| 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
| ||
|
Message-ID: <20250904092432.113c4940@kernel.org>
Date: Thu, 4 Sep 2025 09:24:32 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Eric Dumazet <edumazet@...gle.com>
Cc: "David S . Miller" <davem@...emloft.net>, Paolo Abeni
<pabeni@...hat.com>, Simon Horman <horms@...nel.org>, David Ahern
<dsahern@...nel.org>, Jamie Bainbridge <jamie.bainbridge@...il.com>,
Abhishek Rawal <rawal.abhishek92@...il.com>, netdev@...r.kernel.org,
eric.dumazet@...il.com
Subject: Re: [PATCH net-next 0/3] ipv6: snmp: avoid performance issue with
RATELIMITHOST
On Thu, 4 Sep 2025 13:25:50 +0000 Eric Dumazet wrote:
> Addition of ICMP6_MIB_RATELIMITHOST in commit d0941130c9351
> ("icmp: Add counters for rate limits") introduced a performance
> drop in case of DOS (like receiving UDP packets
> to closed ports).
>
> Per netns ICMP6_MIB_RATELIMITHOST tracking uses per-cpu
> storage and is enough, we do not need per-device and slow tracking
> for this metric.
CI says:
==================================================================
[ 156.608382][ T330] BUG: KASAN: global-out-of-bounds in snmp6_seq_show_item64.constprop.0 (net/ipv6/proc.c:211)
[ 156.608699][ T330] Read of size 8 at addr ffffffffad4d4790 by task connect-deny_ip/330
[ 156.608965][ T330]
[ 156.609064][ T330] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
[ 156.609066][ T330] Call Trace:
[ 156.609069][ T330] <TASK>
[ 156.609071][ T330] dump_stack_lvl (lib/dump_stack.c:123)
[ 156.609077][ T330] print_address_description.constprop.0 (mm/kasan/report.c:379)
[ 156.609084][ T330] ? snmp6_seq_show_item64.constprop.0 (net/ipv6/proc.c:211)
[ 156.609088][ T330] print_report (mm/kasan/report.c:483)
[ 156.609091][ T330] ? snmp6_seq_show_item64.constprop.0 (net/ipv6/proc.c:211)
[ 156.609094][ T330] ? kasan_addr_to_slab (./include/linux/mm.h:1180 mm/kasan/../slab.h:187 mm/kasan/common.c:38)
[ 156.609098][ T330] ? snmp6_seq_show_item64.constprop.0 (net/ipv6/proc.c:211)
[ 156.609101][ T330] kasan_report (mm/kasan/report.c:597)
[ 156.609105][ T330] ? snmp6_seq_show_item64.constprop.0 (net/ipv6/proc.c:211)
[ 156.609111][ T330] snmp6_seq_show_item64.constprop.0 (net/ipv6/proc.c:211)
[ 156.609119][ T330] ? sockstat6_seq_show (net/ipv6/proc.c:202)
[ 156.609137][ T330] ? rcu_is_watching (./include/linux/context_tracking.h:128 kernel/rcu/tree.c:751)
[ 156.609142][ T330] ? trace_kmalloc (./include/trace/events/kmem.h:54 (discriminator 21))
[ 156.609146][ T330] ? __kvmalloc_node_noprof (mm/slub.c:5055)
[ 156.609152][ T330] snmp6_seq_show (net/ipv6/proc.c:224)
[ 156.609155][ T330] seq_read_iter (fs/seq_file.c:231)
[ 156.609165][ T330] seq_read (fs/seq_file.c:163)
[ 156.609169][ T330] ? seq_read_iter (fs/seq_file.c:152)
[ 156.609173][ T330] ? __lock_acquire (kernel/locking/lockdep.c:5237)
