All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v2] media: camss: Allocate camss struct as a managed device resource
@ 2022-05-19  5:14 Vladimir Zapolskiy
  2022-05-19  7:16 ` Hans Verkuil
  0 siblings, 1 reply; 3+ messages in thread
From: Vladimir Zapolskiy @ 2022-05-19  5:14 UTC (permalink / raw)
  To: Robert Foss, Todor Tomov
  Cc: Andy Gross, Bjorn Andersson, Mauro Carvalho Chehab, linux-media,
	linux-arm-msm

The change simplifies driver's probe and remove functions, no functional
change is intended.

Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
---
Changes from v1 to v2:
* rebased on top of media/master

 drivers/media/platform/qcom/camss/camss.c | 33 +++++++----------------
 1 file changed, 10 insertions(+), 23 deletions(-)

diff --git a/drivers/media/platform/qcom/camss/camss.c b/drivers/media/platform/qcom/camss/camss.c
index 79ad82e233cb..2055233af101 100644
--- a/drivers/media/platform/qcom/camss/camss.c
+++ b/drivers/media/platform/qcom/camss/camss.c
@@ -1529,7 +1529,7 @@ static int camss_probe(struct platform_device *pdev)
 	struct camss *camss;
 	int num_subdevs, ret;
 
-	camss = kzalloc(sizeof(*camss), GFP_KERNEL);
+	camss = devm_kzalloc(dev, sizeof(*camss), GFP_KERNEL);
 	if (!camss)
 		return -ENOMEM;
 
@@ -1567,39 +1567,30 @@ static int camss_probe(struct platform_device *pdev)
 		camss->csid_num = 4;
 		camss->vfe_num = 4;
 	} else {
-		ret = -EINVAL;
-		goto err_free;
+		return -EINVAL;
 	}
 
 	camss->csiphy = devm_kcalloc(dev, camss->csiphy_num,
 				     sizeof(*camss->csiphy), GFP_KERNEL);
-	if (!camss->csiphy) {
-		ret = -ENOMEM;
-		goto err_free;
-	}
+	if (!camss->csiphy)
+		return -ENOMEM;
 
 	camss->csid = devm_kcalloc(dev, camss->csid_num, sizeof(*camss->csid),
 				   GFP_KERNEL);
-	if (!camss->csid) {
-		ret = -ENOMEM;
-		goto err_free;
-	}
+	if (!camss->csid)
+		return -ENOMEM;
 
 	if (camss->version == CAMSS_8x16 ||
 	    camss->version == CAMSS_8x96) {
 		camss->ispif = devm_kcalloc(dev, 1, sizeof(*camss->ispif), GFP_KERNEL);
-		if (!camss->ispif) {
-			ret = -ENOMEM;
-			goto err_free;
-		}
+		if (!camss->ispif)
+			return -ENOMEM;
 	}
 
 	camss->vfe = devm_kcalloc(dev, camss->vfe_num, sizeof(*camss->vfe),
 				  GFP_KERNEL);
-	if (!camss->vfe) {
-		ret = -ENOMEM;
-		goto err_free;
-	}
+	if (!camss->vfe)
+		return -ENOMEM;
 
 	v4l2_async_nf_init(&camss->notifier);
 
@@ -1681,8 +1672,6 @@ static int camss_probe(struct platform_device *pdev)
 	v4l2_device_unregister(&camss->v4l2_dev);
 err_cleanup:
 	v4l2_async_nf_cleanup(&camss->notifier);
-err_free:
-	kfree(camss);
 
 	return ret;
 }
@@ -1709,8 +1698,6 @@ void camss_delete(struct camss *camss)
 		device_link_del(camss->genpd_link[i]);
 		dev_pm_domain_detach(camss->genpd[i], true);
 	}
-
-	kfree(camss);
 }
 
 /*
-- 
2.33.0


^ permalink raw reply related	[flat|nested] 3+ messages in thread
* Re: [PATCH v2] media: camss: Allocate camss struct as a managed device resource
  2022-05-19  5:14 [PATCH v2] media: camss: Allocate camss struct as a managed device resource Vladimir Zapolskiy
@ 2022-05-19  7:16 ` Hans Verkuil
  2022-05-23  9:26   ` Robert Foss
  0 siblings, 1 reply; 3+ messages in thread
From: Hans Verkuil @ 2022-05-19  7:16 UTC (permalink / raw)
  To: Vladimir Zapolskiy, Robert Foss, Todor Tomov
  Cc: Andy Gross, Bjorn Andersson, Mauro Carvalho Chehab, linux-media,
	linux-arm-msm

