bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH bpf v3] xdp: Handle device unregister for devmap_hash map type
@ 2019-10-18 19:44 Toke Høiland-Jørgensen
  2019-10-18 21:18 ` Martin Lau
  0 siblings, 1 reply; 3+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-10-18 19:44 UTC (permalink / raw)
  To: daniel, ast
  Cc: Toke Høiland-Jørgensen, bpf, netdev, kafai, Tetsuo Handa

It seems I forgot to add handling of devmap_hash type maps to the device
unregister hook for devmaps. This omission causes devices to not be
properly released, which causes hangs.

Fix this by adding the missing handler.

Fixes: 6f9d451ab1a3 ("xdp: Add devmap_hash map type for looking up devices by hashed index")
Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
---
v3:
  - Use u32 for loop iterator variable
  - Since we're holding the lock we can just iterate with hlist_for_each_entry_safe()
v2:
  - Grab the update lock while walking the map and removing entries.

 kernel/bpf/devmap.c | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
index c0a48f336997..012dbfb0f54b 100644
--- a/kernel/bpf/devmap.c
+++ b/kernel/bpf/devmap.c
@@ -719,6 +719,31 @@ const struct bpf_map_ops dev_map_hash_ops = {
 	.map_check_btf = map_check_no_btf,
 };
 
+static void dev_map_hash_remove_netdev(struct bpf_dtab *dtab,
+				       struct net_device *netdev)
+{
+	unsigned long flags;
+	u32 i;
+
+	spin_lock_irqsave(&dtab->index_lock, flags);
+	for (i = 0; i < dtab->n_buckets; i++) {
+		struct bpf_dtab_netdev *dev;
+		struct hlist_head *head;
+		struct hlist_node *next;
+
+		head = dev_map_index_hash(dtab, i);
+
+		hlist_for_each_entry_safe(dev, next, head, index_hlist) {
+			if (netdev != dev->dev)
+				continue;
+
+			hlist_del_rcu(&dev->index_hlist);
+			call_rcu(&dev->rcu, __dev_map_entry_free);
+		}
+	}
+	spin_unlock_irqrestore(&dtab->index_lock, flags);
+}
+
 static int dev_map_notification(struct notifier_block *notifier,
 				ulong event, void *ptr)
 {
@@ -735,6 +760,11 @@ static int dev_map_notification(struct notifier_block *notifier,
 		 */
 		rcu_read_lock();
 		list_for_each_entry_rcu(dtab, &dev_map_list, list) {
+			if (dtab->map.map_type == BPF_MAP_TYPE_DEVMAP_HASH) {
+				dev_map_hash_remove_netdev(dtab, netdev);
+				continue;
+			}
+
 			for (i = 0; i < dtab->map.max_entries; i++) {
 				struct bpf_dtab_netdev *dev, *odev;
 
-- 
2.23.0


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH bpf v3] xdp: Handle device unregister for devmap_hash map type
  2019-10-18 19:44 [PATCH bpf v3] xdp: Handle device unregister for devmap_hash map type Toke Høiland-Jørgensen
@ 2019-10-18 21:18 ` Martin Lau
  2019-10-19 11:16   ` Toke Høiland-Jørgensen
  0 siblings, 1 reply; 3+ messages in thread
From: Martin Lau @ 2019-10-18 21:18 UTC (permalink / raw)
  To: Toke Høiland-Jørgensen
  Cc: daniel, Alexei Starovoitov, bpf, netdev, Tetsuo Handa

On Fri, Oct 18, 2019 at 09:44:18PM +0200, Toke Høiland-Jørgensen wrote:
> It seems I forgot to add handling of devmap_hash type maps to the device
> unregister hook for devmaps. This omission causes devices to not be
> properly released, which causes hangs.
> 
> Fix this by adding the missing handler.
> 
> Fixes: 6f9d451ab1a3 ("xdp: Add devmap_hash map type for looking up devices by hashed index")
> Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
> ---
> v3:
>   - Use u32 for loop iterator variable
>   - Since we're holding the lock we can just iterate with hlist_for_each_entry_safe()
> v2:
>   - Grab the update lock while walking the map and removing entries.
> 
>  kernel/bpf/devmap.c | 30 ++++++++++++++++++++++++++++++
>  1 file changed, 30 insertions(+)
> 
> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
> index c0a48f336997..012dbfb0f54b 100644
> --- a/kernel/bpf/devmap.c
> +++ b/kernel/bpf/devmap.c
> @@ -719,6 +719,31 @@ const struct bpf_map_ops dev_map_hash_ops = {
>  	.map_check_btf = map_check_no_btf,
>  };
>  
> +static void dev_map_hash_remove_netdev(struct bpf_dtab *dtab,
> +				       struct net_device *netdev)
> +{
> +	unsigned long flags;
> +	u32 i;
> +
> +	spin_lock_irqsave(&dtab->index_lock, flags);
> +	for (i = 0; i < dtab->n_buckets; i++) {
> +		struct bpf_dtab_netdev *dev;
> +		struct hlist_head *head;
> +		struct hlist_node *next;
> +
> +		head = dev_map_index_hash(dtab, i);
> +
> +		hlist_for_each_entry_safe(dev, next, head, index_hlist) {
> +			if (netdev != dev->dev)
> +				continue;
> +
> +			hlist_del_rcu(&dev->index_hlist);
There is another issue...
"dtab->items--;"

