Skip to contentSkip to conversation

New: oops in port_carrier_check

Bugme-new In review Andrew Morton 5 participants 12 messages Feb 7, 2007 – Feb 14, 2007

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

Begin forwarded message:

Date: Wed, 7 Feb 2007 12:41:07 -0800
From: [email protected]
To: [email protected]
Subject: [Bugme-new] [Bug 7962] New: oops in port_carrier_check

http://bugzilla.kernel.org/show_bug.cgi?id=7962

Summary: oops in port_carrier_check
Kernel Version: 2.6.20-rc7
Status: NEW
Severity: normal
Owner: [email protected]
Submitter: [email protected]

While playing with qemu, I got a oops in bridge (and lost keyboard) :

Feb  7 21:20:18 plop kernel: BUG: unable to handle kernel paging request at
virtual address 6b6b6b6b
Feb  7 21:20:18 plop kernel:  printing eip:
Feb  7 21:20:18 plop kernel: *pde = 00000000
Feb  7 21:20:18 plop kernel: Oops: 0000 [#1]
Feb  7 21:20:18 plop kernel: CPU:    0
Feb  7 21:20:19 plop kernel: EIP:    0060:[pg0+814360305/1067136000]    Not
tainted VLI
Feb  7 21:20:19 plop kernel: EIP:    0060:[<f0eed6f1>]    Not tainted VLI
Feb  7 21:20:19 plop kernel: EFLAGS: 00010202   (2.6.20.0.rc7-1mdv #1)
Feb  7 21:20:19 plop kernel: EIP is at port_carrier_check+0x22/0x75 [bridge]
Feb  7 21:20:19 plop kernel: eax: 6b6b6b6b   ebx: 6b6b6b6b   ecx: 00000000  
edx: 00000001
Feb  7 21:20:19 plop kernel: esi: eb99b120   edi: 00000296   ebp: eff0bf58  
esp: eff0bf4c
Feb  7 21:20:19 plop kernel: ds: 007b   es: 007b   ss: 0068
Feb  7 21:20:19 plop kernel: Process events/0 (pid: 4, ti=eff0a000 task=eff09530
task.ti=eff0a000)
Feb  7 21:20:19 plop kernel: Stack: cd566744 eff4e86c 00000296 eff0bf84 c012534a
eff0bf70 00000296 eff0bfa0
Feb  7 21:20:19 plop kernel:        eff0bfac cd566740 f0eed6cf eff4e86c eff03ec8
eff0bfb4 eff0bfc4 c012590d
Feb  7 21:20:19 plop kernel:        00000001 00000000 00000001 00010000 00000000
00000000 eff09530 c0114770
Feb  7 21:20:19 plop kernel: Call Trace:
Feb  7 21:20:19 plop kernel:  [show_trace_log_lvl+26/47]
show_trace_log_lvl+0x1a/0x2f
Feb  7 21:20:19 plop kernel:  [<c010422c>] show_trace_log_lvl+0x1a/0x2f
Feb  7 21:20:19 plop kernel:  [show_stack_log_lvl+155/163]
show_stack_log_lvl+0x9b/0xa3
Feb  7 21:20:19 plop kernel:  [<c01042dc>] show_stack_log_lvl+0x9b/0xa3
Feb  7 21:20:19 plop kernel:  [show_registers+402/616] show_registers+0x192/0x268
Feb  7 21:20:19 plop kernel:  [<c0104476>] show_registers+0x192/0x268
Feb  7 21:20:19 plop kernel:  [die+234/511] die+0xea/0x1ff
Feb  7 21:20:19 plop kernel:  [<c0104636>] die+0xea/0x1ff
Feb  7 21:20:19 plop kernel:  [do_page_fault+1111/1334] do_page_fault+0x457/0x536
Feb  7 21:20:19 plop kernel:  [<c02c0c73>] do_page_fault+0x457/0x536
Feb  7 21:20:19 plop kernel:  [error_code+116/128] error_code+0x74/0x80
Feb  7 21:20:19 plop kernel:  [<c02bf624>] error_code+0x74/0x80
Feb  7 21:20:19 plop kernel:  [run_workqueue+142/333] run_workqueue+0x8e/0x14d
Feb  7 21:20:19 plop kernel:  [<c012534a>] run_workqueue+0x8e/0x14d
Feb  7 21:20:19 plop kernel:  [worker_thread+260/302] worker_thread+0x104/0x12e 
Feb  7 21:20:19 plop kernel:  [<c012590d>] worker_thread+0x104/0x12e
Feb  7 21:20:19 plop kernel:  [kthread+163/206] kthread+0xa3/0xce
Feb  7 21:20:19 plop kernel:  [<c0127e55>] kthread+0xa3/0xce
Feb  7 21:20:19 plop kernel:  [kernel_thread_helper+7/16]
kernel_thread_helper+0x7/0x10
Feb  7 21:20:19 plop kernel:  [<c0103ed7>] kernel_thread_helper+0x7/0x10
Feb  7 21:20:19 plop kernel:  ======================= 
Feb  7 21:20:19 plop kernel: Code: 38 cf 89 d8 5b 5e 5f 5d c3 55 89 e5 57 56 53
8b b0 24 ff ff ff 0f ba 30 00 e8 d3 20 38 cf 8b 9e 40 02 00 00 85 db 74 4c 8b 46
2c <8b> 3b a8 10 75 0a 89 f0 e8 e2 f9 ff ff 89 43 2c 8b 47 30 f6 40 
Feb  7 21:20:19 plop kernel: EIP: [pg0+814360305/1067136000]
port_carrier_check+0x22/0x75 [bridge] SS:ESP 0068:eff0bf4c
Feb  7 21:20:19 plop kernel: EIP: [<f0eed6f1>] port_carrier_check+0x22/0x75

