linux-stable.git/net/core/skmsg.c, branch v5.8.2

bpf, sockmap: RCU dereferenced psock may be used outside RCU block

2020-06-28T15:33:28+00:00

If an ingress verdict program specifies message sizes greater than
skb->len and there is an ENOMEM error due to memory pressure we
may call the rcv_msg handler outside the strp_data_ready() caller
context. This is because on an ENOMEM error the strparser will
retry from a workqueue. The caller currently protects the use of
psock by calling the strp_data_ready() inside a rcu_read_lock/unlock
block.

But, in above workqueue error case the psock is accessed outside
the read_lock/unlock block of the caller. So instead of using
psock directly we must do a look up against the sk again to
ensure the psock is available.

There is an an ugly piece here where we must handle
the case where we paused the strp and removed the psock. On
psock removal we first pause the strparser and then remove
the psock. If the strparser is paused while an skb is
scheduled on the workqueue the skb will be dropped on the
flow and kfree_skb() is called. If the workqueue manages
to get called before we pause the strparser but runs the rcvmsg
callback after the psock is removed we will hit the unlikely
case where we run the sockmap rcvmsg handler but do not have
a psock. For now we will follow strparser logic and drop the
skb on the floor with skb_kfree(). This is ugly because the
data is dropped. To date this has not caused problems in practice
because either the application controlling the sockmap is
coordinating with the datapath so that skbs are "flushed"
before removal or we simply wait for the sock to be closed before
removing it.

This patch fixes the describe RCU bug and dropping the skb doesn't
make things worse. Future patches will improve this by allowing
the normal case where skbs are not merged to skip the strparser
altogether. In practice many (most?) use cases have no need to
merge skbs so its both a code complexity hit as seen above and
a performance issue. For example, in the Cilium case we always
set the strparser up to return sbks 1:1 without any merging and
have avoided above issues.

Fixes: e91de6afa81c1 ("bpf: Fix running sk_skb program types with ktls")
Signed-off-by: John Fastabend
Signed-off-by: Alexei Starovoitov
Acked-by: Martin KaFai Lau
Link: https://lore.kernel.org/bpf/159312679888.18340.15248924071966273998.stgit@john-XPS-13-9370

bpf, sockmap: RCU splat with redirect and strparser error or TLS

2020-06-28T15:33:28+00:00

There are two paths to generate the below RCU splat the first and
most obvious is the result of the BPF verdict program issuing a
redirect on a TLS socket (This is the splat shown below). Unlike
the non-TLS case the caller of the *strp_read() hooks does not
wrap the call in a rcu_read_lock/unlock. Then if the BPF program
issues a redirect action we hit the RCU splat.

However, in the non-TLS socket case the splat appears to be
relatively rare, because the skmsg caller into the strp_data_ready()
is wrapped in a rcu_read_lock/unlock. Shown here,

 static void sk_psock_strp_data_ready(struct sock *sk)
 {
	struct sk_psock *psock;

	rcu_read_lock();
	psock = sk_psock(sk);
	if (likely(psock)) {
		if (tls_sw_has_ctx_rx(sk)) {
			psock->parser.saved_data_ready(sk);
		} else {
			write_lock_bh(&sk->sk_callback_lock);
			strp_data_ready(&psock->parser.strp);
			write_unlock_bh(&sk->sk_callback_lock);
		}
	}
	rcu_read_unlock();
 }

If the above was the only way to run the verdict program we
would be safe. But, there is a case where the strparser may throw an
ENOMEM error while parsing the skb. This is a result of a failed
skb_clone, or alloc_skb_for_msg while building a new merged skb when
the msg length needed spans multiple skbs. This will in turn put the
skb on the strp_wrk workqueue in the strparser code. The skb will
later be dequeued and verdict programs run, but now from a
different context without the rcu_read_lock()/unlock() critical
section in sk_psock_strp_data_ready() shown above. In practice
I have not seen this yet, because as far as I know most users of the
verdict programs are also only working on single skbs. In this case no
merge happens which could trigger the above ENOMEM errors. In addition
the system would need to be under memory pressure. For example, we
can't hit the above case in selftests because we missed having tests
to merge skbs. (Added in later patch)

To fix the below splat extend the rcu_read_lock/unnlock block to
include the call to sk_psock_tls_verdict_apply(). This will fix both
TLS redirect case and non-TLS redirect+error case. Also remove
psock from the sk_psock_tls_verdict_apply() function signature its
not used there.

