[PATCH net-next] bpf: pass sk to helper functions

[PATCH net-next] bpf: pass sk to helper functions

[Willem de Bruijn]

From: Willem de Bruijn <willemb@google.com>

BPF helper functions access socket fields through skb->sk. This is not
set in ingress cgroup and socket filters. The association is only made
in skb_set_owner_r once the filter has accepted the packet. Sk is
available as socket lookup has taken place.

Temporarily set skb->sk to sk in these cases.

Signed-off-by: Willem de Bruijn <willemb@google.com>
---
 kernel/bpf/cgroup.c | 4 +++-
 net/core/filter.c   | 6 +++++-
 2 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
index da0f53690295..a2c387e433a4 100644
--- a/kernel/bpf/cgroup.c
+++ b/kernel/bpf/cgroup.c
@@ -154,7 +154,7 @@ int __cgroup_bpf_update(struct cgroup *cgrp, struct cgroup *parent,
 
 /**
  * __cgroup_bpf_run_filter_skb() - Run a program for packet filtering
- * @sk: The socken sending or receiving traffic
+ * @sk: The socket sending or receiving traffic
  * @skb: The skb that is being sent or received
  * @type: The type of program to be exectuted
  *
@@ -190,9 +190,11 @@ int __cgroup_bpf_run_filter_skb(struct sock *sk,
 	if (prog) {
 		unsigned int offset = skb->data - skb_network_header(skb);
 
+		swap(skb->sk, sk);
 		__skb_push(skb, offset);
 		ret = bpf_prog_run_save_cb(prog, skb) == 1 ? 0 : -EPERM;
 		__skb_pull(skb, offset);
+		swap(skb->sk, sk);
 	}
 
 	rcu_read_unlock();
diff --git a/net/core/filter.c b/net/core/filter.c
index 15e9a81ffebe..979cb508302a 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -92,8 +92,12 @@ int sk_filter_trim_cap(struct sock *sk, struct sk_buff *skb, unsigned int cap)
 	rcu_read_lock();
 	filter = rcu_dereference(sk->sk_filter);
 	if (filter) {
-		unsigned int pkt_len = bpf_prog_run_save_cb(filter->prog, skb);
+		unsigned int pkt_len;
+
+		swap(skb->sk, sk);
+		pkt_len = bpf_prog_run_save_cb(filter->prog, skb);
 		err = pkt_len ? pskb_trim(skb, max(cap, pkt_len)) : -EPERM;
+		swap(skb->sk, sk);
 	}
 	rcu_read_unlock();
 
-- 
2.12.2.715.g7642488e1d-goog

Re: [PATCH net-next] bpf: pass sk to helper functions

[Eric Dumazet]

On Tue, 2017-04-11 at 12:22 -0400, Willem de Bruijn wrote:
> From: Willem de Bruijn <willemb@google.com>
> 
> BPF helper functions access socket fields through skb->sk. This is not
> set in ingress cgroup and socket filters. The association is only made
> in skb_set_owner_r once the filter has accepted the packet. Sk is
> available as socket lookup has taken place.
> 
> Temporarily set skb->sk to sk in these cases.
> 
> Signed-off-by: Willem de Bruijn <willemb@google.com>
> ---
>  kernel/bpf/cgroup.c | 4 +++-
>  net/core/filter.c   | 6 +++++-
>  2 files changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
> index da0f53690295..a2c387e433a4 100644
> --- a/kernel/bpf/cgroup.c
> +++ b/kernel/bpf/cgroup.c
> @@ -154,7 +154,7 @@ int __cgroup_bpf_update(struct cgroup *cgrp, struct cgroup *parent,
>  
>  /**
>   * __cgroup_bpf_run_filter_skb() - Run a program for packet filtering
> - * @sk: The socken sending or receiving traffic
> + * @sk: The socket sending or receiving traffic
>   * @skb: The skb that is being sent or received
>   * @type: The type of program to be exectuted
>   *
> @@ -190,9 +190,11 @@ int __cgroup_bpf_run_filter_skb(struct sock *sk,
>  	if (prog) {
>  		unsigned int offset = skb->data - skb_network_header(skb);
>  

It would be nice to not touch sk. swap() is a bit overkill here I think.

	struct sock *save_sk = skb->sk;

	skb->sk = sk;


> +		swap(skb->sk, sk);
>  		__skb_push(skb, offset);
>  		ret = bpf_prog_run_save_cb(prog, skb) == 1 ? 0 : -EPERM;
>  		__skb_pull(skb, offset);
> +		swap(skb->sk, sk);

	skb->sk = save_sk;

>  	}
>  

Re: [PATCH net-next] bpf: pass sk to helper functions

[Willem de Bruijn]

On Tue, Apr 11, 2017 at 12:51 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Tue, 2017-04-11 at 12:22 -0400, Willem de Bruijn wrote:
>> From: Willem de Bruijn <willemb@google.com>
>>
>> BPF helper functions access socket fields through skb->sk. This is not
>> set in ingress cgroup and socket filters. The association is only made
>> in skb_set_owner_r once the filter has accepted the packet. Sk is
>> available as socket lookup has taken place.
>>
>> Temporarily set skb->sk to sk in these cases.
>>
>> Signed-off-by: Willem de Bruijn <willemb@google.com>
>> ---
>>  kernel/bpf/cgroup.c | 4 +++-
>>  net/core/filter.c   | 6 +++++-
>>  2 files changed, 8 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
>> index da0f53690295..a2c387e433a4 100644
>> --- a/kernel/bpf/cgroup.c
>> +++ b/kernel/bpf/cgroup.c
>> @@ -154,7 +154,7 @@ int __cgroup_bpf_update(struct cgroup *cgrp, struct cgroup *parent,
>>
>>  /**
>>   * __cgroup_bpf_run_filter_skb() - Run a program for packet filtering
>> - * @sk: The socken sending or receiving traffic
>> + * @sk: The socket sending or receiving traffic
>>   * @skb: The skb that is being sent or received
>>   * @type: The type of program to be exectuted
>>   *
>> @@ -190,9 +190,11 @@ int __cgroup_bpf_run_filter_skb(struct sock *sk,
>>       if (prog) {
>>               unsigned int offset = skb->data - skb_network_header(skb);
>>
>
> It would be nice to not touch sk. swap() is a bit overkill here I think.
>
>         struct sock *save_sk = skb->sk;
>
>         skb->sk = sk;
>
>
>> +             swap(skb->sk, sk);
>>               __skb_push(skb, offset);
>>               ret = bpf_prog_run_save_cb(prog, skb) == 1 ? 0 : -EPERM;
>>               __skb_pull(skb, offset);
>> +             swap(skb->sk, sk);
>
>         skb->sk = save_sk;
>
>>       }

Thanks. I'll send a v2.

>>
>
>
>