[bridge] SS:ESP 0068:eff0bf4c

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

#1 lore raw
Quoted 77 lines · On Wed, 7 Feb 2007 12:52:16 -0800 Andrew Morton <[email protected]>…

Begin forwarded message:

Date: Wed, 7 Feb 2007 12:41:07 -0800
From: [email protected]
To: [email protected]
Subject: [Bugme-new] [Bug 7962] New: oops in port_carrier_check

http://bugzilla.kernel.org/show_bug.cgi?id=7962

Summary: oops in port_carrier_check
Kernel Version: 2.6.20-rc7
Status: NEW
Severity: normal
Owner: [email protected]
Submitter: [email protected]

While playing with qemu, I got a oops in bridge (and lost keyboard) :

Feb  7 21:20:18 plop kernel: BUG: unable to handle kernel paging request at
virtual address 6b6b6b6b
Feb  7 21:20:18 plop kernel:  printing eip:
Feb  7 21:20:18 plop kernel: *pde = 00000000
Feb  7 21:20:18 plop kernel: Oops: 0000 [#1]
Feb  7 21:20:18 plop kernel: CPU:    0
Feb  7 21:20:19 plop kernel: EIP:    0060:[pg0+814360305/1067136000]    Not
tainted VLI
Feb  7 21:20:19 plop kernel: EIP:    0060:[<f0eed6f1>]    Not tainted VLI
Feb  7 21:20:19 plop kernel: EFLAGS: 00010202   (2.6.20.0.rc7-1mdv #1)
Feb  7 21:20:19 plop kernel: EIP is at port_carrier_check+0x22/0x75 [bridge]
Feb  7 21:20:19 plop kernel: eax: 6b6b6b6b   ebx: 6b6b6b6b   ecx: 00000000  
edx: 00000001
Feb  7 21:20:19 plop kernel: esi: eb99b120   edi: 00000296   ebp: eff0bf58  
esp: eff0bf4c
Feb  7 21:20:19 plop kernel: ds: 007b   es: 007b   ss: 0068
Feb  7 21:20:19 plop kernel: Process events/0 (pid: 4, ti=eff0a000 task=eff09530
task.ti=eff0a000)
Feb  7 21:20:19 plop kernel: Stack: cd566744 eff4e86c 00000296 eff0bf84 c012534a
eff0bf70 00000296 eff0bfa0
Feb  7 21:20:19 plop kernel:        eff0bfac cd566740 f0eed6cf eff4e86c eff03ec8
eff0bfb4 eff0bfc4 c012590d
Feb  7 21:20:19 plop kernel:        00000001 00000000 00000001 00010000 00000000
00000000 eff09530 c0114770
Feb  7 21:20:19 plop kernel: Call Trace:
Feb  7 21:20:19 plop kernel:  [show_trace_log_lvl+26/47]
show_trace_log_lvl+0x1a/0x2f
Feb  7 21:20:19 plop kernel:  [<c010422c>] show_trace_log_lvl+0x1a/0x2f
Feb  7 21:20:19 plop kernel:  [show_stack_log_lvl+155/163]
show_stack_log_lvl+0x9b/0xa3
Feb  7 21:20:19 plop kernel:  [<c01042dc>] show_stack_log_lvl+0x9b/0xa3
Feb  7 21:20:19 plop kernel:  [show_registers+402/616] show_registers+0x192/0x268
Feb  7 21:20:19 plop kernel:  [<c0104476>] show_registers+0x192/0x268
Feb  7 21:20:19 plop kernel:  [die+234/511] die+0xea/0x1ff
Feb  7 21:20:19 plop kernel:  [<c0104636>] die+0xea/0x1ff
Feb  7 21:20:19 plop kernel:  [do_page_fault+1111/1334] do_page_fault+0x457/0x536
Feb  7 21:20:19 plop kernel:  [<c02c0c73>] do_page_fault+0x457/0x536
Feb  7 21:20:19 plop kernel:  [error_code+116/128] error_code+0x74/0x80
Feb  7 21:20:19 plop kernel:  [<c02bf624>] error_code+0x74/0x80
Feb  7 21:20:19 plop kernel:  [run_workqueue+142/333] run_workqueue+0x8e/0x14d
Feb  7 21:20:19 plop kernel:  [<c012534a>] run_workqueue+0x8e/0x14d
Feb  7 21:20:19 plop kernel:  [worker_thread+260/302] worker_thread+0x104/0x12e 
Feb  7 21:20:19 plop kernel:  [<c012590d>] worker_thread+0x104/0x12e
Feb  7 21:20:19 plop kernel:  [kthread+163/206] kthread+0xa3/0xce
Feb  7 21:20:19 plop kernel:  [<c0127e55>] kthread+0xa3/0xce
Feb  7 21:20:19 plop kernel:  [kernel_thread_helper+7/16]
kernel_thread_helper+0x7/0x10
Feb  7 21:20:19 plop kernel:  [<c0103ed7>] kernel_thread_helper+0x7/0x10
Feb  7 21:20:19 plop kernel:  ======================= 
Feb  7 21:20:19 plop kernel: Code: 38 cf 89 d8 5b 5e 5f 5d c3 55 89 e5 57 56 53
8b b0 24 ff ff ff 0f ba 30 00 e8 d3 20 38 cf 8b 9e 40 02 00 00 85 db 74 4c 8b 46
2c <8b> 3b a8 10 75 0a 89 f0 e8 e2 f9 ff ff 89 43 2c 8b 47 30 f6 40 
Feb  7 21:20:19 plop kernel: EIP: [pg0+814360305/1067136000]
port_carrier_check+0x22/0x75 [bridge] SS:ESP 0068:eff0bf4c
Feb  7 21:20:19 plop kernel: EIP: [<f0eed6f1>] port_carrier_check+0x22/0x75
Quoted 4 lines

[bridge] SS:ESP 0068:eff0bf4c

------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

I wonder if this is work_queue API change fallout.

#2 lore raw
Quoted 2 lines · Stephen Hemminger

On Wed, 7 Feb 2007 12:52:16 -0800
Andrew Morton <[email protected]> wrote:

...