Robert, Vladimir,

On 5/19/22 07:14, Vladimir Zapolskiy wrote:
> The change simplifies driver's probe and remove functions, no functional
> change is intended.
> 
> Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
> ---
> Changes from v1 to v2:
> * rebased on top of media/master
> 
>  drivers/media/platform/qcom/camss/camss.c | 33 +++++++----------------
>  1 file changed, 10 insertions(+), 23 deletions(-)
> 
> diff --git a/drivers/media/platform/qcom/camss/camss.c b/drivers/media/platform/qcom/camss/camss.c
> index 79ad82e233cb..2055233af101 100644
> --- a/drivers/media/platform/qcom/camss/camss.c
> +++ b/drivers/media/platform/qcom/camss/camss.c
> @@ -1529,7 +1529,7 @@ static int camss_probe(struct platform_device *pdev)
>  	struct camss *camss;
>  	int num_subdevs, ret;
>  
> -	camss = kzalloc(sizeof(*camss), GFP_KERNEL);
> +	camss = devm_kzalloc(dev, sizeof(*camss), GFP_KERNEL);

In general it is not a good idea to use devm_*alloc. The problem is that if a
device instance is forcibly unbound, then all the memory allocated with devm_
is immediately freed. But if an application still has a file handle to a device
open, then it can still use that memory.

Now in this case the driver is already using devm_kcalloc, so it doesn't matter,
but it is something to keep in mind. In general, hotpluggable devices cannot use
devm_*alloc.

In general, my recommendation is to not use devm_*alloc for this, but since it
is already in use in this driver, it's better to use devm_*alloc consistently.
Or, alternatively, not use it all. That's up to Robert.

This is just as background information.

Regards,

	Hans

>  	if (!camss)
>  		return -ENOMEM;
>  
> @@ -1567,39 +1567,30 @@ static int camss_probe(struct platform_device *pdev)
>  		camss->csid_num = 4;
>  		camss->vfe_num = 4;
>  	} else {
> -		ret = -EINVAL;
> -		goto err_free;
> +		return -EINVAL;
>  	}
>  
>  	camss->csiphy = devm_kcalloc(dev, camss->csiphy_num,
>  				     sizeof(*camss->csiphy), GFP_KERNEL);
> -	if (!camss->csiphy) {
> -		ret = -ENOMEM;
> -		goto err_free;
> -	}
> +	if (!camss->csiphy)
> +		return -ENOMEM;
>  
>  	camss->csid = devm_kcalloc(dev, camss->csid_num, sizeof(*camss->csid),
>  				   GFP_KERNEL);
> -	if (!camss->csid) {
> -		ret = -ENOMEM;
> -		goto err_free;
> -	}
> +	if (!camss->csid)
> +		return -ENOMEM;
>  
>  	if (camss->version == CAMSS_8x16 ||
>  	    camss->version == CAMSS_8x96) {
>  		camss->ispif = devm_kcalloc(dev, 1, sizeof(*camss->ispif), GFP_KERNEL);
> -		if (!camss->ispif) {
> -			ret = -ENOMEM;
> -			goto err_free;
> -		}
> +		if (!camss->ispif)
> +			return -ENOMEM;
>  	}
>  
>  	camss->vfe = devm_kcalloc(dev, camss->vfe_num, sizeof(*camss->vfe),
>  				  GFP_KERNEL);
> -	if (!camss->vfe) {
> -		ret = -ENOMEM;
> -		goto err_free;
> -	}
> +	if (!camss->vfe)
> +		return -ENOMEM;
>  
>  	v4l2_async_nf_init(&camss->notifier);
>  
> @@ -1681,8 +1672,6 @@ static int camss_probe(struct platform_device *pdev)
>  	v4l2_device_unregister(&camss->v4l2_dev);
>  err_cleanup:
>  	v4l2_async_nf_cleanup(&camss->notifier);
> -err_free:
> -	kfree(camss);
>  
>  	return ret;
>  }
> @@ -1709,8 +1698,6 @@ void camss_delete(struct camss *camss)
>  		device_link_del(camss->genpd_link[i]);
>  		dev_pm_domain_detach(camss->genpd[i], true);
>  	}
> -
> -	kfree(camss);
>  }
>  
>  /*

^ permalink raw reply	[flat|nested] 3+ messages in thread
* Re: [PATCH v2] media: camss: Allocate camss struct as a managed device resource
  2022-05-19  7:16 ` Hans Verkuil
@ 2022-05-23  9:26   ` Robert Foss
  0 siblings, 0 replies; 3+ messages in thread
From: Robert Foss @ 2022-05-23  9:26 UTC (permalink / raw)
  To: Hans Verkuil
  Cc: Vladimir Zapolskiy, Todor Tomov, Andy Gross, Bjorn Andersson,
	Mauro Carvalho Chehab, linux-media, linux-arm-msm