[ 1095.937597] WARNING: suspicious RCU usage
[ 1095.940964] 5.7.0-rc7-02911-g463bac5f1ca79 #1 Tainted: G        W
[ 1095.944363] -----------------------------
[ 1095.947384] include/linux/skmsg.h:284 suspicious rcu_dereference_check() usage!
[ 1095.950866]
[ 1095.950866] other info that might help us debug this:
[ 1095.950866]
[ 1095.957146]
[ 1095.957146] rcu_scheduler_active = 2, debug_locks = 1
[ 1095.961482] 1 lock held by test_sockmap/15970:
[ 1095.964501]  #0: ffff9ea6b25de660 (sk_lock-AF_INET){+.+.}-{0:0}, at: tls_sw_recvmsg+0x13a/0x840 [tls]
[ 1095.968568]
[ 1095.968568] stack backtrace:
[ 1095.975001] CPU: 1 PID: 15970 Comm: test_sockmap Tainted: G        W         5.7.0-rc7-02911-g463bac5f1ca79 #1
[ 1095.977883] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
[ 1095.980519] Call Trace:
[ 1095.982191]  dump_stack+0x8f/0xd0
[ 1095.984040]  sk_psock_skb_redirect+0xa6/0xf0
[ 1095.986073]  sk_psock_tls_strp_read+0x1d8/0x250
[ 1095.988095]  tls_sw_recvmsg+0x714/0x840 [tls]

v2: Improve commit message to identify non-TLS redirect plus error case
    condition as well as more common TLS case. In the process I decided
    doing the rcu_read_unlock followed by the lock/unlock inside branches
    was unnecessarily complex. We can just extend the current rcu block
    and get the same effeective without the shuffling and branching.
    Thanks Martin!

Fixes: e91de6afa81c1 ("bpf: Fix running sk_skb program types with ktls")
Reported-by: Jakub Sitnicki 
Reported-by: kernel test robot 
Signed-off-by: John Fastabend 
Signed-off-by: Alexei Starovoitov 
Acked-by: Martin KaFai Lau 
Acked-by: Jakub Sitnicki 
Link: https://lore.kernel.org/bpf/159312677907.18340.11064813152758406626.stgit@john-XPS-13-9370

bpf: Fix running sk_skb program types with ktls

2020-06-01T21:48:32+00:00

KTLS uses a stream parser to collect TLS messages and send them to
the upper layer tls receive handler. This ensures the tls receiver
has a full TLS header to parse when it is run. However, when a
socket has BPF_SK_SKB_STREAM_VERDICT program attached before KTLS
is enabled we end up with two stream parsers running on the same
socket.

The result is both try to run on the same socket. First the KTLS
stream parser runs and calls read_sock() which will tcp_read_sock
which in turn calls tcp_rcv_skb(). This dequeues the skb from the
sk_receive_queue. When this is done KTLS code then data_ready()
callback which because we stacked KTLS on top of the bpf stream
verdict program has been replaced with sk_psock_start_strp(). This
will in turn kick the stream parser again and eventually do the
same thing KTLS did above calling into tcp_rcv_skb() and dequeuing
a skb from the sk_receive_queue.

At this point the data stream is broke. Part of the stream was
handled by the KTLS side some other bytes may have been handled
by the BPF side. Generally this results in either missing data
or more likely a "Bad Message" complaint from the kTLS receive
handler as the BPF program steals some bytes meant to be in a
TLS header and/or the TLS header length is no longer correct.

We've already broke the idealized model where we can stack ULPs
in any order with generic callbacks on the TX side to handle this.
So in this patch we do the same thing but for RX side. We add
a sk_psock_strp_enabled() helper so TLS can learn a BPF verdict
program is running and add a tls_sw_has_ctx_rx() helper so BPF
side can learn there is a TLS ULP on the socket.

Then on BPF side we omit calling our stream parser to avoid
breaking the data stream for the KTLS receiver. Then on the
KTLS side we call BPF_SK_SKB_STREAM_VERDICT once the KTLS
receiver is done with the packet but before it posts the
msg to userspace. This gives us symmetry between the TX and
RX halfs and IMO makes it usable again. On the TX side we
process packets in this order BPF -> TLS -> TCP and on
the receive side in the reverse order TCP -> TLS -> BPF.

