diff options
| author | Jakub Kicinski <kuba@kernel.org> | 2025-03-25 01:45:32 +0300 |
|---|---|---|
| committer | Jakub Kicinski <kuba@kernel.org> | 2025-03-25 20:06:49 +0300 |
| commit | 310ae9eb2617c62deedef8f121d7ca1ae774fa76 (patch) | |
| tree | 2257965b50c36fe428e5040ef6f040e675bbda35 /include/net/netdev_lock.h | |
| parent | 0a65dcf6249b75c841b4218426b0d246a805c7e0 (diff) | |
| download | linux-310ae9eb2617c62deedef8f121d7ca1ae774fa76.tar.xz | |
net: designate queue -> napi linking as "ops protected"
netdev netlink is the only reader of netdev_{,rx_}queue->napi,
and it already holds netdev->lock. Switch protection of
the writes to netdev->lock to "ops protected".
The expectation will be now that accessing queue->napi
will require netdev->lock for "ops locked" drivers, and
rtnl_lock for all other drivers.
Current "ops locked" drivers don't require any changes.
gve and netdevsim use _locked() helpers right next to
netif_queue_set_napi() so they must be holding the instance
lock. iavf doesn't call it. bnxt is a bit messy but all paths
seem locked.
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250324224537.248800-7-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Diffstat (limited to 'include/net/netdev_lock.h')
| -rw-r--r-- | include/net/netdev_lock.h | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/include/net/netdev_lock.h b/include/net/netdev_lock.h index efd302375ef2..1c0c9a94cc22 100644 --- a/include/net/netdev_lock.h +++ b/include/net/netdev_lock.h @@ -56,6 +56,14 @@ static inline void netdev_ops_assert_locked(const struct net_device *dev) ASSERT_RTNL(); } +static inline void +netdev_ops_assert_locked_or_invisible(const struct net_device *dev) +{ + if (dev->reg_state == NETREG_REGISTERED || + dev->reg_state == NETREG_UNREGISTERING) + netdev_ops_assert_locked(dev); +} + static inline int netdev_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b) { |