Skip to contentSkip to conversation

New: unregister_netdev() should return unregister_netdevice() return code

Bugme-new Applied Andrew Morton 3 participants 5 messages Dec 18, 2006 – Feb 7, 2007 1 message not in the archive

Subsystems (3)
lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw
Quoted 11 lines · On Mon, 18 Dec 2006 14:56:31 -0800 [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=7708

Summary: unregister_netdev() should return unregister_netdevice()
return code
Kernel Version: 2.6.19.1
Status: NEW
Severity: low
Owner: [email protected]
Submitter: [email protected]

Quoted 6 lines

net/core/dev.c:unregister_netdev() function is a wrapper around net/core/dev.c:unregister_netdevice(). The unregister_netdevice() function returns a return code while unregister_netdev() currently does not. For completeness, we should pass the return code from unregister_netdevice() all the way to the caller. unregister_netdev() should not swallow the return code.

Certainly there's some truth in that ;)

Is there some reason why you want to test the unregister_netdev() return value?

#2 lore raw
Quoted 21 lines · On Mon, 18 Dec 2006 15:21:11 -0800 Andrew Morton <[email protected]> wrote:
Quoted 11 lines · On Mon, 18 Dec 2006 14:56:31 -0800 [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=7708

Summary: unregister_netdev() should return unregister_netdevice()
return code
Kernel Version: 2.6.19.1
Status: NEW
Severity: low
Owner: [email protected]
Submitter: [email protected]

Quoted 6 lines

net/core/dev.c:unregister_netdev() function is a wrapper around net/core/dev.c:unregister_netdevice(). The unregister_netdevice() function returns a return code while unregister_netdev() currently does not. For completeness, we should pass the return code from unregister_netdevice() all the way to the caller. unregister_netdev() should not swallow the return code.

Quoted 4 lines

Certainly there's some truth in that ;)

Is there some reason why you want to test the unregister_netdev() return value?

The only return value is -ENODEV, so I would vote for both just being void

Signature · 1 line
Stephen Hemminger <[email protected]>
maintainer
#3 lore raw

From: Stephen Hemminger <[email protected]>
Date: Mon, 18 Dec 2006 15:50:57 -0800

Quoted 10 lines
Quoted 7 lines · On Mon, 18 Dec 2006 15:21:11 -0800 Andrew Morton <[email protected]> wrote:
Quoted 6 lines

net/core/dev.c:unregister_netdev() function is a wrapper around net/core/dev.c:unregister_netdevice(). The unregister_netdevice() function returns a return code while unregister_netdev() currently does not. For completeness, we should pass the return code from unregister_netdevice() all the way to the caller. unregister_netdev() should not swallow the return code.

Quoted 5 lines
Quoted 4 lines

Certainly there's some truth in that ;)

Is there some reason why you want to test the unregister_netdev() return value?

The only return value is -ENODEV, so I would vote for both just being void

Me too.

FWIW, I think bug reports like this are a lot of back-and-forth waste of time. If it's important enough to someone, let them write a god-damn patch for something so amazingly trivial. With the bugzill entry, you have to respond to it, change it's diapers, etc.

#4 lore raw

PATCH 1/1 net: unregister_netdevice as void

There was no real useful information from the unregister_netdevice() return code, the only error occurred in a situation that was a driver bug. So change it to a void function.

