diff mbox series

[v2,net-next] net: fix GSO for SG-enabled devices.

Message ID 861947c2d2d087db82af93c21920ce8147d15490.1611074818.git.pabeni@redhat.com
State New
Headers show
Series [v2,net-next] net: fix GSO for SG-enabled devices. | expand

Commit Message

Paolo Abeni Jan. 19, 2021, 4:56 p.m. UTC
The commit dbd50f238dec ("net: move the hsize check to the else
block in skb_segment") introduced a data corruption for devices
supporting scatter-gather.

The problem boils down to signed/unsigned comparison given
unexpected results: if signed 'hsize' is negative, it will be
considered greater than a positive 'len', which is unsigned.

This commit addresses resorting to the old checks order, so that
'hsize' never has a negative value when compared with 'len'.

v1 -> v2:
 - reorder hsize checks instead of explicit cast (Alex)

Bisected-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Fixes: dbd50f238dec ("net: move the hsize check to the else block in skb_segment")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 net/core/skbuff.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Xin Long Jan. 20, 2021, 2:17 a.m. UTC | #1
On Wed, Jan 20, 2021 at 12:57 AM Paolo Abeni <pabeni@redhat.com> wrote:
>

> The commit dbd50f238dec ("net: move the hsize check to the else

> block in skb_segment") introduced a data corruption for devices

> supporting scatter-gather.

>

> The problem boils down to signed/unsigned comparison given

> unexpected results: if signed 'hsize' is negative, it will be

> considered greater than a positive 'len', which is unsigned.

>

> This commit addresses resorting to the old checks order, so that

> 'hsize' never has a negative value when compared with 'len'.

>

> v1 -> v2:

>  - reorder hsize checks instead of explicit cast (Alex)

>

> Bisected-by: Matthieu Baerts <matthieu.baerts@tessares.net>

> Fixes: dbd50f238dec ("net: move the hsize check to the else block in skb_segment")

> Signed-off-by: Paolo Abeni <pabeni@redhat.com>

Reviewed-by: Xin Long <lucien.xin@gmail.com>

> ---

>  net/core/skbuff.c | 4 ++--

>  1 file changed, 2 insertions(+), 2 deletions(-)

>

> diff --git a/net/core/skbuff.c b/net/core/skbuff.c

> index e835193cabcc3..cf2c4dcf42579 100644

> --- a/net/core/skbuff.c

> +++ b/net/core/skbuff.c

> @@ -3938,10 +3938,10 @@ struct sk_buff *skb_segment(struct sk_buff *head_skb,

>                         skb_release_head_state(nskb);

>                         __skb_push(nskb, doffset);

>                 } else {

> +                       if (hsize < 0)

> +                               hsize = 0;

>                         if (hsize > len || !sg)

>                                 hsize = len;

> -                       else if (hsize < 0)

> -                               hsize = 0;

>

>                         nskb = __alloc_skb(hsize + doffset + headroom,

>                                            GFP_ATOMIC, skb_alloc_rx_flag(head_skb),

> --

> 2.26.2

>

Reviewed-by: Xin Long <lucien.xin@gmail.com>
Jakub Kicinski Jan. 20, 2021, 2:43 a.m. UTC | #2
On Wed, 20 Jan 2021 10:17:44 +0800 Xin Long wrote:
> On Wed, Jan 20, 2021 at 12:57 AM Paolo Abeni <pabeni@redhat.com> wrote:

> >

> > The commit dbd50f238dec ("net: move the hsize check to the else

> > block in skb_segment") introduced a data corruption for devices

> > supporting scatter-gather.

> >

> > The problem boils down to signed/unsigned comparison given

> > unexpected results: if signed 'hsize' is negative, it will be

> > considered greater than a positive 'len', which is unsigned.

> >

> > This commit addresses resorting to the old checks order, so that

> > 'hsize' never has a negative value when compared with 'len'.

> >

> > v1 -> v2:

> >  - reorder hsize checks instead of explicit cast (Alex)

> >

> > Bisected-by: Matthieu Baerts <matthieu.baerts@tessares.net>

> > Fixes: dbd50f238dec ("net: move the hsize check to the else block in skb_segment")

> > Signed-off-by: Paolo Abeni <pabeni@redhat.com>  

> Reviewed-by: Xin Long <lucien.xin@gmail.com>


I'm hitting this as well, so applied, thanks!

> > ---

> Reviewed-by: Xin Long <lucien.xin@gmail.com>


One review tag is enough ;) apparently patchwork doesn't know to dedup
them :S
diff mbox series

Patch

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index e835193cabcc3..cf2c4dcf42579 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -3938,10 +3938,10 @@  struct sk_buff *skb_segment(struct sk_buff *head_skb,
 			skb_release_head_state(nskb);
 			__skb_push(nskb, doffset);
 		} else {
+			if (hsize < 0)
+				hsize = 0;
 			if (hsize > len || !sg)
 				hsize = len;
-			else if (hsize < 0)
-				hsize = 0;
 
 			nskb = __alloc_skb(hsize + doffset + headroom,
 					   GFP_ATOMIC, skb_alloc_rx_flag(head_skb),