CID_SHARPNESS :
drivers/staging/media/atomisp/pci/atomisp_ioctl.c static int atomisp_s_ctrl(struct file *file, void *fh, · line 1360
Quoted 10 lines
drivers/staging/media/atomisp/pci/atomisp_ioctl.c +0
-10-case V4L2_CID_3A_LOCK : -case V4L2_CID_COLORFX_CBCR : -case V4L2_CID_TEST_PATTERN : -case V4L2_CID_TEST_PATTERN_COLOR_R : -case V4L2_CID_TEST_PATTERN_COLOR_GR : -case V4L2_CID_TEST_PATTERN_COLOR_GB : -case V4L2_CID_TEST_PATTERN_COLOR_B : -return v4l2_s_ctrl (NULL , - isp->inputs[asd->input_curr].camera-> - ctrl_handler, control);
Looks like a reasonable clean up from that.
Quoted 104 lines
drivers/staging/media/atomisp/pci/atomisp_ioctl.c +0
-49case V4L2_CID_COLORFX : ret =atomisp_color_effect (asd,1 , &control->value);break ;@@ -1398,7 +1199,6 @@ static int atomisp_s_ctrl(struct file *file, void *fh,static int atomisp_queryctl (struct file *file,void *fh,struct v4l2_queryctrl *qc) { -struct video_device *vdev =video_devdata (file);int i, ret = -EINVAL ;if (qc->id &V4L2_CTRL_FLAG_NEXT_CTRL )@@ -1433,23 +1233,6 @@ static int atomisp_camera_g_ext_ctrls(struct file *file, void *fh, ctrl.id = c->controls[i].id; ctrl.value = c->controls[i].value;switch (ctrl.id) { -case V4L2_CID_EXPOSURE_ABSOLUTE : -case V4L2_CID_EXPOSURE_AUTO : -case V4L2_CID_IRIS_ABSOLUTE : -case V4L2_CID_3A_LOCK : -case V4L2_CID_TEST_PATTERN : -case V4L2_CID_TEST_PATTERN_COLOR_R : -case V4L2_CID_TEST_PATTERN_COLOR_GR : -case V4L2_CID_TEST_PATTERN_COLOR_GB : -case V4L2_CID_TEST_PATTERN_COLOR_B : -/* -* Exposure related control will be handled by sensor -* driver -*/ - ret = -v4l2_g_ctrl (isp->inputs[asd->input_curr].camera-> - ctrl_handler, &ctrl); -break ;case V4L2_CID_FLASH_STATUS :case V4L2_CID_FLASH_INTENSITY :case V4L2_CID_FLASH_TORCH_INTENSITY :@@ -1466,11 +1249,6 @@ static int atomisp_camera_g_ext_ctrls(struct file *file, void *fh,case V4L2_CID_ZOOM_ABSOLUTE : ret =atomisp_digital_zoom (asd,0 , &ctrl.value);break ; -case V4L2_CID_G_SKIP_FRAMES : - ret =v4l2_subdev_call ( - isp->inputs[asd->input_curr].camera, - sensor, g_skip_frames, (u32 *)&ctrl.value); -break ;default : ret = -EINVAL ; }@@ -1528,22 +1306,6 @@ static int atomisp_camera_s_ext_ctrls(struct file *file, void *fh, ctrl.id = c->controls[i].id; ctrl.value = c->controls[i].value;switch (ctrl.id) { -case V4L2_CID_EXPOSURE_ABSOLUTE : -case V4L2_CID_EXPOSURE_AUTO : -case V4L2_CID_EXPOSURE_METERING : -case V4L2_CID_IRIS_ABSOLUTE : -case V4L2_CID_VCM_TIMING : -case V4L2_CID_VCM_SLEW : -case V4L2_CID_3A_LOCK : -case V4L2_CID_TEST_PATTERN : -case V4L2_CID_TEST_PATTERN_COLOR_R : -case V4L2_CID_TEST_PATTERN_COLOR_GR : -case V4L2_CID_TEST_PATTERN_COLOR_GB : -case V4L2_CID_TEST_PATTERN_COLOR_B : - ret =v4l2_s_ctrl (NULL , - isp->inputs[asd->input_curr].camera-> - ctrl_handler, &ctrl); -break ;case V4L2_CID_FLASH_STATUS :case V4L2_CID_FLASH_INTENSITY :case V4L2_CID_FLASH_TORCH_INTENSITY :@@ -1692,7 +1454,6 @@ static long atomisp_vidioc_default(struct file *file, void *fh, bool valid_prio, unsigned int cmd, void *arg) { struct video_device *vdev = video_devdata(file); -struct