netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH net] net: dsa: skb_put_padto() already frees nskb
@ 2017-08-21 19:41 Florian Fainelli
  2017-08-22 18:27 ` David Miller
  0 siblings, 1 reply; 7+ messages in thread
From: Florian Fainelli @ 2017-08-21 19:41 UTC (permalink / raw)
  To: netdev; +Cc: davem, andrew, vivien.didtelot, woojung.huh, Florian Fainelli

skb_put_padto() already frees the passed sk_buff reference upon error,
so calling kfree_skb() on it again is not necessary.

Detected by CoverityScan, CID#1416687 ("USE_AFTER_FREE")

Fixes: e71cb9e00922 ("net: dsa: ksz: fix skb freeing")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
 net/dsa/tag_ksz.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
index de66ca8e6201..107172c82107 100644
--- a/net/dsa/tag_ksz.c
+++ b/net/dsa/tag_ksz.c
@@ -60,10 +60,8 @@ static struct sk_buff *ksz_xmit(struct sk_buff *skb, struct net_device *dev)
 					 skb_transport_header(skb) - skb->head);
 		skb_copy_and_csum_dev(skb, skb_put(nskb, skb->len));
 
-		if (skb_put_padto(nskb, nskb->len + padlen)) {
-			kfree_skb(nskb);
+		if (skb_put_padto(nskb, nskb->len + padlen))
 			return NULL;
-		}
 
 		kfree_skb(skb);
 	}
-- 
2.9.3

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [PATCH net] net: dsa: skb_put_padto() already frees nskb
@ 2017-08-21 19:41 Florian Fainelli
  2017-08-21 19:47 ` Andrew Lunn
  2017-08-21 22:15 ` Woojung.Huh
  0 siblings, 2 replies; 7+ messages in thread
From: Florian Fainelli @ 2017-08-21 19:41 UTC (permalink / raw)
  To: netdev; +Cc: davem, andrew, vivien.didelot, woojung.huh, Florian Fainelli

skb_put_padto() already frees the passed sk_buff reference upon error,
so calling kfree_skb() on it again is not necessary.

Detected by CoverityScan, CID#1416687 ("USE_AFTER_FREE")

Fixes: e71cb9e00922 ("net: dsa: ksz: fix skb freeing")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
 net/dsa/tag_ksz.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
index de66ca8e6201..107172c82107 100644
--- a/net/dsa/tag_ksz.c
+++ b/net/dsa/tag_ksz.c
@@ -60,10 +60,8 @@ static struct sk_buff *ksz_xmit(struct sk_buff *skb, struct net_device *dev)
 					 skb_transport_header(skb) - skb->head);
 		skb_copy_and_csum_dev(skb, skb_put(nskb, skb->len));
 
-		if (skb_put_padto(nskb, nskb->len + padlen)) {
-			kfree_skb(nskb);
+		if (skb_put_padto(nskb, nskb->len + padlen))
 			return NULL;
-		}
 
 		kfree_skb(skb);
 	}
-- 
2.9.3

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* Re: [PATCH net] net: dsa: skb_put_padto() already frees nskb
  2017-08-21 19:41 Florian Fainelli
@ 2017-08-21 19:47 ` Andrew Lunn
  2017-08-21 22:15 ` Woojung.Huh
  1 sibling, 0 replies; 7+ messages in thread
From: Andrew Lunn @ 2017-08-21 19:47 UTC (permalink / raw)
  To: Florian Fainelli; +Cc: netdev, davem, vivien.didelot, woojung.huh

On Mon, Aug 21, 2017 at 12:41:43PM -0700, Florian Fainelli wrote:
> skb_put_padto() already frees the passed sk_buff reference upon error,
> so calling kfree_skb() on it again is not necessary.
> 
> Detected by CoverityScan, CID#1416687 ("USE_AFTER_FREE")
> 
> Fixes: e71cb9e00922 ("net: dsa: ksz: fix skb freeing")
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

^ permalink raw reply	[flat|nested] 7+ messages in thread

* RE: [PATCH net] net: dsa: skb_put_padto() already frees nskb
  2017-08-21 19:41 Florian Fainelli
  2017-08-21 19:47 ` Andrew Lunn
@ 2017-08-21 22:15 ` Woojung.Huh
  2017-08-21 22:24   ` Florian Fainelli
  1 sibling, 1 reply; 7+ messages in thread
From: Woojung.Huh @ 2017-08-21 22:15 UTC (permalink / raw)
  To: f.fainelli, netdev; +Cc: davem, andrew, vivien.didelot

