| Message ID | 20240828-imx290-avail-v2-1-bd320ac8e8fa@skidata.com |
|---|---|
| State | New |
| Headers | show |
| Series | [v2,1/2] media: i2c: imx290: Check for availability in probe() | expand |
Hi Sakari, On Thu, 29 Aug 2024 at 09:05, Sakari Ailus <sakari.ailus@linux.intel.com> wrote: > Hi Benjamin, > > On Wed, Aug 28, 2024 at 08:13:07PM +0200, Benjamin Bara wrote: > > Currently, the V4L2 subdevice is also created when the device is not > > available/connected. In this case, dmesg shows the following: > > > > [ 10.419510] imx290 7-001a: Error writing reg 0x301c: -6 > > [ 10.428981] imx290 7-001a: Error writing reg 0x3020: -6 > > [ 10.442712] imx290 7-001a: Error writing reg 0x3018: -6 > > [ 10.454018] imx290 7-001a: Error writing reg 0x3020: -6 > > > > which seems to come from imx290_ctrl_update() after the subdev init is > > finished. However, as the errors are ignored, the subdev is initialized > > but simply does not work. From userspace perspective, there is no > > visible difference between a working and not-working subdevice (except > > when trying it out or watching for the error message). > > > > This commit adds a simple availability check before starting with the > > subdev initialization to error out instead. > > > > Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com> > > --- > > Changes since v1: > > - define operating/standby mode and use it > > - read out the standby mode during probe and ensure it is standby > > --- > > drivers/media/i2c/imx290.c | 17 +++++++++++++++-- > > 1 file changed, 15 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/media/i2c/imx290.c b/drivers/media/i2c/imx290.c > > index 4150e6e4b9a6..2a869576600c 100644 > > --- a/drivers/media/i2c/imx290.c > > +++ b/drivers/media/i2c/imx290.c > > @@ -29,6 +29,8 @@ > > #include <media/v4l2-subdev.h> > > > > #define IMX290_STANDBY CCI_REG8(0x3000) > > +#define IMX290_STANDBY_OPERATING (0 << 0) > > +#define IMX290_STANDBY_STANDBY (1 << 0) > > #define IMX290_REGHOLD CCI_REG8(0x3001) > > #define IMX290_XMSTA CCI_REG8(0x3002) > > #define IMX290_ADBIT CCI_REG8(0x3005) > > @@ -1016,7 +1018,7 @@ static int imx290_start_streaming(struct imx290 *imx290, > > return ret; > > } > > > > - cci_write(imx290->regmap, IMX290_STANDBY, 0x00, &ret); > > + cci_write(imx290->regmap, IMX290_STANDBY, IMX290_STANDBY_OPERATING, &ret); > > Please run > > ./scripts/checkpatch.pl --strict --max-line-length=80 > > the next time. I'll wrap the line this time. Sorry. I added the strict max-line check to my .gitconfig. "b4 prep --check" now warns me. > > > > msleep(30); > > > > @@ -1029,7 +1031,7 @@ static int imx290_stop_streaming(struct imx290 *imx290) > > { > > int ret = 0; > > > > - cci_write(imx290->regmap, IMX290_STANDBY, 0x01, &ret); > > + cci_write(imx290->regmap, IMX290_STANDBY, IMX290_STANDBY_STANDBY, &ret); > > > > msleep(30); > > > > @@ -1520,6 +1522,7 @@ static int imx290_probe(struct i2c_client *client) > > { > > struct device *dev = &client->dev; > > struct imx290 *imx290; > > + u64 val; > > int ret; > > > > imx290 = devm_kzalloc(dev, sizeof(*imx290), GFP_KERNEL); > > @@ -1580,6 +1583,16 @@ static int imx290_probe(struct i2c_client *client) > > pm_runtime_set_autosuspend_delay(dev, 1000); > > pm_runtime_use_autosuspend(dev); > > > > + /* Make sure the sensor is available before V4L2 subdev init. */ > > + ret = cci_read(imx290->regmap, IMX290_STANDBY, &val, NULL); > > + if (ret) > > + goto err_pm; > > + if (val != IMX290_STANDBY_STANDBY) { > > + dev_err(dev, "Sensor is not in standby mode\n"); > > + ret = -ENODEV; > > + goto err_pm; > > + } > > + > > /* Initialize the V4L2 subdev. */ > > ret = imx290_subdev_init(imx290); > > if (ret) > > > > -- > Kind regards, > > Sakari Ailus thx & regards Benjamin
On Thu, Aug 29, 2024 at 07:05:34AM +0000, Sakari Ailus wrote:
> the next time. I'll wrap the line this time.