---
 include/linux/netdevice.h |    2 +-
 net/core/dev.c            |   15 ++++++---------
 net/ipv4/ip_gre.c         |    3 ++-
 net/ipv4/ipip.c           |    3 ++-
 net/ipv6/ip6_tunnel.c     |    3 ++-
 net/ipv6/sit.c            |    3 ++-
 6 files changed, 15 insertions(+), 14 deletions(-)
include/linux/netdevice.h +1 -1
index 6be767c..e16d66e 100644
@@ -588,7 +588,7 @@ extern int		dev_open(struct net_device *
 extern int		dev_close(struct net_device *dev);
 extern int		dev_queue_xmit(struct sk_buff *skb);
 extern int		register_netdevice(struct net_device *dev);
-extern int		unregister_netdevice(struct net_device *dev);
+extern void		unregister_netdevice(struct net_device *dev);
 extern void		free_netdev(struct net_device *dev);
 extern void		synchronize_net(void);
 extern int 		register_netdevice_notifier(struct notifier_block *nb);
net/core/dev.c +6 -9
index e660cb5..4a8a7fd 100644
@@ -3247,7 +3247,7 @@ void synchronize_net(void) 
  *	unregister_netdev() instead of this.
  */
 
-int unregister_netdevice(struct net_device *dev)
+void unregister_netdevice(struct net_device *dev)
 {
 	struct net_device *d, **dp;
 
@@ -3258,8 +3258,10 @@ int unregister_netdevice(struct net_devi
 	if (dev->reg_state == NETREG_UNINITIALIZED) {
 		printk(KERN_DEBUG "unregister_netdevice: device %s/%p never "
 				  "was registered\n", dev->name, dev);
-		return -ENODEV;
-	}
+		
+		WARN_ON(1);
+		return;
+	}		
 
 	BUG_ON(dev->reg_state != NETREG_REGISTERED);
 
@@ -3280,11 +3282,7 @@ int unregister_netdevice(struct net_devi
 			break;
 		}
 	}
-	if (!d) {
-		printk(KERN_ERR "unregister net_device: '%s' not found\n",
-		       dev->name);
-		return -ENODEV;
-	}
+	BUG_ON(!d);
 
 	dev->reg_state = NETREG_UNREGISTERING;
 
@@ -3316,7 +3314,6 @@ int unregister_netdevice(struct net_devi
 	synchronize_net();
 
 	dev_put(dev);
-	return 0;
 }
 
 /**
net/ipv4/ip_gre.c +2 -1
index 476cb60..51c8350 100644
@@ -1008,7 +1008,8 @@ ipgre_tunnel_ioctl (struct net_device *d
 				goto done;
 			dev = t->dev;
 		}
-		err = unregister_netdevice(dev);
+		unregister_netdevice(dev);
+		err = 0;
 		break;
 
 	default:
net/ipv4/ipip.c +2 -1
index 9d719d6..da8bbd2 100644
@@ -754,7 +754,8 @@ ipip_tunnel_ioctl (struct net_device *de
 				goto done;
 			dev = t->dev;
 		}
-		err = unregister_netdevice(dev);
+		unregister_netdevice(dev);
+		err = 0;
 		break;
 
 	default:
net/ipv6/ip6_tunnel.c +2 -1
index 8d91834..2b9e3bb 100644
@@ -999,7 +999,8 @@ ip6ip6_tnl_ioctl(struct net_device *dev,
 				break;
 			dev = t->dev;
 		}
-		err = unregister_netdevice(dev);
+		err = 0;
+		unregister_netdevice(dev);
 		break;
 	default:
 		err = -EINVAL;
net/ipv6/sit.c +2 -1
index 77b7b09..47cfead 100644
@@ -686,7 +686,8 @@ ipip6_tunnel_ioctl (struct net_device *d
 				goto done;
 			dev = t->dev;
 		}
-		err = unregister_netdevice(dev);
+		unregister_netdevice(dev);
+		err = 0;
 		break;
 
 	default:
Signature · 1 line
1.4.2.4
maintainer Applied
#5 lore raw

Re: [PATCH] net: unregister_netdevice as void

From: Stephen Hemminger <[email protected]>
Date: Mon, 18 Dec 2006 21:05:11 -0800

Quoted 5 lines

There was no real useful information from the unregister_netdevice() return code, the only error occurred in a situation that was a driver bug. So change it to a void function.

Applied to net-2.6.21, thanks Stephen.