- five sensitive data leak fixes (clear stack and heap cryptographic
keys/hashes)
- six file size and cache synchronization fixes (fscache cookie
serialization and truncation handling)
- seven protocol validation and buffer safety fixes (prevent OOB
access and loff_t underflow)
- six metadata and POSIX attribute fixes (proper hard-link counts and
setuid/setgid stripping)
- three DFS cache and unmount fixes (prevent target-hint UAF and
unmount hangs)
- six general client improvements (fix read request leaks, stats
loops, handle servers that don't support O_TMPFILE)
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQTcqRusfSdYROJQwGkpVtNKoQNdYwUCaoy9dgAKCRApVtNKoQNd
Y5zrAP9HRp0z9rLmezHzGoTnF+0WYnkE9xK9pqRDoIjflXPyDAD+IDYzBXTWJpoq
O1+3OiuNGvoF+X46i8xE9voAbnCTDgM=
=ETzr
-----END PGP SIGNATURE-----
Merge tag 'cifs-fixes-7.3-rc1' of https://git.manguebit.org/linux
Pull smb client updates from Paulo Alcantara:
- clear sensitive data after use (stack and heap cryptographic
keys/hashes)
- file size and cache synchronization fixes (fscache cookie
serialization and truncation handling)
- protocol validation and buffer safety fixes (prevent OOB access and
loff_t underflow)
- metadata and POSIX attribute fixes (proper hard-link counts and
setuid/setgid stripping)
- DFS cache and unmount fixes (prevent target-hint UAF and unmount
hangs)
- general client improvements (fix read request leaks, stats loops,
handle servers that don't support O_TMPFILE)
* tag 'cifs-fixes-7.3-rc1' of https://git.manguebit.org/linux: (33 commits)
cifs: fix loff_t underflow in cifs_remap_file_range() when len == 0
smb: client: reject a tree connect response whose byte count is too small
cifs: call pagecache_isize_extended() in cifs_setsize() when extending
smb: client: fix copy-paste error in WSL EA length accounting for $LXDEV
smb: client: remove redundant NULL check before kfree()
smb: client: restore the data_offset bound in is_valid_oplock_break()
cifs: clear tcon after cifsFileInfo_put() in cifs_file_set_size()
smb: client: Avoid leaking sensitive data to the heap in connect.c
smb: client: Clear sensitive stack data in smb1encrypt.c
smb: client: Clear sensitive stack data in cifsencrypt.c
smb: client: Clear sensitive stack and heap data in smb2ops.c
smb: client: Clear sensitive stack data in smb2transport.c
Revert "cifs: remove all cifs files before kill super"
smb: client: fix use-before-check of ReparseDataLength in reparse_buf_ptr()
smb: client: fix ALIGN() overflow in symlink_data() error context loop
smb: client: simplify __build_path_from_dentry_optional_prefix()
smb: client: fix UAF and buffer leak in cifs_check_trans2() for malformed secondary T2
smb: client: fix OOB read/write from unvalidated DataOffset in coalesce_t2()
smb/client: decode reparse metadata using its payload type
smb/client: preserve open info type across compound queries
...
With len == 0 (clone to EOF), the effective length is computed as:
len = src_inode->i_size - off;
If off > i_size, this is a negative loff_t, corrupting the ByteCount
in the FSCTL_DUPLICATE_EXTENTS_TO_FILE request and inverting the range
in filemap_write_and_wait_range(). The existing off >= i_size check
fires only after the ioctl has already been sent.
Snapshot i_size_read() once for both the bounds check and the length
calculation, eliminating the TOCTOU and 32-bit torn-read risk. Reject
off > src_size with -EINVAL. Treat off == src_size as a no-op,
consistent with __generic_remap_file_range_prep().
Fixes: 04b38d6012 ("vfs: pull btrfs clone API to vfs layer")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
CIFSTCon() bounds its strnlen() over the byte area with the server's
ByteCount minus two, which for ByteCount 0 or 1 goes negative as an int
and converts to a huge size_t. The later subtraction wraps the __u16
bytes_left, and that is what bounds cifs_strndup_from_utf16(): a bound of
up to 65535 against a ~16 KB cifs_req_poolp object runs off the end of the
slab object, and the bytes reach userspace through tcon->nativeFileSystem
in /proc/fs/cifs/DebugData.
Reject a byte area too small for what the parser consumes. Two bytes is
the least it can consume, and no conformant response carries fewer. The
new trace point is the 129th smb_eio_trace entry, which __mode(byte)
cannot represent, so the attribute goes with it.
Fixes: cc20c031bb ("cifs: convert CIFSTCon to use new unicode helper functions")
Cc: stable@vger.kernel.org
Signed-off-by: Bryam Vargas <hexlabsecurity@proton.me>
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
cifs_setsize() calls truncate_pagecache() but skips
pagecache_isize_extended() on extension. truncate_setsize() shows
the correct pattern:
i_size_write(inode, newsize);
if (newsize > oldsize)
pagecache_isize_extended(inode, oldsize, newsize);
truncate_pagecache(inode, newsize);
pagecache_isize_extended() zeroes the tail of the page straddling old
EOF. Without it, dirty bytes in that region can be written back to
the server, exposing stale data in the newly extended range.
Cc: stable@vger.kernel.org
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
The LXDEV block in cifs_query_path_info() uses SMB2_WSL_XATTR_MODE_SIZE
(4) instead of SMB2_WSL_XATTR_DEV_SIZE (8), undercounting eas_len by 4
bytes per $LXDEV EA.
eas_len is used only as a zero/non-zero presence flag so there is no
current functional impact, but the value is incorrect and misleading.
Fixes: 97db416045 ("smb: client: parse uid, gid, mode and dev from WSL reparse points")
Cc: stable@vger.kernel.org
Cc: Paulo Alcantara <pc@manguebit.org>
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
kfree() safely handles NULL pointers, so the explicit NULL check
before calling kfree() is unnecessary.
This issue was reported by ifnullfree.cocci.
Signed-off-by: Mohammad Shahid <mdshahid03@gmail.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Commit 83bfbd0bb9 ("cifs: Remove the RFC1002 header from smb_hdr")
changed the quantity this bound is measured against. It used to be
srv->total_read minus the 4-byte RFC1002 preamble that total_read then
included, so it was the SMB message length. The same commit stopped
counting the preamble, and the mechanical substitution to
srv->total_read - srv->pdu_size left an expression that is identically
zero: standard_receive3() reads MID_HEADER_SIZE() bytes and then exactly
pdu_length - MID_HEADER_SIZE() more, adding both to total_read.
len is therefore 0, the subtraction below it wraps, and no __u32
DataOffset can exceed the result, so the check from commit 097f5863b1
("cifs: read overflow in is_valid_oplock_break()") no longer rejects
anything. Use total_read, which is now the message length on its own.
Fixes: 83bfbd0bb9 ("cifs: Remove the RFC1002 header from smb_hdr")
Cc: stable@kernel.org
Signed-off-by: Bryam Vargas <hexlabsecurity@proton.me>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
When the else branch of cifs_file_set_size() finds a writable file handle
via find_writable_file(), it borrows tcon and server from the handle's
tlink, attempts the handle-based set_file_size() RPC, and then releases
the handle with cifsFileInfo_put().
If set_file_size() fails, execution falls through to the path-based
fallback, which reuses the borrowed tcon and server under the
"if (tcon == NULL)" guard. Since tcon is not NULL at that point, the
guard is skipped. If cifsFileInfo_put() dropped the last reference on a
tlink that was already removed from the tlink tree (TCON_LINK_IN_TREE
cleared, as happens during reconnection or session teardown),
cifs_put_tlink() will have freed tcon; the subsequent set_path_size()
call is then a use-after-free.
Setting tcon = NULL after cifsFileInfo_put() causes the existing guard
to take the cifs_sb_tlink() path, which acquires a fresh reference for
the path-based operation or fails cleanly if the session is gone.
Fixes: 110fee6b9b ("smb: client: fix missing timestamp updates with O_TRUNC")
Cc: stable@vger.kernel.org
Cc: Paulo Alcantara <pc@manguebit.com>
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
TCP_Server_Info contains a preauth_sha_hash[] and a cryptkey[] array
that might contain sensitive data. Thus free its memory with
kfree_sensitive() to avoid that we are leaking this information to
the heap.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Make sure to not leak signature data via the stack, clear it
with memzero_explicit() before leaving the function.
