* [PATCH v2] smb: client: fix fiemap wild read on empty truncated response
@ 2026-10-23 11:52 Abdifatah Suruur
2026-10-23 20:24 ` Paulo Alcantara
0 siblings, 1 reply; 2+ messages in thread
From: Abdifatah Suruur @ 2026-10-23 11:52 UTC (permalink / raw)
To: linux-cifs
Cc: Paulo Alcantara, Namjae Jeon, Ronnie Sahlberg, Shyam Prasad N,
Tom Talpey, Bharath SM, samba-technical, linux-kernel
smb3_fiemap() treats -E2BIG from SMB2_ioctl() as a truncated range blob
and loops for the next chunk. If the server reports
STATUS_BUFFER_OVERFLOW with OutputCount == 0, SMB2_ioctl() returns
-E2BIG with out_data == NULL and out_data_len == 0, and the existing
length checks accept the empty result. The loop then evaluates
next = out_data[num - 1].file_offset + ...
with num == 0, reading 16 bytes before NULL.
Reject a truncated range blob that contains no range, so the resume
offset is never computed from an empty response. Also require the last
range in a truncated blob to advance the resume offset without
overflowing, and to stay within the requested range; a malicious or
broken server could otherwise make no progress or underflow len and loop
forever. Free the previous response buffer before retrying.
Reproduced on 6.6.87.2 with a malicious SMB2 server that answers
FSCTL_QUERY_ALLOCATED_RANGES with STATUS_BUFFER_OVERFLOW and
OutputCount == 0; an FS_IOC_FIEMAP ioctl on the mount oopses in
cifs_fiemap() with CR2 = fffffffffffffff8. The code path is unchanged
on mainline.
Cc: stable@vger.kernel.org
Signed-off-by: Abdifatah Suruur <suruurism@gmail.com>
---
v2:
- Validate the resume offset in a truncated range blob (overflow, forward
progress and remaining length) and free the previous response buffer
before retrying, addressing the sashiko review of v1.
fs/smb/client/smb2ops.c | 36 +++++++++++++++++++++++++++++++++---
1 file changed, 33 insertions(+), 3 deletions(-)
diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c
index 20624f5dff9b..56b3b183db4b 100644
--- a/fs/smb/client/smb2ops.c
+++ b/fs/smb/client/smb2ops.c
@@ -4360,6 +4360,15 @@ static int smb3_fiemap(struct cifs_tcon *tcon,
goto out;
}
+ /*
+ * A truncated range blob must contain at least one range, otherwise
+ * there is no offset to resume from and out_data is NULL.
+ */
+ if (!last_blob && out_data_len == 0) {
+ rc = -EINVAL;
+ goto out;
+ }
+
num = out_data_len / sizeof(struct file_allocated_range_buffer);
for (i = 0; i < num; i++) {
flags = 0;
@@ -4380,10 +4389,31 @@ static int smb3_fiemap(struct cifs_tcon *tcon,
}
if (!last_blob) {
- next = le64_to_cpu(out_data[num - 1].file_offset) +
- le64_to_cpu(out_data[num - 1].length);
- len = len - (next - start);
+ u64 last_start, last_len;
+
+ last_start = le64_to_cpu(out_data[num - 1].file_offset);
+ last_len = le64_to_cpu(out_data[num - 1].length);
+
+ /*
+ * A truncated range blob must advance the resume offset and
+ * stay within the requested range. A malicious or broken
+ * server could otherwise overflow the offset, make no
+ * progress, or underflow len and keep us looping forever.
+ */
+ if (check_add_overflow(last_start, last_len, &next) ||
+ next <= start) {
+ rc = -EINVAL;
+ goto out;
+ }
+ if (next - start >= len) {
+ rc = 0;
+ goto out;
+ }
+
+ len -= next - start;
start = next;
+ kfree(out_data);
+ out_data = NULL;
goto again;
}
--
2.53.0
^ permalink raw reply [flat|nested] 2+ messages in thread* Re: [PATCH v2] smb: client: fix fiemap wild read on empty truncated response
2026-10-23 11:52 [PATCH v2] smb: client: fix fiemap wild read on empty truncated response Abdifatah Suruur
@ 2026-10-23 20:24 ` Paulo Alcantara
0 siblings, 0 replies; 2+ messages in thread
From: Paulo Alcantara @ 2026-10-23 20:24 UTC (permalink / raw)
To: Abdifatah Suruur, linux-cifs
Cc: Namjae Jeon, Ronnie Sahlberg, Shyam Prasad N, Tom Talpey,
Bharath SM, samba-technical, linux-kernel
Abdifatah Suruur <suruurism@gmail.com> writes:
> smb3_fiemap() treats -E2BIG from SMB2_ioctl() as a truncated range blob
> and loops for the next chunk. If the server reports
> STATUS_BUFFER_OVERFLOW with OutputCount == 0, SMB2_ioctl() returns
> -E2BIG with out_data == NULL and out_data_len == 0, and the existing
> length checks accept the empty result. The loop then evaluates
>
> next = out_data[num - 1].file_offset + ...
>
> with num == 0, reading 16 bytes before NULL.
>
> Reject a truncated range blob that contains no range, so the resume
> offset is never computed from an empty response. Also require the last
> range in a truncated blob to advance the resume offset without
> overflowing, and to stay within the requested range; a malicious or
> broken server could otherwise make no progress or underflow len and loop
> forever. Free the previous response buffer before retrying.
> ...
Applied.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-10-23 20:24 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-10-23 11:52 [PATCH v2] smb: client: fix fiemap wild read on empty truncated response Abdifatah Suruur
2026-10-23 20:24 ` Paulo Alcantara
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®