* [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl()
@ 2026-04-16 20:04 Michael Bommarito
2026-04-17 2:46 ` Namjae Jeon
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Michael Bommarito @ 2026-04-16 20:04 UTC (permalink / raw)
To: Namjae Jeon, Steve French, linux-cifs
Cc: Sergey Senozhatsky, Tom Talpey, stable
Another one on the smbd side this time. smb_inherit_dacl() trusts
the on-disk num_aces value from the parent directory's DACL xattr
and uses it to size a heap allocation:
aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2, ...);
num_aces is a u16 read from le16_to_cpu(parent_pdacl->num_aces)
without checking that it is consistent with the declared pdacl_size.
An authenticated client that can set a crafted DACL on a parent
directory can declare num_aces = 65535 while providing minimal actual
ACE data. This causes a ~2.6 MB allocation (not kzalloc, so
uninitialized) that the subsequent loop only partially populates, and
may also overflow the three-way size_t multiply on 32-bit kernels.
Additionally, the ACE walk loop uses the weaker
offsetof(struct smb_ace, access_req) minimum size check rather than
the minimum valid on-wire ACE size, and does not reject ACEs whose
declared size is below the minimum.
Reproduced the ACE walk OOB under UML + KASAN by constructing a
12-byte DACL (smb_acl(8) + 4-byte undersized ACE with size=4,
num_aces=1). The old 4-byte guard passes, then reading
ace->access_req at offset 4 within the ACE triggers:
BUG: KASAN: slab-out-of-bounds in kcifs3_test_inherit_dacl_old
Read of size 4 at addr ... by task mount.nfs4/220
Confirmed clean exit without splat after patch applied: the new
16-byte minimum (offsetof(smb_ace, sid) + CIFS_SID_BASE_SIZE)
rejects the undersized ACE before any field read.
Fix by:
1. Validating num_aces against pdacl_size using the same formula
applied in parse_dacl() by commit 1b8b67f3c5e5169535e2
("ksmbd: fix incorrect validation for num_aces field of
smb_acl").
2. Replacing the raw kmalloc(sizeof * num_aces * 2) with
kmalloc_array(num_aces * 2, sizeof(...)) for overflow-safe
allocation.
3. Tightening the per-ACE loop guard to require the minimum valid
ACE size (offsetof(smb_ace, sid) + CIFS_SID_BASE_SIZE) and
rejecting under-sized ACEs, matching the hardening in
smb_check_perm_dacl() and parse_dacl().
Let me know if you want 2/2 instead of this single patch.
Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
---
fs/smb/server/smbacl.c | 29 ++++++++++++++++++++++++-----
1 file changed, 24 insertions(+), 5 deletions(-)
diff --git a/fs/smb/server/smbacl.c b/fs/smb/server/smbacl.c
index d5943256c071..fc4fcd48d6c9 100644
--- a/fs/smb/server/smbacl.c
+++ b/fs/smb/server/smbacl.c
@@ -1105,8 +1105,25 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
goto free_parent_pntsd;
}
- aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2,
- KSMBD_DEFAULT_GFP);
+ aces_size = pdacl_size - sizeof(struct smb_acl);
+
+ /*
+ * Validate num_aces against the DACL payload before allocating.
+ * Each ACE must be at least as large as its fixed-size header
+ * (up to the SID base), so num_aces cannot exceed the payload
+ * divided by the minimum ACE size. This mirrors the check in
+ * parse_dacl() added by commit 1b8b67f3c5e5 ("ksmbd: fix
+ * incorrect validation for num_aces field of smb_acl").