To avoid that we have to introduce "goto"-cleanup here, we re-arrange
the code a little bit (and drop the commented cifs_dump_mem debug
code that looks like a leftover from very early days).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Make sure to not leak hash data via the stack, clear it
with memzero_explicit() before leaving the function.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Make sure to not leak key-related data via the heap or the stack
by using kfree_sensitive() or memzero_explicit() here.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Sensitive data like keys that are stored in stack-local arrays could
be leaked via the stack to the calling functions. There is no known
vulnerability for this right now, but it's good security style to
explicitly zeroize this sensitive material as soon as possible to
avoid that it could be exploited together with other bugs later.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
This reverts commit 6d9a4aaaa8.
First, directly flushing fileinfo_put_wq in that commit cannot guarantee
that all in-flight I/O has run its cleanup_work on system_dfl_wq and
subsequently called queue_work(fileinfo_put_wq, ...). Flushing only the
latter workqueue may therefore miss puts that have not yet been queued, so
the fix is not reliable in the first place. Moreover, this fix flushes
inside cifs_umount(), which means the busy-dentry warning can still be
triggered when umount_check() is called inside kill_anon_super(), because
kill_anon_super() is executed before cifs_umount().
Second, commit 75f5c412fa ("smb: client: fix busy dentry warning on
unmount after DIO") already drains both serverclose_wq and fileinfo_put_wq
in cifs_kill_sb(), before kill_anon_super(). By adding a per-superblock
outstanding-rreq counter, it guarantees that all cleanup_work for this sb
have run, and thus all relevant cfile puts are queued on fileinfo_put_wq
or serverclose_wq.
Third, no path between those drains and cifs_umount() can queue new work
onto either workqueue. In the "cifs_sb->root == NULL" path there are no
file-related workers either, so that case is safe as well.
Therefore the busy-dentry and null-ptr-deref problems cannot arise, and
the flush added by commit 6d9a4aaaa8 ("cifs: remove all cifs files before
kill super") is redundant and can be removed.
Signed-off-by: Zizhi Wo <wozizhi@huawei.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
reparse_buf_ptr() reads buf->ReparseDataLength before checking that
count covers the full fixed header:
buf = (struct reparse_data_buffer *)((u8 *)io + off);
len = sizeof(*buf); /* 8 bytes */
rdlen = le16_to_cpu(buf->ReparseDataLength); /* offset 4, 2 bytes */
if (count < len || count < rdlen + len) /* check comes after */
struct reparse_data_buffer has ReparseDataLength at offset 4. If a
server returns OutputCount < 6, the read at offset 4-5 reaches past
the end of the received data. The off+count bounds against iov_len
were already validated, but that does not protect against count being
smaller than sizeof(*buf).
Split the check: verify count >= sizeof(*buf) before reading
ReparseDataLength, then verify count covers the data region.
Fixes: a158bb66b1 ("smb: client: optimise reparse point querying")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
The check added by commit 7d9a7f1f96 ("smb/client: fix possible
infinite loop and oob read in symlink_data()") compared the post-ALIGN
length against the remaining buffer, but ALIGN() itself can overflow:
for ErrorDataLength near UINT32_MAX (e.g. 0xFFFFFFF9), ALIGN(x, 8)
wraps to 0, so the subsequent bounds check passes, and the loop
advances by zero bytes leaving 'p' pointing into stale data.
Fix by checking the raw ErrorDataLength against the remaining space
before applying ALIGN(), then checking again after. Since raw_len is
bounded by the buffer, raw_len + 7 cannot overflow, so the second check
is an exact post-alignment bounds guard.
Fixes: 76894f3e2f ("cifs: improve symlink handling for smb2+")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Use the convenient 'strreplace()' to simplify
'__build_path_from_dentry_optional_prefix()'.
