linux-toradex.git/include/net, branch v5.3-rc5

net: sched: sample: allow accessing psample_group with rtnl

2019-08-06T21:15:39+00:00

Recently implemented support for sample action in flow_offload infra leads
to following rcu usage warning:

[ 1938.234856] =============================
[ 1938.234858] WARNING: suspicious RCU usage
[ 1938.234863] 5.3.0-rc1+ #574 Not tainted
[ 1938.234866] -----------------------------
[ 1938.234869] include/net/tc_act/tc_sample.h:47 suspicious rcu_dereference_check() usage!
[ 1938.234872]
               other info that might help us debug this:

[ 1938.234875]
               rcu_scheduler_active = 2, debug_locks = 1
[ 1938.234879] 1 lock held by tc/19540:
[ 1938.234881]  #0: 00000000b03cb918 (rtnl_mutex){+.+.}, at: tc_new_tfilter+0x47c/0x970
[ 1938.234900]
               stack backtrace:
[ 1938.234905] CPU: 2 PID: 19540 Comm: tc Not tainted 5.3.0-rc1+ #574
[ 1938.234908] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017
[ 1938.234911] Call Trace:
[ 1938.234922]  dump_stack+0x85/0xc0
[ 1938.234930]  tc_setup_flow_action+0xed5/0x2040
[ 1938.234944]  fl_hw_replace_filter+0x11f/0x2e0 [cls_flower]
[ 1938.234965]  fl_change+0xd24/0x1b30 [cls_flower]
[ 1938.234990]  tc_new_tfilter+0x3e0/0x970
[ 1938.235021]  ? tc_del_tfilter+0x720/0x720
[ 1938.235028]  rtnetlink_rcv_msg+0x389/0x4b0
[ 1938.235038]  ? netlink_deliver_tap+0x95/0x400
[ 1938.235044]  ? rtnl_dellink+0x2d0/0x2d0
[ 1938.235053]  netlink_rcv_skb+0x49/0x110
[ 1938.235063]  netlink_unicast+0x171/0x200
[ 1938.235073]  netlink_sendmsg+0x224/0x3f0
[ 1938.235091]  sock_sendmsg+0x5e/0x60
[ 1938.235097]  ___sys_sendmsg+0x2ae/0x330
[ 1938.235111]  ? __handle_mm_fault+0x12cd/0x19e0
[ 1938.235125]  ? __handle_mm_fault+0x12cd/0x19e0
[ 1938.235138]  ? find_held_lock+0x2b/0x80
[ 1938.235147]  ? do_user_addr_fault+0x22d/0x490
[ 1938.235160]  __sys_sendmsg+0x59/0xa0
[ 1938.235178]  do_syscall_64+0x5c/0xb0
[ 1938.235187]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 1938.235192] RIP: 0033:0x7ff9a4d597b8
[ 1938.235197] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83
 ec 28 89 54
[ 1938.235200] RSP: 002b:00007ffcfe381c48 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
[ 1938.235205] RAX: ffffffffffffffda RBX: 000000005d4497f9 RCX: 00007ff9a4d597b8
[ 1938.235208] RDX: 0000000000000000 RSI: 00007ffcfe381cb0 RDI: 0000000000000003
[ 1938.235211] RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000006
[ 1938.235214] R10: 0000000000404ec2 R11: 0000000000000246 R12: 0000000000000001
[ 1938.235217] R13: 0000000000480640 R14: 0000000000000012 R15: 0000000000000001

Change tcf_sample_psample_group() helper to allow using it from both rtnl
and rcu protected contexts.

Fixes: a7a7be6087b0 ("net/sched: add sample action to the hardware intermediate representation")
Signed-off-by: Vlad Buslov 
Reviewed-by: Pieter Jansen van Vuuren 
Signed-off-by: David S. Miller

net: sched: police: allow accessing police->params with rtnl

2019-08-06T21:15:39+00:00

Recently implemented support for police action in flow_offload infra leads
to following rcu usage warning:

[ 1925.881092] =============================
[ 1925.881094] WARNING: suspicious RCU usage
[ 1925.881098] 5.3.0-rc1+ #574 Not tainted
[ 1925.881100] -----------------------------
[ 1925.881104] include/net/tc_act/tc_police.h:57 suspicious rcu_dereference_check() usage!
[ 1925.881106]
               other info that might help us debug this:

