diff options
author | David S. Miller <davem@davemloft.net> | 2015-07-10 18:16:37 -0700 |
---|---|---|
committer | David S. Miller <davem@davemloft.net> | 2015-07-10 18:16:37 -0700 |
commit | 7234e030194ccc27c48d3bde1ad66e1729c55a7d (patch) | |
tree | 8868fe137839c66fcef2a435d8d1b38fd0cd7533 /net/ipv4/inet_diag.c | |
parent | bridge: fix potential crash in __netdev_pick_tx() (diff) | |
parent | net: call rcu_read_lock early in process_backlog (diff) | |
download | linux-dev-7234e030194ccc27c48d3bde1ad66e1729c55a7d.tar.xz linux-dev-7234e030194ccc27c48d3bde1ad66e1729c55a7d.zip |
Merge branch 'netdev_unregister_races'
Julian Anastasov says:
====================
net: fixes for device unregistration
Test script from Eric W. Biederman can catch a problem
where packets from backlog are processed long after the last
synchronize_net call. This can be reproduced after few tests
if commit 381c759d9916 ("ipv4: Avoid crashing in ip_error")
is reverted for the test. Incoming packets do not hold
reference to device but even if they do, subsystems do not
expect packets to fly during and after the NETDEV_UNREGISTER
event.
The first fix has the cost of netif_running check in fast path.
The second fix calls rcu_read_lock while local IRQ is disabled,
I hope this is not against the rules.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/ipv4/inet_diag.c')
0 files changed, 0 insertions, 0 deletions