hi, should these code is a problem in nfs system clnt.c?

hi, should these code is a problem in nfs system clnt.c?

[linuxer linuxer]

Hi, everyone: 
    I am a newbie, if my question waste your time, I
am sorry for that. 

    In clnt.c file ,call_timeout function:    
    I suggest the code that judge whether the network
link status is down should be added, won't they? 
    I tested it with one Ethernet netcard.

static void
call_timeout(struct rpc_task *task)
{
	struct rpc_clnt	*clnt = task->tk_client;

+       struct netdev * dev;
+       if ((dev = __dev_get_by_name("eth0")) == 
+        NULL){
+		rpc_exit(task, -ENOTCONN);
+		return;
+    	}
+	else{
+   		 if (!netif_carrier_ok(dev)){
+  			rpc_exit(task, -ENOTCONN);
+			return;
+  		 }
+     	}

	if (xprt_adjust_timeout(task->tk_rqstp) == 0) {
		dprintk("RPC: %4d call_timeout (minor)\n",
task->tk_pid);
		goto retry;
	}

	dprintk("RPC: %4d call_timeout (major)\n",
task->tk_pid);
	task->tk_timeouts++;

	if (RPC_IS_SOFT(task)) {
		printk(KERN_NOTICE "%s: server %s not responding,
timed out\n",
				clnt->cl_protname, clnt->cl_server);
		rpc_exit(task, -EIO);
		return;
	}

	if (!(task->tk_flags & RPC_CALL_MAJORSEEN)) {
		task->tk_flags |= RPC_CALL_MAJORSEEN;
		printk(KERN_NOTICE "%s: server %s not responding,
still trying\n",
			clnt->cl_protname, clnt->cl_server);
	}
	rpc_force_rebind(clnt);

retry:
	clnt->cl_stats->rpcretrans++;
	task->tk_action = call_bind;
	task->tk_status = 0;
}



 
____________________________________________________________________________________
Yahoo! Music Unlimited
Access over 1 million songs.
http://music.yahoo.com/unlimited

Re: hi, should these code is a problem in nfs system clnt.c?

[Peter Staubach]

linuxer linuxer wrote:
> Hi, everyone: 
>     I am a newbie, if my question waste your time, I
> am sorry for that. 
>
>     In clnt.c file ,call_timeout function:    
>     I suggest the code that judge whether the network
> link status is down should be added, won't they? 
>     I tested it with one Ethernet netcard.
>
>   

This would not be a good idea.  What happens if the name of the
interface used on the system is not "eth0"?

Also, IP packets can be routed out of any available interface,
so just because one interface (eth0) is down, doesn't mean
that the entire system is networkless.

    Thanx...

       ps

> static void
> call_timeout(struct rpc_task *task)
> {
> 	struct rpc_clnt	*clnt = task->tk_client;
>
> +       struct netdev * dev;
> +       if ((dev = __dev_get_by_name("eth0")) == 
> +        NULL){
> +		rpc_exit(task, -ENOTCONN);
> +		return;
> +    	}
> +	else{
> +   		 if (!netif_carrier_ok(dev)){
> +  			rpc_exit(task, -ENOTCONN);
> +			return;
> +  		 }
> +     	}
>
> 	if (xprt_adjust_timeout(task->tk_rqstp) == 0) {
> 		dprintk("RPC: %4d call_timeout (minor)\n",
> task->tk_pid);
> 		goto retry;
> 	}
>
> 	dprintk("RPC: %4d call_timeout (major)\n",
> task->tk_pid);
> 	task->tk_timeouts++;
>
> 	if (RPC_IS_SOFT(task)) {
> 		printk(KERN_NOTICE "%s: server %s not responding,
> timed out\n",
> 				clnt->cl_protname, clnt->cl_server);
> 		rpc_exit(task, -EIO);
> 		return;
> 	}
>
> 	if (!(task->tk_flags & RPC_CALL_MAJORSEEN)) {
> 		task->tk_flags |= RPC_CALL_MAJORSEEN;
> 		printk(KERN_NOTICE "%s: server %s not responding,
> still trying\n",
> 			clnt->cl_protname, clnt->cl_server);
> 	}
> 	rpc_force_rebind(clnt);
>
> retry:
> 	clnt->cl_stats->rpcretrans++;
> 	task->tk_action = call_bind;
> 	task->tk_status = 0;
> }
>
>
>
>  
> ____________________________________________________________________________________
> Yahoo! Music Unlimited
> Access over 1 million songs.
> http://music.yahoo.com/unlimited
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>   

Re: hi, should these code is a problem in nfs system clnt.c?

[Trond Myklebust]

On Tue, 2006-12-12 at 09:00 -0500, Peter Staubach wrote:
> linuxer linuxer wrote:
> > Hi, everyone: 
> >     I am a newbie, if my question waste your time, I
> > am sorry for that. 
> >
> >     In clnt.c file ,call_timeout function:    
> >     I suggest the code that judge whether the network
> > link status is down should be added, won't they? 
> >     I tested it with one Ethernet netcard.
> >
> >   
> 
> This would not be a good idea.  What happens if the name of the
> interface used on the system is not "eth0"?
> 
> Also, IP packets can be routed out of any available interface,
> so just because one interface (eth0) is down, doesn't mean
> that the entire system is networkless.

Right. Besides, it is also an obvious layering violation. Idea NAKed.

Trond