[patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c

All of lore.kernel.org
 help / color / mirror / Atom feed

* [patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c
       [not found] <20071210011741.039692330@gmail.com>
@ 2007-12-10  1:17 ` Kevin Winchester
  2007-12-12  0:44   ` Andrew Morton
  0 siblings, 1 reply; 4+ messages in thread
From: Kevin Winchester @ 2007-12-10  1:17 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Ingo Molnar, Per Liden, Jon Maloy, Allan Stephens,
	tipc-discussion, linux-kernel, Kevin Winchester

[-- Attachment #1: tipc-socket-sem-to-mutex.patch --]
[-- Type: text/plain, Size: 6700 bytes --]

To: Andrew Morton <akpm@linux-foundation.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Per Liden <per.liden@ericsson.com>
Cc: Jon Maloy <jon.maloy@ericsson.com>
Cc: Allan Stephens <allan.stephens@windriver.com>
Cc: tipc-discussion@lists.sourceforge.net
Cc: linux-kernel@vger.kernel.org

Note also that in the release method, down_interruptible() was being called
without checking the return value.  I converted it to mutex_lock_interruptible()
and made the interrupted case return -ERESTARTSYS, as was done for all other
calls to down_interruptible() in the file.

Signed-off-by: Kevin Winchester <kjwinchester@gmail.com>

---
 net/tipc/socket.c |   56 +++++++++++++++++++++++++++---------------------------
 1 file changed, 28 insertions(+), 28 deletions(-)

Index: v2.6.24-rc4/net/tipc/socket.c
===================================================================
--- v2.6.24-rc4.orig/net/tipc/socket.c
+++ v2.6.24-rc4/net/tipc/socket.c
@@ -63,7 +63,7 @@
 struct tipc_sock {
 	struct sock sk;
 	struct tipc_port *p;
-	struct semaphore sem;
+	struct mutex lock;
 };
 
 #define tipc_sk(sk) ((struct tipc_sock*)sk)
@@ -217,7 +217,7 @@ static int tipc_create(struct net *net, 
 	tsock->p = port;
 	port->usr_handle = tsock;
 
-	init_MUTEX(&tsock->sem);
+	mutex_init(&tsock->lock);
 
 	dbg("sock_create: %x\n",tsock);
 
@@ -253,9 +253,10 @@ static int release(struct socket *sock)
 	dbg("sock_delete: %x\n",tsock);
 	if (!tsock)
 		return 0;
-	down_interruptible(&tsock->sem);
+	if (mutex_lock_interruptible(&tsock->lock))
+		return -ERESTARTSYS;
 	if (!sock->sk) {
-		up(&tsock->sem);
+		mutex_unlock(&tsock->lock);
 		return 0;
 	}
 
@@ -288,7 +289,7 @@ static int release(struct socket *sock)
 		atomic_dec(&tipc_queue_size);
 	}
 
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 
 	sock_put(sk);
 
@@ -315,7 +316,7 @@ static int bind(struct socket *sock, str
 	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
 	int res;
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	if (unlikely(!uaddr_len)) {
@@ -346,7 +347,7 @@ static int bind(struct socket *sock, str
 		res = tipc_withdraw(tsock->p->ref, -addr->scope,
 				    &addr->addr.nameseq);
 exit:
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 
@@ -367,7 +368,7 @@ static int get_name(struct socket *sock,
 	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
 	u32 res;
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	*uaddr_len = sizeof(*addr);
@@ -380,7 +381,7 @@ static int get_name(struct socket *sock,
 		res = tipc_ownidentity(tsock->p->ref, &addr->addr.id);
 	addr->addr.name.domain = 0;
 
