LinuxLists.cc - [PATCH 1/7] media: cec: no need to check return value of debugfs

2020-08-18 13:37:09

by Greg Kroah-Hartman

[permalink] [raw]

Subject: [PATCH 1/7] media: cec: no need to check return value of debugfs_create functions

2020-08-18 13:37:29

by Greg Kroah-Hartman

[permalink] [raw]

Subject: [PATCH 4/7] media: mtk-vpu: no need to check return value of debugfs_create functions

2020-08-18 13:37:33

by Greg Kroah-Hartman

[permalink] [raw]

Subject: [PATCH 6/7] media: radio: si476x: no need to check return value of debugfs_create functions

When calling debugfs functions, there is no need to ever check the
return value. The function can work or not, but the code logic should
never do something different based on this.

Cc: Mauro Carvalho Chehab <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Cc: Alexios Zavras <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Hans Verkuil <[email protected]>
Cc: [email protected]
Cc: [email protected]
Signed-off-by: Greg Kroah-Hartman <[email protected]>
---
drivers/media/radio/radio-si476x.c | 66 ++++++------------------------
1 file changed, 13 insertions(+), 53 deletions(-)

diff --git a/drivers/media/radio/radio-si476x.c b/drivers/media/radio/radio-si476x.c
index b203296de977..485b2d2d977d 100644
--- a/drivers/media/radio/radio-si476x.c
+++ b/drivers/media/radio/radio-si476x.c
@@ -1344,60 +1344,24 @@ static const struct file_operations radio_rsq_primary_fops = {
};

-static int si476x_radio_init_debugfs(struct si476x_radio *radio)
+static void si476x_radio_init_debugfs(struct si476x_radio *radio)
{
- struct dentry *dentry;
- int ret;
+ radio->debugfs = debugfs_create_dir(dev_name(radio->v4l2dev.dev), NULL);

- dentry = debugfs_create_dir(dev_name(radio->v4l2dev.dev), NULL);
- if (IS_ERR(dentry)) {
- ret = PTR_ERR(dentry);
- goto exit;
- }
- radio->debugfs = dentry;
-
- dentry = debugfs_create_file("acf", S_IRUGO,
- radio->debugfs, radio, &radio_acf_fops);
- if (IS_ERR(dentry)) {
- ret = PTR_ERR(dentry);
- goto cleanup;
- }
+ debugfs_create_file("acf", S_IRUGO, radio->debugfs, radio,
+ &radio_acf_fops);

- dentry = debugfs_create_file("rds_blckcnt", S_IRUGO,
- radio->debugfs, radio,
- &radio_rds_blckcnt_fops);
- if (IS_ERR(dentry)) {
- ret = PTR_ERR(dentry);
- goto cleanup;
- }
+ debugfs_create_file("rds_blckcnt", S_IRUGO, radio->debugfs, radio,
+ &radio_rds_blckcnt_fops);

- dentry = debugfs_create_file("agc", S_IRUGO,
- radio->debugfs, radio, &radio_agc_fops);
- if (IS_ERR(dentry)) {
- ret = PTR_ERR(dentry);
- goto cleanup;
- }
+ debugfs_create_file("agc", S_IRUGO, radio->debugfs, radio,
+ &radio_agc_fops);

- dentry = debugfs_create_file("rsq", S_IRUGO,
- radio->debugfs, radio, &radio_rsq_fops);
- if (IS_ERR(dentry)) {
- ret = PTR_ERR(dentry);
- goto cleanup;
- }
+ debugfs_create_file("rsq", S_IRUGO, radio->debugfs, radio,
+ &radio_rsq_fops);

- dentry = debugfs_create_file("rsq_primary", S_IRUGO,
- radio->debugfs, radio,
- &radio_rsq_primary_fops);
- if (IS_ERR(dentry)) {
- ret = PTR_ERR(dentry);
- goto cleanup;
- }
-
- return 0;
-cleanup:
- debugfs_remove_recursive(radio->debugfs);
-exit:
- return ret;
+ debugfs_create_file("rsq_primary", S_IRUGO, radio->debugfs, radio,
+ &radio_rsq_primary_fops);
}