[ 1925.881109]
               rcu_scheduler_active = 2, debug_locks = 1
[ 1925.881112] 1 lock held by tc/18591:
[ 1925.881115]  #0: 00000000b03cb918 (rtnl_mutex){+.+.}, at: tc_new_tfilter+0x47c/0x970
[ 1925.881124]
               stack backtrace:
[ 1925.881127] CPU: 2 PID: 18591 Comm: tc Not tainted 5.3.0-rc1+ #574
[ 1925.881130] Hardware name: Supermicro SYS-2028TP-DECR/X10DRT-P, BIOS 2.0b 03/30/2017
[ 1925.881132] Call Trace:
[ 1925.881138]  dump_stack+0x85/0xc0
[ 1925.881145]  tc_setup_flow_action+0x1771/0x2040
[ 1925.881155]  fl_hw_replace_filter+0x11f/0x2e0 [cls_flower]
[ 1925.881175]  fl_change+0xd24/0x1b30 [cls_flower]
[ 1925.881200]  tc_new_tfilter+0x3e0/0x970
[ 1925.881231]  ? tc_del_tfilter+0x720/0x720
[ 1925.881243]  rtnetlink_rcv_msg+0x389/0x4b0
[ 1925.881250]  ? netlink_deliver_tap+0x95/0x400
[ 1925.881257]  ? rtnl_dellink+0x2d0/0x2d0
[ 1925.881264]  netlink_rcv_skb+0x49/0x110
[ 1925.881275]  netlink_unicast+0x171/0x200
[ 1925.881284]  netlink_sendmsg+0x224/0x3f0
[ 1925.881299]  sock_sendmsg+0x5e/0x60
[ 1925.881305]  ___sys_sendmsg+0x2ae/0x330
[ 1925.881309]  ? task_work_add+0x43/0x50
[ 1925.881314]  ? fput_many+0x45/0x80
[ 1925.881329]  ? __lock_acquire+0x248/0x1930
[ 1925.881342]  ? find_held_lock+0x2b/0x80
[ 1925.881347]  ? task_work_run+0x7b/0xd0
[ 1925.881359]  __sys_sendmsg+0x59/0xa0
[ 1925.881375]  do_syscall_64+0x5c/0xb0
[ 1925.881381]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
[ 1925.881384] RIP: 0033:0x7feb245047b8
[ 1925.881388] Code: 89 02 48 c7 c0 ff ff ff ff eb bb 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 65 8f 0c 00 8b 00 85 c0 75 17 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 48 83
 ec 28 89 54
[ 1925.881391] RSP: 002b:00007ffc2d2a5788 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
[ 1925.881395] RAX: ffffffffffffffda RBX: 000000005d4497ed RCX: 00007feb245047b8
[ 1925.881398] RDX: 0000000000000000 RSI: 00007ffc2d2a57f0 RDI: 0000000000000003
[ 1925.881400] RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000006
[ 1925.881403] R10: 0000000000404ec2 R11: 0000000000000246 R12: 0000000000000001
[ 1925.881406] R13: 0000000000480640 R14: 0000000000000012 R15: 0000000000000001

Change tcf_police_rate_bytes_ps() and tcf_police_tcfp_burst() helpers to
allow using them from both rtnl and rcu protected contexts.

Fixes: 8c8cfc6ed274 ("net/sched: add police action to the hardware intermediate representation")
Signed-off-by: Vlad Buslov 
Reviewed-by: Pieter Jansen van Vuuren 
Signed-off-by: David S. Miller

net/tls: partially revert fix transition through disconnect with close

2019-08-05T20:15:30+00:00

Looks like we were slightly overzealous with the shutdown()
cleanup. Even though the sock->sk_state can reach CLOSED again,
socket->state will not got back to SS_UNCONNECTED once
connections is ESTABLISHED. Meaning we will see EISCONN if
we try to reconnect, and EINVAL if we try to listen.

Only listen sockets can be shutdown() and reused, but since
ESTABLISHED sockets can never be re-connected() or used for
listen() we don't need to try to clean up the ULP state early.

Fixes: 32857cf57f92 ("net/tls: fix transition through disconnect with close")
Signed-off-by: Jakub Kicinski 
Signed-off-by: David S. Miller

Merge tag 'mac80211-for-davem-2019-07-31' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211

2019-07-31T15:51:34+00:00

Johannes Berg says:

====================
Just a few fixes:
 * revert NETIF_F_LLTX usage as it caused problems
 * avoid warning on WMM parameters from AP that are too short
 * fix possible null-ptr dereference in hwsim
 * fix interface combinations with 4-addr and crypto control
====================

Signed-off-by: David S. Miller

{nl,mac}80211: fix interface combinations on crypto controlled devices

2019-07-26T11:50:43+00:00

Commit 33d915d9e8ce ("{nl,mac}80211: allow 4addr AP operation on
crypto controlled devices") has introduced a change which allows
4addr operation on crypto controlled devices (ex: ath10k). This
change has inadvertently impacted the interface combinations logic
on such devices.

General rule is that software interfaces like AP/VLAN should not be
listed under supported interface combinations and should not be
considered during validation of these combinations; because of the
aforementioned change, AP/VLAN interfaces(if present) will be checked
against interfaces supported by the device and blocks valid interface
combinations.

Consider a case where an AP and AP/VLAN are up and running; when a
second AP device is brought up on the same physical device, this AP
will be checked against the AP/VLAN interface (which will not be
part of supported interface combinations of the device) and blocks
second AP to come up.

Add a new API cfg80211_iftype_allowed() to fix the problem, this
API works for all devices with/without SW crypto control.

Signed-off-by: Manikanta Pubbisetty 
Fixes: 33d915d9e8ce ("{nl,mac}80211: allow 4addr AP operation on crypto controlled devices")
Link: https://lore.kernel.org/r/1563779690-9716-1-git-send-email-mpubbise@codeaurora.org
Signed-off-by: Johannes Berg

Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf

2019-07-26T00:35:03+00:00

Alexei Starovoitov says:

====================
pull-request: bpf 2019-07-25

The following pull-request contains BPF updates for your *net* tree.