Florian,

> -----Original Message-----
> From: Florian Fainelli [mailto:f.fainelli@gmail.com]
> Sent: Monday, August 21, 2017 3:42 PM
> To: netdev@vger.kernel.org
> Cc: davem@davemloft.net; andrew@lunn.ch;
> vivien.didelot@savoirfairelinux.com; Woojung Huh - C21699; Florian Fainelli
> Subject: [PATCH net] net: dsa: skb_put_padto() already frees nskb
> 
> skb_put_padto() already frees the passed sk_buff reference upon error,
> so calling kfree_skb() on it again is not necessary.
> 
> Detected by CoverityScan, CID#1416687 ("USE_AFTER_FREE")
> 
> Fixes: e71cb9e00922 ("net: dsa: ksz: fix skb freeing")
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
>  net/dsa/tag_ksz.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
> index de66ca8e6201..107172c82107 100644
> --- a/net/dsa/tag_ksz.c
> +++ b/net/dsa/tag_ksz.c
> @@ -60,10 +60,8 @@ static struct sk_buff *ksz_xmit(struct sk_buff *skb,
> struct net_device *dev)
>  					 skb_transport_header(skb) - skb-
> >head);
>  		skb_copy_and_csum_dev(skb, skb_put(nskb, skb->len));
> 
> -		if (skb_put_padto(nskb, nskb->len + padlen)) {
> -			kfree_skb(nskb);
> +		if (skb_put_padto(nskb, nskb->len + padlen))
>  			return NULL;
> -		}
> 
>  		kfree_skb(skb);
>  	}
> --

Because skb_put_padto() frees skb when it fails,  below lines in e71cb9e00922
("net: dsa: ksz: fix skb freeing") will be an issue to.

	if (skb_tailroom(skb) >= padlen + KSZ_INGRESS_TAG_LEN) {
+		if (skb_put_padto(skb, skb->len + padlen))
+			return NULL;
+

When it fails skb will be freed twice in skb_put_padto() and
caller of dsa_slave_xmit().

Woojung

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH net] net: dsa: skb_put_padto() already frees nskb
  2017-08-21 22:15 ` Woojung.Huh
@ 2017-08-21 22:24   ` Florian Fainelli
  2017-08-22 18:01     ` Woojung.Huh
  0 siblings, 1 reply; 7+ messages in thread
From: Florian Fainelli @ 2017-08-21 22:24 UTC (permalink / raw)
  To: Woojung.Huh, netdev; +Cc: davem, andrew, vivien.didelot

On 08/21/2017 03:15 PM, Woojung.Huh@microchip.com wrote:
> Florian,
> 
>> -----Original Message-----
>> From: Florian Fainelli [mailto:f.fainelli@gmail.com]
>> Sent: Monday, August 21, 2017 3:42 PM
>> To: netdev@vger.kernel.org
>> Cc: davem@davemloft.net; andrew@lunn.ch;
>> vivien.didelot@savoirfairelinux.com; Woojung Huh - C21699; Florian Fainelli
>> Subject: [PATCH net] net: dsa: skb_put_padto() already frees nskb
>>
>> skb_put_padto() already frees the passed sk_buff reference upon error,
>> so calling kfree_skb() on it again is not necessary.
>>
>> Detected by CoverityScan, CID#1416687 ("USE_AFTER_FREE")
>>
>> Fixes: e71cb9e00922 ("net: dsa: ksz: fix skb freeing")
>> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
>> ---
>>  net/dsa/tag_ksz.c | 4 +---
>>  1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/net/dsa/tag_ksz.c b/net/dsa/tag_ksz.c
>> index de66ca8e6201..107172c82107 100644
>> --- a/net/dsa/tag_ksz.c
>> +++ b/net/dsa/tag_ksz.c
>> @@ -60,10 +60,8 @@ static struct sk_buff *ksz_xmit(struct sk_buff *skb,
>> struct net_device *dev)
>>  					 skb_transport_header(skb) - skb-
>>> head);
>>  		skb_copy_and_csum_dev(skb, skb_put(nskb, skb->len));
>>
>> -		if (skb_put_padto(nskb, nskb->len + padlen)) {
>> -			kfree_skb(nskb);
>> +		if (skb_put_padto(nskb, nskb->len + padlen))
>>  			return NULL;
>> -		}
>>
>>  		kfree_skb(skb);
>>  	}
>> --
> 
> Because skb_put_padto() frees skb when it fails,  below lines in e71cb9e00922
> ("net: dsa: ksz: fix skb freeing") will be an issue to.
> 
> 	if (skb_tailroom(skb) >= padlen + KSZ_INGRESS_TAG_LEN) {
> +		if (skb_put_padto(skb, skb->len + padlen))
> +			return NULL;
> +
> 
> When it fails skb will be freed twice in skb_put_padto() and
> caller of dsa_slave_xmit().