@@ -1534,11 +1498,7 @@ static int si476x_radio_probe(struct platform_device *pdev)
goto exit;
}

- rval = si476x_radio_init_debugfs(radio);
- if (rval < 0) {
- dev_err(&pdev->dev, "Could not create debugfs interface\n");
- goto exit;
- }
+ si476x_radio_init_debugfs(radio);

return 0;
exit:
--
2.28.0

2020-08-18 13:38:08

by Greg Kroah-Hartman

[permalink] [raw]

Subject: [PATCH 7/7] media: usb: uvc: no need to check return value of debugfs_create functions

2020-08-18 13:39:26

by Greg Kroah-Hartman

[permalink] [raw]

Subject: [PATCH 5/7] media: sti: no need to check return value of debugfs_create functions

When calling debugfs functions, there is no need to ever check the
return value. The function can work or not, but the code logic should
never do something different based on this.

Cc: Fabien Dessenne <[email protected]>
Cc: Mauro Carvalho Chehab <[email protected]>
Cc: Jean-Christophe Trotin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Signed-off-by: Greg Kroah-Hartman <[email protected]>
---
.../media/platform/sti/bdisp/bdisp-debug.c | 29 ++++---------------
drivers/media/platform/sti/bdisp/bdisp-v4l2.c | 7 +----
drivers/media/platform/sti/bdisp/bdisp.h | 2 +-
drivers/media/platform/sti/hva/hva-debugfs.c | 22 +++-----------
4 files changed, 12 insertions(+), 48 deletions(-)

diff --git a/drivers/media/platform/sti/bdisp/bdisp-debug.c b/drivers/media/platform/sti/bdisp/bdisp-debug.c
index 77ca7517fa3e..2b270093009c 100644
--- a/drivers/media/platform/sti/bdisp/bdisp-debug.c
+++ b/drivers/media/platform/sti/bdisp/bdisp-debug.c
@@ -637,35 +637,18 @@ DEFINE_SHOW_ATTRIBUTE(last_nodes_raw);
DEFINE_SHOW_ATTRIBUTE(last_request);
DEFINE_SHOW_ATTRIBUTE(perf);

-int bdisp_debugfs_create(struct bdisp_dev *bdisp)
+void bdisp_debugfs_create(struct bdisp_dev *bdisp)
{
char dirname[16];

snprintf(dirname, sizeof(dirname), "%s%d", BDISP_NAME, bdisp->id);
bdisp->dbg.debugfs_entry = debugfs_create_dir(dirname, NULL);
- if (!bdisp->dbg.debugfs_entry)
- goto err;

- if (!bdisp_dbg_create_entry(regs))
- goto err;
-
- if (!bdisp_dbg_create_entry(last_nodes))
- goto err;
-
- if (!bdisp_dbg_create_entry(last_nodes_raw))
- goto err;
-
- if (!bdisp_dbg_create_entry(last_request))
- goto err;
-
- if (!bdisp_dbg_create_entry(perf))
- goto err;
-
- return 0;
-
-err:
- bdisp_debugfs_remove(bdisp);
- return -ENOMEM;
+ bdisp_dbg_create_entry(regs);
+ bdisp_dbg_create_entry(last_nodes);
+ bdisp_dbg_create_entry(last_nodes_raw);
+ bdisp_dbg_create_entry(last_request);
+ bdisp_dbg_create_entry(perf);
}

void bdisp_debugfs_remove(struct bdisp_dev *bdisp)
diff --git a/drivers/media/platform/sti/bdisp/bdisp-v4l2.c b/drivers/media/platform/sti/bdisp/bdisp-v4l2.c
index af2d5eb782ce..7d50d6cd073d 100644
--- a/drivers/media/platform/sti/bdisp/bdisp-v4l2.c
+++ b/drivers/media/platform/sti/bdisp/bdisp-v4l2.c
@@ -1360,11 +1360,7 @@ static int bdisp_probe(struct platform_device *pdev)
}

