linux-toradex.git/net/ipv4, branch v5.1-rc6

tcp: tcp_grow_window() needs to respect tcp_space()

2019-04-17T04:47:39+00:00

For some reason, tcp_grow_window() correctly tests if enough room
is present before attempting to increase tp->rcv_ssthresh,
but does not prevent it to grow past tcp_space()

This is causing hard to debug issues, like failing
the (__tcp_select_window(sk) >= tp->rcv_wnd) test
in __tcp_ack_snd_check(), causing ACK delays and possibly
slow flows.

Depending on tcp_rmem[2], MTU, skb->len/skb->truesize ratio,
we can see the problem happening on "netperf -t TCP_RR -- -r 2000,2000"
after about 60 round trips, when the active side no longer sends
immediate acks.

This bug predates git history.

Signed-off-by: Eric Dumazet 
Acked-by: Soheil Hassas Yeganeh 
Acked-by: Neal Cardwell 
Acked-by: Wei Wang 
Signed-off-by: David S. Miller

ipv4: ensure rcu_read_lock() in ipv4_link_failure()

2019-04-14T20:43:17+00:00

fib_compute_spec_dst() needs to be called under rcu protection.

syzbot reported :

WARNING: suspicious RCU usage
5.1.0-rc4+ #165 Not tainted
include/linux/inetdevice.h:220 suspicious rcu_dereference_check() usage!

other info that might help us debug this:

rcu_scheduler_active = 2, debug_locks = 1
1 lock held by swapper/0/0:
 #0: 0000000051b67925 ((&n->timer)){+.-.}, at: lockdep_copy_map include/linux/lockdep.h:170 [inline]
 #0: 0000000051b67925 ((&n->timer)){+.-.}, at: call_timer_fn+0xda/0x720 kernel/time/timer.c:1315

stack backtrace:
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.1.0-rc4+ #165
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x172/0x1f0 lib/dump_stack.c:113
 lockdep_rcu_suspicious+0x153/0x15d kernel/locking/lockdep.c:5162
 __in_dev_get_rcu include/linux/inetdevice.h:220 [inline]
 fib_compute_spec_dst+0xbbd/0x1030 net/ipv4/fib_frontend.c:294
 spec_dst_fill net/ipv4/ip_options.c:245 [inline]
 __ip_options_compile+0x15a7/0x1a10 net/ipv4/ip_options.c:343
 ipv4_link_failure+0x172/0x400 net/ipv4/route.c:1195
 dst_link_failure include/net/dst.h:427 [inline]
 arp_error_report+0xd1/0x1c0 net/ipv4/arp.c:297
 neigh_invalidate+0x24b/0x570 net/core/neighbour.c:995
 neigh_timer_handler+0xc35/0xf30 net/core/neighbour.c:1081
 call_timer_fn+0x190/0x720 kernel/time/timer.c:1325
 expire_timers kernel/time/timer.c:1362 [inline]
 __run_timers kernel/time/timer.c:1681 [inline]
 __run_timers kernel/time/timer.c:1649 [inline]
 run_timer_softirq+0x652/0x1700 kernel/time/timer.c:1694
 __do_softirq+0x266/0x95a kernel/softirq.c:293
 invoke_softirq kernel/softirq.c:374 [inline]
 irq_exit+0x180/0x1d0 kernel/softirq.c:414
 exiting_irq arch/x86/include/asm/apic.h:536 [inline]
 smp_apic_timer_interrupt+0x14a/0x570 arch/x86/kernel/apic/apic.c:1062
 apic_timer_interrupt+0xf/0x20 arch/x86/entry/entry_64.S:807

Fixes: ed0de45a1008 ("ipv4: recompile ip options in ipv4_link_failure")
Signed-off-by: Eric Dumazet 
Reported-by: syzbot 
Cc: Stephen Suryaputra 
Signed-off-by: David S. Miller

ipv4: recompile ip options in ipv4_link_failure

2019-04-13T00:23:46+00:00

Recompile IP options since IPCB may not be valid anymore when
ipv4_link_failure is called from arp_error_report.

