Re: [PATCH] xfrm : lock input tasklet skb queue

[Joerg Vehlow]

Hi,

I already send a patch on 2019-09-09 to this mailing list with a similar 
issue[1].
Sadly no replies, although this is a huge bug in the rt kernel.
I fixed it a bit differently, using smaller locked regions.
You have also propably a bug in your patch, because trans->queue.lock is
no initialized by __skb_queue_head_init (in xfrm_input_init)
Jörg

[1] https://lkml.org/lkml/2019/9/9/111

Am 20.10.2019 um 17:46 schrieb Tom Rix:
On PREEMPT_RT_FULL while running netperf, a corruption
of the skb queue causes an oops.

This appears to be caused by a race condition here
         __skb_queue_tail(&trans->queue, skb);
         tasklet_schedule(&trans->tasklet);
Where the queue is changed before the tasklet is locked by
tasklet_schedule.

The fix is to use the skb queue lock.

Signed-off-by: Tom Rix <t...@redhat.com>
---
  net/xfrm/xfrm_input.c | 11 ++++++++++-
  1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index 9b599ed66d97..226dead86828 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -758,12 +758,16 @@ static void xfrm_trans_reinject(unsigned long data)
      struct xfrm_trans_tasklet *trans = (void *)data;
      struct sk_buff_head queue;
      struct sk_buff *skb;
+    unsigned long flags;

      __skb_queue_head_init(&queue);
+    spin_lock_irqsave(&trans->queue.lock, flags);
      skb_queue_splice_init(&trans->queue, &queue);

      while ((skb = __skb_dequeue(&queue)))
          XFRM_TRANS_SKB_CB(skb)->finish(dev_net(skb->dev), NULL, skb);
+
+    spin_unlock_irqrestore(&trans->queue.lock, flags);
  }

  int xfrm_trans_queue(struct sk_buff *skb,
@@ -771,15 +775,20 @@ int xfrm_trans_queue(struct sk_buff *skb,
                     struct sk_buff *))
  {
      struct xfrm_trans_tasklet *trans;
+    unsigned long flags;

      trans = this_cpu_ptr(&xfrm_trans_tasklet);
+    spin_lock_irqsave(&trans->queue.lock, flags);

-    if (skb_queue_len(&trans->queue) >= netdev_max_backlog)
+    if (skb_queue_len(&trans->queue) >= netdev_max_backlog) {
+        spin_unlock_irqrestore(&trans->queue.lock, flags);
          return -ENOBUFS;
+    }

      XFRM_TRANS_SKB_CB(skb)->finish = finish;
      __skb_queue_tail(&trans->queue, skb);
      tasklet_schedule(&trans->tasklet);
+    spin_unlock_irqrestore(&trans->queue.lock, flags);
      return 0;
  }
  EXPORT_SYMBOL(xfrm_trans_queue);