linux.git/net/core/sock.c, branch v6.10

net: do not leave a dangling sk pointer, when socket creation fails

2024-06-20T08:43:14+00:00

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 and stripped):
==================================================================
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)
Read of size 8 at addr ffff888007110dd8 by task traceroute/299

CPU: 2 PID: 299 Comm: traceroute Tainted: G            E      6.10.0-rc2+ #2
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.2-debian-1.16.2-1 04/01/2014
Call Trace:
 
dump_stack_lvl (lib/dump_stack.c:117 (discriminator 1))
print_report (mm/kasan/report.c:378 mm/kasan/report.c:488)
? __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)
kasan_report (mm/kasan/report.c:603)
? __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)
kasan_check_range (mm/kasan/generic.c:183 mm/kasan/generic.c:189)
__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)
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)
bpf_prog_875642cf11f1d139___sock_release+0x6e/0x8e
bpf_trampoline_6442506592+0x47/0xaf
__sock_release (net/socket.c:652)
__sock_create (net/socket.c:1601)
...
Allocated by task 299 on cpu 2 at 78.328492s:
kasan_save_stack (mm/kasan/common.c:48)
kasan_save_track (mm/kasan/common.c:68)
__kasan_slab_alloc (mm/kasan/common.c:312 mm/kasan/common.c:338)
kmem_cache_alloc_noprof (mm/slub.c:3941 mm/slub.c:4000 mm/slub.c:4007)
sk_prot_alloc (net/core/sock.c:2075)
sk_alloc (net/core/sock.c:2134)
inet_create (net/ipv4/af_inet.c:327 net/ipv4/af_inet.c:252)
__sock_create (net/socket.c:1572)
__sys_socket (net/socket.c:1660 net/socket.c:1644 net/socket.c:1706)
__x64_sys_socket (net/socket.c:1718)
do_syscall_64 (arch/x86/entry/common.c:52 arch/x86/entry/common.c:83)
entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130)

Freed by task 299 on cpu 2 at 78.328502s:
kasan_save_stack (mm/kasan/common.c:48)
kasan_save_track (mm/kasan/common.c:68)
kasan_save_free_info (mm/kasan/generic.c:582)
poison_slab_object (mm/kasan/common.c:242)
__kasan_slab_free (mm/kasan/common.c:256)
kmem_cache_free (mm/slub.c:4437 mm/slub.c:4511)
__sk_destruct (net/core/sock.c:2117 net/core/sock.c:2208)
inet_create (net/ipv4/af_inet.c:397 net/ipv4/af_inet.c:252)
__sock_create (net/socket.c:1572)
__sys_socket (net/socket.c:1660 net/socket.c:1644 net/socket.c:1706)
__x64_sys_socket (net/socket.c:1718)
do_syscall_64 (arch/x86/entry/common.c:52 arch/x86/entry/common.c:83)
entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130)

Fix this by clearing the struct socket reference in sk_common_release() to cover
all protocol families create functions, which may already attached the
reference to the sk object with sock_init_data().

Fixes: c5dbb89fc2ac ("bpf: Expose bpf_get_socket_cookie to tracing programs")
Suggested-by: Kuniyuki Iwashima 
Signed-off-by: Ignat Korchagin 
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/netdev/20240613194047.36478-1-kuniyu@amazon.com/T/
Reviewed-by: Kuniyuki Iwashima 
Reviewed-by: D. Wythe 
Link: https://lore.kernel.org/r/20240617210205.67311-1-ignat@cloudflare.com
Signed-off-by: Paolo Abeni

net: change proto and proto_ops accept type

2024-05-14T00:19:09+00:00

Rather than pass in flags, error pointer, and whether this is a kernel
invocation or not, add a struct proto_accept_arg struct as the argument.
This then holds all of these arguments, and prepares accept for being
able to pass back more information.

No functional changes in this patch.

Acked-by: Jakub Kicinski 
Signed-off-by: Jens Axboe

net: move sysctl_mem_pcpu_rsv to net_hotdata

2024-05-01T01:46:52+00:00

sysctl_mem_pcpu_rsv is used in TCP fast path,
move it to net_hodata for better cache locality.