+ */
+ if (num_aces > aces_size / (offsetof(struct smb_ace, sid) +
+ offsetof(struct smb_sid, sub_auth) +
+ sizeof(__le16))) {
+ rc = -EINVAL;
+ goto free_parent_pntsd;
+ }
+
+ aces_base = kmalloc_array(num_aces * 2, sizeof(struct smb_ace),
+ KSMBD_DEFAULT_GFP);
if (!aces_base) {
rc = -ENOMEM;
goto free_parent_pntsd;
@@ -1115,7 +1132,6 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
aces = (struct smb_ace *)aces_base;
parent_aces = (struct smb_ace *)((char *)parent_pdacl +
sizeof(struct smb_acl));
- aces_size = acl_len - sizeof(struct smb_acl);
if (pntsd_type & DACL_AUTO_INHERITED)
inherited_flags = INHERITED_ACE;
@@ -1123,11 +1139,14 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
for (i = 0; i < num_aces; i++) {
int pace_size;
- if (offsetof(struct smb_ace, access_req) > aces_size)
+ if (aces_size < offsetof(struct smb_ace, sid) +
+ CIFS_SID_BASE_SIZE)
break;
pace_size = le16_to_cpu(parent_aces->size);
- if (pace_size > aces_size)
+ if (pace_size > aces_size ||
+ pace_size < offsetof(struct smb_ace, sid) +
+ CIFS_SID_BASE_SIZE)
break;
aces_size -= pace_size;
--
2.53.0
^ permalink raw reply related [flat|nested] 6+ messages in thread* Re: [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl()
2026-04-16 20:04 [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl() Michael Bommarito
@ 2026-04-17 2:46 ` Namjae Jeon
2026-04-17 2:58 ` Michael Bommarito
2026-04-17 7:07 ` Namjae Jeon
2026-04-17 18:45 ` [PATCH v2] " Michael Bommarito
2 siblings, 1 reply; 6+ messages in thread
From: Namjae Jeon @ 2026-04-17 2:46 UTC (permalink / raw)
To: Michael Bommarito
Cc: Steve French, linux-cifs, Sergey Senozhatsky, Tom Talpey, stable
On Fri, Apr 17, 2026 at 5:05 AM Michael Bommarito
<michael.bommarito@gmail.com> wrote:
>
> Another one on the smbd side this time. smb_inherit_dacl() trusts
> the on-disk num_aces value from the parent directory's DACL xattr
> and uses it to size a heap allocation:
>
> aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2, ...);
>
> num_aces is a u16 read from le16_to_cpu(parent_pdacl->num_aces)
> without checking that it is consistent with the declared pdacl_size.
> An authenticated client that can set a crafted DACL on a parent
> directory can declare num_aces = 65535 while providing minimal actual
> ACE data. This causes a ~2.6 MB allocation (not kzalloc, so
> uninitialized) that the subsequent loop only partially populates, and
> may also overflow the three-way size_t multiply on 32-bit kernels.
>
> Additionally, the ACE walk loop uses the weaker
> offsetof(struct smb_ace, access_req) minimum size check rather than
> the minimum valid on-wire ACE size, and does not reject ACEs whose
> declared size is below the minimum.
>
> Reproduced the ACE walk OOB under UML + KASAN by constructing a
> 12-byte DACL (smb_acl(8) + 4-byte undersized ACE with size=4,
> num_aces=1). The old 4-byte guard passes, then reading
> ace->access_req at offset 4 within the ACE triggers:
>
> BUG: KASAN: slab-out-of-bounds in kcifs3_test_inherit_dacl_old
> Read of size 4 at addr ... by task mount.nfs4/220
There is no kcifs3_test_inherit_dacl_old function in ksmbd. How did
you reproduce the problem?
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl()
2026-04-17 2:46 ` Namjae Jeon
@ 2026-04-17 2:58 ` Michael Bommarito
0 siblings, 0 replies; 6+ messages in thread
From: Michael Bommarito @ 2026-04-17 2:58 UTC (permalink / raw)
To: Namjae Jeon
Cc: Steve French, linux-cifs, Sergey Senozhatsky, Tom Talpey, stable
On Thu, Apr 16, 2026 at 10:47 PM Namjae Jeon <linkinjeon@kernel.org> wrote:
> There is no kcifs3_test_inherit_dacl_old function in ksmbd. How did
> you reproduce the problem?
