From: Jakub Sitnicki <jakub@cloudflare.com>
To: John Fastabend <john.fastabend@gmail.com>
Cc: bpf@vger.kernel.org, netdev@vger.kernel.org,
kernel-team@cloudflare.com, Eric Dumazet <edumazet@google.com>,
Lorenz Bauer <lmb@cloudflare.com>,
Martin KaFai Lau <kafai@fb.com>
Subject: Re: [PATCH bpf-next v2 02/11] net, sk_msg: Annotate lockless access to sk_prot on clone
Date: Tue, 21 Jan 2020 13:42:22 +0100 [thread overview]
Message-ID: <8736c9rnf5.fsf@cloudflare.com> (raw)
In-Reply-To: <874kwqroap.fsf@cloudflare.com>
On Mon, Jan 20, 2020 at 07:11 PM CET, Jakub Sitnicki wrote:
> On Mon, Jan 20, 2020 at 06:00 PM CET, John Fastabend wrote:
>> Jakub Sitnicki wrote:
>>> On Sun, Jan 12, 2020 at 12:14 AM CET, John Fastabend wrote:
>>> > Jakub Sitnicki wrote:
>>> >> sk_msg and ULP frameworks override protocol callbacks pointer in
>>> >> sk->sk_prot, while TCP accesses it locklessly when cloning the listening
>>> >> socket.
>>> >>
>>> >> Once we enable use of listening sockets with sockmap (and hence sk_msg),
>>> >> there can be shared access to sk->sk_prot if socket is getting cloned while
>>> >> being inserted/deleted to/from the sockmap from another CPU. Mark the
>>> >> shared access with READ_ONCE/WRITE_ONCE annotations.
>>> >>
>>> >> Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
>>> >
>>> > In sockmap side I fixed this by wrapping the access in a lock_sock[0]. So
>>> > Do you think this is still needed with that in mind? The bpf_clone call
>>> > is using sk_prot_creater and also setting the newsk's proto field. Even
>>> > if the listening parent sock was being deleted in parallel would that be
>>> > a problem? We don't touch sk_prot_creator from the tear down path. I've
>>> > only scanned the 3..11 patches so maybe the answer is below. If that is
>>> > the case probably an improved commit message would be helpful.
>>>
>>> I think it is needed. Not because of tcp_bpf_clone or that we access
>>> listener's sk_prot_creator from there, if I'm grasping your question.
>>>
>>> Either way I'm glad this came up. Let's go though my reasoning and
>>> verify it. tcp stack accesses the listener sk_prot while cloning it:
>>>
>>> tcp_v4_rcv
>>> sk = __inet_lookup_skb(...)
>>> tcp_check_req(sk)
>>> inet_csk(sk)->icsk_af_ops->syn_recv_sock
>>> tcp_v4_syn_recv_sock
>>> tcp_create_openreq_child
>>> inet_csk_clone_lock
>>> sk_clone_lock
>>> READ_ONCE(sk->sk_prot)
>>>
>>> It grabs a reference to the listener, but doesn't grab the sk_lock.
>>>
>>> On another CPU we can be inserting/removing the listener socket from the
>>> sockmap and writing to its sk_prot. We have the update and the remove
>>> path:
>>>
>>> sock_map_ops->map_update_elem
>>> sock_map_update_elem
>>> sock_map_update_common
>>> sock_map_link_no_progs
>>> tcp_bpf_init
>>> tcp_bpf_update_sk_prot
>>> sk_psock_update_proto
>>> WRITE_ONCE(sk->sk_prot, ops)
>>>
>>> sock_map_ops->map_delete_elem
>>> sock_map_delete_elem
>>> __sock_map_delete
>>> sock_map_unref
>>> sk_psock_put
>>> sk_psock_drop
>>> sk_psock_restore_proto
>>> tcp_update_ulp
>>> WRITE_ONCE(sk->sk_prot, proto)
>>>
>>> Following the guidelines from KTSAN project [0], sk_prot looks like a
>>> candidate for annotating it. At least on these 3 call paths.
>>>
>>> If that sounds correct, I can add it to the patch description.
>>>
>>> Thanks,
>>> -jkbs
>>>
>>> [0] https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONCE
>>
>> Hi Jakub, can push this to bpf tree as well? There is another case
>> already in-kernel where this is needed. If the map is removed while
>> a recvmsg is in flight.
>>
>> tcp_bpf_recvmsg()
>> psock = sk_psock_get(sk) <- refcnt 2
>> lock_sock(sk);
>> ...
>> sock_map_free() <- refcnt 1
>> release_sock(sk)
>> sk_psock_put() <- refcnt 0
>>
>> Then can you add this diff as well I got a bit too carried away
>> with that. If your busy I can do it as well if you want. Thanks!
>
> Hi John, I get the race between map_free and tcp_bpf_recvmsg, and how we
> end up dropping psock on a path where we don't hold the sock lock. What
> a rare case, since we don't destory maps that often usually.
>
> However, I'm not sure I follow where shared lockless access to
> sk->sk_prot is in this case?
