On Tue, Jun 11, 2024 at 5:55 PM Kuniyuki Iwashima kuniyu@amazon.com wrote:
From: Ignat Korchagin ignat@cloudflare.com Date: Tue, 11 Jun 2024 19:47:16 +0100
It is possible to trigger a use-after-free by:
- attaching an fentry probe to __sock_release() and the probe calling the bpf_get_socket_cookie() helper
- running traceroute -I 1.1.1.1 on a freshly booted VM
A KASAN enabled kernel will log something like below (decoded): [ 78.328507][ T299] ================================================================== [ 78.329018][ T299] BUG: KASAN: slab-use-after-free in __sock_gen_cookie (./arch/x86/include/asm/atomic64_64.h:15 ./include/linux/atomic/atomic-arch-fallback.h:2583 ./include/linux/atomic/atomic-instrumented.h:1611 net/core/sock_diag.c:29) [ 78.329366][ T299] Read of size 8 at addr ffff888007110dd8 by task traceroute/299 [ 78.329366][ T299] [ 78.329366][ T299] CPU: 2 PID: 299 Comm: traceroute Tainted: G E 6.10.0-rc2+ #2 [ 78.329366][ T299] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.2-debian-1.16.2-1 04/01/2014 [ 78.329366][ T299] Call Trace: [ 78.329366][ T299] <TASK> [ 78.329366][ T299] dump_stack_lvl (lib/dump_stack.c:117 (discriminator 1)) [ 78.329366][ T299] print_report (mm/kasan/report.c:378 mm/kasan/report.c:488) [ 78.329366][ T299] ? __sock_gen_cookie (./arch/x86/include/asm/atomic64_64.h:15 ./include/linux/atomic/atomic-arch-fallback.h:2583 ./include/linux/atomic/atomic-instrumented.h:1611 net/core/sock_diag.c:29) [ 78.329366][ T299] kasan_report (mm/kasan/report.c:603) [ 78.329366][ T299] ? __sock_gen_cookie (./arch/x86/include/asm/atomic64_64.h:15 ./include/linux/atomic/atomic-arch-fallback.h:2583 ./include/linux/atomic/atomic-instrumented.h:1611 net/core/sock_diag.c:29) [ 78.329366][ T299] kasan_check_range (mm/kasan/generic.c:183 mm/kasan/generic.c:189) [ 78.329366][ T299] __sock_gen_cookie (./arch/x86/include/asm/atomic64_64.h:15 ./include/linux/atomic/atomic-arch-fallback.h:2583 ./include/linux/atomic/atomic-instrumented.h:1611 net/core/sock_diag.c:29) [ 78.329366][ T299] bpf_get_socket_ptr_cookie (./arch/x86/include/asm/preempt.h:94 ./include/linux/sock_diag.h:42 net/core/filter.c:5094 net/core/filter.c:5092) [ 78.329366][ T299] bpf_prog_875642cf11f1d139___sock_release+0x6e/0x8e [ 78.329366][ T299] bpf_trampoline_6442506592+0x47/0xaf [ 78.329366][ T299] __sock_release (net/socket.c:652) [ 78.329366][ T299] __sock_create (net/socket.c:1601) [ 78.329366][ T299] ? srso_return_thunk (arch/x86/lib/retpoline.S:224) [ 78.329366][ T299] __sys_socket (net/socket.c:1660 net/socket.c:1644 net/socket.c:1706) [ 78.329366][ T299] ? __pfx___sys_socket (net/socket.c:1702) [ 78.329366][ T299] ? srso_return_thunk (arch/x86/lib/retpoline.S:224) [ 78.329366][ T299] ? up_read (./arch/x86/include/asm/atomic64_64.h:79 ./include/linux/atomic/atomic-arch-fallback.h:2749 ./include/linux/atomic/atomic-long.h:184 ./include/linux/atomic/atomic-instrumented.h:3317 kernel/locking/rwsem.c:1347 kernel/locking/rwsem.c:1622) [ 78.329366][ T299] ? srso_return_thunk (arch/x86/lib/retpoline.S:224) [ 78.329366][ T299] ? do_user_addr_fault (arch/x86/mm/fault.c:1419) [ 78.329366][ T299] __x64_sys_socket (net/socket.c:1718) [ 78.329366][ T299] ? srso_return_thunk (arch/x86/lib/retpoline.S:224) [ 78.329366][ T299] do_syscall_64 (arch/x86/entry/common.c:52 arch/x86/entry/common.c:83) [ 78.329366][ T299] entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) [ 78.329366][ T299] RIP: 0033:0x7f4022818ca7 [ 78.329366][ T299] Code: 73 01 c3 48 8b 0d 59 71 0c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 b8 29 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 29 71 0c 00 f7 d8 64 89 01 48 All code ======== 0: 73 01 jae 0x3 2: c3 ret 3: 48 8b 0d 59 71 0c 00 mov 0xc7159(%rip),%rcx # 0xc7163 a: f7 d8 neg %eax c: 64 89 01 mov %eax,%fs:(%rcx) f: 48 83 c8 ff or $0xffffffffffffffff,%rax 13: c3 ret 14: 66 2e 0f 1f 84 00 00 cs nopw 0x0(%rax,%rax,1) 1b: 00 00 00 1e: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1) 23: b8 29 00 00 00 mov $0x29,%eax 28: 0f 05 syscall 2a:* 48 3d 01 f0 ff ff cmp $0xfffffffffffff001,%rax <-- trapping instruction 30: 73 01 jae 0x33 32: c3 ret 33: 48 8b 0d 29 71 0c 00 mov 0xc7129(%rip),%rcx # 0xc7163 3a: f7 d8 neg %eax 3c: 64 89 01 mov %eax,%fs:(%rcx) 3f: 48 rex.W