Sorry for the confusing splat. I pulled the pre-fix ACE-walk loop
from smb_inherit_dacl() to simplify the setup in a test module and
re-used names from another harness. _old keeps the original weak
offset guard, _new uses the tightened size, and the patched version
survives. If you want, I can run through a full repro with qemu
tomorrow using the real paths.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl()
2026-04-16 20:04 [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl() Michael Bommarito
2026-04-17 2:46 ` Namjae Jeon
@ 2026-04-17 7:07 ` Namjae Jeon
2026-04-17 18:45 ` [PATCH v2] " Michael Bommarito
2 siblings, 0 replies; 6+ messages in thread
From: Namjae Jeon @ 2026-04-17 7:07 UTC (permalink / raw)
To: Michael Bommarito
Cc: Steve French, linux-cifs, Sergey Senozhatsky, Tom Talpey, stable
> diff --git a/fs/smb/server/smbacl.c b/fs/smb/server/smbacl.c
> index d5943256c071..fc4fcd48d6c9 100644
> --- a/fs/smb/server/smbacl.c
> +++ b/fs/smb/server/smbacl.c
> @@ -1105,8 +1105,25 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
> goto free_parent_pntsd;
> }
>
> - aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2,
> - KSMBD_DEFAULT_GFP);
> + aces_size = pdacl_size - sizeof(struct smb_acl);
> +
> + /*
> + * Validate num_aces against the DACL payload before allocating.
> + * Each ACE must be at least as large as its fixed-size header
> + * (up to the SID base), so num_aces cannot exceed the payload
> + * divided by the minimum ACE size. This mirrors the check in
> + * parse_dacl() added by commit 1b8b67f3c5e5 ("ksmbd: fix
> + * incorrect validation for num_aces field of smb_acl").
> + */
Please remove the specific commit hash and patch name in the comments.
Thanks.
^ permalink raw reply [flat|nested] 6+ messages in thread* [PATCH v2] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl()
2026-04-16 20:04 [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl() Michael Bommarito
2026-04-17 2:46 ` Namjae Jeon
2026-04-17 7:07 ` Namjae Jeon
@ 2026-04-17 18:45 ` Michael Bommarito
2026-04-18 6:28 ` Namjae Jeon
2 siblings, 1 reply; 6+ messages in thread
From: Michael Bommarito @ 2026-04-17 18:45 UTC (permalink / raw)
To: Namjae Jeon, Steve French, linux-cifs
Cc: Sergey Senozhatsky, Tom Talpey, Hyunchul Lee, Ronnie Sahlberg,
stable
smb_inherit_dacl() trusts the on-disk num_aces value from the parent
directory's DACL xattr and uses it to size a heap allocation:
aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2, ...);
num_aces is a u16 read from le16_to_cpu(parent_pdacl->num_aces)
without checking that it is consistent with the declared pdacl_size.
An authenticated client whose parent directory's security.NTACL is
tampered (e.g. via offline xattr corruption or a concurrent path that
bypasses parse_dacl()) can present num_aces = 65535 with minimal
actual ACE data. This causes a ~8 MB allocation (not kzalloc, so
uninitialized) that the subsequent loop only partially populates, and
may also overflow the three-way size_t multiply on 32-bit kernels.
Additionally, the ACE walk loop uses the weaker
offsetof(struct smb_ace, access_req) minimum size check rather than
the minimum valid on-wire ACE size, and does not reject ACEs whose
declared size is below the minimum.
Reproduced on UML + KASAN + LOCKDEP against the real ksmbd code path.
A legitimate mount.cifs client creates a parent directory over SMB
(ksmbd writes a valid security.NTACL xattr), then the NTACL blob on
the backing filesystem is rewritten to set num_aces = 0xFFFF while
keeping the posix_acl_hash bytes intact so ksmbd_vfs_get_sd_xattr()'s
hash check still passes. A subsequent SMB2 CREATE of a child under
that parent drives smb2_open() into smb_inherit_dacl() (share has
"vfs objects = acl_xattr" set), which fails the page allocator:
WARNING: mm/page_alloc.c:5226 at __alloc_frozen_pages_noprof+0x46c/0x9c0
Workqueue: ksmbd-io handle_ksmbd_work
__alloc_frozen_pages_noprof+0x46c/0x9c0
___kmalloc_large_node+0x68/0x130
__kmalloc_large_node_noprof+0x24/0x70
__kmalloc_noprof+0x4c9/0x690
smb_inherit_dacl+0x394/0x2430
smb2_open+0x595d/0xabe0
handle_ksmbd_work+0x3d3/0x1140
With the patch applied the added guard rejects the tampered value
with -EINVAL before any large allocation runs, smb2_open() falls back
to smb2_create_sd_buffer(), and the child is created with a default
SD. No warning, no splat.
