drivers: xen frontend devices should handle missed backend CLOSING

PATCH In review David Vrabel 5 participants 27 messages Oct 18, 2012 – Nov 30, 2012

Subsystems (10)

Networking Network drivers Block layer PCI +6 more subsystems

lore

Share & tools

lore

David Vrabel

author

2012-10-18 10:02 UTC #0 lore raw

cover drivers: xen frontend devices should handle missed backend CLOSING

Subsystem maintainers, you can either pick up the relevant driver patch or ack it to go via Konrad's Xen tree.

The series makes all the Xen frontend drivers handle the backend transitioning to CLOSED without the frontend having previously seen the backend in the CLOSING state.

Backends shouldn't do this but some do. e.g., if the host is XenServer and the toolstack decides to do a forced shutdown of a VBD, then the blkfront may miss the CLOSING transition and the /dev/xvdX device will not be destroyed which prevents it being reused.

I have seen systems that ended up in this state but it's not clear if this was the actual cause. However, I think in general it's a good thing to thing to improve the handling of unexpected state transitions.

David

David Vrabel

author

2012-10-18 10:03 UTC #1 lore raw

PATCH 1/5 xen-netfront: handle backend CLOSED without CLOSING

From: David Vrabel <[email protected]>

Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: "David S. Miller" <[email protected]>

---
 drivers/net/xen-netfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/net/xen-netfront.c +4 -1

index caa0110..40cde51 100644
@@ -1701,7 +1701,6 @@ static void netback_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -1716,6 +1715,10 @@ static void netback_changed(struct xenbus_device *dev,
 		netdev_notify_peers(netdev);
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:03 UTC #2 lore raw

PATCH 1/5 xen-netfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: "David S. Miller" <[email protected]>

---
 drivers/net/xen-netfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/net/xen-netfront.c +4 -1

index caa0110..40cde51 100644
@@ -1701,7 +1701,6 @@ static void netback_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -1716,6 +1715,10 @@ static void netback_changed(struct xenbus_device *dev,
 		netdev_notify_peers(netdev);
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:03 UTC #3 lore raw

PATCH 2/5 xen-blkfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: Jens Axboe <[email protected]>

---
 drivers/block/xen-blkfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/block/xen-blkfront.c +4 -1

index 007db89..5d162b5 100644
@@ -1340,13 +1340,16 @@ static void blkback_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateConnected:
 		blkfront_connect(info);
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's Closing state -- fallthrough */
 	case XenbusStateClosing:
 		blkfront_closing(info);
 		break;

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:04 UTC #4 lore raw

PATCH 2/5 xen-blkfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: Jens Axboe <[email protected]>

---
 drivers/block/xen-blkfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/block/xen-blkfront.c +4 -1

index 007db89..5d162b5 100644
@@ -1340,13 +1340,16 @@ static void blkback_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateConnected:
 		blkfront_connect(info);
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's Closing state -- fallthrough */
 	case XenbusStateClosing:
 		blkfront_closing(info);
 		break;

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:03 UTC #5 lore raw

PATCH 3/5 xen-pcifront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -1

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
 	case XenbusStateInitialising:
 	case XenbusStateInitWait:
 	case XenbusStateInitialised:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateConnected:
 		pcifront_try_connect(pdev);
 		break;
 
+	case XenbusStateClosed:
+		if (xdev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		dev_warn(&xdev->dev, "backend going away!\n");
 		pcifront_try_disconnect(pdev);

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:04 UTC #6 lore raw

PATCH 3/5 xen-pcifront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -1

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
 	case XenbusStateInitialising:
 	case XenbusStateInitWait:
 	case XenbusStateInitialised:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateConnected:
 		pcifront_try_connect(pdev);
 		break;
 
+	case XenbusStateClosed:
+		if (xdev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		dev_warn(&xdev->dev, "backend going away!\n");
 		pcifront_try_disconnect(pdev);

Signature · 1 line

1.7.2.5

Konrad Rzeszutek Wilk

2012-10-19 12:59 UTC #7 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

Quoted 6 lines · David Vrabel

From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

Quoted 7 lines

So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?

Thx.