Code starting with the faulting instruction
0: 48 3d 01 f0 ff ff cmp $0xfffffffffffff001,%rax 6: 73 01 jae 0x9 8: c3 ret 9: 48 8b 0d 29 71 0c 00 mov 0xc7129(%rip),%rcx # 0xc7139 10: f7 d8 neg %eax 12: 64 89 01 mov %eax,%fs:(%rcx) 15: 48 rex.W [ 78.329366][ T299] RSP: 002b:00007ffd57e63db8 EFLAGS: 00000246 ORIG_RAX: 0000000000000029 [ 78.329366][ T299] RAX: ffffffffffffffda RBX: 0000000000000001 RCX: 00007f4022818ca7 [ 78.329366][ T299] RDX: 0000000000000001 RSI: 0000000000000002 RDI: 0000000000000002 [ 78.329366][ T299] RBP: 0000000000000002 R08: 0000000000000000 R09: 0000564be3dc8ec0 [ 78.329366][ T299] R10: 0c41e8ba3f6107df R11: 0000000000000246 R12: 0000564bbab801e0 [ 78.329366][ T299] R13: 0000000000000000 R14: 0000564bbab7db18 R15: 00007f4022934020 [ 78.329366][ T299] </TASK> [ 78.329366][ T299] [ 78.329366][ T299] Allocated by task 299 on cpu 2 at 78.328492s: [ 78.329366][ T299] kasan_save_stack (mm/kasan/common.c:48) [ 78.329366][ T299] kasan_save_track (mm/kasan/common.c:68) [ 78.329366][ T299] __kasan_slab_alloc (mm/kasan/common.c:312 mm/kasan/common.c:338) [ 78.329366][ T299] kmem_cache_alloc_noprof (mm/slub.c:3941 mm/slub.c:4000 mm/slub.c:4007) [ 78.329366][ T299] sk_prot_alloc (net/core/sock.c:2075) [ 78.329366][ T299] sk_alloc (net/core/sock.c:2134) [ 78.329366][ T299] inet_create (net/ipv4/af_inet.c:327 net/ipv4/af_inet.c:252) [ 78.329366][ T299] __sock_create (net/socket.c:1572) [ 78.329366][ T299] __sys_socket (net/socket.c:1660 net/socket.c:1644 net/socket.c:1706) [ 78.329366][ T299] __x64_sys_socket (net/socket.c:1718) [ 78.329366][ T299] do_syscall_64 (arch/x86/entry/common.c:52 arch/x86/entry/common.c:83) [ 78.329366][ T299] entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) [ 78.329366][ T299] [ 78.329366][ T299] Freed by task 299 on cpu 2 at 78.328502s: [ 78.329366][ T299] kasan_save_stack (mm/kasan/common.c:48) [ 78.329366][ T299] kasan_save_track (mm/kasan/common.c:68) [ 78.329366][ T299] kasan_save_free_info (mm/kasan/generic.c:582) [ 78.329366][ T299] poison_slab_object (mm/kasan/common.c:242) [ 78.329366][ T299] __kasan_slab_free (mm/kasan/common.c:256) [ 78.329366][ T299] kmem_cache_free (mm/slub.c:4437 mm/slub.c:4511) [ 78.329366][ T299] __sk_destruct (net/core/sock.c:2117 net/core/sock.c:2208) [ 78.329366][ T299] inet_create (net/ipv4/af_inet.c:397 net/ipv4/af_inet.c:252) [ 78.329366][ T299] __sock_create (net/socket.c:1572) [ 78.329366][ T299] __sys_socket (net/socket.c:1660 net/socket.c:1644 net/socket.c:1706) [ 78.329366][ T299] __x64_sys_socket (net/socket.c:1718) [ 78.329366][ T299] do_syscall_64 (arch/x86/entry/common.c:52 arch/x86/entry/common.c:83) [ 78.329366][ T299] entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) [ 78.329366][ T299] [ 78.329366][ T299] The buggy address belongs to the object at ffff888007110d80 [ 78.329366][ T299] which belongs to the cache PING of size 976 [ 78.329366][ T299] The buggy address is located 88 bytes inside of [ 78.329366][ T299] freed 976-byte region [ffff888007110d80, ffff888007111150) [ 78.329366][ T299] [ 78.329366][ T299] The buggy address belongs to the physical page: [ 78.329366][ T299] page: refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x7110 [ 78.329366][ T299] head: order:3 mapcount:0 entire_mapcount:0 nr_pages_mapped:0 pincount:0 [ 