diff options
| author |   Eric Dumazet <eric.dumazet@gmail.com> | 2010-09-13 12:24:54 +0000 |
|---|---|---|
| committer |   David S. Miller <davem@davemloft.net> | 2010-09-14 14:27:29 -0700 |
| commit | ef885afbf8a37689afc1d9d545e2f3e7a8276c17 (patch) | |
| tree | 8defdc814be3668f906213b5f32729341eec2aa5 /net/core/stream.c | |
| parent | bonding: correctly process non-linear skbs (diff) | |
| download | linux-dev-ef885afbf8a37689afc1d9d545e2f3e7a8276c17.tar.xz linux-dev-ef885afbf8a37689afc1d9d545e2f3e7a8276c17.zip | |
net: use rcu_barrier() in rollback_registered_many
netdev_wait_allrefs() waits that all references to a device vanishes.
It currently uses a _very_ pessimistic 250 ms delay between each probe.
Some users reported that no more than 4 devices can be dismantled per
second, this is a pretty serious problem for some setups.
Most of the time, a refcount is about to be released by an RCU callback,
that is still in flight because rollback_registered_many() uses a
synchronize_rcu() call instead of rcu_barrier(). Problem is visible if
number of online cpus is one, because synchronize_rcu() is then a no op.
time to remove 50 ipip tunnels on a UP machine :
before patch : real 11.910s
after patch : real 1.250s
Reported-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Reported-by: Octavian Purdila <opurdila@ixiacom.com>
Reported-by: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/core/stream.c')
0 files changed, 0 insertions, 0 deletions