Quoted 4 lines
Quoted 4 lines
Feb  7 21:20:18 plop kernel: BUG: unable to handle kernel paging request at
virtual address 6b6b6b6b
Feb  7 21:20:18 plop kernel:  printing eip:
Feb  7 21:20:18 plop kernel: *pde = 00000000
Quoted 8 lines
Quoted 8 lines
Feb  7 21:20:18 plop kernel: Oops: 0000 [#1]
Feb  7 21:20:18 plop kernel: CPU:    0
Feb  7 21:20:19 plop kernel: EIP:    0060:[pg0+814360305/1067136000]    Not
tainted VLI
Feb  7 21:20:19 plop kernel: EIP:    0060:[<f0eed6f1>]    Not tainted VLI
Feb  7 21:20:19 plop kernel: EFLAGS: 00010202   (2.6.20.0.rc7-1mdv #1)
Feb  7 21:20:19 plop kernel: EIP is at port_carrier_check+0x22/0x75 [bridge]
Feb  7 21:20:19 plop kernel: eax: 6b6b6b6b   ebx: 6b6b6b6b   ecx: 00000000  

I think it's caused by pending delayed workqueue trying to use dev after kfree (POISON_FREE in eax, ebx).

Quoted 12 lines

static void port_carrier_check(struct work_struct *work)

{
       struct net_bridge_port *p;
       struct net_device *dev;
       struct net_bridge *br;

       dev = container_of(work, struct net_bridge_port,
                          carrier_check.work)->dev;
       work_release(work);

       rtnl_lock();
       p = dev->br_port;
Quoted 8 lines
       if (!p)
               goto done;
       br = p->br;

       if (netif_carrier_ok(dev))
               p->path_cost = port_cost(dev);

       if (br->dev->flags & IFF_UP) {

My investigation seems to point at this line (p == ebx but not NULL because of mem debugging on, probably).

Regards,
Jarek P.

#3 lore raw
Quoted 36 lines · On Fri, 9 Feb 2007 08:42:11 +0100 Jarek Poplawski <[email protected]> wrote:
Quoted 2 lines · Stephen Hemminger

On Wed, 7 Feb 2007 12:52:16 -0800
Andrew Morton <[email protected]> wrote:

...

Quoted 4 lines
Quoted 4 lines
Feb  7 21:20:18 plop kernel: BUG: unable to handle kernel paging request at
virtual address 6b6b6b6b
Feb  7 21:20:18 plop kernel:  printing eip:
Feb  7 21:20:18 plop kernel: *pde = 00000000
Quoted 8 lines
Quoted 8 lines
Feb  7 21:20:18 plop kernel: Oops: 0000 [#1]
Feb  7 21:20:18 plop kernel: CPU:    0
Feb  7 21:20:19 plop kernel: EIP:    0060:[pg0+814360305/1067136000]    Not
tainted VLI
Feb  7 21:20:19 plop kernel: EIP:    0060:[<f0eed6f1>]    Not tainted VLI
Feb  7 21:20:19 plop kernel: EFLAGS: 00010202   (2.6.20.0.rc7-1mdv #1)
Feb  7 21:20:19 plop kernel: EIP is at port_carrier_check+0x22/0x75 [bridge]
Feb  7 21:20:19 plop kernel: eax: 6b6b6b6b   ebx: 6b6b6b6b   ecx: 00000000  

I think it's caused by pending delayed workqueue trying to use dev after kfree (POISON_FREE in eax, ebx).

Quoted 16 lines

static void port_carrier_check(struct work_struct *work)

{
       struct net_bridge_port *p;
       struct net_device *dev;
       struct net_bridge *br;

       dev = container_of(work, struct net_bridge_port,
                          carrier_check.work)->dev;
       work_release(work);

       rtnl_lock();
       p = dev->br_port;
       if (!p)
               goto done;
       br = p->br;
Quoted 8 lines
Quoted 4 lines
       if (netif_carrier_ok(dev))
               p->path_cost = port_cost(dev);

       if (br->dev->flags & IFF_UP) {

My investigation seems to point at this line (p == ebx but not NULL because of mem debugging on, probably).

The carrier_check is canceled by removal of port from bridge. Perhaps there is something broken in rcu assumptions under Qemu

#4 lore raw

2007/2/9, Stephen Hemminger <[email protected]>:

Quoted 2 lines

The carrier_check is canceled by removal of port from bridge. Perhaps there is something broken in rcu assumptions under Qemu

If that can help:
I started /stopped qemu several times. Maybe I started /stopped qemu several times as I was testing new PXE support in qemu with different virtual nic. Each time, a tun device was created by qemu at startup and added to the bridge, and destroyed on exit.

