Thread (2 messages) flat view 2 messages, 2 authors, 3d ago
WARM3d

[PATCH bpf v2] bpf: Fix NULL pointer dereference in __bpf_sk_storage_map_seq_show

From: Cen Zhang (Microsoft Security FORGE Labs) <hidden>
Date: 2026-09-11 14:03:31
Also in: bpf, lkml
Subsystem: networking [general], the rest · Maintainers: "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Linus Torvalds

Iterating a sk_storage map is a two-stage operation:
bpf_sk_storage_map_seq_find_next() returns a selem, then
__bpf_sk_storage_map_seq_show() uses it. The latter re-reads
selem->local_storage via rcu_dereference() without checking for NULL.
A concurrent socket close can unlink the selem and clear that pointer
between the two stages, causing a NULL dereference of sk_storage->owner.

Oops: general protection fault, probably for non-canonical
   address 0xdffffc0000000011
  net/core/bpf_sk_storage.c:809 __bpf_sk_storage_map_seq_show()
  bpf_seq_read+0x366/0x1120
  vfs_read+0x174/0xa50
  ksys_read+0xfc/0x1d0

Return SEQ_SKIP if the re-read yields NULL. This prevents the dereference
and tells bpf_seq_read() that the stale element was skipped, so it does
not consume an iterator sequence number without running the BPF program.

Fixes: 0be08389c7f2 ("bpf: Switch to bpf_selem_unlink_nofail in bpf_local_storage_{map_free, destroy}")
Reported-by: Xiang Mei (Microsoft) <redacted>
Closes: https://lore.kernel.org/all/20260827051859.45511-1-blbllhy@gmail.com/ (local)
Suggested-by: Amery Hung <redacted>
Suggested-by: Xuanqiang Luo <redacted>
Link: https://lore.kernel.org/all/CAMB2axNFOC9G2RwOCnsWDth83REMWnmPE8gxMwbLYoGusw9miA@mail.gmail.com/ (local)
Link: https://lore.kernel.org/all/c3f2a61d-d5bc-454c-987d-717b5f8c8809@linux.dev/ (local)
Cc: AutonomousCodeSecurity@microsoft.com
Signed-off-by: Cen Zhang (Microsoft Security FORGE Labs) <redacted>
Assisted-by: Copilot (Grok 4.6)
---
Changes in v2:
- Return SEQ_SKIP instead of 0 when the storage disappeared, preserving
  the iterator sequence number for the next valid element.
- Correct the Fixes tag to the commit that switched the destruction path
  to bpf_selem_unlink_nofail().
- Rebase onto the current bpf master branch.

 net/core/bpf_sk_storage.c | 2 ++
 1 file changed, 2 insertions(+)
diff --git a/net/core/bpf_sk_storage.c b/net/core/bpf_sk_storage.c
index 1d295a8769fa..7d02245aa704 100644
--- a/net/core/bpf_sk_storage.c
+++ b/net/core/bpf_sk_storage.c
@@ -806,6 +806,8 @@ static int __bpf_sk_storage_map_seq_show(struct seq_file *seq,
 		ctx.map = info->map;
 		if (selem) {
 			sk_storage = rcu_dereference(selem->local_storage);
+			if (!sk_storage)
+				return SEQ_SKIP;
 			ctx.sk = sk_storage->owner;
 			ctx.value = SDATA(selem)->data;
 		}
base-commit: 15071f2a1263e82150c77eeb1e94dbfc31950a8e
-- 
2.55.0
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help