Refer to the commit 3da1ed7ac398 ("net: avoid use IPCB in cipso_v4_error")
and the commit before that (9ef6b42ad6fd) for a similar issue.

Signed-off-by: Stephen Suryaputra 
Signed-off-by: David S. Miller

dctcp: more accurate tracking of packets delivery

2019-04-12T04:31:03+00:00

After commit e21db6f69a95 ("tcp: track total bytes delivered with ECN CE marks")
core TCP stack does a very good job tracking ECN signals.

The "sender's best estimate of CE information" Yuchung mentioned in his
patch is indeed the best we can do.

DCTCP can use tp->delivered_ce and tp->delivered to not duplicate the logic,
and use the existing best estimate.

This solves some problems, since current DCTCP logic does not deal with losses
and/or GRO or ack aggregation very well.

This also removes a dubious use of inet_csk(sk)->icsk_ack.rcv_mss
(this should have been tp->mss_cache), and a 64 bit divide.

Finally, we can see that the DCTCP logic, calling dctcp_update_alpha() for
every ACK could be done differently, calling it only once per RTT.

Signed-off-by: Eric Dumazet 
Cc: Yuchung Cheng 
Cc: Neal Cardwell 
Cc: Soheil Hassas Yeganeh 
Cc: Florian Westphal 
Cc: Daniel Borkmann 
Cc: Lawrence Brakmo 
Cc: Abdul Kabbani 
Acked-by: Soheil Hassas Yeganeh 
Acked-by: Neal Cardwell 
Signed-off-by: David S. Miller

net: fou: do not use guehdr after iptunnel_pull_offloads in gue_udp_recv

2019-04-11T06:02:23+00:00

gue tunnels run iptunnel_pull_offloads on received skbs. This can
determine a possible use-after-free accessing guehdr pointer since
the packet will be 'uncloned' running pskb_expand_head if it is a
cloned gso skb (e.g if the packet has been sent though a veth device)

Fixes: a09a4c8dd1ec ("tunnels: Remove encapsulation offloads on decap")
Signed-off-by: Lorenzo Bianconi 
Signed-off-by: David S. Miller

net: ip_gre: fix possible use-after-free in erspan_rcv

2019-04-08T23:16:47+00:00

erspan tunnels run __iptunnel_pull_header on received skbs to remove
gre and erspan headers. This can determine a possible use-after-free
accessing pkt_md pointer in erspan_rcv since the packet will be 'uncloned'
running pskb_expand_head if it is a cloned gso skb (e.g if the packet has
been sent though a veth device). Fix it resetting pkt_md pointer after
__iptunnel_pull_header

Fixes: 1d7e2ed22f8d ("net: erspan: refactor existing erspan code")
Signed-off-by: Lorenzo Bianconi 
Signed-off-by: David S. Miller

tcp: Ensure DCTCP reacts to losses

2019-04-04T17:51:16+00:00

RFC8257 §3.5 explicitly states that "A DCTCP sender MUST react to
loss episodes in the same way as conventional TCP".

Currently, Linux DCTCP performs no cwnd reduction when losses
are encountered. Optionally, the dctcp_clamp_alpha_on_loss resets
alpha to its maximal value if a RTO happens. This behavior
is sub-optimal for at least two reasons: i) it ignores losses
triggering fast retransmissions; and ii) it causes unnecessary large
cwnd reduction in the future if the loss was isolated as it resets
the historical term of DCTCP's alpha EWMA to its maximal value (i.e.,
denoting a total congestion). The second reason has an especially
noticeable effect when using DCTCP in high BDP environments, where
alpha normally stays at low values.

This patch replace the clamping of alpha by setting ssthresh to
half of cwnd for both fast retransmissions and RTOs, at most once
per RTT. Consequently, the dctcp_clamp_alpha_on_loss module parameter
has been removed.

The table below shows experimental results where we measured the
drop probability of a PIE AQM (not applying ECN marks) at a
bottleneck in the presence of a single TCP flow with either the
alpha-clamping option enabled or the cwnd halving proposed by this
patch. Results using reno or cubic are given for comparison.

