| 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
| ||
|
Date: Tue, 6 Dec 2022 10:31:10 -0800
From: Kees Cook <keescook@...omium.org>
To: Dmitry Vyukov <dvyukov@...gle.com>
Cc: pepsipu <soopthegoop@...il.com>,
syzbot+fda18eaa8c12534ccb3b@...kaller.appspotmail.com,
Vlastimil Babka <vbabka@...e.cz>,
kasan-dev <kasan-dev@...glegroups.com>,
syzkaller-bugs <syzkaller-bugs@...glegroups.com>,
Andrii Nakryiko <andrii@...nel.org>, ast@...nel.org,
bpf <bpf@...r.kernel.org>,
Daniel Borkmann <daniel@...earbox.net>,
David Miller <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>,
Hao Luo <haoluo@...gle.com>,
Jesper Dangaard Brouer <hawk@...nel.org>,
John Fastabend <john.fastabend@...il.com>, jolsa@...nel.org,
KP Singh <kpsingh@...nel.org>,
Jakub Kicinski <kuba@...nel.org>,
LKML <linux-kernel@...r.kernel.org>, martin.lau@...ux.dev,
netdev <netdev@...r.kernel.org>, Paolo Abeni <pabeni@...hat.com>,
Stanislav Fomichev <sdf@...gle.com>, song@...nel.org,
Yonghong Song <yhs@...com>
Subject: Re: [syzbot] KASAN: slab-out-of-bounds Write in __build_skb_around
On Mon, Dec 5, 2022 at 12:04 AM Dmitry Vyukov <dvyukov@...gle.com> wrote:
>
> On Sun, 4 Dec 2022 at 19:16, pepsipu <soopthegoop@...il.com> wrote:
> >
> > I believe this is a KASAN bug.
> >
> > I made an easier to read version that still triggers KASAN:
> >
> > #define _GNU_SOURCE
> >
> > #include <stdio.h>
> > #include <stdlib.h>
> > #include <string.h>
> > #include <sys/syscall.h>
> > #include <sys/types.h>
> > #include <linux/bpf.h>
> > #include <unistd.h>
> >
> > #include "bpf.h"
> >
> > int main(void)
> > {
> > __u64 insns[] = {
> > (BPF_CALL | BPF_JMP) | ((__u64)0x61 << 32),
> > (BPF_AND | BPF_ALU),
> > (BPF_EXIT | BPF_JMP),
> > };
> > bpf_load_attr_t load_attr = {
> > .prog_type = BPF_PROG_TYPE_CGROUP_SKB,
> > .insn_cnt = sizeof(insns) / sizeof(__u64),
> > .insns = (__u64)insns,
> > .license = (__u64) "GPL",
> > };
> > long prog_fd = syscall(__NR_bpf, BPF_PROG_LOAD, &load_attr, sizeof(bpf_load_attr_t));
> > if (prog_fd == -1)
> > {
> > printf("could not load bpf prog");
> > exit(-1);
> > }
> > bpf_trun_attr_t trun_attr = {
> > .prog_fd = prog_fd,
> > .data_size_in = 0x81,
> > .data_size_out = -1,
> > .data_in = (__u64) "AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA",
> > };
> >
> > syscall(__NR_bpf, BPF_PROG_TEST_RUN, &trun_attr, sizeof(bpf_trun_attr_t));
> > return 0;
> > }
> >
> > It looks like KASAN believes the tail access of SKB's backing buffer, the SKB shared info struct, allocated by bpf_test_init is out-of-bounds.
> > This is likely because when the SKB is setup, in build_skb, the tail is calculated as "data + ksize(data) - sizeof(skb_shared_info)". ksize returns the size of the slab, not the allocation, so the tail is much further past the allocation.
> > However, KASAN is usually supposed to correct for ksize calls by unpoisioning the entire slab it's called on... I'm not sure why this is happening.
>
> Hi,
>
> [+orignal CC list, please keep it in replies, almost none of relevant
> receivers read syzkaller-bugs@ mailing list]
>
> Also +Kees and kasan-dev for ksize.
>
> After the following patch the behavior has changed and KASAN does not
> unpoison the fail of the object:
>
> mm: Make ksize() a reporting-only function
> https://lore.kernel.org/all/20221118035656.gonna.698-kees@kernel.org/
>
> Kees, is this bpf case is a remaining ksize() use that needs to be fixed?
>
Hi, yes, this seems like a missed ksize() usage. I will take a look at
it -- nothing should be using ksize() to resize the allocation any
more: it should either fully allocate the bucket at the start, or use
krealloc().
