[PATCH 0/2] fs-ext4: Adjustments for two function implementations

[PATCH 0/2] fs-ext4: Adjustments for two function implementations

[SF Markus Elfring]

From: Markus Elfring <elfring@users.sourceforge.net>
Date: Sat, 19 Aug 2017 13:35:43 +0200

Two update suggestions were taken into account
from static source code analysis.

Markus Elfring (2):
  Delete an error message for a failed memory allocation in ext4_multi_mount_protect()
  Improve a size determination in two functions

 fs/ext4/dir.c | 2 +-
 fs/ext4/mmp.c | 7 +++----
 2 files changed, 4 insertions(+), 5 deletions(-)

-- 
2.14.0

[PATCH 1/2] ext4: Delete an error message for a failed memory allocation in ext4_multi_mount_protect()

[SF Markus Elfring]

From: Markus Elfring <elfring@users.sourceforge.net>
Date: Sat, 19 Aug 2017 13:04:50 +0200

Omit an extra message for a memory allocation failure in this function.

This issue was detected by using the Coccinelle software.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
---
 fs/ext4/mmp.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c
index eb9835638680..1ce00453f612 100644
--- a/fs/ext4/mmp.c
+++ b/fs/ext4/mmp.c
@@ -371,7 +371,6 @@ int ext4_multi_mount_protect(struct super_block *sb,
-	if (!mmpd_data) {
-		ext4_warning(sb, "not enough memory for mmpd_data");
+	if (!mmpd_data)
 		goto failed;
-	}
+
 	mmpd_data->sb = sb;
 	mmpd_data->bh = bh;
 
-- 
2.14.0

[PATCH 2/2] ext4: Improve a size determination in two functions

[SF Markus Elfring]

From: Markus Elfring <elfring@users.sourceforge.net>
Date: Sat, 19 Aug 2017 13:14:26 +0200

Replace the specification of data structures by pointer dereferences
as the parameter for the operator "sizeof" to make the corresponding size
determination a bit safer according to the Linux coding style convention.

This issue was detected by using the Coccinelle software.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
---
 fs/ext4/dir.c | 2 +-
 fs/ext4/mmp.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
index e8b365000d73..b04e882179c6 100644
--- a/fs/ext4/dir.c
+++ b/fs/ext4/dir.c
@@ -411,7 +411,7 @@ static struct dir_private_info *ext4_htree_create_dir_info(struct file *filp,
 {
 	struct dir_private_info *p;
 
-	p = kzalloc(sizeof(struct dir_private_info), GFP_KERNEL);
+	p = kzalloc(sizeof(*p), GFP_KERNEL);
 	if (!p)
 		return NULL;
 	p->curr_hash = pos2maj_hash(filp, pos);
diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c
index 1ce00453f612..3fa5df9f5573 100644
--- a/fs/ext4/mmp.c
+++ b/fs/ext4/mmp.c
@@ -367,5 +367,5 @@ int ext4_multi_mount_protect(struct super_block *sb,
 		goto failed;
 	}
 
-	mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL);
+	mmpd_data = kmalloc(sizeof(*mmpd_data), GFP_KERNEL);
 	if (!mmpd_data)
-- 
2.14.0

Re: [PATCH 1/2] ext4: Delete an error message for a failed memory allocation in ext4_multi_mount_protect()

[Eric Sandeen]

On 8/19/17 6:47 AM, SF Markus Elfring wrote:
> From: Markus Elfring <elfring@users.sourceforge.net>
> Date: Sat, 19 Aug 2017 13:04:50 +0200
> 
> Omit an extra message for a memory allocation failure in this function.

I might be dense, but what makes this message "extra?"

(I suppose kmalloc squawks too if it fails, but is Coccinelle
now warning about explicit memory allocation failure warnings?)

-Eric
 
> This issue was detected by using the Coccinelle software.
> 
> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
> ---
>  fs/ext4/mmp.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c
> index eb9835638680..1ce00453f612 100644
> --- a/fs/ext4/mmp.c
> +++ b/fs/ext4/mmp.c
> @@ -371,7 +371,6 @@ int ext4_multi_mount_protect(struct super_block *sb,
> -	if (!mmpd_data) {
> -		ext4_warning(sb, "not enough memory for mmpd_data");
> +	if (!mmpd_data)
>  		goto failed;
> -	}
> +
>  	mmpd_data->sb = sb;
>  	mmpd_data->bh = bh;
>  
> 

Re: ext4: Delete an error message for a failed memory allocation in ext4_multi_mount_protect()

[SF Markus Elfring]

>> Omit an extra message for a memory allocation failure in this function.
> 
> I might be dense, but what makes this message "extra?"
> 
> (I suppose kmalloc squawks too if it fails,

Do you find the default allocation failure report sufficient?


