Avoid SKB coalescing if eor bit is set in one of the relevant
SKBs.

Change-Id: I64d4f9874f2e23b3fd03daddccbefed53e098028
Fixes: c134ecb87817 ("tcp: Make use of MSG_EOR in tcp_sendmsg")
Signed-off-by: Ilya Lesokhin <il...@mellanox.com>
---
 net/ipv4/tcp_output.c | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index e9f985e42405..b2bca373f8be 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2027,6 +2027,24 @@ static inline void tcp_mtu_check_reprobe(struct sock *sk)
        }
 }
 
+static bool tcp_can_coalesce_send_queue_head(struct sock *sk, int len)
+{
+       struct sk_buff *skb, *next;
+
+       skb = tcp_send_head(sk);
+       tcp_for_write_queue_from_safe(skb, next, sk) {
+               if (len <= skb->len)
+                       break;
+
+               if (unlikely(TCP_SKB_CB(skb)->eor))
+                       return false;
+
+               len -= skb->len;
+       }
+
+       return true;
+}
+
 /* Create a new MTU probe if we are ready.
  * MTU probe is regularly attempting to increase the path MTU by
  * deliberately sending larger packets.  This discovers routing
@@ -2099,6 +2117,9 @@ static int tcp_mtu_probe(struct sock *sk)
                        return 0;
        }
 
+       if (!tcp_can_coalesce_send_queue_head(sk, probe_size))
+               return -1;
+
        /* We're allowed to probe.  Build it now. */
        nskb = sk_stream_alloc_skb(sk, probe_size, GFP_ATOMIC, false);
        if (!nskb)
@@ -2134,6 +2155,10 @@ static int tcp_mtu_probe(struct sock *sk)
                        /* We've eaten all the data from this skb.
                         * Throw it away. */
                        TCP_SKB_CB(nskb)->tcp_flags |= 
TCP_SKB_CB(skb)->tcp_flags;
+                       /* If this is the last SKB we copy and eor is set
+                        * we need to propagate it to the new skb.
+                        */
+                       TCP_SKB_CB(nskb)->eor = TCP_SKB_CB(skb)->eor;
                        tcp_unlink_write_queue(skb, sk);
                        sk_wmem_free_skb(sk, skb);
                } else {
-- 
Changes from v3:
- Fix coding style

Changes from v2:                                                                
 
- Initialize SKB before tcp_for_write_queue_from_safe                           
 
- Move new logic to its own function
- Add comment about eor propagation
                                                                                
 
Changes from v1:                                                                
 
- Propagate eor when consuming the old SKB 
2.15.0.317.g14c63a9

Reply via email to