Re: [PATCH net 6/6] net/sched: qdisc_destroy() old ingress and clsact Qdiscs before grafting

[Jakub Kicinski]

On Fri,  5 May 2023 17:16:10 -0700 Peilin Ye wrote:
>   Thread 1 creates ingress Qdisc A (containing mini Qdisc a1 and a2), then
>   adds a flower filter X to A.
> 
>   Thread 2 creates another ingress Qdisc B (containing mini Qdisc b1 and
>   b2) to replace A, then adds a flower filter Y to B.
> 
>  Thread 1               A's refcnt   Thread 2
>   RTM_NEWQDISC (A, RTNL-locked)
>    qdisc_create(A)               1
>    qdisc_graft(A)                9
> 
>   RTM_NEWTFILTER (X, RTNL-lockless)
>    __tcf_qdisc_find(A)          10
>    tcf_chain0_head_change(A)
>    mini_qdisc_pair_swap(A) (1st)
>             |
>             |                         RTM_NEWQDISC (B, RTNL-locked)
>            RCU                   2     qdisc_graft(B)
>             |                    1     notify_and_destroy(A)
>             |
>    tcf_block_release(A)          0    RTM_NEWTFILTER (Y, RTNL-lockless)
>    qdisc_destroy(A)                    tcf_chain0_head_change(B)
>    tcf_chain0_head_change_cb_del(A)    mini_qdisc_pair_swap(B) (2nd)
>    mini_qdisc_pair_swap(A) (3rd)                |
>            ...                                 ...
> 
> Here, B calls mini_qdisc_pair_swap(), pointing eth0->miniq_ingress to its
> mini Qdisc, b1.  Then, A calls mini_qdisc_pair_swap() again during
> ingress_destroy(), setting eth0->miniq_ingress to NULL, so ingress packets
> on eth0 will not find filter Y in sch_handle_ingress().
> 
> This is just one of the possible consequences of concurrently accessing
> net_device::miniq_{in,e}gress pointers.  The point is clear, however:
> B's first call to mini_qdisc_pair_swap() should take place after A's
> last call, in qdisc_destroy().

Great analysis, thanks for squashing this bug.

Have you considered creating a fix more localized to the miniq
implementation? It seems that having per-device miniq pointers is
incompatible with using reference counted objects. So miniq is 
a more natural place to solve the problem. Otherwise workarounds 
in the core keep piling up (here qdisc_graft()).

Can we replace the rcu_assign_pointer in (3rd) with a cmpxchg()?
If active qdisc is neither a1 nor a2 we should leave the dev state
alone.