[RFC,net,0/2] Fix rcu_read_lock issues in netdev-genl

Message

Joe Damato Nov. 12, 2024, 6:13 p.m. UTC

Greetings:

Paolo reported a splat [1] when running the new selftest for busy poll.
I confirmed and reproduced this splat locally.

This series proposed 2 patches, which:
  - Patch 1:
    - Adds a helper function to reduce code duplication that sets the
      error, extack, napi, etc.
    - Fixes a similar issue in an older commit and CCs stable as this
      fix could be backported.
  - Patch 2:
    - Uses the helper added in 1 to fix the recently added commit that
      adds netdev_nl_napi_set_doit which is exercised by the selftest
      triggering the splat that Paolo reported.

I retested locally after applying this series and confirmed that the
splat is fixed.

Note: I only CC'd stable on patch 1 because that code goes back a few
releases. patch 2 is fixing code merged very recently that does not yet
appear in any RC and so I've omit the CC for stable there. I've sent
this as an RFC because I am not sure if that's the right thing to do.

Let me know and I'll be happy to re-send (after 24hr) an official
series.

Thanks,
Joe

[1]: https://lore.kernel.org/netdev/[e-mail]/

Joe Damato (2):
  netdev-genl: Hold rcu_read_lock in napi_get
  netdev-genl: Hold rcu_read_lock in napi_set

 net/core/netdev-genl.c | 37 +++++++++++++++++++++++++------------
 1 file changed, 25 insertions(+), 12 deletions(-)


base-commit: a58f00ed24b849d449f7134fd5d86f07090fe2f5