Based on the other review comments I'll wait for v2.
Thanks.
On Thu, Aug 29, 2024 at 12:39:36PM +0000, Sakari Ailus wrote: > On Thu, Aug 29, 2024 at 07:05:34AM +0000, Sakari Ailus wrote: > > the next time. I'll wrap the line this time. > > Based on the other review comments I'll wait for v2. Oh well, wrong e-mail. Please ignore.
Hi Benjamin, Thank you for the patch. On Wed, Aug 28, 2024 at 08:13:07PM +0200, Benjamin Bara wrote: > Currently, the V4L2 subdevice is also created when the device is not > available/connected. In this case, dmesg shows the following: > > [ 10.419510] imx290 7-001a: Error writing reg 0x301c: -6 > [ 10.428981] imx290 7-001a: Error writing reg 0x3020: -6 > [ 10.442712] imx290 7-001a: Error writing reg 0x3018: -6 > [ 10.454018] imx290 7-001a: Error writing reg 0x3020: -6 > > which seems to come from imx290_ctrl_update() after the subdev init is > finished. I think this should also be fixed. There should be no need to write those registers at probe time. Would moving the pm_runtime_mark_last_busy() and pm_runtime_put_autosuspend() calls to just before imx290_subdev_init() help ? > However, as the errors are ignored, the subdev is initialized > but simply does not work. From userspace perspective, there is no > visible difference between a working and not-working subdevice (except > when trying it out or watching for the error message). > > This commit adds a simple availability check before starting with the > subdev initialization to error out instead. > > Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com> > --- > Changes since v1: > - define operating/standby mode and use it > - read out the standby mode during probe and ensure it is standby > --- > drivers/media/i2c/imx290.c | 17 +++++++++++++++-- > 1 file changed, 15 insertions(+), 2 deletions(-) > > diff --git a/drivers/media/i2c/imx290.c b/drivers/media/i2c/imx290.c > index 4150e6e4b9a6..2a869576600c 100644 > --- a/drivers/media/i2c/imx290.c > +++ b/drivers/media/i2c/imx290.c > @@ -29,6 +29,8 @@ > #include <media/v4l2-subdev.h> > > #define IMX290_STANDBY CCI_REG8(0x3000) > +#define IMX290_STANDBY_OPERATING (0 << 0) > +#define IMX290_STANDBY_STANDBY (1 << 0) The datasheet documents the STANDBY field as a single bit, but doesn't mention an OPERATING value. I would match that, drop IMX290_STANDBY_OPERATING and write #define IMX290_STANDBY_STANDBY BIT(0) > #define IMX290_REGHOLD CCI_REG8(0x3001) > #define IMX290_XMSTA CCI_REG8(0x3002) > #define IMX290_ADBIT CCI_REG8(0x3005) > @@ -1016,7 +1018,7 @@ static int imx290_start_streaming(struct imx290 *imx290, > return ret; > } > > - cci_write(imx290->regmap, IMX290_STANDBY, 0x00, &ret); > + cci_write(imx291->regmap, IMX290_STANDBY, IMX290_STANDBY_OPERATING, &ret); This hunk would then be dropped. > > msleep(30); > > @@ -1029,7 +1031,7 @@ static int imx290_stop_streaming(struct imx290 *imx290) > { > int ret = 0; > > - cci_write(imx290->regmap, IMX290_STANDBY, 0x01, &ret); > + cci_write(imx290->regmap, IMX290_STANDBY, IMX290_STANDBY_STANDBY, &ret); And this looks fine. The change isn't mentioned in the commit message though. I wouldn't ask for a v3 just to split this, but as you need to address other issues, it would be nice to have a separate patch in v3. > > msleep(30); > > @@ -1520,6 +1522,7 @@ static int imx290_probe(struct i2c_client *client) > { > struct device *dev = &client->dev; > struct imx290 *imx290; > + u64 val; > int ret; > > imx290 = devm_kzalloc(dev, sizeof(*imx290), GFP_KERNEL); > @@ -1580,6 +1583,16 @@ static int imx290_probe(struct i2c_client *client) > pm_runtime_set_autosuspend_delay(dev, 1000); > pm_runtime_use_autosuspend(dev); > > + /* Make sure the sensor is available before V4L2 subdev init. */ > + ret = cci_read(imx290->regmap, IMX290_STANDBY, &val, NULL); I still wish we had an ID register, but so be it. > + if (ret) Maybe add ret = dev_err_probe(dev, -ENODEV, "Failed to detect sensor\n"); or something similar ? Up to you. > + goto err_pm; > + if (val != IMX290_STANDBY_STANDBY) { I think this check could be dropped though. If the sensor isn't present or doesn't respond to I2C reads for any other reason, cci_read() will fail. > + dev_err(dev, "Sensor is not in standby mode\n"); > + ret = -ENODEV; > + goto err_pm; > + } > + My last concern is about accessing hardware at probe time. There are known cases where this is problematic. They can be split in two categories, systems that exhibit unwanted side effects when powering the sensor up, and