netdev - Re: [PATCH v1 net] tcp: Fix shift-out-of-bounds in dctcp_update

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Fri, 17 May 2024 21:28:22 +0100
From: Simon Horman <horms@...nel.org>
To: Kuniyuki Iwashima <kuniyu@...zon.com>
Cc: "David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>,
	Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
	David Ahern <dsahern@...nel.org>,
	Daniel Borkmann <daniel@...earbox.net>,
	Florian Westphal <fw@...len.de>,
	Glenn Judd <glenn.judd@...ganstanley.com>,
	Kuniyuki Iwashima <kuni1840@...il.com>, netdev@...r.kernel.org,
	syzkaller <syzkaller@...glegroups.com>,
	Yue Sun <samsun1006219@...il.com>,
	xingwei lee <xrivendell7@...il.com>
Subject: Re: [PATCH v1 net] tcp: Fix shift-out-of-bounds in
 dctcp_update_alpha().

On Fri, May 17, 2024 at 06:16:26PM +0900, Kuniyuki Iwashima wrote:
> In dctcp_update_alpha(), we use a module parameter dctcp_shift_g
> as follows:
> 
>   alpha -= min_not_zero(alpha, alpha >> dctcp_shift_g);
>   ...
>   delivered_ce <<= (10 - dctcp_shift_g);
> 
> It seems syzkaller started fuzzing module parameters and triggered
> shift-out-of-bounds [0] by setting 100 to dctcp_shift_g:
> 
>   memcpy((void*)0x20000080,
>          "/sys/module/tcp_dctcp/parameters/dctcp_shift_g\000", 47);
>   res = syscall(__NR_openat, /*fd=*/0xffffffffffffff9cul, /*file=*/0x20000080ul,
>                 /*flags=*/2ul, /*mode=*/0ul);
>   memcpy((void*)0x20000000, "100\000", 4);
>   syscall(__NR_write, /*fd=*/r[0], /*val=*/0x20000000ul, /*len=*/4ul);
> 
> Let's limit the max value of dctcp_shift_g by param_set_uint_minmax().
> 
> With this patch:
> 
>   # echo 10 > /sys/module/tcp_dctcp/parameters/dctcp_shift_g
>   # cat /sys/module/tcp_dctcp/parameters/dctcp_shift_g
>   10
>   # echo 11 > /sys/module/tcp_dctcp/parameters/dctcp_shift_g
>   -bash: echo: write error: Invalid argument
> 
> [0]:
> UBSAN: shift-out-of-bounds in net/ipv4/tcp_dctcp.c:143:12
> shift exponent 100 is too large for 32-bit type 'u32' (aka 'unsigned int')
> CPU: 0 PID: 8083 Comm: syz-executor345 Not tainted 6.9.0-05151-g1b294a1f3561 #2
> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
> 1.13.0-1ubuntu1.1 04/01/2014
> Call Trace:
>  <TASK>
>  __dump_stack lib/dump_stack.c:88 [inline]
>  dump_stack_lvl+0x201/0x300 lib/dump_stack.c:114
>  ubsan_epilogue lib/ubsan.c:231 [inline]
>  __ubsan_handle_shift_out_of_bounds+0x346/0x3a0 lib/ubsan.c:468
>  dctcp_update_alpha+0x540/0x570 net/ipv4/tcp_dctcp.c:143
>  tcp_in_ack_event net/ipv4/tcp_input.c:3802 [inline]
>  tcp_ack+0x17b1/0x3bc0 net/ipv4/tcp_input.c:3948
>  tcp_rcv_state_process+0x57a/0x2290 net/ipv4/tcp_input.c:6711
>  tcp_v4_do_rcv+0x764/0xc40 net/ipv4/tcp_ipv4.c:1937
>  sk_backlog_rcv include/net/sock.h:1106 [inline]
>  __release_sock+0x20f/0x350 net/core/sock.c:2983
>  release_sock+0x61/0x1f0 net/core/sock.c:3549
>  mptcp_subflow_shutdown+0x3d0/0x620 net/mptcp/protocol.c:2907
>  mptcp_check_send_data_fin+0x225/0x410 net/mptcp/protocol.c:2976
>  __mptcp_close+0x238/0xad0 net/mptcp/protocol.c:3072
>  mptcp_close+0x2a/0x1a0 net/mptcp/protocol.c:3127
>  inet_release+0x190/0x1f0 net/ipv4/af_inet.c:437
>  __sock_release net/socket.c:659 [inline]
>  sock_close+0xc0/0x240 net/socket.c:1421
>  __fput+0x41b/0x890 fs/file_table.c:422
>  task_work_run+0x23b/0x300 kernel/task_work.c:180
>  exit_task_work include/linux/task_work.h:38 [inline]
>  do_exit+0x9c8/0x2540 kernel/exit.c:878
>  do_group_exit+0x201/0x2b0 kernel/exit.c:1027
>  __do_sys_exit_group kernel/exit.c:1038 [inline]
>  __se_sys_exit_group kernel/exit.c:1036 [inline]
>  __x64_sys_exit_group+0x3f/0x40 kernel/exit.c:1036
>  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
>  do_syscall_64+0xe4/0x240 arch/x86/entry/common.c:83
>  entry_SYSCALL_64_after_hwframe+0x67/0x6f
> RIP: 0033:0x7f6c2b5005b6
> Code: Unable to access opcode bytes at 0x7f6c2b50058c.
> RSP: 002b:00007ffe883eb948 EFLAGS: 00000246 ORIG_RAX: 00000000000000e7
> RAX: ffffffffffffffda RBX: 00007f6c2b5862f0 RCX: 00007f6c2b5005b6
> RDX: 0000000000000001 RSI: 000000000000003c RDI: 0000000000000001
> RBP: 0000000000000001 R08: 00000000000000e7 R09: ffffffffffffffc0
> R10: 0000000000000006 R11: 0000000000000246 R12: 00007f6c2b5862f0
> R13: 0000000000000001 R14: 0000000000000000 R15: 0000000000000001
>  </TASK>
> 
> Reported-by: syzkaller <syzkaller@...glegroups.com>
> Reported-by: Yue Sun <samsun1006219@...il.com>
> Reported-by: xingwei lee <xrivendell7@...il.com>
> Closes: https://lore.kernel.org/netdev/CAEkJfYNJM=cw-8x7_Vmj1J6uYVCWMbbvD=EFmDPVBGpTsqOxEA@mail.gmail.com/
> Fixes: e3118e8359bb ("net: tcp: add DCTCP congestion control algorithm")
> Signed-off-by: Kuniyuki Iwashima <kuniyu@...zon.com>

Reviewed-by: Simon Horman <horms@...nel.org>