/* Debug */
- ret = bdisp_debugfs_create(bdisp);
- if (ret) {
- dev_err(dev, "failed to create debugfs\n");
- goto err_v4l2;
- }
+ bdisp_debugfs_create(bdisp);

/* Power management */
pm_runtime_enable(dev);
@@ -1401,7 +1397,6 @@ static int bdisp_probe(struct platform_device *pdev)
pm_runtime_put(dev);
err_dbg:
bdisp_debugfs_remove(bdisp);
-err_v4l2:
v4l2_device_unregister(&bdisp->v4l2_dev);
err_clk:
if (!IS_ERR(bdisp->clock))
diff --git a/drivers/media/platform/sti/bdisp/bdisp.h b/drivers/media/platform/sti/bdisp/bdisp.h
index e309cde379ca..3fb009d24791 100644
--- a/drivers/media/platform/sti/bdisp/bdisp.h
+++ b/drivers/media/platform/sti/bdisp/bdisp.h
@@ -209,6 +209,6 @@ int bdisp_hw_get_and_clear_irq(struct bdisp_dev *bdisp);
int bdisp_hw_update(struct bdisp_ctx *ctx);

void bdisp_debugfs_remove(struct bdisp_dev *bdisp);
-int bdisp_debugfs_create(struct bdisp_dev *bdisp);
+void bdisp_debugfs_create(struct bdisp_dev *bdisp);
void bdisp_dbg_perf_begin(struct bdisp_dev *bdisp);
void bdisp_dbg_perf_end(struct bdisp_dev *bdisp);
diff --git a/drivers/media/platform/sti/hva/hva-debugfs.c b/drivers/media/platform/sti/hva/hva-debugfs.c
index 7d12a5b5d914..a86a07b6fbc7 100644
--- a/drivers/media/platform/sti/hva/hva-debugfs.c
+++ b/drivers/media/platform/sti/hva/hva-debugfs.c
@@ -337,25 +337,11 @@ DEFINE_SHOW_ATTRIBUTE(regs);
void hva_debugfs_create(struct hva_dev *hva)
{
hva->dbg.debugfs_entry = debugfs_create_dir(HVA_NAME, NULL);
- if (!hva->dbg.debugfs_entry)
- goto err;

- if (!hva_dbg_create_entry(device))
- goto err;
-
- if (!hva_dbg_create_entry(encoders))
- goto err;
-
- if (!hva_dbg_create_entry(last))
- goto err;
-
- if (!hva_dbg_create_entry(regs))
- goto err;
-
- return;
-
-err:
- hva_debugfs_remove(hva);
+ hva_dbg_create_entry(device);
+ hva_dbg_create_entry(encoders);
+ hva_dbg_create_entry(last);
+ hva_dbg_create_entry(regs);
}

void hva_debugfs_remove(struct hva_dev *hva)
--
2.28.0

2020-08-18 13:41:56

by Greg Kroah-Hartman

[permalink] [raw]

Subject: [PATCH 3/7] media: exynos4-is: no need to check return value of debugfs_create functions

2020-08-19 00:11:38

by Laurent Pinchart

[permalink] [raw]

Subject: Re: [PATCH 7/7] media: usb: uvc: no need to check return value of debugfs_create functions

Hi Greg,

Thank you for the patch.

On Tue, Aug 18, 2020 at 03:36:08PM +0200, Greg Kroah-Hartman wrote:
> When calling debugfs functions, there is no need to ever check the
> return value. The function can work or not, but the code logic should
> never do something different based on this.

Is there no value in warning the user that something went wrong ? Silent
failures are harder to debug.

> Cc: Laurent Pinchart <[email protected]>
> Cc: Mauro Carvalho Chehab <[email protected]>
> Cc: [email protected]
> Cc: [email protected]
> Signed-off-by: Greg Kroah-Hartman <[email protected]>
> ---
> drivers/media/usb/uvc/uvc_debugfs.c | 20 ++++----------------
> 1 file changed, 4 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/media/usb/uvc/uvc_debugfs.c b/drivers/media/usb/uvc/uvc_debugfs.c
> index 2b8af4b54117..1a1258d4ffca 100644
> --- a/drivers/media/usb/uvc/uvc_debugfs.c
> +++ b/drivers/media/usb/uvc/uvc_debugfs.c
> @@ -73,7 +73,6 @@ static struct dentry *uvc_debugfs_root_dir;
> void uvc_debugfs_init_stream(struct uvc_streaming *stream)
> {
> struct usb_device *udev = stream->dev->udev;
> - struct dentry *dent;
> char dir_name[33];
>
> if (uvc_debugfs_root_dir == NULL)
> @@ -82,22 +81,11 @@ void uvc_debugfs_init_stream(struct uvc_streaming *stream)
> snprintf(dir_name, sizeof(dir_name), "%u-%u-%u", udev->bus->busnum,
> udev->devnum, stream->intfnum);
>
> - dent = debugfs_create_dir(dir_name, uvc_debugfs_root_dir);
> - if (IS_ERR_OR_NULL(dent)) {
> - uvc_printk(KERN_INFO, "Unable to create debugfs %s "
> - "directory.\n", dir_name);
> - return;
> - }
> -
> - stream->debugfs_dir = dent;
> + stream->debugfs_dir = debugfs_create_dir(dir_name,
> + uvc_debugfs_root_dir);
>
> - dent = debugfs_create_file("stats", 0444, stream->debugfs_dir,
> - stream, &uvc_debugfs_stats_fops);
> - if (IS_ERR_OR_NULL(dent)) {
> - uvc_printk(KERN_INFO, "Unable to create debugfs stats file.\n");
> - uvc_debugfs_cleanup_stream(stream);
> - return;
> - }
> + debugfs_create_file("stats", 0444, stream->debugfs_dir, stream,
> + &uvc_debugfs_stats_fops);
> }
>
> void uvc_debugfs_cleanup_stream(struct uvc_streaming *stream)

--
Regards,

Laurent Pinchart

2020-09-17 12:37:15

by Laurent Pinchart

[permalink] [raw]

Subject: Re: [PATCH 7/7] media: usb: uvc: no need to check return value of debugfs_create functions

Hi Greg,

On Wed, Aug 19, 2020 at 02:47:19AM +0300, Laurent Pinchart wrote:
> Hi Greg,
>
> Thank you for the patch.
>
> On Tue, Aug 18, 2020 at 03:36:08PM +0200, Greg Kroah-Hartman wrote:
> > When calling debugfs functions, there is no need to ever check the
> > return value. The function can work or not, but the code logic should
> > never do something different based on this.
>
> Is there no value in warning the user that something went wrong ? Silent
> failures are harder to debug.

Could yous share your opinion about this ?