atomisp_device *isp =video_get_drvdata (vdev);struct atomisp_sub_device *asd =atomisp_to_video_pipe (vdev)->asd;int err;@@ -1839,11 +1600,6 @@ static long atomisp_vidioc_default(struct file *file, void *fh, err =atomisp_fixed_pattern_table (asd, arg);break ; -case ATOMISP_IOC_S_EXPOSURE : - err =v4l2_subdev_call (isp->inputs[asd->input_curr].camera, - core, ioctl, cmd, arg); -break ; -case ATOMISP_IOC_S_ISP_SHD_TAB : err =atomisp_set_shading_table (asd, arg);break ;@@ -1860,10 +1616,6 @@ static long atomisp_vidioc_default(struct file *file, void *fh, err =atomisp_set_parameters (vdev, arg);break ; -case ATOMISP_IOC_EXT_ISP_CTRL : - err =v4l2_subdev_call (isp->inputs[asd->input_curr].camera, - core, ioctl, cmd, arg); -break ;case ATOMISP_IOC_EXP_ID_UNLOCK : err =atomisp_exp_id_unlock (asd, arg);break ;Signature · 1 line
2.43.0
PATCH 5/9 media: atomisp: Stop setting sd->devnode for the ATOMISP_SUBDEV v4l2-subdev
In reply to Hans de GoedeStop setting sd->devnode for the ATOMISP_SUBDEV v4l2-subdev, so that a proper /dev/v4l-subdev# gets created for it.
--- drivers/staging/media/atomisp/pci/atomisp_subdev.c | 1 - 1 file changed, 1 deletion(-)
drivers/staging/media/atomisp/pci/atomisp_subdev.c +0 -1
@@ -802,7 +802,6 @@ static int isp_subdev_init_entities(struct atomisp_sub_device *asd)sprintf (sd->name,"ATOMISP_SUBDEV" );v4l2_set_subdevdata (sd, asd); sd->flags |=V4L2_SUBDEV_FL_HAS_EVENTS |V4L2_SUBDEV_FL_HAS_DEVNODE ; - sd->devnode = &asd->video_out.vdev; pads[ATOMISP_SUBDEV_PAD_SINK ].flags =MEDIA_PAD_FL_SINK ; pads[ATOMISP_SUBDEV_PAD_SOURCE ].flags =MEDIA_PAD_FL_SOURCE ;
Signature · 1 line
2.43.0
PATCH 6/9 media: atomisp: Add DMABUF support
In reply to Hans de GoedeAdd DMABUF support and while at it drop userptr support.
Now that atomisp has been ported to videobuf2 this is trivial.
--- drivers/staging/media/atomisp/pci/atomisp_ioctl.c | 1 + drivers/staging/media/atomisp/pci/atomisp_subdev.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-)
drivers/staging/media/atomisp/pci/atomisp_ioctl.c +1 -0
@@ -1666,6 +1666,7 @@ const struct v4l2_ioctl_ops atomisp_ioctl_ops = { .vidioc_querybuf = vb2_ioctl_querybuf, .vidioc_qbuf = atomisp_qbuf_wrapper, .vidioc_dqbuf = atomisp_dqbuf_wrapper, + .vidioc_expbuf = vb2_ioctl_expbuf, .vidioc_streamon = vb2_ioctl_streamon, .vidioc_streamoff = vb2_ioctl_streamoff, .vidioc_default = atomisp_vidioc_default,
drivers/staging/media/atomisp/pci/atomisp_subdev.c +1 -1
@@ -765,7 +765,7 @@ static int atomisp_init_subdev_pipe(struct atomisp_sub_device *asd,/* Init videobuf2 queue structure */ pipe->vb_queue.type =V4L2_BUF_TYPE_VIDEO_CAPTURE ; - pipe->vb_queue.io_modes =VB2_MMAP |VB2_USERPTR ; + pipe->vb_queue.io_modes =VB2_MMAP |VB2_DMABUF ; pipe->vb_queue.buf_struct_size =sizeof (struct ia_css_frame ); pipe->vb_queue.ops = &atomisp_vb2_ops; pipe->vb_queue.mem_ops = &vb2_vmalloc_memops;
Signature · 1 line
2.43.0
Re: [PATCH 6/9] media: atomisp: Add DMABUF support
Quoting Hans de Goede (2024-02-17 11:24:35)
Quoted 3 lines
Add DMABUF support and while at it drop userptr support.
Now that atomisp has been ported to videobuf2 this is trivial.
Seems 'too' easy, Nice.
Quoted 35 lines
Signed-off-by Hans de Goede <[email protected]>--- drivers/staging/media/atomisp/pci/atomisp_ioctl.c | 1 + drivers/staging/media/atomisp/pci/atomisp_subdev.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-)drivers/staging/media/atomisp/pci/atomisp_ioctl.c +1
-0@@ -1666,6 +1666,7 @@ const struct v4l2_ioctl_ops atomisp_ioctl_ops = { .vidioc_querybuf = vb2_ioctl_querybuf, .vidioc_qbuf = atomisp_qbuf_wrapper, .vidioc_dqbuf = atomisp_dqbuf_wrapper, + .vidioc_expbuf = vb2_ioctl_expbuf, .vidioc_streamon = vb2_ioctl_streamon, .vidioc_streamoff = vb2_ioctl_streamoff, .vidioc_default = atomisp_vidioc_default,drivers/staging/media/atomisp/pci/atomisp_subdev.c +1
-1@@ -765,7 +765,7 @@ static int atomisp_init_subdev_pipe(struct atomisp_sub_device *asd,/* Init videobuf2 queue structure */ pipe->vb_queue.type =V4L2_BUF_TYPE_VIDEO_CAPTURE ; - pipe->vb_queue.io_modes =VB2_MMAP |VB2_USERPTR ; + pipe->vb_queue.io_modes =VB2_MMAP |VB2_DMABUF ; pipe->vb_queue.buf_struct_size =sizeof (struct ia_css_frame ); pipe->vb_queue.ops = &atomisp_vb2_ops; pipe->vb_queue.mem_ops = &vb2_vmalloc_memops;Signature · 1 line
2.43.0
PATCH 7/9 media: atomisp: Change ISP subdev name to "ATOM ISP"
In reply to Hans de GoedeChange the generic "ATOMISP_SUBDEV" name to "ATOM ISP" to make clear that this is the subdev for the ISP itself.