systems where the sensor can't be accessed at probe time. The two issues I can think of in the first category is devices that have a camera privacy light that could cause worries among users if it flashes at boot time, and devices that agressively optimize boot time. In the second category, I know that some people use camera serdes (FPD-Link, GMSL, ...) that are controlled by userspace. As they should instead use kernel drivers for those components, upstream may not care too much about this use case. Another issue I was told about was a device booting in temperatures that were too low for the camera to operate, which then needed half an hour to heat the device enclosure before the sensor and serdes could be accessed. That's a bit extreme, but it sounds like a valid use case to me. What do we do with those cases ? Detecting devices at probe time does have value, so I think it should be a policy decision. We may want to convey some of that information through DT properties (I'm not sure what would be acceptable there though). In any case, that's quite a bit of yak shaving, so I'm inclined to accept this series (or rather its next version), given that quite a few other camera sensor drivers detect the device at probe time. I would however like feedback on the problem to try and find a good solution. > /* Initialize the V4L2 subdev. */ > ret = imx290_subdev_init(imx290); > if (ret) >
Hi Manivannan, On Thu, Aug 29, 2024 at 10:02:47PM +0530, Manivannan Sadhasivam wrote: > On Thu, Aug 29, 2024 at 04:19:09PM +0300, Laurent Pinchart wrote: > > Hi Laurent, > > [...] > > > > + dev_err(dev, "Sensor is not in standby mode\n"); > > > + ret = -ENODEV; > > > + goto err_pm; > > > + } > > > + > > > > My last concern is about accessing hardware at probe time. There are > > known cases where this is problematic. They can be split in two > > categories, systems that exhibit unwanted side effects when powering the > > sensor up, and systems where the sensor can't be accessed at probe time. > > > > The two issues I can think of in the first category is devices that have > > a camera privacy light that could cause worries among users if it > > flashes at boot time, and devices that agressively optimize boot time. > > > > In the second category, I know that some people use camera serdes > > (FPD-Link, GMSL, ...) that are controlled by userspace. As they should > > instead use kernel drivers for those components, upstream may not care > > too much about this use case. Another issue I was told about was a > > device booting in temperatures that were too low for the camera to > > operate, which then needed half an hour to heat the device enclosure > > before the sensor and serdes could be accessed. That's a bit extreme, > > but it sounds like a valid use case to me. > > > > What do we do with those cases ? Detecting devices at probe time does > > have value, so I think it should be a policy decision. We may want to > > convey some of that information through DT properties (I'm not sure what > > would be acceptable there though). In any case, that's quite a bit of > > yak shaving, so I'm inclined to accept this series (or rather its next > > version), given that quite a few other camera sensor drivers detect the > > device at probe time. I would however like feedback on the problem to > > try and find a good solution. > > Most of the issues you mentioned applies to other hardware peripherals also IMO. > And it is common for the drivers to read registers and make sure the device is > detected on the bus during probe(). That's true. I think the problem affects different device types differently though, and this may (or may not) call for different solutions. > If an usecase doesn't want to read the > registers during probe time, then they _should_not_ build the driver as built-in > rather make it as a loadable module and load it whenever necessary. This applies > to boot time optimization as well. For most of the use cases I listed I agree with you. One exception is the privacy light issue. Regardless of when the camera sensor driver is loaded, powering the device at probe time will flash the privacy light. Doing so later than boot time would probably make the issue even worse, I would worry more if I saw my webcam privacy light flashing at a random point after boot time. > A DT property wouldn't be feasible as DT is supposed to describe the hardware, > not the usecase. I think that rule is typically slightly relaxed, by allowing in DT system descriptions, not just hardware descriptions. Otherwise we wouldn't allow things like reserved memory ranges. Describing that a camera sensor has a privacy light, in a way that would allow drivers to avoid powering up the device at probe time without requiring much duplicated code in all drivers, would in my opinion be an acceptable DT usage.