Fix by:
1. Validating num_aces against pdacl_size using the same formula
applied in parse_dacl().
2. Replacing the raw kmalloc(sizeof * num_aces * 2) with
kmalloc_array(num_aces * 2, sizeof(...)) for overflow-safe
allocation.
3. Tightening the per-ACE loop guard to require the minimum valid
ACE size (offsetof(smb_ace, sid) + CIFS_SID_BASE_SIZE) and
rejecting under-sized ACEs, matching the hardening in
smb_check_perm_dacl() and parse_dacl().
v1 -> v2:
- Replace the synthetic test-module splat in the changelog with a
real-path UML + KASAN reproduction driven through mount.cifs and
SMB2 CREATE; Namjae flagged the kcifs3_test_inherit_dacl_old name
in v1 since it does not exist in ksmbd.
- Drop the commit-hash citation from the code comment per Namjae's
review; keep the parse_dacl() pointer.
Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
---
fs/smb/server/smbacl.c | 28 +++++++++++++++++++++++-----
1 file changed, 23 insertions(+), 5 deletions(-)
diff --git a/fs/smb/server/smbacl.c b/fs/smb/server/smbacl.c
index d5943256c071..4a341c1f6630 100644
--- a/fs/smb/server/smbacl.c
+++ b/fs/smb/server/smbacl.c
@@ -1105,8 +1105,24 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
goto free_parent_pntsd;
}
- aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2,
- KSMBD_DEFAULT_GFP);
+ aces_size = pdacl_size - sizeof(struct smb_acl);
+
+ /*
+ * Validate num_aces against the DACL payload before allocating.
+ * Each ACE must be at least as large as its fixed-size header
+ * (up to the SID base), so num_aces cannot exceed the payload
+ * divided by the minimum ACE size. This mirrors the existing
+ * check in parse_dacl().
+ */
+ if (num_aces > aces_size / (offsetof(struct smb_ace, sid) +
+ offsetof(struct smb_sid, sub_auth) +
+ sizeof(__le16))) {
+ rc = -EINVAL;
+ goto free_parent_pntsd;
+ }
+
+ aces_base = kmalloc_array(num_aces * 2, sizeof(struct smb_ace),
+ KSMBD_DEFAULT_GFP);
if (!aces_base) {
rc = -ENOMEM;
goto free_parent_pntsd;
@@ -1115,7 +1131,6 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
aces = (struct smb_ace *)aces_base;
parent_aces = (struct smb_ace *)((char *)parent_pdacl +
sizeof(struct smb_acl));
- aces_size = acl_len - sizeof(struct smb_acl);
if (pntsd_type & DACL_AUTO_INHERITED)
inherited_flags = INHERITED_ACE;
@@ -1123,11 +1138,14 @@ int smb_inherit_dacl(struct ksmbd_conn *conn,
for (i = 0; i < num_aces; i++) {
int pace_size;
- if (offsetof(struct smb_ace, access_req) > aces_size)
+ if (aces_size < offsetof(struct smb_ace, sid) +
+ CIFS_SID_BASE_SIZE)
break;
pace_size = le16_to_cpu(parent_aces->size);
- if (pace_size > aces_size)
+ if (pace_size > aces_size ||
+ pace_size < offsetof(struct smb_ace, sid) +
+ CIFS_SID_BASE_SIZE)
break;
aces_size -= pace_size;
--
2.53.0
^ permalink raw reply related [flat|nested] 6+ messages in thread* Re: [PATCH v2] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl()
2026-04-17 18:45 ` [PATCH v2] " Michael Bommarito
@ 2026-04-18 6:28 ` Namjae Jeon
0 siblings, 0 replies; 6+ messages in thread
From: Namjae Jeon @ 2026-04-18 6:28 UTC (permalink / raw)
To: Michael Bommarito
Cc: Steve French, linux-cifs, Sergey Senozhatsky, Tom Talpey,
Hyunchul Lee, Ronnie Sahlberg, stable
On Sat, Apr 18, 2026 at 3:46 AM Michael Bommarito
<michael.bommarito@gmail.com> wrote:
>
> smb_inherit_dacl() trusts the on-disk num_aces value from the parent
> directory's DACL xattr and uses it to size a heap allocation:
>
> aces_base = kmalloc(sizeof(struct smb_ace) * num_aces * 2, ...);
>
> num_aces is a u16 read from le16_to_cpu(parent_pdacl->num_aces)
> without checking that it is consistent with the declared pdacl_size.
