nfp: Remove rcu_read_lock() around XDP program invocation
authorToke Høiland-Jørgensen <toke@redhat.com>
Thu, 24 Jun 2021 16:06:04 +0000 (18:06 +0200)
committerDaniel Borkmann <daniel@iogearbox.net>
Thu, 24 Jun 2021 17:45:26 +0000 (19:45 +0200)
commitd5789621b658369b21bd13446bab8102cf75df65
treebc918a2b0cdaa5b22c9bcf5c0bac6b2c3a9ccdda
parentc4411b371c104e65efb531ebd4d8892c568e3a29
nfp: Remove rcu_read_lock() around XDP program invocation

The nfp driver has rcu_read_lock()/rcu_read_unlock() pairs around XDP
program invocations. However, the actual lifetime of the objects referred
by the XDP program invocation is longer, all the way through to the call to
xdp_do_flush(), making the scope of the rcu_read_lock() too small.

While this is not actually an issue for the nfp driver because it doesn't
support XDP_REDIRECT (and thus doesn't call xdp_do_flush()), the
rcu_read_lock() is still unneeded. And With the addition of RCU annotations
to the XDP_REDIRECT map types that take bh execution into account, lockdep
even understands this to be safe, so there's really no reason to keep it
around.

Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Cc: oss-drivers@netronome.com
Link: https://lore.kernel.org/bpf/20210624160609.292325-15-toke@redhat.com
drivers/net/ethernet/netronome/nfp/nfp_net_common.c