On Thu, Aug 29, 2024 at 07:48:43PM +0300, Laurent Pinchart wrote: > Hi Manivannan, > > On Thu, Aug 29, 2024 at 10:02:47PM +0530, Manivannan Sadhasivam wrote: > > On Thu, Aug 29, 2024 at 04:19:09PM +0300, Laurent Pinchart wrote: > > > > Hi Laurent, > > > > [...] > > > > > > + dev_err(dev, "Sensor is not in standby mode\n"); > > > > + ret = -ENODEV; > > > > + goto err_pm; > > > > + } > > > > + > > > > > > My last concern is about accessing hardware at probe time. There are > > > known cases where this is problematic. They can be split in two > > > categories, systems that exhibit unwanted side effects when powering the > > > sensor up, and systems where the sensor can't be accessed at probe time. > > > > > > The two issues I can think of in the first category is devices that have > > > a camera privacy light that could cause worries among users if it > > > flashes at boot time, and devices that agressively optimize boot time. > > > > > > In the second category, I know that some people use camera serdes > > > (FPD-Link, GMSL, ...) that are controlled by userspace. As they should > > > instead use kernel drivers for those components, upstream may not care > > > too much about this use case. Another issue I was told about was a > > > device booting in temperatures that were too low for the camera to > > > operate, which then needed half an hour to heat the device enclosure > > > before the sensor and serdes could be accessed. That's a bit extreme, > > > but it sounds like a valid use case to me. > > > > > > What do we do with those cases ? Detecting devices at probe time does > > > have value, so I think it should be a policy decision. We may want to > > > convey some of that information through DT properties (I'm not sure what > > > would be acceptable there though). In any case, that's quite a bit of > > > yak shaving, so I'm inclined to accept this series (or rather its next > > > version), given that quite a few other camera sensor drivers detect the > > > device at probe time. I would however like feedback on the problem to > > > try and find a good solution. > > > > Most of the issues you mentioned applies to other hardware peripherals also IMO. > > And it is common for the drivers to read registers and make sure the device is > > detected on the bus during probe(). > > That's true. I think the problem affects different device types > differently though, and this may (or may not) call for different > solutions. > > > If an usecase doesn't want to read the > > registers during probe time, then they _should_not_ build the driver as built-in > > rather make it as a loadable module and load it whenever necessary. This applies > > to boot time optimization as well. > > For most of the use cases I listed I agree with you. One exception is > the privacy light issue. Regardless of when the camera sensor driver is > loaded, powering the device at probe time will flash the privacy light. > Doing so later than boot time would probably make the issue even worse, > I would worry more if I saw my webcam privacy light flashing at a random > point after boot time. > I'm not familiar with the privacy light feature in camera sensors, but is there no way to prevent it from enabling by default? If that's not possible, it makes sense to disable it using a DT property as it is a hardware feature. > > A DT property wouldn't be feasible as DT is supposed to describe the hardware, > > not the usecase. > > I think that rule is typically slightly relaxed, by allowing in DT > system descriptions, not just hardware descriptions. Otherwise we > wouldn't allow things like reserved memory ranges. Describing that a > camera sensor has a privacy light, in a way that would allow drivers to > avoid powering up the device at probe time without requiring much > duplicated code in all drivers, would in my opinion be an acceptable DT > usage. > Well, I agree with you. As I said above, privacy light is a hardware feature, so we can enable/disable it using a DT property. My comment about the DT property applies only to detecting the devices during probe time, which is a driver/usecase dependent feature. - Mani