You are right, I am not sure what is the best way to fix tag_ksz.c other
than somehow open coding skb_put_padto() minus the freeing on error part?
-- 
Florian

^ permalink raw reply	[flat|nested] 7+ messages in thread

* RE: [PATCH net] net: dsa: skb_put_padto() already frees nskb
  2017-08-21 22:24   ` Florian Fainelli
@ 2017-08-22 18:01     ` Woojung.Huh
  0 siblings, 0 replies; 7+ messages in thread
From: Woojung.Huh @ 2017-08-22 18:01 UTC (permalink / raw)
  To: f.fainelli, netdev; +Cc: davem, andrew, vivien.didelot

> > Because skb_put_padto() frees skb when it fails,  below lines in
> e71cb9e00922
> > ("net: dsa: ksz: fix skb freeing") will be an issue to.
> >
> > 	if (skb_tailroom(skb) >= padlen + KSZ_INGRESS_TAG_LEN) {
> > +		if (skb_put_padto(skb, skb->len + padlen))
> > +			return NULL;
> > +
> >
> > When it fails skb will be freed twice in skb_put_padto() and
> > caller of dsa_slave_xmit().
> 
> You are right, I am not sure what is the best way to fix tag_ksz.c other
> than somehow open coding skb_put_padto() minus the freeing on error part?

Agree. May need to go back to 1st submission without skb_put_padto().
Will check more if any other ideas.

Thanks.
Woojung

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH net] net: dsa: skb_put_padto() already frees nskb
  2017-08-21 19:41 [PATCH net] net: dsa: skb_put_padto() already frees nskb Florian Fainelli
@ 2017-08-22 18:27 ` David Miller
  0 siblings, 0 replies; 7+ messages in thread
From: David Miller @ 2017-08-22 18:27 UTC (permalink / raw)
  To: f.fainelli; +Cc: netdev, andrew, vivien.didtelot, woojung.huh

From: Florian Fainelli <f.fainelli@gmail.com>
Date: Mon, 21 Aug 2017 12:41:31 -0700

> skb_put_padto() already frees the passed sk_buff reference upon error,
> so calling kfree_skb() on it again is not necessary.
> 
> Detected by CoverityScan, CID#1416687 ("USE_AFTER_FREE")
> 
> Fixes: e71cb9e00922 ("net: dsa: ksz: fix skb freeing")
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>

I know it seems like a lot of work, but with appropriate wrappers we
can control the freeing that skb_pad() does at the deepest part of
this call chain.

int __skb_pad(struct sk_buff *skb, int pad, bool free_skb_on_err);
static inline int skb_pad(struct sk_buff *skb, int pad)
{
	return __skb_pad(skb, pad, true);
}

static inline int __skb_put_padto(struct sk_buff *skb, unsigned int len,
				  bool free_skb_on_err)
{
	unsigned int size = skb->len;

	if (unlikely(size < len)) {
		len -= size;
		if (__skb_pad(skb, len, free_skb_on_err))
			return -ENOMEM;
		__skb_put(skb, len);
	}
	return 0;
}

static inline int skb_put_padto(struct sk_buff *skb, unsigned int len)
{
	return __skb_put_padto(skb, len, true);
}

And then here in the ksz_xmit() code, invoke __skb_put_padto() with the
boolean set appropriately.

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2017-08-22 18:27 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-08-21 19:41 [PATCH net] net: dsa: skb_put_padto() already frees nskb Florian Fainelli
2017-08-22 18:27 ` David Miller
  -- strict thread matches above, loose matches on Subject: below --
2017-08-21 19:41 Florian Fainelli
2017-08-21 19:47 ` Andrew Lunn
2017-08-21 22:15 ` Woojung.Huh
2017-08-21 22:24   ` Florian Fainelli
2017-08-22 18:01     ` Woojung.Huh

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).