Date:   Tue, 29 Jun 2021 01:15:01 +0800
From:   Phi Nguyen <phind.uet@...il.com>
To:     Neal Cardwell <ncardwell@...gle.com>
Cc:     Eric Dumazet <edumazet@...gle.com>,
        David Miller <davem@...emloft.net>,
        Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
        David Ahern <dsahern@...nel.org>,
        Jakub Kicinski <kuba@...nel.org>,
        Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Andrii Nakryiko <andrii@...nel.org>,
        Martin KaFai Lau <kafai@...com>,
        Song Liu <songliubraving@...com>,
        John Fastabend <john.fastabend@...il.com>, kpsingh@...nel.org,
        netdev <netdev@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>, bpf <bpf@...r.kernel.org>,
        linux-kernel-mentees@...ts.linuxfoundation.org,
        syzbot+f1e24a0594d4e3a895d3@...kaller.appspotmail.com,
        Yuchung Cheng <ycheng@...gle.com>
Subject: Re: [PATCH] tcp: Do not reset the icsk_ca_initialized in
 tcp_init_transfer.

On 6/29/2021 12:24 AM, Neal Cardwell wrote:

> Thanks.
> 
> Can you also please provide a summary of the event sequence that
> triggers the bug? Based on your Reported-by tag, I guess this is based
> on the syzbot reproducer:
> 
>   https://groups.google.com/g/syzkaller-bugs/c/VbHoSsBz0hk/m/cOxOoTgPCAAJ
> 
> but perhaps you can give a summary of the event sequence that causes
> the bug? Is it that the call:
> 
> setsockopt$inet_tcp_TCP_CONGESTION(r0, 0x6, 0xd,
> &(0x7f0000000000)='cdg\x00', 0x4)
> 
> initializes the CC and happens before the connection is established,
> and then when the connection is established, the line that sets:
>    icsk->icsk_ca_initialized = 0;
> is incorrect, causing the CC to be initialized again without first
> calling the cleanup code that deallocates the CDG-allocated memory?
> 
> thanks,
> neal
> 

Hi Neal,

The gdb stack trace that lead to init_transfer_input() is as bellow, the 
current sock state is TCP_SYN_RECV.

