Skip to contentSkip to conversation

net: do not leave dangling sk pointer in inet_create()/inet6_create()

PATCH In review Ignat Korchagin 2 participants 5 messages Jun 11, 2024 – Jun 14, 2024

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: do not leave dangling sk pointer in inet_create()/inet6_create()

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")
Fixes 610236587600 ("bpf: Add new cgroup attach type to enable sock modifications")
---
 net/ipv4/af_inet.c  | 3 +++
 net/ipv6/af_inet6.c | 3 +++
 2 files changed, 6 insertions(+)
net/ipv4/af_inet.c +3 -0
index b24d74616637..db53701db29e 100644
@@ -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;
 		}
 	}
@@ -386,6 +387,7 @@ static int inet_create(struct net *net, struct socket *sock, int protocol,
 		err = sk->sk_prot->init(sk);
 		if (err) {
 			sk_common_release(sk);
+			sock->sk = NULL;
 			goto out;
 		}
 	}
@@ -394,6 +396,7 @@ static int inet_create(struct net *net, struct socket *sock, int protocol,
 		err = BPF_CGROUP_RUN_PROG_INET_SOCK(sk);
 		if (err) {
 			sk_common_release(sk);
+			sock->sk = NULL;
 			goto out;
 		}
 	}
net/ipv6/af_inet6.c +3 -0
index 8041dc181bd4..6d5ebb2af928 100644
@@ -254,6 +254,7 @@ static int inet6_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;
 		}
 	}
@@ -261,6 +262,7 @@ static int inet6_create(struct net *net, struct socket *sock, int protocol,
 		err = sk->sk_prot->init(sk);
 		if (err) {
 			sk_common_release(sk);
+			sock->sk = NULL;
 			goto out;
 		}
 	}
@@ -269,6 +271,7 @@ static int inet6_create(struct net *net, struct socket *sock, int protocol,
 		err = BPF_CGROUP_RUN_PROG_INET_SOCK(sk);
 		if (err) {
 			sk_common_release(sk);
+			sock->sk = NULL;
 			goto out;
 		}
 	}
Signature · 1 line
2.39.2
#1 lore raw

From: Ignat Korchagin <[email protected]>
Date: Tue, 11 Jun 2024 19:47:16 +0100

Quoted 135 lines

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
Quoted 4 lines

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.

Quoted 13 lines
Fixes 610236587600 ("bpf: Add new cgroup attach type to enable sock modifications")
---
 net/ipv4/af_inet.c  | 3 +++
 net/ipv6/af_inet6.c | 3 +++
 2 files changed, 6 insertions(+)
net/ipv4/af_inet.c +0 -0
index b24d74616637..db53701db29e 100644

net/ipv4/af_inet.c static int inet_create(struct net *net, struct socket *sock, int protocol, · line 384

Quoted 7 lines
net/ipv4/af_inet.c +1 -0
@@ -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.

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.

If so, how about clearing sock->sk in sk_common_release() ?

---8<---

net/core/sock.c +3 -0
index 8629f9aecf91..bbc94954d9bf 100644
@@ -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<---

#2 lore raw
Quoted 138 lines · Kuniyuki Iwashima

From: Ignat Korchagin <[email protected]>
Date: Tue, 11 Jun 2024 19:47:16 +0100

Quoted 135 lines

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
Quoted 7 lines
Quoted 4 lines

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.

Quoted 25 lines
Quoted 14 lines
Fixes 610236587600 ("bpf: Add new cgroup attach type to enable sock modifications")
---
 net/ipv4/af_inet.c  | 3 +++
 net/ipv6/af_inet6.c | 3 +++
 2 files changed, 6 insertions(+)
net/ipv4/af_inet.c +0 -0
index b24d74616637..db53701db29e 100644

net/ipv4/af_inet.c static int inet_create(struct net *net, struct socket *sock, int protocol, · line 384

Quoted 7 lines
net/ipv4/af_inet.c +1 -0
@@ -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.

Quoted 1 line

Same for IPv6.

OK

Quoted 3 lines

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.

Quoted 1 line

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.

