diff mbox series

[for-next] rdma_rxe: fixed bug in rxe_rcv_mcast_pkt

Message ID 20201008203651.256958-1-rpearson@hpe.com (mailing list archive)
State Accepted
Delegated to: Jason Gunthorpe
Headers show
Series [for-next] rdma_rxe: fixed bug in rxe_rcv_mcast_pkt | expand

Commit Message

Bob Pearson Oct. 8, 2020, 8:36 p.m. UTC
- The changes referenced below replaced sbk_clone by
    taking additional references, passing the skb along and
    then freeing the skb. This deleted the packets before
    they could be processed and additionally passed bad data
    in each packet. Since pkt is stored in skb->cb
    changing pkt->qp changed it for all the packets.
  - Replace skb_get by sbk_clone in rxe_rcv_mcast_pkt for
    cases where multiple QPs are receiving multicast packets
    on the same address.
  - Delete kfree_skb because the packets need to live until
    they have been processed by each QP. They are freed later.

Signed-off-by: Bob Pearson <rpearson@hpe.com>
Fixes: 86af61764151 ("IB/rxe: remove unnecessary skb_clone")
Fixes: fe896ceb5772 ("IB/rxe: replace refcount_inc with skb_get")
---
 drivers/infiniband/sw/rxe/rxe_recv.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

Comments

Jason Gunthorpe Oct. 8, 2020, 11:33 p.m. UTC | #1
On Thu, Oct 08, 2020 at 03:36:52PM -0500, Bob Pearson wrote:
> - The changes referenced below replaced sbk_clone by
>     taking additional references, passing the skb along and
>     then freeing the skb. This deleted the packets before
>     they could be processed and additionally passed bad data
>     in each packet. Since pkt is stored in skb->cb
>     changing pkt->qp changed it for all the packets.
>   - Replace skb_get by sbk_clone in rxe_rcv_mcast_pkt for
>     cases where multiple QPs are receiving multicast packets
>     on the same address.
>   - Delete kfree_skb because the packets need to live until
>     they have been processed by each QP. They are freed later.
> 
> Signed-off-by: Bob Pearson <rpearson@hpe.com>
> Fixes: 86af61764151 ("IB/rxe: remove unnecessary skb_clone")
> Fixes: fe896ceb5772 ("IB/rxe: replace refcount_inc with skb_get")

Fixes lines go before the Signed-off-by

Makes sense to me, applied to for-next

Thanks,
Jasno
Bob Pearson Oct. 9, 2020, 5:02 p.m. UTC | #2
On 10/8/20 6:33 PM, Jason Gunthorpe wrote:
> On Thu, Oct 08, 2020 at 03:36:52PM -0500, Bob Pearson wrote:
>> - The changes referenced below replaced sbk_clone by
>>     taking additional references, passing the skb along and
>>     then freeing the skb. This deleted the packets before
>>     they could be processed and additionally passed bad data
>>     in each packet. Since pkt is stored in skb->cb
>>     changing pkt->qp changed it for all the packets.
>>   - Replace skb_get by sbk_clone in rxe_rcv_mcast_pkt for
>>     cases where multiple QPs are receiving multicast packets
>>     on the same address.
>>   - Delete kfree_skb because the packets need to live until
>>     they have been processed by each QP. They are freed later.
>>
>> Signed-off-by: Bob Pearson <rpearson@hpe.com>
>> Fixes: 86af61764151 ("IB/rxe: remove unnecessary skb_clone")
>> Fixes: fe896ceb5772 ("IB/rxe: replace refcount_inc with skb_get")
> 
> Fixes lines go before the Signed-off-by
> 
> Makes sense to me, applied to for-next
> 
> Thanks,
> Jasno
> 
Thanks.
diff mbox series

Patch

diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c
index a9be4efba5f3..6b78e2ff6709 100644
--- a/drivers/infiniband/sw/rxe/rxe_recv.c
+++ b/drivers/infiniband/sw/rxe/rxe_recv.c
@@ -233,6 +233,8 @@  static void rxe_rcv_mcast_pkt(struct rxe_dev *rxe, struct sk_buff *skb)
 	struct rxe_mc_elem *mce;
 	struct rxe_qp *qp;
 	union ib_gid dgid;
+	struct sk_buff *per_qp_skb;
+	struct rxe_pkt_info *per_qp_pkt;
 	int err;
 
 	if (skb->protocol == htons(ETH_P_IP))
@@ -261,21 +263,26 @@  static void rxe_rcv_mcast_pkt(struct rxe_dev *rxe, struct sk_buff *skb)
 		if (err)
 			continue;
 
-		/* if *not* the last qp in the list
-		 * increase the users of the skb then post to the next qp
+		/* for all but the last qp create a new clone of the
+		 * skb and pass to the qp.
 		 */
 		if (mce->qp_list.next != &mcg->qp_list)
-			skb_get(skb);
+			per_qp_skb = skb_clone(skb, GFP_ATOMIC);
+		else
+			per_qp_skb = skb;
 
-		pkt->qp = qp;
+		per_qp_pkt = SKB_TO_PKT(per_qp_skb);
+		per_qp_pkt->qp = qp;
 		rxe_add_ref(qp);
-		rxe_rcv_pkt(pkt, skb);
+		rxe_rcv_pkt(per_qp_pkt, per_qp_skb);
 	}
 
 	spin_unlock_bh(&mcg->mcg_lock);
 
 	rxe_drop_ref(mcg);	/* drop ref from rxe_pool_get_key. */
 
+	return;
+
 err1:
 	kfree_skb(skb);
 }