> +			call_rcu(&dev->rcu, __dev_map_entry_free);
> +		}
> +	}
> +	spin_unlock_irqrestore(&dtab->index_lock, flags);
> +}
> +
>  static int dev_map_notification(struct notifier_block *notifier,
>  				ulong event, void *ptr)
>  {
> @@ -735,6 +760,11 @@ static int dev_map_notification(struct notifier_block *notifier,
>  		 */
>  		rcu_read_lock();
>  		list_for_each_entry_rcu(dtab, &dev_map_list, list) {
> +			if (dtab->map.map_type == BPF_MAP_TYPE_DEVMAP_HASH) {
> +				dev_map_hash_remove_netdev(dtab, netdev);
> +				continue;
> +			}
> +
>  			for (i = 0; i < dtab->map.max_entries; i++) {
>  				struct bpf_dtab_netdev *dev, *odev;
>  
> -- 
> 2.23.0
> 

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH bpf v3] xdp: Handle device unregister for devmap_hash map type
  2019-10-18 21:18 ` Martin Lau
@ 2019-10-19 11:16   ` Toke Høiland-Jørgensen
  0 siblings, 0 replies; 3+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-10-19 11:16 UTC (permalink / raw)
  To: Martin Lau; +Cc: daniel, Alexei Starovoitov, bpf, netdev, Tetsuo Handa

Martin Lau <kafai@fb.com> writes:

> On Fri, Oct 18, 2019 at 09:44:18PM +0200, Toke Høiland-Jørgensen wrote:
>> It seems I forgot to add handling of devmap_hash type maps to the device
>> unregister hook for devmaps. This omission causes devices to not be
>> properly released, which causes hangs.
>> 
>> Fix this by adding the missing handler.
>> 
>> Fixes: 6f9d451ab1a3 ("xdp: Add devmap_hash map type for looking up devices by hashed index")
>> Reported-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
>> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
>> ---
>> v3:
>>   - Use u32 for loop iterator variable
>>   - Since we're holding the lock we can just iterate with hlist_for_each_entry_safe()
>> v2:
>>   - Grab the update lock while walking the map and removing entries.
>> 
>>  kernel/bpf/devmap.c | 30 ++++++++++++++++++++++++++++++
>>  1 file changed, 30 insertions(+)
>> 
>> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
>> index c0a48f336997..012dbfb0f54b 100644
>> --- a/kernel/bpf/devmap.c
>> +++ b/kernel/bpf/devmap.c
>> @@ -719,6 +719,31 @@ const struct bpf_map_ops dev_map_hash_ops = {
>>  	.map_check_btf = map_check_no_btf,
>>  };
>>  
>> +static void dev_map_hash_remove_netdev(struct bpf_dtab *dtab,
>> +				       struct net_device *netdev)
>> +{
>> +	unsigned long flags;
>> +	u32 i;
>> +
>> +	spin_lock_irqsave(&dtab->index_lock, flags);
>> +	for (i = 0; i < dtab->n_buckets; i++) {
>> +		struct bpf_dtab_netdev *dev;
>> +		struct hlist_head *head;
>> +		struct hlist_node *next;
>> +
>> +		head = dev_map_index_hash(dtab, i);
>> +
>> +		hlist_for_each_entry_safe(dev, next, head, index_hlist) {
>> +			if (netdev != dev->dev)
>> +				continue;
>> +
>> +			hlist_del_rcu(&dev->index_hlist);
> There is another issue...
> "dtab->items--;"

Right, of course. Thanks for your diligence in spotting my silly
mistakes! Will send a v4 :)

-Toke


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2019-10-19 11:16 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-10-18 19:44 [PATCH bpf v3] xdp: Handle device unregister for devmap_hash map type Toke Høiland-Jørgensen
2019-10-18 21:18 ` Martin Lau
2019-10-19 11:16   ` Toke Høiland-Jørgensen

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).