[ 156.609180][ T330] ? __mutex_trylock_common (./arch/x86/include/asm/atomic64_64.h:101 ./include/linux/atomic/atomic-arch-fallback.h:4296 ./include/linux/atomic/atomic-long.h:1482 ./include/linux/atomic/atomic-instrumented.h:4458 kernel/locking/mutex.c:113)
[ 156.609184][ T330] ? rcu_is_watching (./include/linux/context_tracking.h:128 kernel/rcu/tree.c:751)
[ 156.609189][ T330] proc_reg_read (fs/proc/inode.c:308 fs/proc/inode.c:320)
[ 156.609196][ T330] vfs_read (fs/read_write.c:570)
[ 156.609201][ T330] ? fdget_pos (fs/file.c:1235)
[ 156.609206][ T330] ? ww_mutex_lock (kernel/locking/mutex.c:759)
[ 156.609212][ T330] ? kernel_read (fs/read_write.c:553)
[ 156.609215][ T330] ? __lock_release (kernel/locking/lockdep.c:5536)
[ 156.609220][ T330] ? __fget_files (./include/linux/rcupdate.h:341 ./include/linux/rcupdate.h:871 fs/file.c:1072)
[ 156.609223][ T330] ? __fget_files (fs/file.c:1075)
[ 156.609230][ T330] ksys_read (fs/read_write.c:715)
[ 156.609234][ T330] ? vfs_write (fs/read_write.c:705)
[ 156.609240][ T330] do_syscall_64 (arch/x86/entry/syscall_64.c:63 arch/x86/entry/syscall_64.c:94)
[ 156.609244][ T330] entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130)
[ 156.609248][ T330] RIP: 0033:0x7f5190d4a2cc
[ 156.609252][ T330] Code: ec 28 48 89 54 24 18 48 89 74 24 10 89 7c 24 08 e8 29 8b f8 ff 48 8b 54 24 18 48 8b 74 24 10 41 89 c0 8b 7c 24 08 31 c0 0f 05 <48> 3d 00 f0 ff ff 77 34 44 89 c7 48 89 44 24 08 e8 7f 8b f8 ff 48
All code
========
0: ec in (%dx),%al
1: 28 48 89 sub %cl,-0x77(%rax)
4: 54 push %rsp
5: 24 18 and $0x18,%al
7: 48 89 74 24 10 mov %rsi,0x10(%rsp)
c: 89 7c 24 08 mov %edi,0x8(%rsp)
10: e8 29 8b f8 ff call 0xfffffffffff88b3e
15: 48 8b 54 24 18 mov 0x18(%rsp),%rdx
1a: 48 8b 74 24 10 mov 0x10(%rsp),%rsi
1f: 41 89 c0 mov %eax,%r8d
22: 8b 7c 24 08 mov 0x8(%rsp),%edi
26: 31 c0 xor %eax,%eax
28: 0f 05 syscall
2a:* 48 3d 00 f0 ff ff cmp $0xfffffffffffff000,%rax <-- trapping instruction
30: 77 34 ja 0x66
32: 44 89 c7 mov %r8d,%edi
35: 48 89 44 24 08 mov %rax,0x8(%rsp)
3a: e8 7f 8b f8 ff call 0xfffffffffff88bbe
3f: 48 rex.W
Code starting with the faulting instruction
===========================================
0: 48 3d 00 f0 ff ff cmp $0xfffffffffffff000,%rax
6: 77 34 ja 0x3c
8: 44 89 c7 mov %r8d,%edi
b: 48 89 44 24 08 mov %rax,0x8(%rsp)
10: e8 7f 8b f8 ff call 0xfffffffffff88b94
15: 48 rex.W
[ 156.609255][ T330] RSP: 002b:00007f518bffe090 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
[ 156.609259][ T330] RAX: ffffffffffffffda RBX: 00007f5184000f30 RCX: 00007f5190d4a2cc
[ 156.609261][ T330] RDX: 0000000000000400 RSI: 00007f5184001190 RDI: 000000000000000b
[ 156.609263][ T330] RBP: 00007f5190e425c0 R08: 0000000000000000 R09: 0000000000000000
[ 156.609265][ T330] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
Powered by blists - more mailing lists