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 
@@ -477,7 +478,7 @@ static int send_msg(struct kiocb *iocb, 
 		}
 	}
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	if (needs_conn) {
@@ -523,7 +524,7 @@ static int send_msg(struct kiocb *iocb, 
 		}
 		if (likely(res != -ELINKCONG)) {
 exit:
-			up(&tsock->sem);
+			mutex_unlock(&tsock->lock);
 			return res;
 		}
 		if (m->msg_flags & MSG_DONTWAIT) {
@@ -562,9 +563,8 @@ static int send_packet(struct kiocb *ioc
 	if (unlikely(dest))
 		return send_msg(iocb, sock, m, total_len);
 
-	if (down_interruptible(&tsock->sem)) {
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
-	}
 
 	do {
 		if (unlikely(sock->state != SS_CONNECTED)) {
@@ -578,7 +578,7 @@ static int send_packet(struct kiocb *ioc
 		res = tipc_send(tsock->p->ref, m->msg_iovlen, m->msg_iov);
 		if (likely(res != -ELINKCONG)) {
 exit:
-			up(&tsock->sem);
+			mutex_unlock(&tsock->lock);
 			return res;
 		}
 		if (m->msg_flags & MSG_DONTWAIT) {
@@ -846,7 +846,7 @@ static int recv_msg(struct kiocb *iocb, 
 
 	/* Look for a message in receive queue; wait if necessary */
 
-	if (unlikely(down_interruptible(&tsock->sem)))
+	if (unlikely(mutex_lock_interruptible(&tsock->lock)))
 		return -ERESTARTSYS;
 
 restart:
@@ -930,7 +930,7 @@ restart:
 		advance_queue(tsock);
 	}
 exit:
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 
@@ -981,7 +981,7 @@ static int recv_stream(struct kiocb *ioc
 
 	/* Look for a message in receive queue; wait if necessary */
 
-	if (unlikely(down_interruptible(&tsock->sem)))
+	if (unlikely(mutex_lock_interruptible(&tsock->lock)))
 		return -ERESTARTSYS;
 
 restart:
@@ -1077,7 +1077,7 @@ restart:
 		goto restart;
 
 exit:
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return sz_copied ? sz_copied : res;
 }
 
@@ -1293,7 +1293,7 @@ static int connect(struct socket *sock, 
 	   return res;
    }
 
-   if (down_interruptible(&tsock->sem))
+   if (mutex_lock_interruptible(&tsock->lock))
 	   return -ERESTARTSYS;
 
    /* Wait for destination's 'ACK' response */
@@ -1317,7 +1317,7 @@ static int connect(struct socket *sock, 
 	   sock->state = SS_DISCONNECTING;
    }
 
-   up(&tsock->sem);
+   mutex_unlock(&tsock->lock);
    return res;
 }
 
@@ -1365,7 +1365,7 @@ static int accept(struct socket *sock, s
 		     (flags & O_NONBLOCK)))
 		return -EWOULDBLOCK;
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	if (wait_event_interruptible(*sock->sk->sk_sleep,
@@ -1412,7 +1412,7 @@ static int accept(struct socket *sock, s
 		}
 	}
 exit:
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 
@@ -1434,7 +1434,7 @@ static int shutdown(struct socket *sock,
 
 	/* Could return -EINVAL for an invalid "how", but why bother? */
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	sock_lock(tsock);
@@ -1484,7 +1484,7 @@ restart:
 
 	sock_unlock(tsock);
 
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 
@@ -1518,7 +1518,7 @@ static int setsockopt(struct socket *soc
 	if ((res = get_user(value, (u32 __user *)ov)))
 		return res;
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	switch (opt) {
@@ -1541,7 +1541,7 @@ static int setsockopt(struct socket *soc
 		res = -EINVAL;
 	}
 
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 
@@ -1574,7 +1574,7 @@ static int getsockopt(struct socket *soc
 	if ((res = get_user(len, ol)))
 		return res;
 
-	if (down_interruptible(&tsock->sem))
+	if (mutex_lock_interruptible(&tsock->lock))
 		return -ERESTARTSYS;
 
 	switch (opt) {
@@ -1607,7 +1607,7 @@ static int getsockopt(struct socket *soc
 		res = put_user(sizeof(value), ol);
 	}
 