> but is Coccinelle now warning about explicit memory allocation failure warnings?)

This software tool can help to find source code places for further
development considerations. Would you like to clarify a corresponding
search pattern a bit more?

Regards,
Markus

Re: [PATCH 2/2] ext4: Improve a size determination in two functions

[Eric Sandeen]

On 8/19/17 6:48 AM, SF Markus Elfring wrote:
> From: Markus Elfring <elfring@users.sourceforge.net>
> Date: Sat, 19 Aug 2017 13:14:26 +0200
> 
> Replace the specification of data structures by pointer dereferences
> as the parameter for the operator "sizeof" to make the corresponding size
> determination a bit safer according to the Linux coding style convention.
> 
> This issue was detected by using the Coccinelle software.
> 
> Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>

Looks good,

Reviewed-by: Eric Sandeen <sandeen@redhat.com>

> ---
>  fs/ext4/dir.c | 2 +-
>  fs/ext4/mmp.c | 2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c
> index e8b365000d73..b04e882179c6 100644
> --- a/fs/ext4/dir.c
> +++ b/fs/ext4/dir.c
> @@ -411,7 +411,7 @@ static struct dir_private_info *ext4_htree_create_dir_info(struct file *filp,
>  {
>  	struct dir_private_info *p;
>  
> -	p = kzalloc(sizeof(struct dir_private_info), GFP_KERNEL);
> +	p = kzalloc(sizeof(*p), GFP_KERNEL);
>  	if (!p)
>  		return NULL;
>  	p->curr_hash = pos2maj_hash(filp, pos);
> diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c
> index 1ce00453f612..3fa5df9f5573 100644
> --- a/fs/ext4/mmp.c
> +++ b/fs/ext4/mmp.c
> @@ -367,5 +367,5 @@ int ext4_multi_mount_protect(struct super_block *sb,
>  		goto failed;
>  	}
>  
> -	mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL);
> +	mmpd_data = kmalloc(sizeof(*mmpd_data), GFP_KERNEL);
>  	if (!mmpd_data)
> 

Re: ext4: Delete an error message for a failed memory allocation in ext4_multi_mount_protect()

[Theodore Ts'o]

On Sat, Aug 19, 2017 at 08:00:31PM +0200, SF Markus Elfring wrote:
> >> Omit an extra message for a memory allocation failure in this function.
> > 
> > I might be dense, but what makes this message "extra?"
> > 
> > (I suppose kmalloc squawks too if it fails,
> 
> Do you find the default allocation failure report sufficient?

>From a helpdesk reporting situation, having a more specific message
when there is a MMP failure causing the mount to fail is definitely
useful.

So, NACK.

					- Ted

Re: [PATCH 2/2] ext4: Improve a size determination in two functions

[Theodore Ts'o]

On Sat, Aug 19, 2017 at 08:49:20PM -0500, Eric Sandeen wrote:
> On 8/19/17 6:48 AM, SF Markus Elfring wrote:
> > From: Markus Elfring <elfring@users.sourceforge.net>
> > Date: Sat, 19 Aug 2017 13:14:26 +0200
> > 
> > Replace the specification of data structures by pointer dereferences
> > as the parameter for the operator "sizeof" to make the corresponding size
> > determination a bit safer according to the Linux coding style convention.
> > 
> > This issue was detected by using the Coccinelle software.
> > 
> > Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
> 
> Looks good,
> 
> Reviewed-by: Eric Sandeen <sandeen@redhat.com>

Thanks, applied.

					- Ted

Re: [PATCH 1/2] ext4: Delete an error message for a failed memory allocation in ext4_multi_mount_protect()

[Dan Carpenter]

On Sat, Aug 19, 2017 at 12:08:29PM -0500, Eric Sandeen wrote:
> On 8/19/17 6:47 AM, SF Markus Elfring wrote:
> > From: Markus Elfring <elfring@users.sourceforge.net>
> > Date: Sat, 19 Aug 2017 13:04:50 +0200
> > 
> > Omit an extra message for a memory allocation failure in this function.
> 
> I might be dense, but what makes this message "extra?"
> 
> (I suppose kmalloc squawks too if it fails, but is Coccinelle
> now warning about explicit memory allocation failure warnings?)
> 

Yeah...  Checkpatch complains that the kmalloc squawks is enough.
"WARNING: Possible unnecessary 'out of memory' message".  This
allocation is small so it's guaranteed to succeed in current kernels.

regards,
dan carpenter