Skip to contentSkip to conversation

net_sched: make qdisc_reset() smaller

PATCH Applied Eric Dumazet 5 participants 5 messages Apr 14, 2022 – Apr 17, 2022

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

PATCH 1/1 net_sched: make qdisc_reset() smaller

From: Eric Dumazet <[email protected]>

For some unknown reason qdisc_reset() is using
a convoluted way of freeing two lists of skbs.

Use __skb_queue_purge() instead.

---
 net/sched/sch_generic.c | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)
net/sched/sch_generic.c +2 -10
index 5bab9f8b8f453526185c3b6df57065450b1e3d89..dba0b3e24af5e84f7116ae9b6fdb6f66b01a896c 100644
@@ -1019,22 +1019,14 @@ EXPORT_SYMBOL(qdisc_create_dflt);
 void qdisc_reset(struct Qdisc *qdisc)
 {
 	const struct Qdisc_ops *ops = qdisc->ops;
-	struct sk_buff *skb, *tmp;
 
 	trace_qdisc_reset(qdisc);
 
 	if (ops->reset)
 		ops->reset(qdisc);
 
-	skb_queue_walk_safe(&qdisc->gso_skb, skb, tmp) {
-		__skb_unlink(skb, &qdisc->gso_skb);
-		kfree_skb_list(skb);
-	}
-
-	skb_queue_walk_safe(&qdisc->skb_bad_txq, skb, tmp) {
-		__skb_unlink(skb, &qdisc->skb_bad_txq);
-		kfree_skb_list(skb);
-	}
+	__skb_queue_purge(&qdisc->gso_skb);
+	__skb_queue_purge(&qdisc->skb_bad_txq);
 
 	qdisc->q.qlen = 0;
 	qdisc->qstats.backlog = 0;
Signature · 1 line
2.35.1.1178.g4f1659d476-goog
maintainer Acked-by on v1 1/1
#1 lore raw
Quoted 29 lines · Eric Dumazet

From: Eric Dumazet <[email protected]>

For some unknown reason qdisc_reset() is using
a convoluted way of freeing two lists of skbs.

Use __skb_queue_purge() instead.

---
  net/sched/sch_generic.c | 12 ++----------
  1 file changed, 2 insertions(+), 10 deletions(-)
net/sched/sch_generic.c +0 -3
index 5bab9f8b8f453526185c3b6df57065450b1e3d89..dba0b3e24af5e84f7116ae9b6fdb6f66b01a896c 100644
@@ -1019,22 +1019,14 @@ EXPORT_SYMBOL(qdisc_create_dflt);
  void qdisc_reset(struct Qdisc *qdisc)
  {
  	const struct Qdisc_ops *ops = qdisc->ops;
-	struct sk_buff *skb, *tmp;
 
  	trace_qdisc_reset(qdisc);
 
  	if (ops->reset)
  		ops->reset(qdisc);
 
-	skb_queue_walk_safe(&qdisc->gso_skb, skb, tmp) {
-		__skb_unlink(skb, &qdisc->gso_skb);

net/sched/sch_generic.c EXPORT_SYMBOL(qdisc_create_dflt); · line 1030

Quoted 10 lines
net/sched/sch_generic.c +2 -7
-		kfree_skb_list(skb);
-	}
-
-	skb_queue_walk_safe(&qdisc->skb_bad_txq, skb, tmp) {
-		__skb_unlink(skb, &qdisc->skb_bad_txq);
-		kfree_skb_list(skb);
-	}
+	__skb_queue_purge(&qdisc->gso_skb);
+	__skb_queue_purge(&qdisc->skb_bad_txq);
 

cheers,
jamal

#2 lore raw
In reply to Eric Dumazet
This patch was applied to netdev/net-next.git (master) by Jakub Kicinski <[email protected]>:

Hello:

This patch was applied to netdev/net-next.git (master) by Jakub Kicinski <[email protected]>:

Quoted 8 lines · On Wed, 13 Apr 2022 18:10:04 -0700 you wrote:

From: Eric Dumazet <[email protected]>

For some unknown reason qdisc_reset() is using
a convoluted way of freeing two lists of skbs.

Use __skb_queue_purge() instead.

[...]

Here is the summary with links:
- [net-next] net_sched: make qdisc_reset() smaller

    https://git.kernel.org/netdev/net-next/c/c9a40d1c87e9

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
#3 lore raw
In reply to Eric Dumazet
Quoted 7 lines · Eric Dumazet

From: Eric Dumazet <[email protected]>

For some unknown reason qdisc_reset() is using
a convoluted way of freeing two lists of skbs.

Use __skb_queue_purge() instead.

[...]

net/sched/sch_generic.c EXPORT_SYMBOL(qdisc_create_dflt); · line 1031

Quoted 4 lines
net/sched/sch_generic.c +0 -3
-       skb_queue_walk_safe(&qdisc->gso_skb, skb, tmp) {
-               __skb_unlink(skb, &qdisc->gso_skb);
-               kfree_skb_list(skb);

Isn't it precisely because of this kfree_skb_list()?

Thanks.

#4 lore raw
Quoted 8 lines · Cong Wang
Quoted 6 lines · Eric Dumazet

From: Eric Dumazet <[email protected]>

For some unknown reason qdisc_reset() is using
a convoluted way of freeing two lists of skbs.

Quoted 8 lines
Quoted 1 line

Use __skb_queue_purge() instead.

[...]

net/sched/sch_generic.c EXPORT_SYMBOL(qdisc_create_dflt); · line 1031

Quoted 4 lines
net/sched/sch_generic.c +0 -3
-       skb_queue_walk_safe(&qdisc->gso_skb, skb, tmp) {
-               __skb_unlink(skb, &qdisc->gso_skb);
-               kfree_skb_list(skb);

Isn't it precisely because of this kfree_skb_list()?

Note the __skb_unlink(skb, &qdisc->gso_skb) which happens at the line above.

__skb_unlink(...)
{
...
skb->next = skb->prev = NULL;
...
}

This means skb->next is NULL, thus there is no list of skb attached to @skb ?

kfree_skb_list(skb) is thus a convoluted way to call kfree_skb(skb)

It seems to me that this construct had been copy/pasted from elsewhere.