> > Cc: Laurent Pinchart <[email protected]>
> > Cc: Mauro Carvalho Chehab <[email protected]>
> > Cc: [email protected]
> > Cc: [email protected]
> > Signed-off-by: Greg Kroah-Hartman <[email protected]>
> > ---
> > drivers/media/usb/uvc/uvc_debugfs.c | 20 ++++----------------
> > 1 file changed, 4 insertions(+), 16 deletions(-)
> >
> > diff --git a/drivers/media/usb/uvc/uvc_debugfs.c b/drivers/media/usb/uvc/uvc_debugfs.c
> > index 2b8af4b54117..1a1258d4ffca 100644
> > --- a/drivers/media/usb/uvc/uvc_debugfs.c
> > +++ b/drivers/media/usb/uvc/uvc_debugfs.c
> > @@ -73,7 +73,6 @@ static struct dentry *uvc_debugfs_root_dir;
> > void uvc_debugfs_init_stream(struct uvc_streaming *stream)
> > {
> > struct usb_device *udev = stream->dev->udev;
> > - struct dentry *dent;
> > char dir_name[33];
> >
> > if (uvc_debugfs_root_dir == NULL)
> > @@ -82,22 +81,11 @@ void uvc_debugfs_init_stream(struct uvc_streaming *stream)
> > snprintf(dir_name, sizeof(dir_name), "%u-%u-%u", udev->bus->busnum,
> > udev->devnum, stream->intfnum);
> >
> > - dent = debugfs_create_dir(dir_name, uvc_debugfs_root_dir);
> > - if (IS_ERR_OR_NULL(dent)) {
> > - uvc_printk(KERN_INFO, "Unable to create debugfs %s "
> > - "directory.\n", dir_name);
> > - return;
> > - }
> > -
> > - stream->debugfs_dir = dent;
> > + stream->debugfs_dir = debugfs_create_dir(dir_name,
> > + uvc_debugfs_root_dir);
> >
> > - dent = debugfs_create_file("stats", 0444, stream->debugfs_dir,
> > - stream, &uvc_debugfs_stats_fops);
> > - if (IS_ERR_OR_NULL(dent)) {
> > - uvc_printk(KERN_INFO, "Unable to create debugfs stats file.\n");
> > - uvc_debugfs_cleanup_stream(stream);
> > - return;
> > - }
> > + debugfs_create_file("stats", 0444, stream->debugfs_dir, stream,
> > + &uvc_debugfs_stats_fops);
> > }
> >
> > void uvc_debugfs_cleanup_stream(struct uvc_streaming *stream)

--
Regards,

Laurent Pinchart

2020-09-17 12:42:07

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH 7/7] media: usb: uvc: no need to check return value of debugfs_create functions

On Thu, Sep 17, 2020 at 03:25:50PM +0300, Laurent Pinchart wrote:
> Hi Greg,
>
> On Wed, Aug 19, 2020 at 02:47:19AM +0300, Laurent Pinchart wrote:
> > Hi Greg,
> >
> > Thank you for the patch.
> >
> > On Tue, Aug 18, 2020 at 03:36:08PM +0200, Greg Kroah-Hartman wrote:
> > > When calling debugfs functions, there is no need to ever check the
> > > return value. The function can work or not, but the code logic should
> > > never do something different based on this.
> >
> > Is there no value in warning the user that something went wrong ? Silent
> > failures are harder to debug.
>
> Could yous share your opinion about this ?

For debugfs, this isn't an issue, what can a user do with something like
"debugfs isn't working? What does that mean???"

And if we _really_ want warnings like this, it should go into the
debugfs core, not require this to be done for every debugfs user, right?

debugfs is just there for kernel developers to help debug things, it's
not a dependancy on any userspace functionality, so if it works or not
should not be an issue for any user.

Unless that user is a kernel developer of course :)

thanks,

greg k-h

2020-09-17 12:46:07

by Laurent Pinchart

[permalink] [raw]

Subject: Re: [PATCH 7/7] media: usb: uvc: no need to check return value of debugfs_create functions

