[PATCH v2 5/6] net: dsa: add missing calls in dsa_switch_destroy

Add missing netif_carrier_off and phy_disconnect calls to the
dsa_switch_destroy function to make sure the netdev and phy
ressources are clean before complete removal.

Signed-off-by: Frode Isaksen <redacted>
Signed-off-by: Neil Armstrong <redacted>
---
 net/dsa/dsa.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index 597a462..11452e4 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c

@@ -454,7 +454,9 @@ static void dsa_switch_destroy(struct dsa_switch *ds)
 		if (!ds->ports[port])
 			continue;

+		netif_carrier_off(ds->ports[port]);
 		unregister_netdev(ds->ports[port]);
+		phy_disconnect(p->phy);
 		free_netdev(ds->ports[port]);
 	}

-- 
1.9.1

Hi Neil,

[auto build test ERROR on net/master -- if it's inappropriate base, please suggest rules for selecting the more suitable base]

url:    https://github.com/0day-ci/linux/commits/Neil-Armstrong/Further-fix-for-dsa-unbinding/20151029-212633
config: x86_64-randconfig-x010-201543 (attached as .config)
reproduce:
        # save the attached .config to linux build tree
        make ARCH=x86_64 

Note: the linux-review/Neil-Armstrong/Further-fix-for-dsa-unbinding/20151029-212633 HEAD 3c957c00de125eeffe9f50eda7cf94a98855f615 builds fine.
      It only hurts bisectibility.

All errors (new ones prefixed by >>):

   net/dsa/dsa.c: In function 'dsa_switch_destroy':

quoted

net/dsa/dsa.c:459:18: error: 'p' undeclared (first use in this function)

      phy_disconnect(p->phy);
                     ^
   net/dsa/dsa.c:459:18: note: each undeclared identifier is reported only once for each function it appears in

vim +/p +459 net/dsa/dsa.c

   453	
   454			if (!ds->ports[port])
   455				continue;
   456	
   457			netif_carrier_off(ds->ports[port]);
   458			unregister_netdev(ds->ports[port]);
 > 459			phy_disconnect(p->phy);
   460			free_netdev(ds->ports[port]);
   461		}
   462	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

On Thu, Oct 29, 2015 at 02:23:25PM +0100, Neil Armstrong wrote:

Once you make it actually compile....

I'm not sure this is safe. The loop above this one has just destroyed
some phys, and now you are potentially disconnecting a phy you just
destroyed, causing it to be accessed?

I would suggest you first fix the ordering in dsa_switch_destroy()
and then add the missing netif_carrier_off() and phy_disconnect().

Thanks
	Andrew

On 10/29/2015 03:00 PM, Andrew Lunn wrote:

On Thu, Oct 29, 2015 at 02:23:25PM +0100, Neil Armstrong wrote:

quoted

+		netif_carrier_off(ds->ports[port]);
 		unregister_netdev(ds->ports[port]);
+		phy_disconnect(p->phy);
 		free_netdev(ds->ports[port]);
 	}

Once you make it actually compile....

I'm not sure this is safe. The loop above this one has just destroyed
some phys, and now you are potentially disconnecting a phy you just
destroyed, causing it to be accessed?

I would suggest you first fix the ordering in dsa_switch_destroy()
and then add the missing netif_carrier_off() and phy_disconnect().

Thanks
	Andrew

Yes, you are right, I will submit a cleaned up version.
I forgot the fixed phy case actually.

Thanks,
Neil