Signed-off-by: Dmitry Antipov <dmantipov@yandex.ru>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
When a valid primary TRANSACT2 response has been received (mid->resp_buf
set, mid->multiRsp true) and a subsequent secondary response causes
cifs_check_trans2() to return false -- either because the SMB header is
invalid (malformed != 0) or because check2ndT2() rejects the PDU --
handle_mid() overwrites mid->resp_buf with the new buffer (leaking the
primary buffer) and, because mid->multiRsp is set, skips the
server->smallbuf/bigbuf NULL-out. When the user thread frees
mid->resp_buf, server->smallbuf or server->bigbuf is left dangling; the
demux thread reuses it for the next packet, resulting in a use-after-free.
Combine both early-exit conditions and, when mid->multiRsp is already
set, abort the pending transaction inline: set multiEnd, call
dequeue_mid() with malformed=true, and return true so handle_mid() exits
without touching mid->resp_buf or the server buffer pointers.
Fixes: 316cf94a91 ("CIFS: Move trans2 processing to ops struct")
Cc: stable@vger.kernel.org # cifs_check_trans2() is in smb1ops.c on kernels < 7.0
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
coalesce_t2() computes data pointers directly from server-supplied
DataOffset fields with no validation against buffer bounds:
data_area_of_tgt = (char *)&pSMBt->hdr.Protocol +
get_unaligned_le16(&pSMBt->t2_rsp.DataOffset);
data_area_of_src = (char *)&pSMBs->hdr.Protocol +
get_unaligned_le16(&pSMBs->t2_rsp.DataOffset);
data_area_of_tgt += total_in_tgt;
...
memcpy(data_area_of_tgt, data_area_of_src, total_in_src);
A small DataOffset can push a pointer below the actual byte area,
overwriting header fields; a large one can push it past the buffer
end, causing out-of-bounds heap reads (source) or writes (target).
The BCC overflow guard does not prevent this: BCC reflects how much
data is present, while DataOffset controls where in the buffer it
starts.
The "validate target area" comment present since the function was
first written in 2005 was a placeholder that was never implemented.
Add lower- and upper-bound checks for both data pointers before the
memcpy, and before any target header fields are modified.
Fixes: e4eb295d38 ("[PATCH] cifs: Handle multiple response transact2 part 1 of 2")
Cc: stable@vger.kernel.org
Reported-by: Shen Yongchao <grayhat@foxmail.com>
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
cifs_open_info_data stores FILE_ALL_INFORMATION and SMB3 POSIX query
information in a union. reparse_info_to_fattr() selects a union member
from the mount mode, while several directory checks always read
fi.Attributes.
The metadata can instead come from an SMB2 CREATE response on a POSIX
mount, or from a POSIX query while processing a reparse point. In those
cases the mount mode and hard-coded fi accesses select the wrong union
member.
See the procedures below:
cifs_nt_open
smb2_open_file
SMB2_open
data->fi = SMB2 CREATE response
data->contains_posix_file_info = false
cifs_get_inode_info
reparse_info_to_fattr
if (tcon->posix_extensions) // true
smb311_posix_info_to_fattr
data->posix_fi // wrong union member
smb311_posix_get_fattr
smb2_query_path_info
smb2_compound_op
data->posix_fi = SMB3 POSIX query response
data->contains_posix_file_info = true
reparse_info_to_fattr
data->fi.Attributes // wrong union member
Add a common DOS attribute accessor and use contains_posix_file_info
both for attribute reads and for the final fattr conversion.
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
contains_posix_file_info describes the metadata stored in the
fi/posix_fi union. GET_REPARSE and QUERY_WSL_EA do not update that
union, so clearing the flag while processing those responses can make
POSIX metadata look like FILE_ALL_INFORMATION.
Set the flag when CREATE or a validated query response actually
populates the union, and leave it unchanged for auxiliary compound
operations. This also avoids changing the type when a query fails
before copying any metadata.
The issue can be reproduced against a Samba server with SMB3 UNIX
extensions enabled:
mount -t cifs //<server>/<share> /mnt/cifs \
-o vers=3.1.1,posix,reparse=nfs,actimeo=0
mkfifo /mnt/cifs/test-fifo
umount /mnt/cifs
mount -t cifs //<server>/<share> /mnt/cifs \
-o vers=3.1.1,posix,reparse=nfs,actimeo=0
stat -c '%F %s' /mnt/cifs/test-fifo
Before this change, stat reports "fifo 1024" although the server-side
EOF is zero. After this change, it reports "fifo 0".