>
> Perhaps between drop path:
>
> sk_psock_put
> sk_psock_drop
> sk_psock_restore_proto
> WRITE_ONCE(sk->sk_prot, proto)
>
> ... and update path where we grab sk_callback_lock a little too late,
> that is after updating the proto?
>
> sock_map_update_common
> sock_map_link
> tcp_bpf_init
> tcp_bpf_update_sk_prot
> sk_psock_update_proto
> WRITE_ONCE(sk->sk_prot, ops)
>
> I'm getting v3 ready to post, so happy to help you spin these bits.
> I'll need to do it with a fresh head tomorrow, though.
>
> If I don't see any patches from you hit the ML, I'll split out the
> chunks that annotate sk_prot access in sk_psock_{retore,update}_proto
> and post them together with the revert you suggested below.
I've sent out the partial revert you wanted:
https://lore.kernel.org/netdev/20200121123147.706666-1-jakub@cloudflare.com/T/#u
But otherwise didn't manage to convince myself that we need to annotate
access to sk_prot with READ/WRITE_ONCE in sk_psock_update/restore_proto.
Instead, I believe we might need to the extend critical section that
grabs sk_callback_lock in sock_map_link over tcp_bpf_init/reinit() to
serialize the writers.
Unless I'm missing the point here and you had some other race in mind?
-jkbs
next prev parent reply other threads:[~2020-01-21 12:42 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-01-10 10:50 [PATCH bpf-next v2 00/11] Extend SOCKMAP to store listening sockets Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 01/11] bpf, sk_msg: Don't reset saved sock proto on restore Jakub Sitnicki
2020-01-11 22:50 ` John Fastabend
2020-01-10 10:50 ` [PATCH bpf-next v2 02/11] net, sk_msg: Annotate lockless access to sk_prot on clone Jakub Sitnicki
2020-01-11 23:14 ` John Fastabend
2020-01-13 15:09 ` Jakub Sitnicki
2020-01-14 3:14 ` John Fastabend
2020-01-20 17:00 ` John Fastabend
2020-01-20 18:11 ` Jakub Sitnicki
2020-01-21 12:42 ` Jakub Sitnicki [this message]
2020-01-10 10:50 ` [PATCH bpf-next v2 03/11] net, sk_msg: Clear sk_user_data pointer on clone if tagged Jakub Sitnicki
2020-01-11 23:38 ` John Fastabend
2020-01-12 12:55 ` kbuild test robot
2020-01-13 20:15 ` Martin Lau
2020-01-14 16:04 ` Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 04/11] tcp_bpf: Don't let child socket inherit parent protocol ops on copy Jakub Sitnicki
2020-01-11 2:42 ` kbuild test robot
2020-01-11 3:02 ` kbuild test robot
2020-01-11 23:48 ` John Fastabend
2020-01-13 22:31 ` Jakub Sitnicki
2020-01-13 22:23 ` Martin Lau
2020-01-13 22:42 ` Jakub Sitnicki
2020-01-13 23:23 ` Martin Lau
2020-01-10 10:50 ` [PATCH bpf-next v2 05/11] bpf, sockmap: Allow inserting listening TCP sockets into sockmap Jakub Sitnicki
2020-01-11 23:59 ` John Fastabend
2020-01-13 15:48 ` Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 06/11] bpf, sockmap: Don't set up sockmap progs for listening sockets Jakub Sitnicki
2020-01-12 0:51 ` John Fastabend
2020-01-12 1:07 ` John Fastabend
2020-01-13 17:59 ` Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 07/11] bpf, sockmap: Return socket cookie on lookup from syscall Jakub Sitnicki
2020-01-12 0:56 ` John Fastabend
2020-01-13 23:12 ` Martin Lau
2020-01-14 3:16 ` John Fastabend
2020-01-14 15:48 ` Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 08/11] bpf, sockmap: Let all kernel-land lookup values in SOCKMAP Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 09/11] bpf: Allow selecting reuseport socket from a SOCKMAP Jakub Sitnicki
2020-01-12 1:00 ` John Fastabend
2020-01-13 23:45 ` Martin Lau
2020-01-15 12:41 ` Jakub Sitnicki
2020-01-13 23:51 ` Martin Lau
2020-01-15 12:57 ` Jakub Sitnicki
2020-01-10 10:50 ` [PATCH bpf-next v2 10/11] selftests/bpf: Extend SK_REUSEPORT tests to cover SOCKMAP Jakub Sitnicki
2020-01-12 1:01 ` John Fastabend
2020-01-10 10:50 ` [PATCH bpf-next v2 11/11] selftests/bpf: Tests for SOCKMAP holding listening sockets Jakub Sitnicki
2020-01-12 1:06 ` John Fastabend
2020-01-13 15:58 ` Jakub Sitnicki
2020-01-11 0:18 ` [PATCH bpf-next v2 00/11] Extend SOCKMAP to store " Alexei Starovoitov
2020-01-11 22:47 ` John Fastabend
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=8736c9rnf5.fsf@cloudflare.com \
--to=jakub@cloudflare.com \
--cc=bpf@vger.kernel.org \
--cc=edumazet@google.com \
--cc=john.fastabend@gmail.com \
--cc=kafai@fb.com \
--cc=kernel-team@cloudflare.com \
--cc=lmb@cloudflare.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).