[PATCH 1/1] ovpn: use kmalloc_array() for array space allocation

[PATCH 1/1] ovpn: use kmalloc_array() for array space allocation

[chuguangqing]

Replace kmalloc(size * sizeof) with kmalloc_array() for safer memory
allocation and overflow prevention.

Signed-off-by: chuguangqing <chuguangqing@inspur.com>
---
 drivers/net/ovpn/crypto_aead.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ovpn/crypto_aead.c b/drivers/net/ovpn/crypto_aead.c
index 2cca759feffa..8274c3ae8d0b 100644
--- a/drivers/net/ovpn/crypto_aead.c
+++ b/drivers/net/ovpn/crypto_aead.c
@@ -72,8 +72,8 @@ int ovpn_aead_encrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
 		return -ENOSPC;
 
 	/* sg may be required by async crypto */
-	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
-				       (nfrags + 2), GFP_ATOMIC);
+	ovpn_skb_cb(skb)->sg = kmalloc_array((nfrags + 2), sizeof(*ovpn_skb_cb(skb)->sg),
+					     GFP_ATOMIC);
 	if (unlikely(!ovpn_skb_cb(skb)->sg))
 		return -ENOMEM;
 
@@ -185,8 +185,8 @@ int ovpn_aead_decrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
 		return -ENOSPC;
 
 	/* sg may be required by async crypto */
-	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
-				       (nfrags + 2), GFP_ATOMIC);
+	ovpn_skb_cb(skb)->sg = kmalloc_array((nfrags + 2), sizeof(*ovpn_skb_cb(skb)->sg),
+					     GFP_ATOMIC);
 	if (unlikely(!ovpn_skb_cb(skb)->sg))
 		return -ENOMEM;
 
-- 
2.43.5

Re: [PATCH 1/1] ovpn: use kmalloc_array() for array space allocation

[Simon Horman]

On Mon, Sep 01, 2025 at 07:21:36PM +0800, chuguangqing wrote:
> Replace kmalloc(size * sizeof) with kmalloc_array() for safer memory
> allocation and overflow prevention.
> 
> Signed-off-by: chuguangqing <chuguangqing@inspur.com>
> ---
>  drivers/net/ovpn/crypto_aead.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/ovpn/crypto_aead.c b/drivers/net/ovpn/crypto_aead.c
> index 2cca759feffa..8274c3ae8d0b 100644
> --- a/drivers/net/ovpn/crypto_aead.c
> +++ b/drivers/net/ovpn/crypto_aead.c
> @@ -72,8 +72,8 @@ int ovpn_aead_encrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
>  		return -ENOSPC;
>  
>  	/* sg may be required by async crypto */
> -	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
> -				       (nfrags + 2), GFP_ATOMIC);
> +	ovpn_skb_cb(skb)->sg = kmalloc_array((nfrags + 2), sizeof(*ovpn_skb_cb(skb)->sg),
> +					     GFP_ATOMIC);

I think this could benefit from:
a) Removal of unnecessary parentheses
b) Line wrapping to 80 columns wide or less,
   as is still preferred for Networking code

(Completely untested!)

	ovpn_skb_cb(skb)->sg = kmalloc_array(nfrags + 2,
					     sizeof(*ovpn_skb_cb(skb)->sg),
					     GFP_ATOMIC);

>  	if (unlikely(!ovpn_skb_cb(skb)->sg))
>  		return -ENOMEM;
>  
> @@ -185,8 +185,8 @@ int ovpn_aead_decrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
>  		return -ENOSPC;
>  
>  	/* sg may be required by async crypto */
> -	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
> -				       (nfrags + 2), GFP_ATOMIC);
> +	ovpn_skb_cb(skb)->sg = kmalloc_array((nfrags + 2), sizeof(*ovpn_skb_cb(skb)->sg),
> +					     GFP_ATOMIC);

Likewise here.

>  	if (unlikely(!ovpn_skb_cb(skb)->sg))
>  		return -ENOMEM;
>  

-- 
pw-bot: changes-requested

PATCH v2 Re: Re: [PATCH 1/1] ovpn: use kmalloc_array() for array space allocation

[chuguangqing]

v1 -> v2:
Unnecessary parentheses have been removed and line alignment has been adjusted to 80 characters in width.

> 
> On Mon, Sep 01, 2025 at 07:21:36PM +0800, chuguangqing wrote:
> > Replace kmalloc(size * sizeof) with kmalloc_array() for safer memory
> > allocation and overflow prevention.
> >
> > Signed-off-by: chuguangqing <chuguangqing@inspur.com>
> > ---
> >  drivers/net/ovpn/crypto_aead.c | 8 ++++----
> >  1 file changed, 4 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/net/ovpn/crypto_aead.c
> > b/drivers/net/ovpn/crypto_aead.c index 2cca759feffa..8274c3ae8d0b
> > 100644
> > --- a/drivers/net/ovpn/crypto_aead.c
> > +++ b/drivers/net/ovpn/crypto_aead.c
> > @@ -72,8 +72,8 @@ int ovpn_aead_encrypt(struct ovpn_peer *peer, struct
> ovpn_crypto_key_slot *ks,
> >  		return -ENOSPC;
> >
> >  	/* sg may be required by async crypto */
> > -	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
> > -				       (nfrags + 2), GFP_ATOMIC);
> > +	ovpn_skb_cb(skb)->sg = kmalloc_array((nfrags + 2),
> sizeof(*ovpn_skb_cb(skb)->sg),
> > +					     GFP_ATOMIC);
> 
> I think this could benefit from:
> a) Removal of unnecessary parentheses
> b) Line wrapping to 80 columns wide or less,
>    as is still preferred for Networking code
> 
> (Completely untested!)
> 
> 	ovpn_skb_cb(skb)->sg = kmalloc_array(nfrags + 2,
> 					     sizeof(*ovpn_skb_cb(skb)->sg),
> 					     GFP_ATOMIC);
> 
> >  	if (unlikely(!ovpn_skb_cb(skb)->sg))
> >  		return -ENOMEM;
> >
> > @@ -185,8 +185,8 @@ int ovpn_aead_decrypt(struct ovpn_peer *peer,
> struct ovpn_crypto_key_slot *ks,
> >  		return -ENOSPC;
> >
> >  	/* sg may be required by async crypto */
> > -	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
> > -				       (nfrags + 2), GFP_ATOMIC);
> > +	ovpn_skb_cb(skb)->sg = kmalloc_array((nfrags + 2),
> sizeof(*ovpn_skb_cb(skb)->sg),
> > +					     GFP_ATOMIC);
> 
> Likewise here.
> 
> >  	if (unlikely(!ovpn_skb_cb(skb)->sg))
> >  		return -ENOMEM;
> >
> 
> --
> pw-bot: changes-requested


chuguangqing (1):
  ovpn: use kmalloc_array() for array space allocation

 drivers/net/ovpn/crypto_aead.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

-- 
2.43.5

[PATCH v2 1/1] ovpn: use kmalloc_array() for array space allocation

[chuguangqing]

Replace kmalloc(size * sizeof) with kmalloc_array() for safer memory
allocation and overflow prevention.

Signed-off-by: chuguangqing <chuguangqing@inspur.com>
---
 drivers/net/ovpn/crypto_aead.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ovpn/crypto_aead.c b/drivers/net/ovpn/crypto_aead.c
index 2cca759feffa..da0ce4f348e6 100644
--- a/drivers/net/ovpn/crypto_aead.c
+++ b/drivers/net/ovpn/crypto_aead.c
@@ -72,8 +72,9 @@ int ovpn_aead_encrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
 		return -ENOSPC;
 
 	/* sg may be required by async crypto */
-	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
-				       (nfrags + 2), GFP_ATOMIC);
+	ovpn_skb_cb(skb)->sg = kmalloc_array(nfrags + 2,
+					     sizeof(*ovpn_skb_cb(skb)->sg),
+					     GFP_ATOMIC);
 	if (unlikely(!ovpn_skb_cb(skb)->sg))
 		return -ENOMEM;
 