On Thu, 19 May 2022 at 09:16, Hans Verkuil <hverkuil@xs4all.nl> wrote:
>
> Robert, Vladimir,
>
> On 5/19/22 07:14, Vladimir Zapolskiy wrote:
> > The change simplifies driver's probe and remove functions, no functional
> > change is intended.
> >
> > Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@linaro.org>
> > ---
> > Changes from v1 to v2:
> > * rebased on top of media/master
> >
> >  drivers/media/platform/qcom/camss/camss.c | 33 +++++++----------------
> >  1 file changed, 10 insertions(+), 23 deletions(-)
> >
> > diff --git a/drivers/media/platform/qcom/camss/camss.c b/drivers/media/platform/qcom/camss/camss.c
> > index 79ad82e233cb..2055233af101 100644
> > --- a/drivers/media/platform/qcom/camss/camss.c
> > +++ b/drivers/media/platform/qcom/camss/camss.c
> > @@ -1529,7 +1529,7 @@ static int camss_probe(struct platform_device *pdev)
> >       struct camss *camss;
> >       int num_subdevs, ret;
> >
> > -     camss = kzalloc(sizeof(*camss), GFP_KERNEL);
> > +     camss = devm_kzalloc(dev, sizeof(*camss), GFP_KERNEL);
>
> In general it is not a good idea to use devm_*alloc. The problem is that if a
> device instance is forcibly unbound, then all the memory allocated with devm_
> is immediately freed. But if an application still has a file handle to a device
> open, then it can still use that memory.
>
> Now in this case the driver is already using devm_kcalloc, so it doesn't matter,
> but it is something to keep in mind. In general, hotpluggable devices cannot use
> devm_*alloc.
>
> In general, my recommendation is to not use devm_*alloc for this, but since it
> is already in use in this driver, it's better to use devm_*alloc consistently.
> Or, alternatively, not use it all. That's up to Robert.
>
> This is just as background information.


Thanks for explaining the nuances Hans, that's very helpful.

I think this patch is ok, since it doesn't make the situation any worse,
and that CSI camera sensors are very likely to be hotplugged.

>
> Regards,
>
>         Hans
>
> >       if (!camss)
> >               return -ENOMEM;
> >
> > @@ -1567,39 +1567,30 @@ static int camss_probe(struct platform_device *pdev)
> >               camss->csid_num = 4;
> >               camss->vfe_num = 4;
> >       } else {
> > -             ret = -EINVAL;
> > -             goto err_free;
> > +             return -EINVAL;
> >       }
> >
> >       camss->csiphy = devm_kcalloc(dev, camss->csiphy_num,
> >                                    sizeof(*camss->csiphy), GFP_KERNEL);
> > -     if (!camss->csiphy) {
> > -             ret = -ENOMEM;
> > -             goto err_free;
> > -     }
> > +     if (!camss->csiphy)
> > +             return -ENOMEM;
> >
> >       camss->csid = devm_kcalloc(dev, camss->csid_num, sizeof(*camss->csid),
> >                                  GFP_KERNEL);
> > -     if (!camss->csid) {
> > -             ret = -ENOMEM;
> > -             goto err_free;
> > -     }
> > +     if (!camss->csid)
> > +             return -ENOMEM;
> >
> >       if (camss->version == CAMSS_8x16 ||
> >           camss->version == CAMSS_8x96) {
> >               camss->ispif = devm_kcalloc(dev, 1, sizeof(*camss->ispif), GFP_KERNEL);
> > -             if (!camss->ispif) {
> > -                     ret = -ENOMEM;
> > -                     goto err_free;
> > -             }
> > +             if (!camss->ispif)
> > +                     return -ENOMEM;
> >       }
> >
> >       camss->vfe = devm_kcalloc(dev, camss->vfe_num, sizeof(*camss->vfe),
> >                                 GFP_KERNEL);
> > -     if (!camss->vfe) {
> > -             ret = -ENOMEM;
> > -             goto err_free;
> > -     }
> > +     if (!camss->vfe)
> > +             return -ENOMEM;
> >
> >       v4l2_async_nf_init(&camss->notifier);
> >
> > @@ -1681,8 +1672,6 @@ static int camss_probe(struct platform_device *pdev)
> >       v4l2_device_unregister(&camss->v4l2_dev);
> >  err_cleanup:
> >       v4l2_async_nf_cleanup(&camss->notifier);
> > -err_free:
> > -     kfree(camss);
> >
> >       return ret;
> >  }
> > @@ -1709,8 +1698,6 @@ void camss_delete(struct camss *camss)
> >               device_link_del(camss->genpd_link[i]);
> >               dev_pm_domain_detach(camss->genpd[i], true);
> >       }
> > -
> > -     kfree(camss);
> >  }
> >
> >  /*

Reviewed-by: Robert Foss <robert.foss@linaro.org>

^ permalink raw reply	[flat|nested] 3+ messages in thread
end of thread, other threads:[~2022-05-23  9:27 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-19  5:14 [PATCH v2] media: camss: Allocate camss struct as a managed device resource Vladimir Zapolskiy
2022-05-19  7:16 ` Hans Verkuil
2022-05-23  9:26   ` Robert Foss

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.