Quoted 28 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -1

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
 	case XenbusStateInitialising:
 	case XenbusStateInitWait:
 	case XenbusStateInitialised:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateConnected:
 		pcifront_try_connect(pdev);
 		break;
 
+	case XenbusStateClosed:
+		if (xdev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		dev_warn(&xdev->dev, "backend going away!\n");
 		pcifront_try_disconnect(pdev);

Signature · 1 line

1.7.2.5

Konrad Rzeszutek Wilk

2012-10-19 13:11 UTC #8 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

In reply to David Vrabel

Quoted 6 lines · David Vrabel

From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

Quoted 7 lines

So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?

Thx.

Quoted 28 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -1

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
 	case XenbusStateInitialising:
 	case XenbusStateInitWait:
 	case XenbusStateInitialised:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateConnected:
 		pcifront_try_connect(pdev);
 		break;
 
+	case XenbusStateClosed:
+		if (xdev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		dev_warn(&xdev->dev, "backend going away!\n");
 		pcifront_try_disconnect(pdev);

Signature · 1 line

1.7.2.5

Bjorn Helgaas

maintainer

2012-11-30 18:41 UTC #9 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

Quoted 15 lines · Konrad Rzeszutek Wilk

Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

Quoted 3 lines

Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?

Sorry, I missed this. I can put it in my -next branch for the v3.8 merge window. Would that work for you?

Quoted 28 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -2

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
      case XenbusStateInitialising:
      case XenbusStateInitWait:
      case XenbusStateInitialised:
-     case XenbusStateClosed:
              break;
 
      case XenbusStateConnected:
              pcifront_try_connect(pdev);
              break;
 
+     case XenbusStateClosed:
+             if (xdev->state == XenbusStateClosed)
+                     break;
+             /* Missed the backend's CLOSING state -- fallthrough */
      case XenbusStateClosing:
              dev_warn(&xdev->dev, "backend going away!\n");
              pcifront_try_disconnect(pdev);
--

1.7.2.5

Bjorn Helgaas

maintainer

2012-11-30 18:42 UTC #10 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

In reply to Konrad Rzeszutek Wilk

Quoted 15 lines · Konrad Rzeszutek Wilk

Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

Quoted 3 lines

Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?

Sorry, I missed this. I can put it in my -next branch for the v3.8 merge window. Would that work for you?

Quoted 28 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -2

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
      case XenbusStateInitialising:
      case XenbusStateInitWait:
      case XenbusStateInitialised:
-     case XenbusStateClosed:
              break;
 
      case XenbusStateConnected:
              pcifront_try_connect(pdev);
              break;
 
+     case XenbusStateClosed:
+             if (xdev->state == XenbusStateClosed)
+                     break;
+             /* Missed the backend's CLOSING state -- fallthrough */
      case XenbusStateClosing:
              dev_warn(&xdev->dev, "backend going away!\n");
              pcifront_try_disconnect(pdev);
--

1.7.2.5

Bjorn Helgaas

maintainer

2012-11-30 21:42 UTC #11 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

Quoted 17 lines · Bjorn Helgaas

Quoted 15 lines · Konrad Rzeszutek Wilk
Quoted 13 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

Quoted 6 lines

Quoted 3 lines
Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?
Sorry, I missed this. I can put it in my -next branch for the v3.8 merge window. Would that work for you?

I put this in my -next branch, so we'll at least have a chance of making a linux-next cycle before v3.7 pops.

Quoted 28 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -2

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
      case XenbusStateInitialising:
      case XenbusStateInitWait:
      case XenbusStateInitialised:
-     case XenbusStateClosed:
              break;
 
      case XenbusStateConnected:
              pcifront_try_connect(pdev);
              break;
 
+     case XenbusStateClosed:
+             if (xdev->state == XenbusStateClosed)
+                     break;
+             /* Missed the backend's CLOSING state -- fallthrough */
      case XenbusStateClosing:
              dev_warn(&xdev->dev, "backend going away!\n");
              pcifront_try_disconnect(pdev);
--

1.7.2.5

Bjorn Helgaas

maintainer

2012-11-30 21:42 UTC #12 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

In reply to Bjorn Helgaas

Quoted 17 lines · Bjorn Helgaas

Quoted 15 lines · Konrad Rzeszutek Wilk
Quoted 13 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>

Quoted 6 lines

Quoted 3 lines
Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?
Sorry, I missed this. I can put it in my -next branch for the v3.8 merge window. Would that work for you?

I put this in my -next branch, so we'll at least have a chance of making a linux-next cycle before v3.7 pops.

Quoted 28 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -2

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
      case XenbusStateInitialising:
      case XenbusStateInitWait:
      case XenbusStateInitialised:
-     case XenbusStateClosed:
              break;
 
      case XenbusStateConnected:
              pcifront_try_connect(pdev);
              break;
 
+     case XenbusStateClosed:
+             if (xdev->state == XenbusStateClosed)
+                     break;
+             /* Missed the backend's CLOSING state -- fallthrough */
      case XenbusStateClosing:
              dev_warn(&xdev->dev, "backend going away!\n");
              pcifront_try_disconnect(pdev);
--

1.7.2.5

Konrad Rzeszutek Wilk

2012-11-30 22:39 UTC #13 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

Quoted 22 lines · Bjorn Helgaas

Quoted 21 lines · Bjorn Helgaas
Quoted 18 lines · Konrad Rzeszutek Wilk
Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>
Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?

Quoted 5 lines

Quoted 2 lines
Sorry, I missed this. I can put it in my -next branch for the v3.8 merge window. Would that work for you?
I put this in my -next branch, so we'll at least have a chance of making a linux-next cycle before v3.7 pops.

Great. Thx!

Quoted 29 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -2

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
      case XenbusStateInitialising:
      case XenbusStateInitWait:
      case XenbusStateInitialised:
-     case XenbusStateClosed:
              break;
 
      case XenbusStateConnected:
              pcifront_try_connect(pdev);
              break;
 
+     case XenbusStateClosed:
+             if (xdev->state == XenbusStateClosed)
+                     break;
+             /* Missed the backend's CLOSING state -- fallthrough */
      case XenbusStateClosing:
              dev_warn(&xdev->dev, "backend going away!\n");
              pcifront_try_disconnect(pdev);
--

1.7.2.5

Konrad Rzeszutek Wilk

2012-11-30 22:39 UTC #14 lore raw

Re: [PATCH 3/5] xen-pcifront: handle backend CLOSED without CLOSING

In reply to Bjorn Helgaas

Quoted 22 lines · Bjorn Helgaas

Quoted 21 lines · Bjorn Helgaas
Quoted 18 lines · Konrad Rzeszutek Wilk
Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Bjorn Helgaas <[email protected]>
Bjorn, do you want me to prep a git pull with this patch or can I have your Ack to put it my tree and have it part of my git pull to Linus?

Quoted 5 lines

Quoted 2 lines
Sorry, I missed this. I can put it in my -next branch for the v3.8 merge window. Would that work for you?
I put this in my -next branch, so we'll at least have a chance of making a linux-next cycle before v3.7 pops.

Great. Thx!

Quoted 29 lines

---
 drivers/pci/xen-pcifront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/pci/xen-pcifront.c +4 -2

index 0aab85a..a0c7312 100644
@@ -1068,13 +1068,16 @@ static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
      case XenbusStateInitialising:
      case XenbusStateInitWait:
      case XenbusStateInitialised:
-     case XenbusStateClosed:
              break;
 
      case XenbusStateConnected:
              pcifront_try_connect(pdev);
              break;
 
+     case XenbusStateClosed:
+             if (xdev->state == XenbusStateClosed)
+                     break;
+             /* Missed the backend's CLOSING state -- fallthrough */
      case XenbusStateClosing:
              dev_warn(&xdev->dev, "backend going away!\n");
              pcifront_try_disconnect(pdev);
--

1.7.2.5

David Vrabel

author

2012-10-18 10:03 UTC #15 lore raw

PATCH 4/5 xen-fbfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: Florian Tobias Schandinat <[email protected]>

---
 drivers/video/xen-fbfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/video/xen-fbfront.c +4 -1

index b7f5173..917bb56 100644
@@ -641,7 +641,6 @@ static void xenfb_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -670,6 +669,10 @@ InitWait:
 		info->feature_resize = val;
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state = XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Konrad Rzeszutek Wilk

2012-10-19 13:00 UTC #16 lore raw

Re: [PATCH 4/5] xen-fbfront: handle backend CLOSED without CLOSING

Quoted 6 lines · David Vrabel

From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

Quoted 7 lines

So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Florian Tobias Schandinat <[email protected]>

Hey Florian,

Should I prep a git pull for you with this or would it be OK if I just have your Ack to put this in my git pull for Linus?

Thanks!

Quoted 29 lines

---
 drivers/video/xen-fbfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/video/xen-fbfront.c +4 -1

index b7f5173..917bb56 100644
@@ -641,7 +641,6 @@ static void xenfb_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -670,6 +669,10 @@ InitWait:
 		info->feature_resize = val;
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state = XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Florian Tobias Schandinat

Acked-by

2012-10-23 04:50 UTC #17 lore raw

Re: [PATCH 4/5] xen-fbfront: handle backend CLOSED without CLOSING

Hi Konrad,

Quoted 15 lines · Konrad Rzeszutek Wilk

Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Florian Tobias Schandinat <[email protected]>

Quoted 4 lines

Hey Florian,
Should I prep a git pull for you with this or would it be OK if I just have your Ack to put this in my git pull for Linus?

Feel free to take it and add

Acked-by Florian Tobias Schandinat <[email protected]>

Best regards,

Florian Tobias Schandinat

Quoted 32 lines

Thanks!

Quoted 29 lines

---
 drivers/video/xen-fbfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/video/xen-fbfront.c +4 -1

index b7f5173..917bb56 100644
@@ -641,7 +641,6 @@ static void xenfb_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -670,6 +669,10 @@ InitWait:
 		info->feature_resize = val;
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state = XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Florian Tobias Schandinat

Acked-by

2012-10-23 04:50 UTC #18 lore raw

Re: [PATCH 4/5] xen-fbfront: handle backend CLOSED without CLOSING

In reply to Konrad Rzeszutek Wilk

Hi Konrad,

Quoted 15 lines · Konrad Rzeszutek Wilk

Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Florian Tobias Schandinat <[email protected]>

Quoted 4 lines

Hey Florian,
Should I prep a git pull for you with this or would it be OK if I just have your Ack to put this in my git pull for Linus?

Feel free to take it and add

Acked-by Florian Tobias Schandinat <[email protected]>

Best regards,

Florian Tobias Schandinat

Quoted 32 lines

Thanks!

Quoted 29 lines

---
 drivers/video/xen-fbfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/video/xen-fbfront.c +4 -1

index b7f5173..917bb56 100644
@@ -641,7 +641,6 @@ static void xenfb_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -670,6 +669,10 @@ InitWait:
 		info->feature_resize = val;
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Konrad Rzeszutek Wilk

2012-10-19 13:00 UTC #19 lore raw

Re: [PATCH 4/5] xen-fbfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

Quoted 6 lines · David Vrabel

From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

Quoted 7 lines

So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Florian Tobias Schandinat <[email protected]>

Hey Florian,

Should I prep a git pull for you with this or would it be OK if I just have your Ack to put this in my git pull for Linus?

Thanks!

Quoted 29 lines

---
 drivers/video/xen-fbfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/video/xen-fbfront.c +4 -1

index b7f5173..917bb56 100644
@@ -641,7 +641,6 @@ static void xenfb_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -670,6 +669,10 @@ InitWait:
 		info->feature_resize = val;
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:03 UTC #20 lore raw

PATCH 4/5 xen-fbfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: Florian Tobias Schandinat <[email protected]>

---
 drivers/video/xen-fbfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/video/xen-fbfront.c +4 -1

index b7f5173..917bb56 100644
@@ -641,7 +641,6 @@ static void xenfb_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -670,6 +669,10 @@ InitWait:
 		info->feature_resize = val;
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

David Vrabel

author

2012-10-18 10:03 UTC #21 lore raw

PATCH 5/5 xen-kbdfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: Dmitry Torokhov <[email protected]>

---
 drivers/input/misc/xen-kbdfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/input/misc/xen-kbdfront.c +4 -1

index 02ca868..6f7d990 100644
@@ -311,7 +311,6 @@ static void xenkbd_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -350,6 +349,10 @@ InitWait:
 
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Konrad Rzeszutek Wilk

2012-10-19 13:00 UTC #22 lore raw

Re: [PATCH 5/5] xen-kbdfront: handle backend CLOSED without CLOSING

Quoted 6 lines · David Vrabel

From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

Quoted 7 lines

So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Dmitry Torokhov <[email protected]>

Hey Dmitry,

Should I prep a git pull for you for this or are you OK giving an Ack for me to put this patch in my git pull for Linus?

Thx.

Quoted 29 lines

---
 drivers/input/misc/xen-kbdfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/input/misc/xen-kbdfront.c +4 -1

index 02ca868..6f7d990 100644
@@ -311,7 +311,6 @@ static void xenkbd_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -350,6 +349,10 @@ InitWait:
 
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Konrad Rzeszutek Wilk

2012-10-19 13:13 UTC #23 lore raw

Re: [PATCH 5/5] xen-kbdfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

Quoted 6 lines · David Vrabel

From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.

Quoted 7 lines

So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Dmitry Torokhov <[email protected]>

Hey Dmitry,

Should I prep a git pull for you for this or are you OK giving an Ack for me to put this patch in my git pull for Linus?

Thx.

Quoted 29 lines

---
 drivers/input/misc/xen-kbdfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/input/misc/xen-kbdfront.c +4 -1

index 02ca868..6f7d990 100644
@@ -311,7 +311,6 @@ static void xenkbd_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -350,6 +349,10 @@ InitWait:
 
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Dmitry Torokhov

maintainer Acked-by

2012-10-19 16:34 UTC #24 lore raw

Re: [PATCH 5/5] xen-kbdfront: handle backend CLOSED without CLOSING

Quoted 15 lines · Konrad Rzeszutek Wilk

Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Dmitry Torokhov <[email protected]>

Quoted 4 lines

Hey Dmitry,
Should I prep a git pull for you for this or are you OK giving an Ack for me to put this patch in my git pull for Linus?

Sure, please merge with the rest through your tree.

Acked-by Dmitry Torokhov <[email protected]>

Thanks!

Quoted 31 lines

Thx.

Quoted 29 lines

---
 drivers/input/misc/xen-kbdfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/input/misc/xen-kbdfront.c +4 -1

index 02ca868..6f7d990 100644
@@ -311,7 +311,6 @@ static void xenkbd_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -350,6 +349,10 @@ InitWait:
 
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Signature · 1 line

Dmitry

Dmitry Torokhov

maintainer Acked-by

2012-10-19 16:34 UTC #25 lore raw

Re: [PATCH 5/5] xen-kbdfront: handle backend CLOSED without CLOSING

In reply to Konrad Rzeszutek Wilk

Quoted 15 lines · Konrad Rzeszutek Wilk

Quoted 6 lines · David Vrabel
From: David Vrabel <[email protected]>
Backend drivers shouldn't transistion to CLOSED unless the frontend is CLOSED. If a backend does transition to CLOSED too soon then the frontend may not see the CLOSING state and will not properly shutdown.
Quoted 7 lines
So, treat an unexpected backend CLOSED state the same as CLOSING.
Signed-off-by David Vrabel <[email protected]>
Acked-by Konrad Rzeszutek Wilk <[email protected]>
---
Cc: [email protected]
Cc: Dmitry Torokhov <[email protected]>

Quoted 4 lines

Hey Dmitry,
Should I prep a git pull for you for this or are you OK giving an Ack for me to put this patch in my git pull for Linus?

Sure, please merge with the rest through your tree.

Acked-by Dmitry Torokhov <[email protected]>

Thanks!

Quoted 31 lines

Thx.

Quoted 29 lines

---
 drivers/input/misc/xen-kbdfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/input/misc/xen-kbdfront.c +4 -1

index 02ca868..6f7d990 100644
@@ -311,7 +311,6 @@ static void xenkbd_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -350,6 +349,10 @@ InitWait:
 
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5

Signature · 1 line

Dmitry

David Vrabel

author

2012-10-18 10:03 UTC #26 lore raw

PATCH 5/5 xen-kbdfront: handle backend CLOSED without CLOSING

In reply to David Vrabel

From: David Vrabel <[email protected]>

So, treat an unexpected backend CLOSED state the same as CLOSING.

Signed-off-by David Vrabel <[email protected]>

Acked-by Konrad Rzeszutek Wilk <[email protected]>

---
Cc: [email protected]
Cc: Dmitry Torokhov <[email protected]>

---
 drivers/input/misc/xen-kbdfront.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

drivers/input/misc/xen-kbdfront.c +4 -1

index 02ca868..6f7d990 100644
@@ -311,7 +311,6 @@ static void xenkbd_backend_changed(struct xenbus_device *dev,
 	case XenbusStateReconfiguring:
 	case XenbusStateReconfigured:
 	case XenbusStateUnknown:
-	case XenbusStateClosed:
 		break;
 
 	case XenbusStateInitWait:

@@ -350,6 +349,10 @@ InitWait:
 
 		break;
 
+	case XenbusStateClosed:
+		if (dev->state == XenbusStateClosed)
+			break;
+		/* Missed the backend's CLOSING state -- fallthrough */
 	case XenbusStateClosing:
 		xenbus_frontend_closed(dev);
 		break;

Signature · 1 line

1.7.2.5