On Fri, Aug 30, 2024 at 05:55:29PM +0530, Manivannan Sadhasivam wrote: > On Fri, Aug 30, 2024 at 12:25:26PM +0300, Laurent Pinchart wrote: > > On Fri, Aug 30, 2024 at 02:01:07PM +0530, Manivannan Sadhasivam wrote: > > > On Thu, Aug 29, 2024 at 07:48:43PM +0300, Laurent Pinchart wrote: > > > > On Thu, Aug 29, 2024 at 10:02:47PM +0530, Manivannan Sadhasivam wrote: > > > > > On Thu, Aug 29, 2024 at 04:19:09PM +0300, Laurent Pinchart wrote: > > > > > > > > > > Hi Laurent, > > > > > > > > > > [...] > > > > > > > > > > > > + dev_err(dev, "Sensor is not in standby mode\n"); > > > > > > > + ret = -ENODEV; > > > > > > > + goto err_pm; > > > > > > > + } > > > > > > > + > > > > > > > > > > > > My last concern is about accessing hardware at probe time. There are > > > > > > known cases where this is problematic. They can be split in two > > > > > > categories, systems that exhibit unwanted side effects when powering the > > > > > > sensor up, and systems where the sensor can't be accessed at probe time. > > > > > > > > > > > > The two issues I can think of in the first category is devices that have > > > > > > a camera privacy light that could cause worries among users if it > > > > > > flashes at boot time, and devices that agressively optimize boot time. > > > > > > > > > > > > In the second category, I know that some people use camera serdes > > > > > > (FPD-Link, GMSL, ...) that are controlled by userspace. As they should > > > > > > instead use kernel drivers for those components, upstream may not care > > > > > > too much about this use case. Another issue I was told about was a > > > > > > device booting in temperatures that were too low for the camera to > > > > > > operate, which then needed half an hour to heat the device enclosure > > > > > > before the sensor and serdes could be accessed. That's a bit extreme, > > > > > > but it sounds like a valid use case to me. > > > > > > > > > > > > What do we do with those cases ? Detecting devices at probe time does > > > > > > have value, so I think it should be a policy decision. We may want to > > > > > > convey some of that information through DT properties (I'm not sure what > > > > > > would be acceptable there though). In any case, that's quite a bit of > > > > > > yak shaving, so I'm inclined to accept this series (or rather its next > > > > > > version), given that quite a few other camera sensor drivers detect the > > > > > > device at probe time. I would however like feedback on the problem to > > > > > > try and find a good solution. > > > > > > > > > > Most of the issues you mentioned applies to other hardware peripherals also IMO. > > > > > And it is common for the drivers to read registers and make sure the device is > > > > > detected on the bus during probe(). > > > > > > > > That's true. I think the problem affects different device types > > > > differently though, and this may (or may not) call for different > > > > solutions. > > > > > > > > > If an usecase doesn't want to read the > > > > > registers during probe time, then they _should_not_ build the driver as built-in > > > > > rather make it as a loadable module and load it whenever necessary. This applies > > > > > to boot time optimization as well. > > > > > > > > For most of the use cases I listed I agree with you. One exception is > > > > the privacy light issue. Regardless of when the camera sensor driver is > > > > loaded, powering the device at probe time will flash the privacy light. > > > > Doing so later than boot time would probably make the issue even worse, > > > > I would worry more if I saw my webcam privacy light flashing at a random > > > > point after boot time. > > > > > > I'm not familiar with the privacy light feature in camera sensors, but is there > > > no way to prevent it from enabling by default? If that's not possible, it makes > > > sense to disable it using a DT property as it is a hardware feature. > > > > The whole point of the privacy light is that it shouldn't be possible to > > disable it by software. Otherwise malicious software could try to work > > around it. On many devices it is hardwired to one of the camera sensor's > > power supplies. > > Ah okay, please forgive my ignorance here. But still I'm not sure about the DT > usage. Maybe it is best to send out a bindings patch and see what the > maintainers have to say? Yes, that's a good path forward. I don't think it should block this patch though.
diff --git a/drivers/media/i2c/imx290.c b/drivers/media/i2c/imx290.c index 4150e6e4b9a6..2a869576600c 100644 --- a/drivers/media/i2c/imx290.c +++ b/drivers/media/i2c/imx290.c @@ -29,6 +29,8 @@ #include <media/v4l2-subdev.h> #define IMX290_STANDBY CCI_REG8(0x3000) +#define IMX290_STANDBY_OPERATING (0 << 0) +#define IMX290_STANDBY_STANDBY (1 << 0) #define IMX290_REGHOLD CCI_REG8(0x3001) #define IMX290_XMSTA CCI_REG8(0x3002) #define IMX290_ADBIT CCI_REG8(0x3005) @@ -1016,7 +1018,7 @@ static int imx290_start_streaming(struct imx290 *imx290, return ret; } - cci_write(imx290->regmap, IMX290_STANDBY, 0x00, &ret); + cci_write(imx290->regmap, IMX290_STANDBY, IMX290_STANDBY_OPERATING, &ret); msleep(30); @@ -1029,7 +1031,7 @@ static int imx290_stop_streaming(struct imx290 *imx290) { int ret = 0; - cci_write(imx290->regmap, IMX290_STANDBY, 0x01, &ret); + cci_write(imx290->regmap, IMX290_STANDBY, IMX290_STANDBY_STANDBY, &ret); msleep(30); @@ -1520,6 +1522,7 @@ static int imx290_probe(struct i2c_client *client) { struct device *dev = &client->dev; struct imx290 *imx290; + u64 val; int ret; imx290 = devm_kzalloc(dev, sizeof(*imx290), GFP_KERNEL); @@ -1580,6 +1583,16 @@ static int imx290_probe(struct i2c_client *client) pm_runtime_set_autosuspend_delay(dev, 1000); pm_runtime_use_autosuspend(dev); + /* Make sure the sensor is available before V4L2 subdev init. */ + ret = cci_read(imx290->regmap, IMX290_STANDBY, &val, NULL); + if (ret) + goto err_pm; + if (val != IMX290_STANDBY_STANDBY) { + dev_err(dev, "Sensor is not in standby mode\n"); + ret = -ENODEV; + goto err_pm; + } + /* Initialize the V4L2 subdev. */ ret = imx290_subdev_init(imx290); if (ret)
Currently, the V4L2 subdevice is also created when the device is not available/connected. In this case, dmesg shows the following: [ 10.419510] imx290 7-001a: Error writing reg 0x301c: -6 [ 10.428981] imx290 7-001a: Error writing reg 0x3020: -6 [ 10.442712] imx290 7-001a: Error writing reg 0x3018: -6 [ 10.454018] imx290 7-001a: Error writing reg 0x3020: -6 which seems to come from imx290_ctrl_update() after the subdev init is finished. However, as the errors are ignored, the subdev is initialized but simply does not work. From userspace perspective, there is no visible difference between a working and not-working subdevice (except when trying it out or watching for the error message). This commit adds a simple availability check before starting with the subdev initialization to error out instead. Signed-off-by: Benjamin Bara <benjamin.bara@skidata.com> --- Changes since v1: - define operating/standby mode and use it - read out the standby mode during probe and ensure it is standby --- drivers/media/i2c/imx290.c | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-)