* [IGMP Discuss] Should we use one lock for struct ip_mc_list ?
@ 2022-09-08 3:40 Hangbin Liu
0 siblings, 0 replies; only message in thread
From: Hangbin Liu @ 2022-09-08 3:40 UTC (permalink / raw)
To: netdev
Cc: David S. Miller, Hideaki YOSHIFUJI, David Ahern, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, Liu Jian, Taehee Yoo, Jiri Benc
Hi,
When reviewing commit 23d2b94043ca ("igmp: Add ip_mc_list lock in
ip_check_mc_rcu"). Jiri pointed that struct ip_mc_list is protected by
different lock.
In function ip_check_mc_rcu() and ip_mc_del_src(), the struct ip_mc_list
in in_dev->mc_list is protected by the lock of struct ip_mc_list itself.
But in function igmpv3_send_cr, the ip_mc_list in in_dev->mc_tomb is
protected by in_dev->mc_tomb_lock.
This is no clear doc about what is protected by ip_mc_list->lock.
Is it OK for a single field that be protected by different locks?
Taehee did an update for MLD by using one mc_lock in commit 63ed8de4be81
("mld: add mc_lock for protecting per-interface mld data"). Should we also
do this on IGMP?
Thanks
Hangbin
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2022-09-08 3:40 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2022-09-08 3:40 [IGMP Discuss] Should we use one lock for struct ip_mc_list ? Hangbin Liu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox