| 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: <53444aeb-aa22-4076-95dd-f079861567c9@stanley.mountain>
Date: Mon, 4 Nov 2024 10:23:32 +0300
From: Dan Carpenter <dan.carpenter@...aro.org>
To: oe-kbuild@...ts.linux.dev, mrpre <mrpre@....com>,
yonghong.song@...ux.dev, john.fastabend@...il.com,
martin.lau@...nel.org, edumazet@...gle.com, jakub@...udflare.com,
davem@...emloft.net, dsahern@...nel.org, kuba@...nel.org,
pabeni@...hat.com, netdev@...r.kernel.org, bpf@...r.kernel.org,
linux-kernel@...r.kernel.org
Cc: lkp@...el.com, oe-kbuild-all@...ts.linux.dev, mrpre <mrpre@....com>
Subject: Re: [PATCH v2 1/2] bpf: Introduce cpu affinity for sockmap
Hi mrpre,
kernel test robot noticed the following build warnings:
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/mrpre/bpf-Introduce-cpu-affinity-for-sockmap/20241102-001844
base: https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git master
patch link: https://lore.kernel.org/r/20241101161624.568527-2-mrpre%40163.com
patch subject: [PATCH v2 1/2] bpf: Introduce cpu affinity for sockmap
config: i386-randconfig-141-20241102 (https://download.01.org/0day-ci/archive/20241103/202411030036.PSKG1pW3-lkp@intel.com/config)
compiler: clang version 19.1.3 (https://github.com/llvm/llvm-project ab51eccf88f5321e7c60591c5546b254b6afab99)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@...el.com>
| Reported-by: Dan Carpenter <dan.carpenter@...aro.org>
| Closes: https://lore.kernel.org/r/202411030036.PSKG1pW3-lkp@intel.com/
smatch warnings:
net/core/sock_map.c:511 sock_map_update_common() warn: variable dereferenced before check 'psock' (see line 492)
vim +/psock +511 net/core/sock_map.c
604326b41a6fb9 Daniel Borkmann 2018-10-13 467 static int sock_map_update_common(struct bpf_map *map, u32 idx,
ffed654afa8dc1 mrpre 2024-11-02 468 struct sock *sk, u64 flags, s32 target_cpu)
604326b41a6fb9 Daniel Borkmann 2018-10-13 469 {
604326b41a6fb9 Daniel Borkmann 2018-10-13 470 struct bpf_stab *stab = container_of(map, struct bpf_stab, map);
604326b41a6fb9 Daniel Borkmann 2018-10-13 471 struct sk_psock_link *link;
604326b41a6fb9 Daniel Borkmann 2018-10-13 472 struct sk_psock *psock;
604326b41a6fb9 Daniel Borkmann 2018-10-13 473 struct sock *osk;
604326b41a6fb9 Daniel Borkmann 2018-10-13 474 int ret;
604326b41a6fb9 Daniel Borkmann 2018-10-13 475
604326b41a6fb9 Daniel Borkmann 2018-10-13 476 WARN_ON_ONCE(!rcu_read_lock_held());
604326b41a6fb9 Daniel Borkmann 2018-10-13 477 if (unlikely(flags > BPF_EXIST))
604326b41a6fb9 Daniel Borkmann 2018-10-13 478 return -EINVAL;
604326b41a6fb9 Daniel Borkmann 2018-10-13 479 if (unlikely(idx >= map->max_entries))
604326b41a6fb9 Daniel Borkmann 2018-10-13 480 return -E2BIG;
604326b41a6fb9 Daniel Borkmann 2018-10-13 481
604326b41a6fb9 Daniel Borkmann 2018-10-13 482 link = sk_psock_init_link();
604326b41a6fb9 Daniel Borkmann 2018-10-13 483 if (!link)
604326b41a6fb9 Daniel Borkmann 2018-10-13 484 return -ENOMEM;
604326b41a6fb9 Daniel Borkmann 2018-10-13 485
2004fdbd8a2b56 Cong Wang 2021-03-30 486 ret = sock_map_link(map, sk);
604326b41a6fb9 Daniel Borkmann 2018-10-13 487 if (ret < 0)
604326b41a6fb9 Daniel Borkmann 2018-10-13 488 goto out_free;
604326b41a6fb9 Daniel Borkmann 2018-10-13 489
604326b41a6fb9 Daniel Borkmann 2018-10-13 490 psock = sk_psock(sk);
604326b41a6fb9 Daniel Borkmann 2018-10-13 491 WARN_ON_ONCE(!psock);
ffed654afa8dc1 mrpre 2024-11-02 @492 psock->target_cpu = target_cpu;
^^^^^^^^^^^^^^^^^
The patch adds an unchecked dereference
35d2b7ffffc1d9 John Fastabend 2023-08-29 493 spin_lock_bh(&stab->lock);
604326b41a6fb9 Daniel Borkmann 2018-10-13 494 osk = stab->sks[idx];
604326b41a6fb9 Daniel Borkmann 2018-10-13 495 if (osk && flags == BPF_NOEXIST) {
604326b41a6fb9 Daniel Borkmann 2018-10-13 496 ret = -EEXIST;
604326b41a6fb9 Daniel Borkmann 2018-10-13 497 goto out_unlock;
604326b41a6fb9 Daniel Borkmann 2018-10-13 498 } else if (!osk && flags == BPF_EXIST) {
604326b41a6fb9 Daniel Borkmann 2018-10-13 499 ret = -ENOENT;
604326b41a6fb9 Daniel Borkmann 2018-10-13 500 goto out_unlock;
604326b41a6fb9 Daniel Borkmann 2018-10-13 501 }
604326b41a6fb9 Daniel Borkmann 2018-10-13 502
604326b41a6fb9 Daniel Borkmann 2018-10-13 503 sock_map_add_link(psock, link, map, &stab->sks[idx]);
This also dereferences psock btw.
604326b41a6fb9 Daniel Borkmann 2018-10-13 504 stab->sks[idx] = sk;
604326b41a6fb9 Daniel Borkmann 2018-10-13 505 if (osk)
604326b41a6fb9 Daniel Borkmann 2018-10-13 506 sock_map_unref(osk, &stab->sks[idx]);
35d2b7ffffc1d9 John Fastabend 2023-08-29 507 spin_unlock_bh(&stab->lock);
604326b41a6fb9 Daniel Borkmann 2018-10-13 508 return 0;
604326b41a6fb9 Daniel Borkmann 2018-10-13 509 out_unlock:
35d2b7ffffc1d9 John Fastabend 2023-08-29 510 spin_unlock_bh(&stab->lock);
604326b41a6fb9 Daniel Borkmann 2018-10-13 @511 if (psock)
^^^^^
Probably after 6 years of not triggering the WARN_ON_ONCE() on line 490, we can
remove this check?
604326b41a6fb9 Daniel Borkmann 2018-10-13 512 sk_psock_put(sk, psock);
604326b41a6fb9 Daniel Borkmann 2018-10-13 513 out_free:
604326b41a6fb9 Daniel Borkmann 2018-10-13 514 sk_psock_free_link(link);
604326b41a6fb9 Daniel Borkmann 2018-10-13 515 return ret;
604326b41a6fb9 Daniel Borkmann 2018-10-13 516 }
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
Powered by blists - more mailing lists