Hi Greg,

On Thu, Sep 17, 2020 at 02:34:26PM +0200, Greg Kroah-Hartman wrote:
> On Thu, Sep 17, 2020 at 03:25:50PM +0300, Laurent Pinchart wrote:
> > On Wed, Aug 19, 2020 at 02:47:19AM +0300, Laurent Pinchart wrote:
> > > On Tue, Aug 18, 2020 at 03:36:08PM +0200, Greg Kroah-Hartman wrote:
> > > > When calling debugfs functions, there is no need to ever check the
> > > > return value. The function can work or not, but the code logic should
> > > > never do something different based on this.
> > >
> > > Is there no value in warning the user that something went wrong ? Silent
> > > failures are harder to debug.
> >
> > Could yous share your opinion about this ?
>
> For debugfs, this isn't an issue, what can a user do with something like
> "debugfs isn't working? What does that mean???"
>
> And if we _really_ want warnings like this, it should go into the
> debugfs core, not require this to be done for every debugfs user, right?
>
> debugfs is just there for kernel developers to help debug things, it's
> not a dependancy on any userspace functionality, so if it works or not
> should not be an issue for any user.
>
> Unless that user is a kernel developer of course :)

Exactly my point :-)

I'm fine moving the error message to the debugfs core itself instead of
duplicating it in drivers. Maybe it's already there though, I haven't
checked. Not printing any message isn't a great idea in my opinion, it
makes debugging more difficult. I can't count the number of times where
I've had to add printk's and recompile the kernel to debug issues that
really should have generated at least a dev_dbg().

--
Regards,

Laurent Pinchart

2020-09-17 12:51:38

by Greg Kroah-Hartman

[permalink] [raw]

Subject: Re: [PATCH 7/7] media: usb: uvc: no need to check return value of debugfs_create functions

On Thu, Sep 17, 2020 at 03:37:57PM +0300, Laurent Pinchart wrote:
> Hi Greg,
>
> On Thu, Sep 17, 2020 at 02:34:26PM +0200, Greg Kroah-Hartman wrote:
> > On Thu, Sep 17, 2020 at 03:25:50PM +0300, Laurent Pinchart wrote:
> > > On Wed, Aug 19, 2020 at 02:47:19AM +0300, Laurent Pinchart wrote:
> > > > On Tue, Aug 18, 2020 at 03:36:08PM +0200, Greg Kroah-Hartman wrote:
> > > > > When calling debugfs functions, there is no need to ever check the
> > > > > return value. The function can work or not, but the code logic should
> > > > > never do something different based on this.
> > > >
> > > > Is there no value in warning the user that something went wrong ? Silent
> > > > failures are harder to debug.
> > >
> > > Could yous share your opinion about this ?
> >
> > For debugfs, this isn't an issue, what can a user do with something like
> > "debugfs isn't working? What does that mean???"
> >
> > And if we _really_ want warnings like this, it should go into the
> > debugfs core, not require this to be done for every debugfs user, right?
> >
> > debugfs is just there for kernel developers to help debug things, it's
> > not a dependancy on any userspace functionality, so if it works or not
> > should not be an issue for any user.
> >
> > Unless that user is a kernel developer of course :)
>
> Exactly my point :-)
>
> I'm fine moving the error message to the debugfs core itself instead of
> duplicating it in drivers. Maybe it's already there though, I haven't
> checked. Not printing any message isn't a great idea in my opinion, it
> makes debugging more difficult. I can't count the number of times where
> I've had to add printk's and recompile the kernel to debug issues that
> really should have generated at least a dev_dbg().

There are a lot of error messages that debugfs will print out if it can
not create a file:

inode.c 329 pr_err("Unable to pin filesystem for file '%s'\n", name);
inode.c 348 pr_err("Directory '%s' with parent '%s' already present!\n",
inode.c 351 pr_err("File '%s' in directory '%s' already present!\n",
inode.c 402 pr_err("out of free dentries, can not create file '%s'\n",
inode.c 563 pr_err("out of free dentries, can not create directory '%s'\n",
inode.c 610 pr_err("out of free dentries, can not create automount '%s'\n",
inode.c 668 pr_err("out of free dentries, can not create symlink '%s'\n",

So I think you are safe here. If we are missing any, I'll gladly add
them.

Also given that you've never noticed this being a real error, means it's
probably not an issue :)

thanks,

greg k-h