Skip to contentSkip to conversation

af_packet: relax BUG statement in tpacket_destruct_skb

PATCH In review Daniel Borkmann 2 participants 3 messages Aug 9, 2012 – Aug 11, 2012

Subsystems (3)
Networking Network drivers Packet sockets
lore
Share & tools
lore

Daniel Borkmann

author
#0 lore raw

PATCH 1/1 af_packet: relax BUG statement in tpacket_destruct_skb

Here's a quote of the comment about the BUG macro from asm-generic/bug.h:

Don't use BUG() or BUG_ON() unless there's really no way out; one example might be detecting data structure corruption in the middle of an operation that can't be backed out of. If the (sub)system can somehow continue operating, perhaps with reduced functionality, it's probably not BUG-worthy.

If you're tempted to BUG(), think again: is completely giving up really the *only* solution? There are usually better options, where users don't need to reboot ASAP and can mostly shut down cleanly.

In our case, the status flag of a ring buffer slot is managed from both sides, the kernel space and the user space. This means that even though the kernel side might work as expected, the user space screws up and changes this flag right between the send(2) is triggered when the flag is changed to TP_STATUS_SENDING and a given skb is destructed after some time. Then, this will hit the BUG macro. Instead, we relax this condition with a WARN_ON_ONCE macro, so that the user is aware of this situation. I've tested it and the system still behaves /stable/, so in accordance with the above comment, we should rather relax this behavior with a warning.

Signed-off-by Daniel Borkmann <[email protected]>
---
 net/packet/af_packet.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
net/packet/af_packet.c +1 -1
index ceaca7c..4def36f 100644
@@ -1936,7 +1936,7 @@ static void tpacket_destruct_skb(struct sk_buff *skb)
 
 	if (likely(po->tx_ring.pg_vec)) {
 		ph = skb_shinfo(skb)->destructor_arg;
-		BUG_ON(__packet_get_status(po, ph) != TP_STATUS_SENDING);
+		WARN_ON_ONCE(__packet_get_status(po, ph) != TP_STATUS_SENDING);
 		BUG_ON(atomic_read(&po->tx_ring.pending) == 0);
 		atomic_dec(&po->tx_ring.pending);
 		__packet_set_status(po, ph, TP_STATUS_AVAILABLE);
maintainer
#1 lore raw

From: Daniel Borkmann <[email protected]>
Date: Thu, 09 Aug 2012 18:39:05 +0200

Quoted 22 lines

Here's a quote of the comment about the BUG macro from asm-generic/bug.h:

Don't use BUG() or BUG_ON() unless there's really no way out; one example might be detecting data structure corruption in the middle of an operation that can't be backed out of. If the (sub)system can somehow continue operating, perhaps with reduced functionality, it's probably not BUG-worthy.

If you're tempted to BUG(), think again: is completely giving up really the *only* solution? There are usually better options, where users don't need to reboot ASAP and can mostly shut down cleanly.

In our case, the status flag of a ring buffer slot is managed from both sides, the kernel space and the user space. This means that even though the kernel side might work as expected, the user space screws up and changes this flag right between the send(2) is triggered when the flag is changed to TP_STATUS_SENDING and a given skb is destructed after some time. Then, this will hit the BUG macro. Instead, we relax this condition with a WARN_ON_ONCE macro, so that the user is aware of this situation. I've tested it and the system still behaves /stable/, so in accordance with the above comment, we should rather relax this behavior with a warning.

Quoted 1 line
Signed-off-by Daniel Borkmann <[email protected]>

I would like this check to simply be deleted completely.

As you said, it's a user changable value, therefore we cannot use it for kernel side internal consistency checks at all.

author
#2 lore raw
Quoted 25 lines · David Miller

From: Daniel Borkmann <[email protected]>
Date: Thu, 09 Aug 2012 18:39:05 +0200

Quoted 22 lines

Here's a quote of the comment about the BUG macro from asm-generic/bug.h:

Don't use BUG() or BUG_ON() unless there's really no way out; one example might be detecting data structure corruption in the middle of an operation that can't be backed out of. If the (sub)system can somehow continue operating, perhaps with reduced functionality, it's probably not BUG-worthy.

If you're tempted to BUG(), think again: is completely giving up really the *only* solution? There are usually better options, where users don't need to reboot ASAP and can mostly shut down cleanly.

In our case, the status flag of a ring buffer slot is managed from both sides, the kernel space and the user space. This means that even though the kernel side might work as expected, the user space screws up and changes this flag right between the send(2) is triggered when the flag is changed to TP_STATUS_SENDING and a given skb is destructed after some time. Then, this will hit the BUG macro. Instead, we relax this condition with a WARN_ON_ONCE macro, so that the user is aware of this situation. I've tested it and the system still behaves /stable/, so in accordance with the above comment, we should rather relax this behavior with a warning.

Quoted 6 lines
Quoted 1 line
Signed-off-by Daniel Borkmann <[email protected]>

I would like this check to simply be deleted completely.

As you said, it's a user changable value, therefore we cannot use it for kernel side internal consistency checks at all.

Thanks for the feedback, I will resend a proper patch!

Best,
Daniel