--- drivers/staging/media/atomisp/pci/atomisp_subdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/media/atomisp/pci/atomisp_subdev.c +1 -1
@@ -799,7 +799,7 @@ static int isp_subdev_init_entities(struct atomisp_sub_device *asd)int ret;v4l2_subdev_init (sd, &isp_subdev_v4l2_ops); -sprintf (sd->name,"ATOMISP_SUBDEV" ); +sprintf (sd->name,"ATOM ISP" );v4l2_set_subdevdata (sd, asd); sd->flags |=V4L2_SUBDEV_FL_HAS_EVENTS |V4L2_SUBDEV_FL_HAS_DEVNODE ;
Signature · 1 line
2.43.0
Re: [PATCH 7/9] media: atomisp: Change ISP subdev name to "ATOM ISP"
Quoting Hans de Goede (2024-02-17 11:24:36)
Quoted 13 lines
Change the generic "
ATOMISP_SUBDEV" name to "ATOM ISP" to make clear that this is the subdev for the ISP itself.Signed-off-by Hans de Goede <[email protected]>--- drivers/staging/media/atomisp/pci/atomisp_subdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/staging/media/atomisp/pci/atomisp_subdev.c +0
-0
drivers/staging/media/atomisp/pci/atomisp_subdev.c static int isp_subdev_init_entities(struct atomisp_sub_device *asd) · line 802
Quoted 5 lines
drivers/staging/media/atomisp/pci/atomisp_subdev.c +1
-1@@ -799,7 +799,7 @@ static int isp_subdev_init_entities(struct atomisp_sub_device *asd)int ret;v4l2_subdev_init (sd, &isp_subdev_v4l2_ops); -sprintf (sd->name,"ATOMISP_SUBDEV" ); +sprintf (sd->name,"ATOM ISP" );
Pure bikeshedding, but I'd probably lower the shouting to just
"Atom ISP"
Either way saying it's a subdev on a subdev isn't much of a value add so
Quoted 6 lines
drivers/staging/media/atomisp/pci/atomisp_subdev.c +0
-0v4l2_set_subdevdata (sd, asd); sd->flags |=V4L2_SUBDEV_FL_HAS_EVENTS |V4L2_SUBDEV_FL_HAS_DEVNODE ;Signature · 1 line
2.43.0
Re: [PATCH 7/9] media: atomisp: Change ISP subdev name to "ATOM ISP"
Quoted 1 line · Kieran Bingham
Quoting Hans de Goede (2024-02-17 11:24:36)
...
Quoted 2 lines
Pure bikeshedding, but I'd probably lower the shouting to just
"Atom ISP"
You beat me to it, +1.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Re: [PATCH 7/9] media: atomisp: Change ISP subdev name to "ATOM ISP"
In reply to Kieran BinghamHi,
Quoted 16 lines · Kieran Bingham
Quoting Hans de Goede (2024-02-17 11:24:36)
Quoted 15 lines
Change the generic "
ATOMISP_SUBDEV" name to "ATOM ISP" to make clear that this is the subdev for the ISP itself.Signed-off-by Hans de Goede <[email protected]>--- drivers/staging/media/atomisp/pci/atomisp_subdev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/staging/media/atomisp/pci/atomisp_subdev.c +0
-0@@ -799,7 +799,7 @@ static int isp_subdev_init_entities(struct atomisp_sub_device *asd)int ret;
Quoted 6 lines
drivers/staging/media/atomisp/pci/atomisp_subdev.c static int isp_subdev_init_entities(struct atomisp_sub_device *asd) · line 802
Quoted 3 lines
drivers/staging/media/atomisp/pci/atomisp_subdev.c +1
-1v4l2_subdev_init (sd, &isp_subdev_v4l2_ops); -sprintf (sd->name,"ATOMISP_SUBDEV" ); +sprintf (sd->name,"ATOM ISP" );Pure bikeshedding, but I'd probably lower the shouting to just
"Atom ISP"
Works for me, I've fixed this in my personal tree.
Quoted 3 lines
Either way saying it's a subdev on a subdev isn't much of a value add so
Reviewed-by Kieran Bingham <[email protected]>
Thank you for this and all the other reviews.
Regards,
Hans
PATCH 8/9 media: atomisp: Make MC link from ISP to /dev/video# output node immutable
In reply to Hans de GoedeThe link from the ISP's source pad to the /dev/video# output v4l2-dev is always enabled and immutable, mark it as such.
--- drivers/staging/media/atomisp/pci/atomisp_v4l2.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/staging/media/atomisp/pci/atomisp_v4l2.c +2 -1
@@ -1090,7 +1090,8 @@ int atomisp_register_device_nodes(struct atomisp_device *isp)return err; err =media_create_pad_link (&isp->asd.subdev.entity,ATOMISP_SUBDEV_PAD_SOURCE , - &isp->asd.video_out.vdev.entity,0 ,0 ); + &isp->asd.video_out.vdev.entity,0 , +MEDIA_LNK_FL_ENABLED |MEDIA_LNK_FL_IMMUTABLE );if (err)return err;
Signature · 1 line
2.43.0
Re: [PATCH 8/9] media: atomisp: Make MC link from ISP to /dev/video# output node immutable
Quoting Hans de Goede (2024-02-17 11:24:37)
Quoted 2 lines
The link from the ISP's source pad to the /dev/video# output v4l2-dev is always enabled and immutable, mark it as such.