-Kees
> > On Monday, November 28, 2022 at 5:42:31 AM UTC-8 syzbot wrote:
> >>
> >> Hello,
> >>
> >> syzbot found the following issue on:
> >>
> >> HEAD commit: c35bd4e42885 Add linux-next specific files for 20221124
> >> git tree: linux-next
> >> console+strace: https://syzkaller.appspot.com/x/log.txt?x=15e5d7e5880000
> >> kernel config: https://syzkaller.appspot.com/x/.config?x=11e19c740a0b2926
> >> dashboard link: https://syzkaller.appspot.com/bug?extid=fda18eaa8c12534ccb3b
> >> compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> >> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=1096f205880000
> >> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=10b2d68d880000
> >>
> >> Downloadable assets:
> >> disk image: https://storage.googleapis.com/syzbot-assets/968fee464d14/disk-c35bd4e4.raw.xz
> >> vmlinux: https://storage.googleapis.com/syzbot-assets/4f46fe801b5b/vmlinux-c35bd4e4.xz
> >> kernel image: https://storage.googleapis.com/syzbot-assets/c2cdf8fb264e/bzImage-c35bd4e4.xz
> >>
> >> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> >> Reported-by: syzbot+fda18e...@...kaller.appspotmail.com
> >>
> >> ==================================================================
> >> BUG: KASAN: slab-out-of-bounds in __build_skb_around+0x235/0x340 net/core/skbuff.c:294
> >> Write of size 32 at addr ffff88802aa172c0 by task syz-executor413/5295
> >>
> >> CPU: 0 PID: 5295 Comm: syz-executor413 Not tainted 6.1.0-rc6-next-20221124-syzkaller #0
> >> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022
> >> Call Trace:
> >> <TASK>
> >> __dump_stack lib/dump_stack.c:88 [inline]
> >> dump_stack_lvl+0xd1/0x138 lib/dump_stack.c:106
> >> print_address_description mm/kasan/report.c:253 [inline]
> >> print_report+0x15e/0x45d mm/kasan/report.c:364
> >> kasan_report+0xbf/0x1f0 mm/kasan/report.c:464
> >> check_region_inline mm/kasan/generic.c:183 [inline]
> >> kasan_check_range+0x141/0x190 mm/kasan/generic.c:189
> >> memset+0x24/0x50 mm/kasan/shadow.c:44
> >> __build_skb_around+0x235/0x340 net/core/skbuff.c:294
> >> __build_skb+0x4f/0x60 net/core/skbuff.c:328
> >> build_skb+0x22/0x280 net/core/skbuff.c:340
> >> bpf_prog_test_run_skb+0x343/0x1e10 net/bpf/test_run.c:1131
> >> bpf_prog_test_run kernel/bpf/syscall.c:3644 [inline]
> >> __sys_bpf+0x1599/0x4ff0 kernel/bpf/syscall.c:4997
> >> __do_sys_bpf kernel/bpf/syscall.c:5083 [inline]
> >> __se_sys_bpf kernel/bpf/syscall.c:5081 [inline]
> >> __x64_sys_bpf+0x79/0xc0 kernel/bpf/syscall.c:5081
> >> do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> >> do_syscall_64+0x39/0xb0 arch/x86/entry/common.c:80
> >> entry_SYSCALL_64_after_hwframe+0x63/0xcd
> >> RIP: 0033:0x7f30de9aad19
> >> Code: 28 c3 e8 2a 14 00 00 66 2e 0f 1f 84 00 00 00 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 c0 ff ff ff f7 d8 64 89 01 48
> >> RSP: 002b:00007ffeaee34318 EFLAGS: 00000246 ORIG_RAX: 0000000000000141
> >> RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f30de9aad19
> >> RDX: 0000000000000028 RSI: 0000000020000180 RDI: 000000000000000a
> >> RBP: 00007f30de96eec0 R08: 0000000000000000 R09: 0000000000000000
> >> R10: 0000000000000000 R11: 0000000000000246 R12: 00007f30de96ef50
> >> R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
> >> </TASK>
> >>
> >> Allocated by task 5295:
> >> kasan_save_stack+0x22/0x40 mm/kasan/common.c:45
> >> kasan_set_track+0x25/0x30 mm/kasan/common.c:52
> >> ____kasan_kmalloc mm/kasan/common.c:376 [inline]
> >> ____kasan_kmalloc mm/kasan/common.c:335 [inline]
> >> __kasan_kmalloc+0xa5/0xb0 mm/kasan/common.c:385
> >> kasan_kmalloc include/linux/kasan.h:212 [inline]
> >> __do_kmalloc_node mm/slab_common.c:955 [inline]
> >> __kmalloc+0x5a/0xd0 mm/slab_common.c:968
> >> kmalloc include/linux/slab.h:575 [inline]
> >> kzalloc include/linux/slab.h:711 [inline]
> >> bpf_test_init.isra.0+0xa5/0x150 net/bpf/test_run.c:778
> >> bpf_prog_test_run_skb+0x22e/0x1e10 net/bpf/test_run.c:1097
> >> bpf_prog_test_run kernel/bpf/syscall.c:3644 [inline]
> >> __sys_bpf+0x1599/0x4ff0 kernel/bpf/syscall.c:4997
> >> __do_sys_bpf kernel/bpf/syscall.c:5083 [inline]
> >> __se_sys_bpf kernel/bpf/syscall.c:5081 [inline]
> >> __x64_sys_bpf+0x79/0xc0 kernel/bpf/syscall.c:5081
> >> do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> >> do_syscall_64+0x39/0xb0 arch/x86/entry/common.c:80
> >> entry_SYSCALL_64_after_hwframe+0x63/0xcd
--
Kees Cook
Powered by blists - more mailing lists