#0  tcp_cdg_init (sk=sk@...ry=0xffff8880275fa080) at net/ipv4/tcp_cdg.c:380
#1  0xffffffff83a3c129 in tcp_init_congestion_control 
(sk=sk@...ry=0xffff8880275fa080) at net/ipv4/tcp_cong.c:183
#2  0xffffffff83a21312 in tcp_init_transfer 
(sk=sk@...ry=0xffff8880275fa080, bpf_op=bpf_op@...ry=5, 
skb=skb@...ry=0xffff888027704400) at net/ipv4/tcp_input.c:5928
#3  0xffffffff83a22a55 in tcp_rcv_state_process 
(sk=sk@...ry=0xffff8880275fa080, skb=skb@...ry=0xffff888027704400) at 
net/ipv4/tcp_input.c:6416
#4  0xffffffff83a35f8a in tcp_v4_do_rcv (sk=sk@...ry=0xffff8880275fa080, 
skb=skb@...ry=0xffff888027704400) at net/ipv4/tcp_ipv4.c:1716
#5  0xffffffff83a39300 in tcp_v4_rcv (skb=0xffff888027704400) at 
net/ipv4/tcp_ipv4.c:2087
#6  0xffffffff839eaf57 in ip_protocol_deliver_rcu 
(net=net@...ry=0xffff888022858180, skb=skb@...ry=0xffff888027704400, 
protocol=<optimized out>) at net/ipv4/ip_input.c:204
#7  0xffffffff839eb266 in ip_local_deliver_finish 
(net=0xffff888022858180, sk=<optimized out>, skb=0xffff888027704400) at 
./include/linux/skbuff.h:2544
#8  0xffffffff839eb34a in NF_HOOK (sk=0x0 <fixed_percpu_data>, pf=2 
'\002', hook=1, in=<optimized out>, out=0x0 <fixed_percpu_data>, 
okfn=0xffffffff839eb1f0 <ip_local_deliver_finish>,
     skb=0xffff888027704400, net=0xffff888022858180) at 
./include/linux/netfilter.h:307
#9  NF_HOOK (pf=2 '\002', sk=0x0 <fixed_percpu_data>, out=0x0 
<fixed_percpu_data>, okfn=0xffffffff839eb1f0 <ip_local_deliver_finish>, 
in=<optimized out>, skb=0xffff888027704400,
     net=0xffff888022858180, hook=1) at ./include/linux/netfilter.h:301
#10 ip_local_deliver (skb=0xffff888027704400) at net/ipv4/ip_input.c:252
#11 0xffffffff839ea58e in dst_input (skb=0xffff888027704400) at 
./include/linux/skbuff.h:980
#12 ip_rcv_finish (net=net@...ry=0xffff888022858180, sk=sk@...ry=0x0 
<fixed_percpu_data>, skb=skb@...ry=0xffff888027704400) at 
net/ipv4/ip_input.c:429
#13 0xffffffff839eb4fa in NF_HOOK (sk=0x0 <fixed_percpu_data>, pf=2 
'\002', hook=0, in=0xffff888022a20000, out=0x0 <fixed_percpu_data>, 
okfn=0xffffffff839ea4b0 <ip_rcv_finish>, skb=0xffff888027704400,
     net=0xffff888022858180) at ./include/linux/netfilter.h:307
#14 NF_HOOK (pf=2 '\002', sk=0x0 <fixed_percpu_data>, out=0x0 
<fixed_percpu_data>, okfn=0xffffffff839ea4b0 <ip_rcv_finish>, 
in=0xffff888022a20000, skb=0xffff888027704400, net=0xffff888022858180,
     hook=0) at ./include/linux/netfilter.h:301
#15 ip_rcv (skb=0xffff888027704400, dev=0xffff888022a20000, 
pt=<optimized out>, orig_dev=<optimized out>) at net/ipv4/ip_input.c:540
#16 0xffffffff83712acf in __netif_receive_skb_one_core (skb=<optimized 
out>, skb@...ry=0xffff888027704400, pfmemalloc=pfmemalloc@...ry=false) 
at net/core/dev.c:5482
#17 0xffffffff83712b59 in __netif_receive_skb (skb=0xffff888027704400) 
at net/core/dev.c:5596
#18 0xffffffff83712f09 in process_backlog (napi=0xffff88803ec2c8d0, 
quota=64) at net/core/dev.c:6460
#19 0xffffffff837159f2 in __napi_poll (n=n@...ry=0xffff88803ec2c8d0, 
repoll=repoll@...ry=0xffffc900007b7e0f) at net/core/dev.c:7015
#20 0xffffffff837161b2 in napi_poll (repoll=0xffffc900007b7e20, 
n=0xffff88803ec2c8d0) at net/core/dev.c:7082
#21 net_rx_action (h=<optimized out>) at net/core/dev.c:7169
#22 0xffffffff84600114 in __do_softirq () at kernel/softirq.c:558
#23 0xffffffff81241506 in run_ksoftirqd (cpu=<optimized out>) at 
kernel/softirq.c:920
#24 run_ksoftirqd (cpu=<optimized out>) at kernel/softirq.c:912
#25 0xffffffff8127ec83 in smpboot_thread_fn (data=0xffff888008f70240) at 
kernel/smpboot.c:164
#26 0xffffffff81274475 in kthread (_create=<optimized out>) at 
kernel/kthread.c:319
#27 0xffffffff8100230f in ret_from_fork () at arch/x86/entry/entry_64.S:295
#28 0x0000000000000000 in ?? ()

Best Regards.

Powered by blists - more mailing lists