Revision 580e4273d7a883ececfefa692c1f96bdbacb99b5 authored by Cong Wang on 02 October 2020, 19:13:34 UTC, committed by David S. Miller on 04 October 2020, 21:53:06 UTC
Although we take RTNL on dump path, it is possible to skip RTNL on insertion path. So the following race condition is possible: rtnl_lock() // no rtnl lock mutex_lock(&idrinfo->lock); // insert ERR_PTR(-EBUSY) mutex_unlock(&idrinfo->lock); tc_dump_action() rtnl_unlock() So we have to skip those temporary -EBUSY entries on dump path too. Reported-and-tested-by: syzbot+b47bc4f247856fb4d9e1@syzkaller.appspotmail.com Fixes: 0fedc63fadf0 ("net_sched: commit action insertions together") Cc: Vlad Buslov <vladbu@mellanox.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Cc: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
1 parent 9a9e774
File | Mode | Size |
---|---|---|
heaps | ||
Kconfig | -rw-r--r-- | 2.3 KB |
Makefile | -rw-r--r-- | 472 bytes |
dma-buf.c | -rw-r--r-- | 37.9 KB |
dma-fence-array.c | -rw-r--r-- | 5.7 KB |
dma-fence-chain.c | -rw-r--r-- | 6.6 KB |
dma-fence.c | -rw-r--r-- | 26.7 KB |
dma-heap.c | -rw-r--r-- | 6.7 KB |
dma-resv.c | -rw-r--r-- | 16.2 KB |
selftest.c | -rw-r--r-- | 3.0 KB |
selftest.h | -rw-r--r-- | 549 bytes |
selftests.h | -rw-r--r-- | 584 bytes |
seqno-fence.c | -rw-r--r-- | 1.8 KB |
st-dma-fence-chain.c | -rw-r--r-- | 13.6 KB |
st-dma-fence.c | -rw-r--r-- | 10.4 KB |
sw_sync.c | -rw-r--r-- | 9.5 KB |
sync_debug.c | -rw-r--r-- | 4.6 KB |
sync_debug.h | -rw-r--r-- | 1.8 KB |
sync_file.c | -rw-r--r-- | 11.0 KB |
sync_trace.h | -rw-r--r-- | 774 bytes |
udmabuf.c | -rw-r--r-- | 7.5 KB |
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...