-	up(&tsock->sem);
+	mutex_unlock(&tsock->lock);
 	return res;
 }
 

-- 

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c
  2007-12-10  1:17 ` [patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c Kevin Winchester
@ 2007-12-12  0:44   ` Andrew Morton
  2007-12-12 19:24       ` David Miller
  0 siblings, 1 reply; 4+ messages in thread
From: Andrew Morton @ 2007-12-12  0:44 UTC (permalink / raw)
  To: Kevin Winchester
  Cc: mingo, per.liden, jon.maloy, allan.stephens, tipc-discussion,
	linux-kernel, kjwinchester, netdev, David S. Miller

On Sun, 09 Dec 2007 21:17:42 -0400
Kevin Winchester <kjwinchester@gmail.com> wrote:

> Note also that in the release method, down_interruptible() was being called
> without checking the return value.  I converted it to mutex_lock_interruptible()
> and made the interrupted case return -ERESTARTSYS, as was done for all other
> calls to down_interruptible() in the file.

That's an outright bug.

static int release(struct socket *sock)
{
	struct tipc_sock *tsock = tipc_sk(sock->sk);
	struct sock *sk = sock->sk;
	int res = TIPC_OK;
	struct sk_buff *buf;

	dbg("sock_delete: %x\n",tsock);
	if (!tsock)
		return 0;
	down_interruptible(&tsock->sem);
	if (!sock->sk) {
		up(&tsock->sem);
		return 0;
	}

	...

	up(&tsock->sem);

	...	
}

So if the calling process has signal_pending(), down_interruptible() will
return without having downed the semaphore and then we merrily proceed to
do up() on it, so a subsequent down() won't actually take the lock and
things will deteriorate from there.

So I'd propose this:

--- a/net/tipc/socket.c~a
+++ a/net/tipc/socket.c
@@ -253,7 +253,7 @@ static int release(struct socket *sock)
 	dbg("sock_delete: %x\n",tsock);
 	if (!tsock)
 		return 0;
-	down_interruptible(&tsock->sem);
+	down(&tsock->sem);
 	if (!sock->sk) {
 		up(&tsock->sem);
 		return 0;
_

as a for-2.6.24 bugfix.  And for 2.6.23.  But someone who knows what
they're doing should take a look at this...


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c
  2007-12-12  0:44   ` Andrew Morton
@ 2007-12-12 19:24       ` David Miller
  0 siblings, 0 replies; 4+ messages in thread
From: David Miller @ 2007-12-12 19:24 UTC (permalink / raw)
  To: akpm
  Cc: kjwinchester, mingo, per.liden, jon.maloy, allan.stephens,
	tipc-discussion, linux-kernel, netdev

From: Andrew Morton <akpm@linux-foundation.org>
Date: Tue, 11 Dec 2007 16:44:45 -0800

> So I'd propose this:

I've applied this to net-2.6, thanks Andrew.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c
@ 2007-12-12 19:24       ` David Miller
  0 siblings, 0 replies; 4+ messages in thread
From: David Miller @ 2007-12-12 19:24 UTC (permalink / raw)
  To: akpm
  Cc: jon.maloy, per.liden, netdev, linux-kernel, tipc-discussion,
	allan.stephens, kjwinchester, mingo

From: Andrew Morton <akpm@linux-foundation.org>
Date: Tue, 11 Dec 2007 16:44:45 -0800

> So I'd propose this:

I've applied this to net-2.6, thanks Andrew.

-------------------------------------------------------------------------
SF.Net email is sponsored by:
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services
for just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2007-12-12 19:24 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
     [not found] <20071210011741.039692330@gmail.com>
2007-12-10  1:17 ` [patch 1/1] Convert the semaphore to a mutex in net/tipc/socket.c Kevin Winchester
2007-12-12  0:44   ` Andrew Morton
2007-12-12 19:24     ` David Miller
2007-12-12 19:24       ` David Miller

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.