Quoted 22 lines
Signed-off-by Hans de Goede <[email protected]>--- drivers/staging/media/atomisp/pci/atomisp_v4l2.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)drivers/staging/media/atomisp/pci/atomisp_v4l2.c +2
-1@@ -1090,7 +1090,8 @@ int atomisp_register_device_nodes(struct atomisp_device *isp)return err; err =media_create_pad_link (&isp->asd.subdev.entity,ATOMISP_SUBDEV_PAD_SOURCE , - &isp->asd.video_out.vdev.entity,0 ,0 ); + &isp->asd.video_out.vdev.entity,0 , +MEDIA_LNK_FL_ENABLED |MEDIA_LNK_FL_IMMUTABLE );if (err)return err;Signature · 1 line
2.43.0
PATCH 9/9 media: atomisp: Implement link_setup op for ISP subdev MC entity
In reply to Hans de GoedeThe atomisp driver's Android heritage makes it weird in that even though it uses MC + subdev-s it is designed to primarily be controlled through its /dev/video# node.
It implements s_input on /dev/video# to select which sensor to use, while ignoring setup_link calls to enable a link to another sensor.
Add support for selecting the active sensor the MC way by adding setup_link support.
The implementation is a bit convoluted due to the atomisp driver's heritage.
--- .../media/atomisp/pci/atomisp_subdev.c | 52 +++++++++++++++++++ 1 file changed, 52 insertions(+)
drivers/staging/media/atomisp/pci/atomisp_subdev.c +52 -0
@@ -30,6 +30,7 @@#include "atomisp_common.h" #include "atomisp_compat.h" #include "atomisp_fops.h" +#include "atomisp_ioctl.h" #include "atomisp_internal.h" const struct atomisp_in_fmt_conv atomisp_in_fmt_conv[] = {
@@ -634,8 +635,59 @@ static void isp_subdev_init_params(struct atomisp_sub_device *asd) }/* media operations */ +static int atomisp_link_setup (struct media_entity *entity, +const struct media_pad *local, +const struct media_pad *remote,u32 flags) +{ + struct v4l2_subdev *sd = container_of(entity, struct v4l2_subdev, + entity); +struct atomisp_sub_device *asd =v4l2_get_subdevdata (sd); +struct atomisp_device *isp = asd->isp; +int i, csi_idx, ret; + +/* ISP's source is immutable */ +if (local != &asd->pads[ATOMISP_SUBDEV_PAD_SINK ]) { +v4l2_err (sd,"Error pad %d does not support changing flags\n" , + local->index); +return -EINVAL ; + } + +for (csi_idx =0 ; csi_idx <ATOMISP_CAMERA_NR_PORTS ; csi_idx++) { +if (&isp->csi2_port[csi_idx].pads[CSI2_PAD_SOURCE ] == remote) +break ; + } + +if (csi_idx ==ATOMISP_CAMERA_NR_PORTS ) { +v4l2_err (sd,"Error cannot find CSI receiver for remote pad\n" ); +return -EINVAL ; + } + +/* Ignore disables, input_curr should only be updated on enables */ +if (!(flags &MEDIA_LNK_FL_ENABLED )) +return 0 ; + +for (i =0 ; i < isp->input_cnt; i++) { +if (isp->inputs[i].camera == isp->sensor_subdevs[csi_idx]) +break ; + } + +if (i == isp->input_cnt) { +v4l2_err (sd,"Error no sensor for CSI receiver %d\n" , csi_idx); +return -EINVAL ; + } + +mutex_lock (&isp->mutex); + ret =atomisp_pipe_check (&asd->video_out, true); +if (ret ==0 ) + asd->input_curr = i; +mutex_unlock (&isp->mutex); + +return ret; +} +static const struct media_entity_operations isp_subdev_media_ops = { .link_validate = v4l2_subdev_link_validate, + .link_setup = atomisp_link_setup,/* .set_power = v4l2_subdev_set_power, */ };
Signature · 1 line
2.43.0
Re: [PATCH 9/9] media: atomisp: Implement link_setup op for ISP subdev MC entity
Quoted 10 lines · Hans de Goede
The atomisp driver's Android heritage makes it weird in that even though it uses MC + subdev-s it is designed to primarily be controlled through its /dev/video# node.
It implements
s_inputon /dev/video# to select which sensor to use, while ignoringsetup_linkcalls to enable a link to another sensor.Add support for selecting the active sensor the MC way by adding
setup_linksupport.
->link_setup()
Quoted 2 lines
The implementation is a bit convoluted due to the atomisp driver's heritage.
...
Quoted 5 lines
#include "
atomisp_common.h"
#include "atomisp_compat.h"
#include "atomisp_fops.h"
+#include "atomisp_ioctl.h"
#include "atomisp_internal.h"
Hmm... Perhaps keep it ordered?
...
Quoted 1 line
+
mutex_lock(&isp->mutex);
Side note: Are you planning to use cleanup.h at some point?
Quoted 6 lines
drivers/staging/media/atomisp/pci/atomisp_subdev.c +6
-0+ ret =atomisp_pipe_check (&asd->video_out, true); +if (ret ==0 ) + asd->input_curr = i; +mutex_unlock (&isp->mutex); + +return ret;
Signature · 2 lines
With Best Regards, Andy Shevchenko
Re: [PATCH 9/9] media: atomisp: Implement link_setup op for ISP subdev MC entity
Hi,
Quoted 19 lines · Andy Shevchenko
Quoted 10 lines · Hans de Goede
The atomisp driver's Android heritage makes it weird in that even though it uses MC + subdev-s it is designed to primarily be controlled through its /dev/video# node.
It implements
s_inputon /dev/video# to select which sensor to use, while ignoringsetup_linkcalls to enable a link to another sensor.Add support for selecting the active sensor the MC way by adding
setup_linksupport.->link_setup()
Quoted 2 lines
The implementation is a bit convoluted due to the atomisp driver's heritage.
...
Quoted 7 lines
Quoted 5 lines
#include "
atomisp_common.h"
#include "atomisp_compat.h"
#include "atomisp_fops.h"
+#include "atomisp_ioctl.h"
#include "atomisp_internal.h"Hmm... Perhaps keep it ordered?
Yeah my bad, I intended to keep it ordered but I somehow got it wrong. this and the commit msg remark are both fixed in my personal tree now.
Quoted 6 lines
...
Quoted 1 line
+
mutex_lock(&isp->mutex);Side note: Are you planning to use
cleanup.hat some point?
Maybe. I have no objections against. For functions needing e.g. heap memory only locally it certainly makes sense.
Regards,
Hans
Quoted 9 lines
Quoted 7 lines
drivers/staging/media/atomisp/pci/atomisp_subdev.c +6
-0+ ret =atomisp_pipe_check (&asd->video_out, true); +if (ret ==0 ) + asd->input_curr = i; +mutex_unlock (&isp->mutex); + +return ret;
Quoted 8 lines · Hans de Goede
Hi All,
Here is a series of cleanups (prep work) + some small(ish) changes to make the atomisp driver work with libcamera.
This has been tested with libcamera's simple pipelinehandler together with an ov2680 sensor.
With some minor comments,
Quoted 4 lines
p.s.
Andy, I known I still need to do some small cleanups from my last pull-requests. I have this on my TODO list :)
Hope you will find time to address. Thanks for maintaining this driver!
Signature · 2 lines
With Best Regards, Andy Shevchenko
Hi Andy,
Quoted 4 lines · Andy Shevchenko
Quoted 3 lines · Hans de Goede
Hi All,
Quoted 8 lines
Quoted 5 lines
Here is a series of cleanups (prep work) + some small(ish) changes to make the atomisp driver work with libcamera.
This has been tested with libcamera's simple pipelinehandler together with an ov2680 sensor.
With some minor comments,
Reviewed-by Andy Shevchenko <[email protected]>
Thank you for all the reviews.
Regards,
Hans