Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1636846 > unrolled thread

[PATCH] pinctrl: imx: Check for memory allocation failure

Started byChristophe JAILLET <christophe.jaillet@wanadoo.fr>
First post2017-05-06 10:30 +0200
Last post2017-05-09 14:40 +0200
Articles 4 — 3 participants

Back to article view | Back to linux.kernel

Contents

  [PATCH] pinctrl: imx: Check for memory allocation failure Christophe JAILLET <christophe.jaillet@wanadoo.fr> - 2017-05-06 10:30 +0200
    Re: [PATCH] pinctrl: imx: Check for memory allocation failure Stafford Horne <shorne@gmail.com> - 2017-05-06 21:50 +0200
      Re: [PATCH] pinctrl: imx: Check for memory allocation failure Dan Carpenter <dan.carpenter@oracle.com> - 2017-05-09 09:30 +0200
        Re: [PATCH] pinctrl: imx: Check for memory allocation failure Stafford Horne <shorne@gmail.com> - 2017-05-09 14:40 +0200

#1636846 — [PATCH] pinctrl: imx: Check for memory allocation failure

FromChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Date2017-05-06 10:30 +0200
Subject[PATCH] pinctrl: imx: Check for memory allocation failure
Message-ID<tE3gd-5Fp-3@gated-at.bofh.it>
If 'devm_kzalloc' fails, a NULL pointer will be dereferenced.
Return -ENOMEM instead, as done for the other memory allocation just a
few lines below.
BTW, change the 'devm_kzalloc' into a 'devm_kcalloc'.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
---
 drivers/pinctrl/freescale/pinctrl-imx.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/pinctrl/freescale/pinctrl-imx.c b/drivers/pinctrl/freescale/pinctrl-imx.c
index 74bd90dfd7b1..90a946c028ff 100644
--- a/drivers/pinctrl/freescale/pinctrl-imx.c
+++ b/drivers/pinctrl/freescale/pinctrl-imx.c
@@ -581,9 +581,10 @@ static int imx_pinctrl_parse_functions(struct device_node *np,
 		dev_err(info->dev, "no groups defined in %s\n", np->full_name);
 		return -EINVAL;
 	}
-	func->group_names = devm_kzalloc(info->dev,
-					 func->num_group_names *
+	func->group_names = devm_kcalloc(info->dev, func->num_group_names,
 					 sizeof(char *), GFP_KERNEL);
+	if (!func->group_names)
+		return -ENOMEM;
 
 	for_each_child_of_node(np, child) {
 		func->group_names[i] = child->name;
-- 
2.11.0

[toc] | [next] | [standalone]

#1636972

FromStafford Horne <shorne@gmail.com>
Date2017-05-06 21:50 +0200
Message-ID<tEdSh-3MB-5@gated-at.bofh.it>
In reply to#1636846
Hi Christophe,

On Sat, May 06, 2017 at 10:23:59AM +0200, Christophe JAILLET wrote:
> If 'devm_kzalloc' fails, a NULL pointer will be dereferenced.
> Return -ENOMEM instead, as done for the other memory allocation just a
> few lines below.

This looks fine.

> BTW, change the 'devm_kzalloc' into a 'devm_kcalloc'.

Any reason for the devm_kcalloc change?  It looks like the next for loop
does set all of the group_name values.

-Stafford

> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
> ---
>  drivers/pinctrl/freescale/pinctrl-imx.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/pinctrl/freescale/pinctrl-imx.c b/drivers/pinctrl/freescale/pinctrl-imx.c
> index 74bd90dfd7b1..90a946c028ff 100644
> --- a/drivers/pinctrl/freescale/pinctrl-imx.c
> +++ b/drivers/pinctrl/freescale/pinctrl-imx.c
> @@ -581,9 +581,10 @@ static int imx_pinctrl_parse_functions(struct device_node *np,
>  		dev_err(info->dev, "no groups defined in %s\n", np->full_name);
>  		return -EINVAL;
>  	}
> -	func->group_names = devm_kzalloc(info->dev,
> -					 func->num_group_names *
> +	func->group_names = devm_kcalloc(info->dev, func->num_group_names,
>  					 sizeof(char *), GFP_KERNEL);
> +	if (!func->group_names)
> +		return -ENOMEM;
>  
>  	for_each_child_of_node(np, child) {
>  		func->group_names[i] = child->name;
> -- 
> 2.11.0
>

[toc] | [prev] | [next] | [standalone]

#1637923

FromDan Carpenter <dan.carpenter@oracle.com>
Date2017-05-09 09:30 +0200
Message-ID<tF7KN-6Yd-3@gated-at.bofh.it>
In reply to#1636972
On Sun, May 07, 2017 at 04:40:38AM +0900, Stafford Horne wrote:
> Hi Christophe,
> 
> On Sat, May 06, 2017 at 10:23:59AM +0200, Christophe JAILLET wrote:
> > If 'devm_kzalloc' fails, a NULL pointer will be dereferenced.
> > Return -ENOMEM instead, as done for the other memory allocation just a
> > few lines below.
> 
> This looks fine.
> 
> > BTW, change the 'devm_kzalloc' into a 'devm_kcalloc'.
> 
> Any reason for the devm_kcalloc change?  It looks like the next for loop
> does set all of the group_name values.
> 

The advantage of kcalloc() over kzalloc() is the integer overflow
checking.  There is kmalloc_array() if we don't need to zero the memory.

regards,
dan carpenter

[toc] | [prev] | [next] | [standalone]

#1638092

FromStafford Horne <shorne@gmail.com>
Date2017-05-09 14:40 +0200
Message-ID<tFcAO-1J6-11@gated-at.bofh.it>
In reply to#1637923
On Tue, May 09, 2017 at 10:27:20AM +0300, Dan Carpenter wrote:
> On Sun, May 07, 2017 at 04:40:38AM +0900, Stafford Horne wrote:
> > Hi Christophe,
> > 
> > On Sat, May 06, 2017 at 10:23:59AM +0200, Christophe JAILLET wrote:
> > > If 'devm_kzalloc' fails, a NULL pointer will be dereferenced.
> > > Return -ENOMEM instead, as done for the other memory allocation just a
> > > few lines below.
> > 
> > This looks fine.
> > 
> > > BTW, change the 'devm_kzalloc' into a 'devm_kcalloc'.
> > 
> > Any reason for the devm_kcalloc change?  It looks like the next for loop
> > does set all of the group_name values.
> > 
> 
> The advantage of kcalloc() over kzalloc() is the integer overflow
> checking.  There is kmalloc_array() if we don't need to zero the memory.

Right, usually its good to have a reason why in the commit log. i.e.

 BTW, change the 'devm_kzalloc' into a 'devm_kcalloc' for overflow checking.

Or switch to devm_kmalloc_array() and say: for overflow checking and no
need for zeroing.

-Stafford

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web