78.329366][ T299] flags: 0x1ffff800000040(head|node=0|zone=1|lastcpupid=0x1ffff) [ 78.329366][ T299] page_type: 0xffffefff(slab) [ 78.329366][ T299] raw: 001ffff800000040 ffff888002f328c0 dead000000000122 0000000000000000 [ 78.329366][ T299] raw: 0000000000000000 00000000801c001c 00000001ffffefff 0000000000000000 [ 78.329366][ T299] head: 001ffff800000040 ffff888002f328c0 dead000000000122 0000000000000000 [ 78.329366][ T299] head: 0000000000000000 00000000801c001c 00000001ffffefff 0000000000000000 [ 78.329366][ T299] head: 001ffff800000003 ffffea00001c4401 ffffffffffffffff 0000000000000000 [ 78.329366][ T299] head: 0000000000000008 0000000000000000 00000000ffffffff 0000000000000000 [ 78.329366][ T299] page dumped because: kasan: bad access detected [ 78.329366][ T299] [ 78.329366][ T299] Memory state around the buggy address: [ 78.329366][ T299] ffff888007110c80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc [ 78.329366][ T299] ffff888007110d00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc [ 78.329366][ T299] >ffff888007110d80: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 78.329366][ T299] ^ [ 78.329366][ T299] ffff888007110e00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 78.329366][ T299] ffff888007110e80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 78.329366][ T299] ================================================================== [ 78.366431][ T299] Disabling lock debugging due to kernel taint
Fix this by ensuring the error path of inet_create()/inet6_create do not leave a dangling sk pointer after sk was released.
Fixes: 086c653f5862 ("sock: struct proto hash function may error")
I think this tag is wrong as bpf_get_socket_cookie() does not exist at that time.
OK - will probably replace the tag with the commit that added it.
Fixes: 610236587600 ("bpf: Add new cgroup attach type to enable sock modifications") Cc: stable@vger.kernel.org Signed-off-by: Ignat Korchagin ignat@cloudflare.com
net/ipv4/af_inet.c | 3 +++ net/ipv6/af_inet6.c | 3 +++ 2 files changed, 6 insertions(+)
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index b24d74616637..db53701db29e 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -378,6 +378,7 @@ static int inet_create(struct net *net, struct socket *sock, int protocol, err = sk->sk_prot->hash(sk); if (err) { sk_common_release(sk);
sock->sk = NULL; goto out; } }You can add a new label and call sk_common_release() and set NULL to sock->sk there, then reuse it for other two places.
Same for IPv6.
OK
And curious if bpf_get_socket_cookie() can be called any socket family to trigger the splat. e.g. ieee802154_create() seems to have the same bug.
Just judging from the code - yes, indeed.
If so, how about clearing sock->sk in sk_common_release() ?
This was my first thought, but I was a bit put off by the fact that sk_common_release() is called from many places and the sk object itself is reference counted. So not every call to sk_common_release() seems to actually free the sk object. Secondly, I was put off by this comment (which I don't fully understand TBH) [1]
On the other hand - in inet/inet6_create() we definitely know that the object would be freed, because we just created that.
But if someone more familiar with the code confirms it is better/possible to do in sk_common_release(), I'm happy to adjust and it would be cleaner indeed.
---8<--- diff --git a/net/core/sock.c b/net/core/sock.c index 8629f9aecf91..bbc94954d9bf 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3754,6 +3754,9 @@ void sk_common_release(struct sock *sk) * until the last reference will be released. */
if (sk->sk_socket)sk->sk_socket->sk = NULL;sock_orphan(sk); xfrm_sk_free_policy(sk);---8<---
[1]: https://elixir.bootlin.com/linux/v6.10-rc3/source/include/net/sock.h#L1985