#5 lore raw
In reply to Stephen Hemminger
Quoted 39 lines · Stephen Hemminger
Quoted 36 lines · On Fri, 9 Feb 2007 08:42:11 +0100 Jarek Poplawski <[email protected]> wrote:
Quoted 2 lines · Stephen Hemminger

On Wed, 7 Feb 2007 12:52:16 -0800
Andrew Morton <[email protected]> wrote:

...

Quoted 4 lines
Quoted 4 lines
Feb  7 21:20:18 plop kernel: BUG: unable to handle kernel paging request at
virtual address 6b6b6b6b
Feb  7 21:20:18 plop kernel:  printing eip:
Feb  7 21:20:18 plop kernel: *pde = 00000000
Quoted 8 lines
Quoted 8 lines
Feb  7 21:20:18 plop kernel: Oops: 0000 [#1]
Feb  7 21:20:18 plop kernel: CPU:    0
Feb  7 21:20:19 plop kernel: EIP:    0060:[pg0+814360305/1067136000]    Not
tainted VLI
Feb  7 21:20:19 plop kernel: EIP:    0060:[<f0eed6f1>]    Not tainted VLI
Feb  7 21:20:19 plop kernel: EFLAGS: 00010202   (2.6.20.0.rc7-1mdv #1)
Feb  7 21:20:19 plop kernel: EIP is at port_carrier_check+0x22/0x75 [bridge]
Feb  7 21:20:19 plop kernel: eax: 6b6b6b6b   ebx: 6b6b6b6b   ecx: 00000000  

I think it's caused by pending delayed workqueue trying to use dev after kfree (POISON_FREE in eax, ebx).

Quoted 16 lines

static void port_carrier_check(struct work_struct *work)

{
       struct net_bridge_port *p;
       struct net_device *dev;
       struct net_bridge *br;

       dev = container_of(work, struct net_bridge_port,
                          carrier_check.work)->dev;
       work_release(work);

       rtnl_lock();
       p = dev->br_port;
       if (!p)
               goto done;
       br = p->br;
Quoted 7 lines
Quoted 7 lines
Quoted 4 lines
       if (netif_carrier_ok(dev))
               p->path_cost = port_cost(dev);

       if (br->dev->flags & IFF_UP) {

My investigation seems to point at this line (p == ebx but not NULL because of mem debugging on, probably).

Sorry, I overpasted. This is the line:

--> br = p->br;

Quoted 2 lines

The carrier_check is canceled by removal of port from bridge. Perhaps there is something broken in rcu assumptions under Qemu

If you mean this:

Quoted 4 lines

static void del_nbp(struct net_bridge_port *p)
{
...

       cancel_delayed_work(&p->carrier_check);

it's not sufficient. According to workqueue.h:

Quoted 6 lines

/*
* Kill off a pending schedule_delayed_work(). Note that the work callback
* function may still be running on return from cancel_delayed_work(). Run
* flush_scheduled_work() to wait on it.
*/
static inline int cancel_delayed_work(struct delayed_work *work)

I can't see how rcu could help here with this pointer to dev passed on to delayed_work (out of any rcu block).

IMHO dev_hold/dev_put (or something alike) is needed here.

Regards,
Jarek P.

#6 lore raw

PATCH 1/1 br_if: oops in port_carrier_check

Here is my patch proposal for testing. If it doesn't work - forget about it. (Prepared with 2.6.20-git6 but could be applied to 2.6.20 also.)

Jarek P.

dev_hold/dev_put added to prevent dev kfree during port_carrier_check runnig, while dev and port are removed.