@@ -185,8 +186,9 @@ int ovpn_aead_decrypt(struct ovpn_peer *peer, struct ovpn_crypto_key_slot *ks,
 		return -ENOSPC;
 
 	/* sg may be required by async crypto */
-	ovpn_skb_cb(skb)->sg = kmalloc(sizeof(*ovpn_skb_cb(skb)->sg) *
-				       (nfrags + 2), GFP_ATOMIC);
+	ovpn_skb_cb(skb)->sg = kmalloc_array(nfrags + 2,
+					     sizeof(*ovpn_skb_cb(skb)->sg),
+					     GFP_ATOMIC);
 	if (unlikely(!ovpn_skb_cb(skb)->sg))
 		return -ENOMEM;
 
-- 
2.43.5

Re: [PATCH v2 1/1] ovpn: use kmalloc_array() for array space allocation

[Markus Elfring]

> Signed-off-by: chuguangqing <chuguangqing@inspur.com>

Would the personal name usually deviate a bit from the email identifier
according to the Developer's Certificate of Origin?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.17-rc4#n436

Regards,
Markus

Re: [PATCH v2 1/1] ovpn: use kmalloc_array() for array space allocation

[Antonio Quartulli]

Hello and thanks a lot for contribution!

This is indeed a nice improvement, however, we've been working on 
changing this part of the code.

We already have a pending commit in the queue (to be sent to net-next 
soonish):

https://github.com/mandelbitdev/ovpn-net-next/commit/9b62844193de705502fdf4a693dfe0f6c7d94f13

With the mentioned patch, the line you are changing does not exist 
anymore, therefore I don't think it makes sense to merge your patch 
first and then removing it again in the next commit.

Sorry about this!

Please feel free to stick around and submit more patches in the future!

Regards,


-- 
Antonio Quartulli
OpenVPN Inc.

Re: [PATCH v2 1/1] ovpn: use kmalloc_array() for array space allocation

[Gary Chu(楚光庆)]

[-- Attachment #1: Type: text/plain, Size: 804 bytes --]

Hi Markus，
	First of all, thank you for your reminder. According to your suggestion, the format "Chu Guangqing" should be used. However, in line with our company's signature conventions and my previous contributions to the kernel community, I have been using "chuguangqing". Therefore, I have to continue using this signature. The signature should not be changed frequently.

> 
> > Signed-off-by: chuguangqing <chuguangqing@inspur.com>
> 
> Would the personal name usually deviate a bit from the email identifier
> according to the Developer's Certificate of Origin?
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Docume
> ntation/process/submitting-patches.rst?h=v6.17-rc4#n436
> 
> Regards,
> Markus

Best regards,
Chu Guangqing
<chuguangqing@inspur.com>

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 3931 bytes --]

Re: [PATCH v2 1/1] ovpn: use kmalloc_array() for array space allocation

[chuguangqing]

Hi Markus，
	First of all, thank you for your reminder. According to your suggestion, the format "Chu Guangqing" should be used. However, in line with our company's signature conventions and my previous contributions to the kernel community, I have been using "chuguangqing". Therefore, I have to continue using this signature. The signature should not be changed frequently.

> 
> > Signed-off-by: chuguangqing <chuguangqing@inspur.com>
> 
> Would the personal name usually deviate a bit from the email identifier
> according to the Developer's Certificate of Origin?
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Docume
> ntation/process/submitting-patches.rst?h=v6.17-rc4#n436
> 
> Regards,
> Markus

Best regards,
Chu Guangqing
<chuguangqing@inspur.com>

Re: [v2] ovpn: use kmalloc_array() for array space allocation

[Markus Elfring]

> According to your suggestion, the format "Chu Guangqing" should be used.
…

Would you like to acknowledge that your personal name is actually different from
the shown email identifier?


…> The signature should not be changed frequently.

I imagine that some Linux contributors would appreciate an appropriate Developer's Certificate of Origin.
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.17-rc4#n436

Regards,
Markus