IB/IPoIB: Replace vlan_rwsem with the netdev instance lock
authorCosmin Ratiu <cratiu@nvidia.com>
Wed, 21 May 2025 12:08:59 +0000 (15:08 +0300)
committerJakub Kicinski <kuba@kernel.org>
Thu, 22 May 2025 16:15:04 +0000 (09:15 -0700)
commit463e5176969795a1d80205842e210c539c16fc81
tree749d367b6734d0033b37c5065b07a8dcb057a7f1
parent5f85120e7462c9fc51cb9c52bc6e81b13d4c93e1
IB/IPoIB: Replace vlan_rwsem with the netdev instance lock

vlan_rwsem was added more than a decade ago to work around a deadlock
involving the original mutex being acquired twice, once from the wq.
Subsequent changes then tweaked it to partially protect access to
ipoib_dev_priv->child_intfs together with the RTNL. Flushing the wq
synchronously was also since then refactored to happen separately.

This semaphore unfortunately prevents updating ipoib to work with
devices that require the netdev lock, because of lock ordering issues
between RTNL, vlan_rwsem and the netdev instance locks of parent and
child devices.

To uncomplicate things, this commit replaces vlan_rwsem with the netdev
instance lock of the parent device. Both parent child_intfs list and the
children's list membership in it require holding the parent netdev
instance lock.

All call paths were carefully reviewed and no-longer-needed ASSERT_RTNL
calls were dropped. Some non-trivial changes:
- ipoib_match_gid_pkey_addr() now only acquires the instance lock and
  iterates through child_intfs for the first level of recursion (the
  parent), as it's not possible to have multiple levels of nested
  subinterfaces.
- ipoib_open() and ipoib_stop() schedule tasks on the global workqueue
  to open/stop child interfaces to avoid potentially acquiring nested
  netdev instance locks. To avoid the device going away between the task
  scheduling and execution, netdev_hold/netdev_put are used.

Signed-off-by: Cosmin Ratiu <cratiu@nvidia.com>
Reviewed-by: Carolina Jubran <cjubran@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/1747829342-1018757-3-git-send-email-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
drivers/infiniband/ulp/ipoib/ipoib.h
drivers/infiniband/ulp/ipoib/ipoib_ib.c
drivers/infiniband/ulp/ipoib/ipoib_main.c
drivers/infiniband/ulp/ipoib/ipoib_vlan.c