Submitted and tested by: Pascal Terjan <[email protected]>

---

net/bridge/br_if.c +6 -2
diff -Nurp linux-2.6.20-git6-/net/bridge/br_if.c linux-2.6.20-git6/net/bridge/br_if.c
@@ -108,6 +108,7 @@ static void port_carrier_check(struct wo
 		spin_unlock_bh(&br->lock);
 	}
 done:
+	dev_put(dev);
 	rtnl_unlock();
 }
 
@@ -161,7 +162,8 @@ static void del_nbp(struct net_bridge_po
 
 	dev_set_promiscuity(dev, -1);
 
-	cancel_delayed_work(&p->carrier_check);
+	if (cancel_delayed_work(&p->carrier_check))
+		dev_put(dev);
 
 	spin_lock_bh(&br->lock);
 	br_stp_disable_port(p);
@@ -444,7 +446,9 @@ int br_add_if(struct net_bridge *br, str
 	spin_lock_bh(&br->lock);
 	br_stp_recalculate_bridge_id(br);
 	br_features_recompute(br);
-	schedule_delayed_work(&p->carrier_check, BR_PORT_DEBOUNCE);
+	if (schedule_delayed_work(&p->carrier_check, BR_PORT_DEBOUNCE))
+		dev_hold(dev);
+
 	spin_unlock_bh(&br->lock);
 
 	dev_set_mtu(br->dev, br_min_mtu(br));
#7 lore raw

Re: [PATCH][NET][BRIDGE] br_if: oops in port_carrier_check

Quoted 11 lines · On Mon, 12 Feb 2007 11:28:48 +0100 Jarek Poplawski <[email protected]> wrote:

Here is my patch proposal for testing. If it doesn't work - forget about it. (Prepared with 2.6.20-git6 but could be applied to 2.6.20 also.)

Jarek P.

dev_hold/dev_put added to prevent dev kfree during port_carrier_check runnig, while dev and port are removed.

No, holding the reference just stops the kfree, it doesn't stop the device from being removed from the port.

#8 lore raw

PATCH 1/1 br_if: oops in port_carrier_check

Quoted 11 lines · Stephen Hemminger
Quoted 8 lines · On Mon, 12 Feb 2007 11:28:48 +0100 Jarek Poplawski <[email protected]> wrote:

Here is my patch proposal for testing. If it doesn't work - forget about it. (Prepared with 2.6.20-git6 but could be applied to 2.6.20 also.)

Jarek P.

Quoted 6 lines
Quoted 3 lines

dev_hold/dev_put added to prevent dev kfree during port_carrier_check runnig, while dev and port are removed.

No, holding the reference just stops the kfree, it doesn't stop the device from being removed from the port.

But I wrote above it is only to prevent the kfree.

Quoted 4 lines
       p = dev->br_port;
       if (!p)
               goto done;
       br = p->br;

Then p is NULL here and we goto done.

Sorry, but the first version was wrong (incomplete). Below I attach a new proposal.

---

net/bridge/br_if.c +6 -2
diff -Nurp linux-2.6.20-git7-/net/bridge/br_if.c linux-2.6.20-git7/net/bridge/br_if.c
@@ -108,6 +108,7 @@ static void port_carrier_check(struct wo
 		spin_unlock_bh(&br->lock);
 	}
 done:
+	dev_put(dev);
 	rtnl_unlock();
 }
 
@@ -161,7 +162,8 @@ static void del_nbp(struct net_bridge_po
 
 	dev_set_promiscuity(dev, -1);
 
-	cancel_delayed_work(&p->carrier_check);
+	if (cancel_delayed_work(&p->carrier_check))
+		dev_put(dev);
 
 	spin_lock_bh(&br->lock);
 	br_stp_disable_port(p);
@@ -444,7 +446,9 @@ int br_add_if(struct net_bridge *br, str
 	spin_lock_bh(&br->lock);
 	br_stp_recalculate_bridge_id(br);
 	br_features_recompute(br);
-	schedule_delayed_work(&p->carrier_check, BR_PORT_DEBOUNCE);
+	if (schedule_delayed_work(&p->carrier_check, BR_PORT_DEBOUNCE))
+		dev_hold(dev);
+
 	spin_unlock_bh(&br->lock);
 
 	dev_set_mtu(br->dev, br_min_mtu(br));
net/bridge/br_notify.c +3 -1
diff -Nurp linux-2.6.20-git7-/net/bridge/br_notify.c linux-2.6.20-git7/net/bridge/br_notify.c
@@ -56,7 +56,9 @@ static int br_device_event(struct notifi
 