Discovered while testing OpenSSL 3.0 Alpha2.0 release.

Fixes: d829e9c4112b5 ("tls: convert to generic sk_msg interface")
Signed-off-by: John Fastabend
Signed-off-by: Alexei Starovoitov
Link: https://lore.kernel.org/bpf/159079361946.5745.605854335665044485.stgit@john-Precision-5820-Tower
Signed-off-by: Alexei Starovoitov

bpf: Refactor sockmap redirect code so its easy to reuse

2020-06-01T21:48:32+00:00

We will need this block of code called from tls context shortly
lets refactor the redirect logic so its easy to use. This also
cleans up the switch stmt so we have fewer fallthrough cases.

No logic changes are intended.

Fixes: d829e9c4112b5 ("tls: convert to generic sk_msg interface")
Signed-off-by: John Fastabend 
Signed-off-by: Alexei Starovoitov 
Reviewed-by: Jakub Sitnicki 
Acked-by: Song Liu 
Link: https://lore.kernel.org/bpf/159079360110.5745.7024009076049029819.stgit@john-Precision-5820-Tower
Signed-off-by: Alexei Starovoitov

bpf: Use bpf_prog_run_pin_on_cpu() at simple call sites.

2020-02-25T00:20:09+00:00

All of these cases are strictly of the form:

	preempt_disable();
	BPF_PROG_RUN(...);
	preempt_enable();

Replace this with bpf_prog_run_pin_on_cpu() which wraps BPF_PROG_RUN()
with:

	migrate_disable();
	BPF_PROG_RUN(...);
	migrate_enable();

On non RT enabled kernels this maps to preempt_disable/enable() and on RT
enabled kernels this solely prevents migration, which is sufficient as
there is no requirement to prevent reentrancy to any BPF program from a
preempting task. The only requirement is that the program stays on the same
CPU.

Therefore, this is a trivially correct transformation.

The seccomp loop does not need protection over the loop. It only needs
protection per BPF filter program

[ tglx: Converted to bpf_prog_run_pin_on_cpu() ]

Signed-off-by: David S. Miller 
Signed-off-by: Thomas Gleixner 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20200224145643.691493094@linutronix.de

net, sk_msg: Clear sk_user_data pointer on clone if tagged

2020-02-21T21:29:45+00:00

sk_user_data can hold a pointer to an object that is not intended to be
shared between the parent socket and the child that gets a pointer copy on
clone. This is the case when sk_user_data points at reference-counted
object, like struct sk_psock.

One way to resolve it is to tag the pointer with a no-copy flag by
repurposing its lowest bit. Based on the bit-flag value we clear the child
sk_user_data pointer after cloning the parent socket.

The no-copy flag is stored in the pointer itself as opposed to externally,
say in socket flags, to guarantee that the pointer and the flag are copied
from parent to child socket in an atomic fashion. Parent socket state is
subject to change while copying, we don't hold any locks at that time.

This approach relies on an assumption that sk_user_data holds a pointer to
an object aligned at least 2 bytes. A manual audit of existing users of
rcu_dereference_sk_user_data helper confirms our assumption.

Also, an RCU-protected sk_user_data is not likely to hold a pointer to a
char value or a pathological case of "struct { char c; }". To be safe, warn
when the flag-bit is set when setting sk_user_data to catch any future
misuses.

It is worth considering why clearing sk_user_data unconditionally is not an
option. There exist users, DRBD, NVMe, and Xen drivers being among them,
that rely on the pointer being copied when cloning the listening socket.

Potentially we could distinguish these users by checking if the listening
socket has been created in kernel-space via sock_create_kern, and hence has
sk_kern_sock flag set. However, this is not the case for NVMe and Xen
drivers, which create sockets without marking them as belonging to the
kernel.

Signed-off-by: Jakub Sitnicki 
Signed-off-by: Daniel Borkmann 
Acked-by: John Fastabend 
Acked-by: Martin KaFai Lau 
Link: https://lore.kernel.org/bpf/20200218171023.844439-3-jakub@cloudflare.com

net, sk_msg: Don't check if sock is locked when tearing down psock

2020-01-22T19:30:20+00:00

As John Fastabend reports [0], psock state tear-down can happen on receive
path *after* unlocking the socket, if the only other psock user, that is
sockmap or sockhash, releases its psock reference before tcp_bpf_recvmsg
does so:

 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

Remove the lockdep check for socket lock in psock tear-down that got
introduced in 7e81a3530206 ("bpf: Sockmap, ensure sock lock held during
tear down").

[0] https://lore.kernel.org/netdev/5e25dc995d7d_74082aaee6e465b441@john-XPS-13-9370.notmuch/

Fixes: 7e81a3530206 ("bpf: Sockmap, ensure sock lock held during tear down")
Reported-by: syzbot+d73682fcf7fee6982fe3@syzkaller.appspotmail.com
Suggested-by: John Fastabend 
Signed-off-by: Jakub Sitnicki 
Acked-by: John Fastabend 
Acked-by: Daniel Borkmann 
Signed-off-by: David S. Miller

bpf: Sockmap, ensure sock lock held during tear down

2020-01-15T22:26:13+00:00

The sock_map_free() and sock_hash_free() paths used to delete sockmap
and sockhash maps walk the maps and destroy psock and bpf state associated
with the socks in the map. When done the socks no longer have BPF programs
attached and will function normally. This can happen while the socks in
the map are still "live" meaning data may be sent/received during the walk.

Currently, though we don't take the sock_lock when the psock and bpf state
is removed through this path. Specifically, this means we can be writing
into the ops structure pointers such as sendmsg, sendpage, recvmsg, etc.
while they are also being called from the networking side. This is not
safe, we never used proper READ_ONCE/WRITE_ONCE semantics here if we
believed it was safe. Further its not clear to me its even a good idea
to try and do this on "live" sockets while networking side might also
be using the socket. Instead of trying to reason about using the socks
from both sides lets realize that every use case I'm aware of rarely
deletes maps, in fact kubernetes/Cilium case builds map at init and
never tears it down except on errors. So lets do the simple fix and
grab sock lock.

This patch wraps sock deletes from maps in sock lock and adds some
annotations so we catch any other cases easier.

Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: John Fastabend 
Signed-off-by: Daniel Borkmann 
Acked-by: Song Liu 
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/bpf/20200111061206.8028-3-john.fastabend@gmail.com

net: skmsg: fix TLS 1.3 crash with full sk_msg

2019-11-29T06:40:29+00:00

TLS 1.3 started using the entry at the end of the SG array
for chaining-in the single byte content type entry. This mostly
works:

[ E E E E E E . . ]
  ^           ^
   start       end

                 E < content type
               /
[ E E E E E E C . ]
  ^           ^
   start       end

(Where E denotes a populated SG entry; C denotes a chaining entry.)

If the array is full, however, the end will point to the start:

[ E E E E E E E E ]
  ^
   start
   end

And we end up overwriting the start:

    E < content type
   /
[ C E E E E E E E ]
  ^
   start
   end

The sg array is supposed to be a circular buffer with start and
end markers pointing anywhere. In case where start > end
(i.e. the circular buffer has "wrapped") there is an extra entry
reserved at the end to chain the two halves together.

[ E E E E E E . . l ]

(Where l is the reserved entry for "looping" back to front.

As suggested by John, let's reserve another entry for chaining
SG entries after the main circular buffer. Note that this entry
has to be pointed to by the end entry so its position is not fixed.

Examples of full messages:

[ E E E E E E E E . l ]
  ^               ^
   start           end

   <---------------.
[ E E . E E E E E E l ]
      ^ ^
   end   start

Now the end will always point to an unused entry, so TLS 1.3
can always use it.

Fixes: 130b392c6cd6 ("net: tls: Add tls 1.3 support")
Signed-off-by: Jakub Kicinski 
Reviewed-by: Simon Horman 
Signed-off-by: David S. Miller

bpf: skmsg, fix potential psock NULL pointer dereference

2019-11-21T21:43:45+00:00

Report from Dan Carpenter,

 net/core/skmsg.c:792 sk_psock_write_space()
 error: we previously assumed 'psock' could be null (see line 790)

 net/core/skmsg.c
   789 psock = sk_psock(sk);
   790 if (likely(psock && sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)))
 Check for NULL
   791 schedule_work(&psock->work);
   792 write_space = psock->saved_write_space;
                     ^^^^^^^^^^^^^^^^^^^^^^^^
   793          rcu_read_unlock();
   794          write_space(sk);

Ensure psock dereference on line 792 only occurs if psock is not null.

Reported-by: Dan Carpenter 
Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: John Fastabend 
Signed-off-by: David S. Miller