                          |  Link   |   RTT    |    Drop
                 TCP CC   |  speed  | base+AQM | probability
        ==================|=========|==========|============
                    CUBIC |  40Mbps |  7+20ms  |    0.21%
                     RENO |         |          |    0.19%
        DCTCP-CLAMP-ALPHA |         |          |   25.80%
         DCTCP-HALVE-CWND |         |          |    0.22%
        ------------------|---------|----------|------------
                    CUBIC | 100Mbps |  7+20ms  |    0.03%
                     RENO |         |          |    0.02%
        DCTCP-CLAMP-ALPHA |         |          |   23.30%
         DCTCP-HALVE-CWND |         |          |    0.04%
        ------------------|---------|----------|------------
                    CUBIC | 800Mbps |   1+1ms  |    0.04%
                     RENO |         |          |    0.05%
        DCTCP-CLAMP-ALPHA |         |          |   18.70%
         DCTCP-HALVE-CWND |         |          |    0.06%

We see that, without halving its cwnd for all source of losses,
DCTCP drives the AQM to large drop probabilities in order to keep
the queue length under control (i.e., it repeatedly faces RTOs).
Instead, if DCTCP reacts to all source of losses, it can then be
controlled by the AQM using similar drop levels than cubic or reno.

Signed-off-by: Koen De Schepper 
Signed-off-by: Olivier Tilmans 
Cc: Bob Briscoe 
Cc: Lawrence Brakmo 
Cc: Florian Westphal 
Cc: Daniel Borkmann 
Cc: Yuchung Cheng 
Cc: Neal Cardwell 
Cc: Eric Dumazet 
Cc: Andrew Shewmaker 
Cc: Glenn Judd 
Acked-by: Florian Westphal 
Acked-by: Neal Cardwell 
Acked-by: Daniel Borkmann 
Signed-off-by: David S. Miller

vrf: check accept_source_route on the original netdevice

2019-04-01T17:44:58+00:00

Configuration check to accept source route IP options should be made on
the incoming netdevice when the skb->dev is an l3mdev master. The route
lookup for the source route next hop also needs the incoming netdev.

v2->v3:
- Simplify by passing the original netdevice down the stack (per David
  Ahern).

Signed-off-by: Stephen Suryaputra 
Reviewed-by: David Ahern 
Signed-off-by: David S. Miller

tcp: fix a potential NULL pointer dereference in tcp_sk_exit

2019-04-01T17:11:41+00:00

 When tcp_sk_init() failed in inet_ctl_sock_create(),
 'net->ipv4.tcp_congestion_control' will be left
 uninitialized, but tcp_sk_exit() hasn't check for
 that.

 This patch add checking on 'net->ipv4.tcp_congestion_control'
 in tcp_sk_exit() to prevent NULL-ptr dereference.

Fixes: 6670e1524477 ("tcp: Namespace-ify sysctl_tcp_default_congestion_control")
Signed-off-by: Dust Li 
Signed-off-by: Eric Dumazet 
Signed-off-by: David S. Miller

tcp: Don't access TCP_SKB_CB before initializing it

2019-03-11T22:36:49+00:00

Since commit eeea10b83a13 ("tcp: add
tcp_v4_fill_cb()/tcp_v4_restore_cb()"), tcp_vX_fill_cb is only called
after tcp_filter(). That means, TCP_SKB_CB(skb)->end_seq still points to
the IP-part of the cb.

We thus should not mock with it, as this can trigger bugs (thanks
syzkaller):
[   12.349396] ==================================================================
[   12.350188] BUG: KASAN: slab-out-of-bounds in ip6_datagram_recv_specific_ctl+0x19b3/0x1a20
[   12.351035] Read of size 1 at addr ffff88006adbc208 by task test_ip6_datagr/1799

Setting end_seq is actually no more necessary in tcp_filter as it gets
initialized later on in tcp_vX_fill_cb.

Cc: Eric Dumazet 
Fixes: eeea10b83a13 ("tcp: add tcp_v4_fill_cb()/tcp_v4_restore_cb()")
Signed-off-by: Christoph Paasch 
Signed-off-by: Eric Dumazet 
Signed-off-by: David S. Miller