RE: sock_buff is NULL !

["Jeff Haran" <jharan@xxxxxxxxxxxxxx>]

 

 

From: Kesava Srinivas [mailto:vunnavafuture@xxxxxxxxx]
Sent: Monday, November 21, 2011 11:32 AM
To: Jeff Haran
Cc: kernelnewbies@xxxxxxxxxxxxxxxxx
Subject: Re: sock_buff is NULL !

 

Thanks Jeff for the reply. Yes; after changing the double pointer to a Normal Pointer., I can see that ., "if" block was skipped.

-Thnx,
VKS

On Tue, Nov 22, 2011 at 12:50 AM, Jeff Haran <jharan@xxxxxxxxxxxxxx> wrote:

 

 

From: kernelnewbies-bounces@xxxxxxxxxxxxxxxxx [mailto:kernelnewbies-bounces@xxxxxxxxxxxxxxxxx] On Behalf Of Kesava Srinivas
Sent: Monday, November 21, 2011 11:03 AM
To: kernelnewbies@xxxxxxxxxxxxxxxxx
Subject: sock_buff is NULL !

 

HI Team,

Trying to make this code work; but some how it is entering the if condition mentioned below.sock_buff is still a NULL Pointer.I can see that Function was registered with the Ipv4 Hooks. Any ideas/scenarios where the sock_buff remains as NULL Pointer??? Testing this code by trying to do ssh with this machine.

Any Help is really appreciated !!

#########
static struct nf_hook_ops netfilter_ops;
struct sk_buff *sock_buff;

unsigned int main_hook(unsigned int hooknum,struct sk_buff **skb,const struct net_device *in,const struct net_device *out,int (*okfn)(struct sk_buff*)) {
 sock_buff = *skb;
        if(!sock_buff){
           printk(KERN_ERR "Entered unexpected loop \n");
               return NF_ACCEPT;
         }/* Check for sk_buff NULL */
}
int init_module()
{
        netfilter_ops.hook              =       (nf_hookfn *) main_hook;
        netfilter_ops.hooknum           =       NF_INET_PRE_ROUTING;
        netfilter_ops.pf                =       PF_INET;
        netfilter_ops.priority          =       NF_IP_PRI_FIRST;
        nf_register_hook(&netfilter_ops);
                                                                                                                                                     
}
######################


-Thnx,
VKS

 

Which kernel version are you using? The function prototype of a netfilter hook does vary from kernel version to kernel version, but in the kernels I’ve seen the skb parameter is a typed sk_buff *skb, not sk_buff **skb. In other words, you get passed a pointer to the sk_buff, not the address of a pointer to an sk_buff.

 

Jeff Haran

 

--
-VKS
Masters,Computer Science&Engineering.

I am a bit surprised that you didn’t get a compiler warning when you built this. Type checking is your friend. 8^)

Jeff Haran

 

_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@xxxxxxxxxxxxxxxxx
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies