[PATCH net] net: Clear old fragment checksum value in napi_get_frags

[PATCH net] net: Clear old fragment checksum value in napi_get_frags

[Mohammad Heib]

In certain cases, napi_get_frags() returns an skb that points to an old
received fragment, This skb may have its skb->ip_summed, csum, and other
fields set from previous fragment handling.

Some network drivers set skb->ip_summed to either CHECKSUM_COMPLETE or
CHECKSUM_UNNECESSARY when getting skb from napi_get_frags(), while
others only set skb->ip_summed when RX checksum offload is enabled on
the device, and do not set any value for skb->ip_summed when hardware
checksum offload is disabled, assuming that the skb->ip_summed
initiated to zero by napi_get_frags.

This inconsistency sometimes leads to checksum validation issues in the
upper layers of the network stack.

To resolve this, this patch clears the skb->ip_summed value for each skb
returned by napi_get_frags(), ensuring that the caller is responsible
for setting the correct checksum status. This eliminates potential
checksum validation issues caused by improper handling of
skb->ip_summed.

Fixes: 76620aafd66f ("gro: New frags interface to avoid copying shinfo")
Signed-off-by: Mohammad Heib <mheib@redhat.com>
---
 net/core/gro.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/core/gro.c b/net/core/gro.c
index 78b320b63174..e98007d8f26f 100644
--- a/net/core/gro.c
+++ b/net/core/gro.c
@@ -675,6 +675,8 @@ struct sk_buff *napi_get_frags(struct napi_struct *napi)
 			napi->skb = skb;
 			skb_mark_napi_id(skb, napi);
 		}
+	} else {
+		skb->ip_summed = CHECKSUM_NONE;
 	}
 	return skb;
 }
-- 
2.48.1

Re: [PATCH net] net: Clear old fragment checksum value in napi_get_frags

[Jakub Kicinski]

On Fri, 21 Feb 2025 18:14:05 +0200 Mohammad Heib wrote:
> In certain cases, napi_get_frags() returns an skb that points to an old
> received fragment, This skb may have its skb->ip_summed, csum, and other
> fields set from previous fragment handling.
> 
> Some network drivers set skb->ip_summed to either CHECKSUM_COMPLETE or
> CHECKSUM_UNNECESSARY when getting skb from napi_get_frags(), while
> others only set skb->ip_summed when RX checksum offload is enabled on
> the device, and do not set any value for skb->ip_summed when hardware
> checksum offload is disabled, assuming that the skb->ip_summed
> initiated to zero by napi_get_frags.
> 
> This inconsistency sometimes leads to checksum validation issues in the
> upper layers of the network stack.
> 
> To resolve this, this patch clears the skb->ip_summed value for each skb
> returned by napi_get_frags(), ensuring that the caller is responsible
> for setting the correct checksum status. This eliminates potential
> checksum validation issues caused by improper handling of
> skb->ip_summed.

Could you give an example of a driver where this may happen?
Otherwise the commit message reads too hypothetical.

> Fixes: 76620aafd66f ("gro: New frags interface to avoid copying shinfo")
> Signed-off-by: Mohammad Heib <mheib@redhat.com>
> ---
>  net/core/gro.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/net/core/gro.c b/net/core/gro.c
> index 78b320b63174..e98007d8f26f 100644
> --- a/net/core/gro.c
> +++ b/net/core/gro.c
> @@ -675,6 +675,8 @@ struct sk_buff *napi_get_frags(struct napi_struct *napi)
>  			napi->skb = skb;
>  			skb_mark_napi_id(skb, napi);
>  		}
> +	} else {
> +		skb->ip_summed = CHECKSUM_NONE;
>  	}
>  	return skb;
>  }

I think this belongs in napi_reuse_skb(), doesn't it ?

Please make sure you CC maintainers on v2, especially Eric.
-- 
pw-bot: cr

Re: [PATCH net] net: Clear old fragment checksum value in napi_get_frags

[Mohammad Heib]

On Mon, Feb 24, 2025 at 02:34:30PM -0800, Jakub Kicinski wrote:
> On Fri, 21 Feb 2025 18:14:05 +0200 Mohammad Heib wrote:
> > In certain cases, napi_get_frags() returns an skb that points to an old
> > received fragment, This skb may have its skb->ip_summed, csum, and other
> > fields set from previous fragment handling.
> > 
> > Some network drivers set skb->ip_summed to either CHECKSUM_COMPLETE or
> > CHECKSUM_UNNECESSARY when getting skb from napi_get_frags(), while
> > others only set skb->ip_summed when RX checksum offload is enabled on
> > the device, and do not set any value for skb->ip_summed when hardware
> > checksum offload is disabled, assuming that the skb->ip_summed
> > initiated to zero by napi_get_frags.
> > 
> > This inconsistency sometimes leads to checksum validation issues in the
> > upper layers of the network stack.
> > 
> > To resolve this, this patch clears the skb->ip_summed value for each skb
> > returned by napi_get_frags(), ensuring that the caller is responsible
> > for setting the correct checksum status. This eliminates potential
> > checksum validation issues caused by improper handling of
> > skb->ip_summed.
> 
> Could you give an example of a driver where this may happen?
> Otherwise the commit message reads too hypothetical.
> 
Hi Jakub,
Thank you so much for your review.
all the comments were addressed in v3.

Thanks,
> > Fixes: 76620aafd66f ("gro: New frags interface to avoid copying shinfo")
> > Signed-off-by: Mohammad Heib <mheib@redhat.com>
> > ---
> >  net/core/gro.c | 2 ++
> >  1 file changed, 2 insertions(+)
> > 
> > diff --git a/net/core/gro.c b/net/core/gro.c
> > index 78b320b63174..e98007d8f26f 100644
> > --- a/net/core/gro.c
> > +++ b/net/core/gro.c
> > @@ -675,6 +675,8 @@ struct sk_buff *napi_get_frags(struct napi_struct *napi)
> >  			napi->skb = skb;
> >  			skb_mark_napi_id(skb, napi);
> >  		}
> > +	} else {
> > +		skb->ip_summed = CHECKSUM_NONE;
> >  	}
> >  	return skb;
> >  }
> 
> I think this belongs in napi_reuse_skb(), doesn't it ?
> 
> Please make sure you CC maintainers on v2, especially Eric.
> -- 
> pw-bot: cr
>