The main changes are:

1) fix segfault in libbpf, from Andrii.

2) fix gso_segs access, from Eric.

3) tls/sockmap fixes, from Jakub and John.
====================

Signed-off-by: David S. Miller

bpf: sockmap/tls, close can race with map free

2019-07-22T14:04:17+00:00

When a map free is called and in parallel a socket is closed we
have two paths that can potentially reset the socket prot ops, the
bpf close() path and the map free path. This creates a problem
with which prot ops should be used from the socket closed side.

If the map_free side completes first then we want to call the
original lowest level ops. However, if the tls path runs first
we want to call the sockmap ops. Additionally there was no locking
around prot updates in TLS code paths so the prot ops could
be changed multiple times once from TLS path and again from sockmap
side potentially leaving ops pointed at either TLS or sockmap
when psock and/or tls context have already been destroyed.

To fix this race first only update ops inside callback lock
so that TLS, sockmap and lowest level all agree on prot state.
Second and a ULP callback update() so that lower layers can
inform the upper layer when they are being removed allowing the
upper layer to reset prot ops.

This gets us close to allowing sockmap and tls to be stacked
in arbitrary order but will save that patch for *next trees.

v4:
- make sure we don't free things for device;
- remove the checks which swap the callbacks back
only if TLS is at the top.

Reported-by: syzbot+06537213db7ba2745c4a@syzkaller.appspotmail.com
Fixes: 02c558b2d5d6 ("bpf: sockmap, support for msg_peek in sk_msg with redirect ingress")
Signed-off-by: John Fastabend
Signed-off-by: Jakub Kicinski
Reviewed-by: Dirk van der Merwe
Signed-off-by: Daniel Borkmann

net/tls: fix transition through disconnect with close

2019-07-22T14:04:17+00:00

It is possible (via shutdown()) for TCP socks to go through TCP_CLOSE
state via tcp_disconnect() without actually calling tcp_close which
would then call the tls close callback. Because of this a user could
disconnect a socket then put it in a LISTEN state which would break
our assumptions about sockets always being ESTABLISHED state.

More directly because close() can call unhash() and unhash is
implemented by sockmap if a sockmap socket has TLS enabled we can
incorrectly destroy the psock from unhash() and then call its close
handler again. But because the psock (sockmap socket representation)
is already destroyed we call close handler in sk->prot. However,
in some cases (TLS BASE/BASE case) this will still point at the
sockmap close handler resulting in a circular call and crash reported
by syzbot.

To fix both above issues implement the unhash() routine for TLS.

v4:
 - add note about tls offload still needing the fix;
 - move sk_proto to the cold cache line;
 - split TX context free into "release" and "free",
   otherwise the GC work itself is in already freed
   memory;
 - more TX before RX for consistency;
 - reuse tls_ctx_free();
 - schedule the GC work after we're done with context
   to avoid UAF;
 - don't set the unhash in all modes, all modes "inherit"
   TLS_BASE's callbacks anyway;
 - disable the unhash hook for TLS_HW.

Fixes: 3c4d7559159bf ("tls: kernel TLS support")
Reported-by: Eric Dumazet 
Signed-off-by: John Fastabend 
Signed-off-by: Jakub Kicinski 
Signed-off-by: Daniel Borkmann

net/tls: remove sock unlock/lock around strp_done()

2019-07-22T14:04:16+00:00

The tls close() callback currently drops the sock lock to call
strp_done(). Split up the RX cleanup into stopping the strparser
and releasing most resources, syncing strparser and finally
freeing the context.

To avoid the need for a strp_done() call on the cleanup path
of device offload make sure we don't arm the strparser until
we are sure init will be successful.

Signed-off-by: John Fastabend 
Signed-off-by: Jakub Kicinski 
Reviewed-by: Dirk van der Merwe 
Signed-off-by: Daniel Borkmann

net/tls: remove close callback sock unlock/lock around TX work flush

2019-07-22T14:04:16+00:00

The tls close() callback currently drops the sock lock, makes a
cancel_delayed_work_sync() call, and then relocks the sock.

By restructuring the code we can avoid droping lock and then
reclaiming it. To simplify this we do the following,

 tls_sk_proto_close
 set_bit(CLOSING)
 set_bit(SCHEDULE)
 cancel_delay_work_sync() <- cancel workqueue
 lock_sock(sk)
 ...
 release_sock(sk)
 strp_done()

Setting the CLOSING bit prevents the SCHEDULE bit from being
cleared by any workqueue items e.g. if one happens to be
scheduled and run between when we set SCHEDULE bit and cancel
work. Then because SCHEDULE bit is set now no new work will
be scheduled.

Tested with net selftests and bpf selftests.

Signed-off-by: John Fastabend 
Signed-off-by: Jakub Kicinski 
Reviewed-by: Dirk van der Merwe 
Signed-off-by: Daniel Borkmann