diff mbox series

[net-next,1/2] udp: Refactor udp_read_skb()

Message ID 03db9765fe1ef0f61bfc87fc68b5a95b4126aa4e.1663143016.git.peilin.ye@bytedance.com (mailing list archive)
State Superseded
Delegated to: Netdev Maintainers
Headers show
Series [net-next,1/2] udp: Refactor udp_read_skb() | expand

Checks

Context Check Description
netdev/tree_selection success Clearly marked for net-next
netdev/apply fail Patch does not apply to net-next

Commit Message

Peilin Ye Sept. 14, 2022, 8:15 a.m. UTC 
From: Peilin Ye <peilin.ye@bytedance.com>

Delete the unnecessary while loop in udp_read_skb() for readability.
Additionally, since recv_actor() cannot return a value greater than
skb->len (see sk_psock_verdict_recv()), remove the redundant check.

Suggested-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Peilin Ye <peilin.ye@bytedance.com>
---
Depends on:

"[PATCH net v2] net: Use WARN_ON_ONCE() in {tcp,udp}_read_skb()"
https://lore.kernel.org/all/20220913184016.16095-1-yepeilin.cs@gmail.com/

Thanks,
Peilin Ye

 net/ipv4/udp.c | 46 +++++++++++++++++-----------------------------
 1 file changed, 17 insertions(+), 29 deletions(-)

Comments

Jakub Kicinski Sept. 21, 2022, 12:38 a.m. UTC | #1 
On Wed, 14 Sep 2022 01:15:30 -0700 Peilin Ye wrote:
> Delete the unnecessary while loop in udp_read_skb() for readability.
> Additionally, since recv_actor() cannot return a value greater than
> skb->len (see sk_psock_verdict_recv()), remove the redundant check.

These don't apply cleanly, please rebase?
Jakub Kicinski Sept. 21, 2022, 12:40 a.m. UTC | #2 
On Tue, 20 Sep 2022 17:38:59 -0700 Jakub Kicinski wrote:
> On Wed, 14 Sep 2022 01:15:30 -0700 Peilin Ye wrote:
> > Delete the unnecessary while loop in udp_read_skb() for readability.
> > Additionally, since recv_actor() cannot return a value greater than
> > skb->len (see sk_psock_verdict_recv()), remove the redundant check.  
> 
> These don't apply cleanly, please rebase?

Ah, it's the WARN_ON_ONCE() change. In that case please resend after
net is merged into net-next (Thu evening).
Peilin Ye Sept. 21, 2022, 1:05 a.m. UTC | #3 
On Tue, Sep 20, 2022 at 05:40:52PM -0700, Jakub Kicinski wrote:
> On Tue, 20 Sep 2022 17:38:59 -0700 Jakub Kicinski wrote:
> > On Wed, 14 Sep 2022 01:15:30 -0700 Peilin Ye wrote:
> > > Delete the unnecessary while loop in udp_read_skb() for readability.
> > > Additionally, since recv_actor() cannot return a value greater than
> > > skb->len (see sk_psock_verdict_recv()), remove the redundant check.  
> > 
> > These don't apply cleanly, please rebase?
> 
> Ah, it's the WARN_ON_ONCE() change. In that case please resend after
> net is merged into net-next (Thu evening).

Sure, but only the TCP part was merged [1] into net.  I just sent the
UDP part again [2], and will resend this patchset after both [1] and [2]
are merged into net-next.  Thanks!

[1] https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/commit/?id=96628951869c0dedf0377adca01c8675172d8639
[2] https://lore.kernel.org/netdev/20220921005915.2697-1-yepeilin.cs@gmail.com/T/#u

Peilin Ye
diff mbox series

Patch

diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 560d9eadeaa5..d63118ce5900 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1801,41 +1801,29 @@  EXPORT_SYMBOL(__skb_recv_udp);
 
 int udp_read_skb(struct sock *sk, skb_read_actor_t recv_actor)
 {
-	int copied = 0;
-
-	while (1) {
-		struct sk_buff *skb;
-		int err, used;
-
-		skb = skb_recv_udp(sk, MSG_DONTWAIT, &err);
-		if (!skb)
-			return err;
+	struct sk_buff *skb;
+	int err, copied;
 
-		if (udp_lib_checksum_complete(skb)) {
-			__UDP_INC_STATS(sock_net(sk), UDP_MIB_CSUMERRORS,
-					IS_UDPLITE(sk));
-			__UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS,
-					IS_UDPLITE(sk));
-			atomic_inc(&sk->sk_drops);
-			kfree_skb(skb);
-			continue;
-		}
+try_again:
+	skb = skb_recv_udp(sk, MSG_DONTWAIT, &err);
+	if (!skb)
+		return err;
 
-		WARN_ON_ONCE(!skb_set_owner_sk_safe(skb, sk));
-		used = recv_actor(sk, skb);
-		if (used <= 0) {
-			if (!copied)
-				copied = used;
-			kfree_skb(skb);
-			break;
-		} else if (used <= skb->len) {
-			copied += used;
-		}
+	if (udp_lib_checksum_complete(skb)) {
+		int is_udplite = IS_UDPLITE(sk);
+		struct net *net = sock_net(sk);
 
+		__UDP_INC_STATS(net, UDP_MIB_CSUMERRORS, is_udplite);
+		__UDP_INC_STATS(net, UDP_MIB_INERRORS, is_udplite);
+		atomic_inc(&sk->sk_drops);
 		kfree_skb(skb);
-		break;
+		goto try_again;
 	}
 
+	WARN_ON_ONCE(!skb_set_owner_sk_safe(skb, sk));
+	copied = recv_actor(sk, skb);
+	kfree_skb(skb);
+
 	return copied;
 }
 EXPORT_SYMBOL(udp_read_skb);