Signed-off-by: Eric Dumazet 
Reviewed-by: David Ahern 
Link: https://lore.kernel.org/r/20240429134025.1233626-6-edumazet@google.com
Signed-off-by: Jakub Kicinski

net: add

2024-05-01T01:46:52+00:00

Move some proto memory definitions out of 

Very few files need them, and following patch
will include  from 

Signed-off-by: Eric Dumazet 
Reviewed-by: David Ahern 
Link: https://lore.kernel.org/r/20240429134025.1233626-5-edumazet@google.com
Signed-off-by: Jakub Kicinski

net: skbuff: generalize the skb->decrypted bit

2024-04-06T16:34:31+00:00

The ->decrypted bit can be reused for other crypto protocols.
Remove the direct dependency on TLS, add helpers to clean up
the ifdefs leaking out everywhere.

Signed-off-by: Jakub Kicinski 
Reviewed-by: David Ahern 
Signed-off-by: David S. Miller

net: add sk_wake_async_rcu() helper

2024-03-29T22:03:11+00:00

While looking at UDP receive performance, I saw sk_wake_async()
was no longer inlined.

This matters at least on AMD Zen1-4 platforms (see SRSO)

This might be because rcu_read_lock() and rcu_read_unlock()
are no longer nops in recent kernels ?

Add sk_wake_async_rcu() variant, which must be called from
contexts already holding rcu lock.

As SOCK_FASYNC is deprecated in modern days, use unlikely()
to give a hint to the compiler.

sk_wake_async_rcu() is properly inlined from
__udp_enqueue_schedule_skb() and sock_def_readable().

Signed-off-by: Eric Dumazet 
Link: https://lore.kernel.org/r/20240328144032.1864988-5-edumazet@google.com
Signed-off-by: Jakub Kicinski

net: mark racy access on sk->sk_rcvbuf

2024-03-25T14:46:59+00:00

sk->sk_rcvbuf in __sock_queue_rcv_skb() and __sk_receive_skb() can be
changed by other threads. Mark this as benign using READ_ONCE().

This patch is aimed at reducing the number of benign races reported by
KCSAN in order to focus future debugging effort on harmful races.

Signed-off-by: linke li 
Reviewed-by: Eric Dumazet 
Signed-off-by: David S. Miller

sock: Use unsafe_memcpy() for sock_copy()

2024-03-06T02:35:12+00:00

While testing for places where zero-sized destinations were still showing
up in the kernel, sock_copy() and inet_reqsk_clone() were found, which
are using very specific memcpy() offsets for both avoiding a portion of
struct sock, and copying beyond the end of it (since struct sock is really
just a common header before the protocol-specific allocation). Instead
of trying to unravel this historical lack of container_of(), just switch
to unsafe_memcpy(), since that's effectively what was happening already
(memcpy() wasn't checking 0-sized destinations while the code base was
being converted away from fake flexible arrays).

Avoid the following false positive warning with future changes to
CONFIG_FORTIFY_SOURCE:

  memcpy: detected field-spanning write (size 3068) of destination "&nsk->__sk_common.skc_dontcopy_end" at net/core/sock.c:2057 (size 0)

Signed-off-by: Kees Cook 
Reviewed-by: Simon Horman 
Link: https://lore.kernel.org/r/20240304212928.make.772-kees@kernel.org
Signed-off-by: Jakub Kicinski

net: make SK_MEMORY_PCPU_RESERV tunable

2024-02-28T09:23:08+00:00

This patch adds /proc/sys/net/core/mem_pcpu_rsv sysctl file,
to make SK_MEMORY_PCPU_RESERV tunable.

Commit 3cd3399dd7a8 ("net: implement per-cpu reserves for
memory_allocated") introduced per-cpu forward alloc cache:

"Implement a per-cpu cache of +1/-1 MB, to reduce number
of changes to sk->sk_prot->memory_allocated, which
would otherwise be cause of false sharing."

sk_prot->memory_allocated points to global atomic variable:
atomic_long_t tcp_memory_allocated ____cacheline_aligned_in_smp;

If increasing the per-cpu cache size from 1MB to e.g. 16MB,
changes to sk->sk_prot->memory_allocated can be further reduced.
Performance may be improved on system with many cores.

Signed-off-by: Adam Li 
Reviewed-by: Christoph Lameter (Ampere) 
Signed-off-by: David S. Miller

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

2024-02-22T23:29:26+00:00

Cross-merge networking fixes after downstream PR.

Conflicts:

net/ipv4/udp.c
  f796feabb9f5 ("udp: add local "peek offset enabled" flag")
  56667da7399e ("net: implement lockless setsockopt(SO_PEEK_OFF)")

Adjacent changes:

net/unix/garbage.c
  aa82ac51d633 ("af_unix: Drop oob_skb ref before purging queue in GC.")
  11498715f266 ("af_unix: Remove io_uring code for GC.")

Signed-off-by: Jakub Kicinski