> An authenticated client whose parent directory's security.NTACL is
> tampered (e.g. via offline xattr corruption or a concurrent path that
> bypasses parse_dacl()) can present num_aces = 65535 with minimal
> actual ACE data. This causes a ~8 MB allocation (not kzalloc, so
> uninitialized) that the subsequent loop only partially populates, and
> may also overflow the three-way size_t multiply on 32-bit kernels.
>
> Additionally, the ACE walk loop uses the weaker
> offsetof(struct smb_ace, access_req) minimum size check rather than
> the minimum valid on-wire ACE size, and does not reject ACEs whose
> declared size is below the minimum.
>
> Reproduced on UML + KASAN + LOCKDEP against the real ksmbd code path.
> A legitimate mount.cifs client creates a parent directory over SMB
> (ksmbd writes a valid security.NTACL xattr), then the NTACL blob on
> the backing filesystem is rewritten to set num_aces = 0xFFFF while
> keeping the posix_acl_hash bytes intact so ksmbd_vfs_get_sd_xattr()'s
> hash check still passes. A subsequent SMB2 CREATE of a child under
> that parent drives smb2_open() into smb_inherit_dacl() (share has
> "vfs objects = acl_xattr" set), which fails the page allocator:
>
> WARNING: mm/page_alloc.c:5226 at __alloc_frozen_pages_noprof+0x46c/0x9c0
> Workqueue: ksmbd-io handle_ksmbd_work
> __alloc_frozen_pages_noprof+0x46c/0x9c0
> ___kmalloc_large_node+0x68/0x130
> __kmalloc_large_node_noprof+0x24/0x70
> __kmalloc_noprof+0x4c9/0x690
> smb_inherit_dacl+0x394/0x2430
> smb2_open+0x595d/0xabe0
> handle_ksmbd_work+0x3d3/0x1140
>
> With the patch applied the added guard rejects the tampered value
> with -EINVAL before any large allocation runs, smb2_open() falls back
> to smb2_create_sd_buffer(), and the child is created with a default
> SD. No warning, no splat.
>
> Fix by:
>
> 1. Validating num_aces against pdacl_size using the same formula
> applied in parse_dacl().
>
> 2. Replacing the raw kmalloc(sizeof * num_aces * 2) with
> kmalloc_array(num_aces * 2, sizeof(...)) for overflow-safe
> allocation.
>
> 3. Tightening the per-ACE loop guard to require the minimum valid
> ACE size (offsetof(smb_ace, sid) + CIFS_SID_BASE_SIZE) and
> rejecting under-sized ACEs, matching the hardening in
> smb_check_perm_dacl() and parse_dacl().
>
> v1 -> v2:
> - Replace the synthetic test-module splat in the changelog with a
> real-path UML + KASAN reproduction driven through mount.cifs and
> SMB2 CREATE; Namjae flagged the kcifs3_test_inherit_dacl_old name
> in v1 since it does not exist in ksmbd.
> - Drop the commit-hash citation from the code comment per Namjae's
> review; keep the parse_dacl() pointer.
>
> Fixes: e2f34481b24d ("cifsd: add server-side procedures for SMB3")
> Cc: stable@vger.kernel.org
> Assisted-by: Claude:claude-opus-4-6
> Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Applied it to #ksmbd-for-next-next.
Thanks!
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2026-04-18 6:28 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-04-16 20:04 [PATCH] ksmbd: validate num_aces and harden ACE walk in smb_inherit_dacl() Michael Bommarito
2026-04-17 2:46 ` Namjae Jeon
2026-04-17 2:58 ` Michael Bommarito
2026-04-17 7:07 ` Namjae Jeon
2026-04-17 18:45 ` [PATCH v2] " Michael Bommarito
2026-04-18 6:28 ` Namjae Jeon
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox