virtio-net: fix race between ndo_open() and virtio_device_ready()
Subsystems (4)
PATCH 1/1 virtio-net: fix race between ndo_open() and virtio_device_ready()
We used to call virtio_device_ready() after netdev registration. This cause a race between ndo_open() and virtio_device_ready(): if ndo_open() is called before virtio_device_ready(), the driver may start to use the device before DRIVER_OK which violates the spec.
Fixing this by switching to use register_netdevice() and protect the virtio_device_ready() with rtnl_lock() to make sure ndo_open() can only be called after virtio_device_ready().
--- drivers/net/virtio_net.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/net/virtio_net.c +7 -1
@@ -3655,14 +3655,20 @@ static int virtnet_probe(struct virtio_device *vdev)if (vi->has_rss || vi->has_rss_hash_report)virtnet_init_default_rss (vi); - err =register_netdev (dev); +/* serialize netdev register + virtio_device_ready() with ndo_open() */ +rtnl_lock (); + + err =register_netdevice (dev);if (err) {pr_debug ("virtio_net: registering device failed\n" ); +rtnl_unlock (); goto free_failover; }virtio_device_ready (vdev); +rtnl_unlock (); + err =virtnet_cpu_notif_add (vi);if (err) {pr_debug ("virtio_net: registering cpu notifier failed\n" );
Signature · 6 lines
2.25.1 _______________________________________________ Virtualization mailing list [email protected] https://lists.linuxfoundation.org/mailman/listinfo/virtualization
Michael S. Tsirkin
maintainerQuoted 31 lines · Jason Wang
We used to call
virtio_device_ready()after netdev registration. This cause a race betweenndo_open()andvirtio_device_ready(): ifndo_open()is called beforevirtio_device_ready(), the driver may start to use the device beforeDRIVER_OKwhich violates the spec.Fixing this by switching to use
register_netdevice()and protect thevirtio_device_ready()withrtnl_lock()to make surendo_open()can only be called aftervirtio_device_ready().Fixes 4baf1e33d0842 ("virtio_net: enable VQs early")Signed-off-by Jason Wang <[email protected]>--- drivers/net/virtio_net.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/net/virtio_net.c +5
-1@@ -3655,14 +3655,20 @@ static int virtnet_probe(struct virtio_device *vdev)if (vi->has_rss || vi->has_rss_hash_report)virtnet_init_default_rss (vi); - err =register_netdev (dev); +/* serialize netdev register + virtio_device_ready() with ndo_open() */ +rtnl_lock (); + + err =register_netdevice (dev);if (err) {pr_debug ("virtio_net: registering device failed\n" ); +rtnl_unlock ();
drivers/net/virtio_net.c static int virtnet_probe(struct virtio_device *vdev) · line 3674
Quoted 10 lines
drivers/net/virtio_net.c +2
-0goto free_failover; }virtio_device_ready (vdev); +rtnl_unlock (); + err =virtnet_cpu_notif_add (vi);if (err) {pr_debug ("virtio_net: registering cpu notifier failed\n" );
Looks good but then don't we have the same issue when removing the device?
Actually I looked at virtnet_remove and I see
unregister_netdev(vi->dev);
net_failover_destroy(vi->failover);
remove_vq_common(vi); <- this will reset the devicea window here?
Really, I think what we had originally was a better idea - instead of dropping interrupts they were delayed and when driver is ready to accept them it just enables them. We just need to make sure driver does not wait for interrupts before enabling them.
And I suspect we need to make this opt-in on a per driver basis.
Quoted 2 lines
Signature · 1 line
2.25.1
Quoted 48 lines · Michael S. Tsirkin
Quoted 31 lines · Jason Wang
We used to call
virtio_device_ready()after netdev registration. This cause a race betweenndo_open()andvirtio_device_ready(): ifndo_open()is called beforevirtio_device_ready(), the driver may start to use the device beforeDRIVER_OKwhich violates the spec.Fixing this by switching to use
register_netdevice()and protect thevirtio_device_ready()withrtnl_lock()to make surendo_open()can only be called aftervirtio_device_ready().Fixes 4baf1e33d0842 ("virtio_net: enable VQs early")Signed-off-by Jason Wang <[email protected]>--- drivers/net/virtio_net.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/net/virtio_net.c +5
-1@@ -3655,14 +3655,20 @@ static int virtnet_probe(struct virtio_device *vdev)if (vi->has_rss || vi->has_rss_hash_report)virtnet_init_default_rss (vi); - err =register_netdev (dev); +/* serialize netdev register + virtio_device_ready() with ndo_open() */ +rtnl_lock (); + + err =register_netdevice (dev);if (err) {pr_debug ("virtio_net: registering device failed\n" ); +rtnl_unlock ();drivers/net/virtio_net.c static int virtnet_probe(struct virtio_device *vdev) · line 3674
Quoted 10 lines
drivers/net/virtio_net.c +2
-0goto free_failover; }virtio_device_ready (vdev); +rtnl_unlock (); + err =virtnet_cpu_notif_add (vi);if (err) {pr_debug ("virtio_net: registering cpu notifier failed\n" );Looks good but then don't we have the same issue when removing the device?
Quoted 8 lines
Actually I looked at
virtnet_removeand I seeunregister_netdev(vi->dev); net_failover_destroy(vi->failover); remove_vq_common(vi); <- this will reset the devicea window here?
Probably. For safety, we probably need to reset before unregistering.
Quoted 5 lines
Really, I think what we had originally was a better idea - instead of dropping interrupts they were delayed and when driver is ready to accept them it just enables them.
The problem is that it works only on some specific setup:
- doesn't work on shared IRQ
- doesn't work on some specific driver e.g virtio-blk
Quoted 5 lines
We just need to make sure driver does not wait for interrupts before enabling them.
And I suspect we need to make this opt-in on a per driver basis.
Exactly.
Thanks
Quoted 6 lines
Quoted 5 lines
--
2.25.1
_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
Michael S. Tsirkin
maintainerQuoted 52 lines · Jason Wang
Quoted 51 lines · Michael S. Tsirkin
Quoted 31 lines · Jason Wang
We used to call
virtio_device_ready()after netdev registration. This cause a race betweenndo_open()andvirtio_device_ready(): ifndo_open()is called beforevirtio_device_ready(), the driver may start to use the device beforeDRIVER_OKwhich violates the spec.Fixing this by switching to use
register_netdevice()and protect thevirtio_device_ready()withrtnl_lock()to make surendo_open()can only be called aftervirtio_device_ready().Fixes 4baf1e33d0842 ("virtio_net: enable VQs early")Signed-off-by Jason Wang <[email protected]>--- drivers/net/virtio_net.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/net/virtio_net.c +5
-1@@ -3655,14 +3655,20 @@ static int virtnet_probe(struct virtio_device *vdev)if (vi->has_rss || vi->has_rss_hash_report)virtnet_init_default_rss (vi); - err =register_netdev (dev); +/* serialize netdev register + virtio_device_ready() with ndo_open() */ +rtnl_lock (); + + err =register_netdevice (dev);if (err) {pr_debug ("virtio_net: registering device failed\n" ); +rtnl_unlock ();drivers/net/virtio_net.c static int virtnet_probe(struct virtio_device *vdev) · line 3674
Quoted 10 lines
drivers/net/virtio_net.c +2
-0goto free_failover; }virtio_device_ready (vdev); +rtnl_unlock (); + err =virtnet_cpu_notif_add (vi);if (err) {pr_debug ("virtio_net: registering cpu notifier failed\n" );Looks good but then don't we have the same issue when removing the device?
Actually I looked at
virtnet_removeand I seeunregister_netdev(vi->dev);
Quoted 7 lines
Quoted 5 lines
net_failover_destroy(vi->failover);
remove_vq_common(vi); <- this will reset the devicea window here?
Probably. For safety, we probably need to reset before unregistering.
careful not to create new races, let's analyse this one to be sure first.
Quoted 10 lines
Quoted 5 lines
Really, I think what we had originally was a better idea - instead of dropping interrupts they were delayed and when driver is ready to accept them it just enables them.
The problem is that it works only on some specific setup:
- doesn't work on shared IRQ
- doesn't work on some specific driver e.g virtio-blk
can some core irq work fix that?
Quoted 16 lines
Quoted 5 lines
We just need to make sure driver does not wait for interrupts before enabling them.
And I suspect we need to make this opt-in on a per driver basis.
Exactly.
Thanks
Quoted 6 lines
Quoted 5 lines
--
2.25.1
Quoted 53 lines · Michael S. Tsirkin
Quoted 52 lines · Jason Wang
Quoted 51 lines · Michael S. Tsirkin
Quoted 31 lines · Jason Wang
We used to call
virtio_device_ready()after netdev registration. This cause a race betweenndo_open()andvirtio_device_ready(): ifndo_open()is called beforevirtio_device_ready(), the driver may start to use the device beforeDRIVER_OKwhich violates the spec.Fixing this by switching to use
register_netdevice()and protect thevirtio_device_ready()withrtnl_lock()to make surendo_open()can only be called aftervirtio_device_ready().Fixes 4baf1e33d0842 ("virtio_net: enable VQs early")Signed-off-by Jason Wang <[email protected]>--- drivers/net/virtio_net.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/net/virtio_net.c +5
-1@@ -3655,14 +3655,20 @@ static int virtnet_probe(struct virtio_device *vdev)if (vi->has_rss || vi->has_rss_hash_report)virtnet_init_default_rss (vi); - err =register_netdev (dev); +/* serialize netdev register + virtio_device_ready() with ndo_open() */ +rtnl_lock (); + + err =register_netdevice (dev);if (err) {pr_debug ("virtio_net: registering device failed\n" ); +rtnl_unlock ();drivers/net/virtio_net.c static int virtnet_probe(struct virtio_device *vdev) · line 3674
Quoted 10 lines
drivers/net/virtio_net.c +2
-0goto free_failover; }virtio_device_ready (vdev); +rtnl_unlock (); + err =virtnet_cpu_notif_add (vi);if (err) {pr_debug ("virtio_net: registering cpu notifier failed\n" );Looks good but then don't we have the same issue when removing the device?
Actually I looked at
virtnet_removeand I seeunregister_netdev(vi->dev); net_failover_destroy(vi->failover);
Quoted 7 lines
Quoted 4 lines
Quoted 3 lines
remove_vq_common(vi); <- this will reset the devicea window here?
Probably. For safety, we probably need to reset before unregistering.
careful not to create new races, let's analyse this one to be sure first.
Yes, if we do that, there could be an infinite wait in ctrl commands.
So we are probably fine here since unregister_netdev() will make sure (otherwise it should be a bug of unregister_netdev()):
1) NAPI is disabled (and synced) so no new NAPI could be enabled by the callbacks
2) TX is disabled (and synced) so the qdisc could not be scheduled even if skb_xmit_done() is called between the window
Quoted 10 lines
Quoted 9 lines
Quoted 5 lines
Really, I think what we had originally was a better idea - instead of dropping interrupts they were delayed and when driver is ready to accept them it just enables them.
The problem is that it works only on some specific setup:
- doesn't work on shared IRQ
- doesn't work on some specific driver e.g virtio-blkcan some core irq work fix that?
Not sure. At least for the shared IRQ part, there's no way to disable a specific handler currently. More below.
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
We just need to make sure driver does not wait for interrupts before enabling them.
This only help for the case:
1) the virtio_device_ready() is called after subsystem initialization/registration
2) the driver use rx interrupt
It doesn't solve the race between subsystem registration/initialization and virtio_device_ready() or the case when the virtio_device_ready() needs to be called before subsystem registration.
Thanks
Quoted 11 lines
Quoted 11 lines
Quoted 3 lines
And I suspect we need to make this opt-in on a per driver basis.
Exactly.
Thanks
Quoted 4 lines
Quoted 4 lines
--
2.25.1
Quoted 6 lines · Jason Wang
We used to call
virtio_device_ready()after netdev registration. This cause a race betweenndo_open()andvirtio_device_ready(): ifndo_open()is called beforevirtio_device_ready(), the driver may start to use the device beforeDRIVER_OKwhich violates the spec.
Quoted 6 lines
Fixing this by switching to use
register_netdevice()and protect thevirtio_device_ready()withrtnl_lock()to make surendo_open()can only be called aftervirtio_device_ready().Fixes 4baf1e33d0842 ("virtio_net: enable VQs early")Signed-off-by Jason Wang <[email protected]>
Ok, I think we're fine with this. So I will repost against -net.
If we spot issues with unregistering, we can use a separate patch for that.
Thanks
Quoted 33 lines
--- drivers/net/virtio_net.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/net/virtio_net.c +7
-2@@ -3655,14 +3655,20 @@ static int virtnet_probe(struct virtio_device *vdev)if (vi->has_rss || vi->has_rss_hash_report)virtnet_init_default_rss (vi); - err =register_netdev (dev); +/* serialize netdev register + virtio_device_ready() with ndo_open() */ +rtnl_lock (); + + err =register_netdevice (dev);if (err) {pr_debug ("virtio_net: registering device failed\n" ); +rtnl_unlock (); goto free_failover; }virtio_device_ready (vdev); +rtnl_unlock (); + err =virtnet_cpu_notif_add (vi);if (err) {pr_debug ("virtio_net: registering cpu notifier failed\n" ); --2.25.1