Quoted 6 lines

---8<---

net/core/sock.c +0 -0
index 8629f9aecf91..bbc94954d9bf 100644

net/core/sock.c void sk_common_release(struct sock *sk)

Quoted 10 lines
net/core/sock.c +3 -1
@@ -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

#3 lore raw

From: Ignat Korchagin <[email protected]>
Date: Wed, 12 Jun 2024 14:22:36 -0400

Quoted 4 lines
Quoted 3 lines

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.

Quoted 8 lines

Just judging from the code - yes, indeed.

Quoted 1 line

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.

sk_common_release() is called

1. when we fail to create a socket (socket() or accept() syscall)
2. when we release the last refcount of the socket's file descriptor (basically close() syscall)

The issue only happens at 1. because we clear sock->sk at 2. in __sock_release() after calling sock->ops->release().

So, we need not take care of these callers of sk_common_release().

- inet_release
- ->close()
- udp_lib_close
- ping_close
- raw_close
- rawv6_close
- l2tp_ip_close
- l2tp_ip6_close
- sctp_close
- ieee802154_sock_release
- ->close()
- raw_close
- dgram_close
- mctp_release
- ->close()
- mctp_sk_close
- pn_socket_release
- ->close()
- pn_sock_close
- pep_sock_close

Then, the rest of the callers are:

- __sock_create
- pf->create()
- inet_create
- inet6_create
- ieee802154_create
- smc_create
- __smc_create

- setsockopt(TCP_ULP)
- smc_ulp_init
- __smc_create

- sctp_accept
- sctp_v4_create_accept_sk
- sctp_v6_create_accept_sk

we need not care about sctp_v[46]_create_accept_sk() because they don't set sock->sk for the socket; we don't pass sock to sock_init_data(NULL, newsk) before calling sk_common_release().

__sock_create() path and SMC's ULP path have the same issue, and sk_common_release() releases the last refcount of struct sock there.

So, I think we can set NULL to sock->sk in sk_common_release().

Quoted 28 lines

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.

Quoted 6 lines

---8<---

net/core/sock.c +0 -0
index 8629f9aecf91..bbc94954d9bf 100644

net/core/sock.c void sk_common_release(struct sock *sk)

Quoted 10 lines
net/core/sock.c +3 -1
@@ -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

#4 lore raw
Quoted 66 lines · Kuniyuki Iwashima

From: Ignat Korchagin <[email protected]>
Date: Wed, 12 Jun 2024 14:22:36 -0400

Quoted 4 lines
Quoted 3 lines

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.

Quoted 8 lines

Just judging from the code - yes, indeed.

Quoted 1 line

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.

sk_common_release() is called

1. when we fail to create a socket (socket() or accept() syscall)
2. when we release the last refcount of the socket's file descriptor (basically close() syscall)

The issue only happens at 1. because we clear sock->sk at 2. in __sock_release() after calling sock->ops->release().

So, we need not take care of these callers of sk_common_release().

- inet_release
- ->close()
- udp_lib_close
- ping_close
- raw_close
- rawv6_close
- l2tp_ip_close
- l2tp_ip6_close
- sctp_close
- ieee802154_sock_release
- ->close()
- raw_close
- dgram_close
- mctp_release
- ->close()
- mctp_sk_close
- pn_socket_release
- ->close()
- pn_sock_close
- pep_sock_close

Then, the rest of the callers are:

- __sock_create
- pf->create()
- inet_create
- inet6_create
- ieee802154_create
- smc_create
- __smc_create

- setsockopt(TCP_ULP)
- smc_ulp_init
- __smc_create

- sctp_accept
- sctp_v4_create_accept_sk
- sctp_v6_create_accept_sk

Quoted 8 lines

we need not care about sctp_v[46]_create_accept_sk() because they don't set sock->sk for the socket; we don't pass sock to sock_init_data(NULL, newsk) before calling sk_common_release().

__sock_create() path and SMC's ULP path have the same issue, and sk_common_release() releases the last refcount of struct sock there.

So, I think we can set NULL to sock->sk in sk_common_release().

Thanks for the explanation. Makes sense. I'll spin up a v2 with this (and try to test it as well).

Quoted 30 lines
Quoted 29 lines

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.

Quoted 6 lines

---8<---

net/core/sock.c +0 -0
index 8629f9aecf91..bbc94954d9bf 100644

net/core/sock.c void sk_common_release(struct sock *sk)

Quoted 10 lines
net/core/sock.c +3 -1
@@ -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