 	case NETDEV_CHANGE:
 		if (br->dev->flags & IFF_UP)
-			schedule_delayed_work(&p->carrier_check, BR_PORT_DEBOUNCE);
+			if (schedule_delayed_work(&p->carrier_check,
+						BR_PORT_DEBOUNCE))
+				dev_hold(dev);
 		break;
 
 	case NETDEV_FEAT_CHANGE:
#9 lore raw

Re: [PATCH][NET][BRIDGE] br_if: oops in port_carrier_check

Quoted 26 lines · On Tue, 13 Feb 2007 07:26:51 +0100 Jarek Poplawski <[email protected]> wrote:
Quoted 11 lines · Stephen Hemminger
Quoted 8 lines · On Mon, 12 Feb 2007 11:28:48 +0100 Jarek Poplawski <[email protected]> wrote:

Here is my patch proposal for testing. If it doesn't work - forget about it. (Prepared with 2.6.20-git6 but could be applied to 2.6.20 also.)

Jarek P.

Quoted 6 lines
Quoted 3 lines

dev_hold/dev_put added to prevent dev kfree during port_carrier_check runnig, while dev and port are removed.

No, holding the reference just stops the kfree, it doesn't stop the device from being removed from the port.

But I wrote above it is only to prevent the kfree.

Quoted 4 lines
       p = dev->br_port;
       if (!p)
               goto done;
       br = p->br;
Quoted 7 lines

Then p is NULL here and we goto done.

Sorry, but the first version was wrong (incomplete). Below I attach a new proposal.

Yes, this looks correct. If carrier_check is pending the ref count gets cleared by cancel. If carrier work is waiting for rtnl, then it cleans itself up.

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

Re: [PATCH][NET][BRIDGE] br_if: oops in port_carrier_check

From: Stephen Hemminger <[email protected]>
Date: Tue, 13 Feb 2007 11:55:29 -0800

Quoted 35 lines
Quoted 32 lines · On Tue, 13 Feb 2007 07:26:51 +0100 Jarek Poplawski <[email protected]> wrote:
Quoted 11 lines · Stephen Hemminger
Quoted 8 lines · On Mon, 12 Feb 2007 11:28:48 +0100 Jarek Poplawski <[email protected]> wrote:

Here is my patch proposal for testing. If it doesn't work - forget about it. (Prepared with 2.6.20-git6 but could be applied to 2.6.20 also.)

Jarek P.

Quoted 6 lines
Quoted 3 lines

dev_hold/dev_put added to prevent dev kfree during port_carrier_check runnig, while dev and port are removed.

No, holding the reference just stops the kfree, it doesn't stop the device from being removed from the port.

But I wrote above it is only to prevent the kfree.

Quoted 4 lines
       p = dev->br_port;
       if (!p)
               goto done;
       br = p->br;

Then p is NULL here and we goto done.

Sorry, but the first version was wrong (incomplete). Below I attach a new proposal.

Quoted 5 lines
Quoted 1 line

Yes, this looks correct. If carrier_check is pending the ref count gets cleared by cancel. If carrier work is waiting for rtnl, then it cleans itself up.

I've applied this patch, thanks everyone.

Stephen, do we want this in -stable?

#11 lore raw

Re: [PATCH][NET][BRIDGE] br_if: oops in port_carrier_check

On Tue, Feb 13, 2007 at 12:35:53PM -0800, David Miller wrote:
...

Quoted 3 lines

I've applied this patch, thanks everyone.

Stephen, do we want this in -stable?

I got this info it went trough -mm too:

...

Quoted 3 lines

From: [email protected]
Subject: - br_if-oops-in-port_carrier_check.patch removed from -mm tree
To: [email protected], [email protected]

...

Quoted 6 lines

The patch titled
br_if: oops in port_carrier_check
has been removed from the -mm tree. Its filename was
br_if-oops-in-port_carrier_check.patch

This patch was dropped because it was merged into mainline or a subsystem tree

Regards,
Jarek P.