Fixes: 9df23801c8 ("smb311: failure to open files of length 1040 when mounting with SMB3.1.1 POSIX extensions")
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Several SMB1 fallback and open responses do not provide the hard link
count. The SMB2 create-only query fallback has the same limitation.
These paths currently leave a zero link count or synthesize a value of
one and then expose it as authoritative metadata.
Mark those results with unknown_nlink so existing inodes keep their
cached link count and new inodes receive the usual sane default.
This was tested against Samba with "server min protocol = NT1". Mount
the share using SMB1 with Unix extensions disabled:
mount -t cifs //<server>/<share> /mnt/cifs \
-o username=<user>,vers=1.0,nounix
Create three names for the same inode and cache its real link count:
TESTDIR=/mnt/cifs/nlink-repro-$$
mkdir "$TESTDIR"
touch "$TESTDIR/file1"
ln "$TESTDIR/file1" "$TESTDIR/file2"
ln "$TESTDIR/file1" "$TESTDIR/file3"
stat -c 'before open: %h' "$TESTDIR/file1"
Open the file and read the link count through the open descriptor:
exec 3<"$TESTDIR/file1"
stat -Lc 'after open: %h' /proc/$$/fd/3
exec 3<&-
Clean up the test files:
rm -f "$TESTDIR/file1" "$TESTDIR/file2" "$TESTDIR/file3"
rmdir "$TESTDIR"
Before this change, the two stat commands report 3 and 1 because the
SMB1 open response overwrites the known link count. With this change,
both commands report 3.
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
The code to clear the 'fastest_cmd' statistics has a typo that
repeatedly clears the stat for cmd 0, rather than iterating
through each cmd. Fix the typo (0->i).
Fixes: 433b8dd767 ("SMB3: Track total time spent on roundtrips for each SMB3 command")
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
sequencing, SMB3 request replay and encryption, Apple Time Machine
interoperability, protocol-compatibility fixes validated with smbtorture,
security hardening, SMB Direct transport support, connection reliability,
and other correctness improvements.
New features:
- Implement the SMB2 command sequence window.
Enforce the credit-based MessageId range for each connection, rejecting
out-of-window, duplicate, and wrapped sequence numbers. This prevents
invalid requests and same-channel replays from being processed.
- Add SMB3 request replay support.
SMB3 clients may resend requests with SMB2_FLAGS_REPLAY_OPERATION after
a channel disconnect when the original response was lost. Track the
required channel and open state to safely handle durable CREATE replays
and make oplock, lease, and lock replays idempotent, avoiding duplicate
state changes and improving multichannel reconnect reliability.
- Add opt-in Apple Time Machine support.
Implement the AAPL negotiation and related Finder, stream, COPYCHUNK,
sparse-file, CHANGE_NOTIFY, and RPC compatibility required for Time
Machine shares, allowing macOS backupd to use ksmbd for backups.
- Add per-share SMB3 encryption support.
Allow individual shares to require SMB3 encryption by advertising
SMB2_SHAREFLAG_ENCRYPT_DATA in TREE_CONNECT responses and rejecting
unencrypted tree connects and plaintext requests for protected shares.
- Add SMB Direct RDMA encryption support.
Extend SMB Direct to support SMB3 encrypted payloads over RDMA, with
transform negotiation and encryption/decryption for RDMA READ/WRITE.
Other changes:
- Parse and retain AppInstanceVersion contexts, enforce version ordering,
close older active handles for newer takeovers, and reject invalid or
unversioned opens according to the SMB2 semantics.
- Accept durable reconnect requests that omit VolatileFileId when the
persistent ID and reconnect context identify the handle, while
continuing to reject explicit volatile-ID mismatches.
- Fix SMB2/SMB3 protocol validation and security issues, including request
offsets, file and object IDs, IPC responses, output buffer sizes,
SMB3.1.1 binding validation, signing-required handling, durable handles,
ACLs, maximal access, and security information.
- Fix heap out-of-bounds accesses, use-after-free bugs, memory leaks,
invalid pointer dereferences, and sensitive-data lifetime issues in
authentication, Kerberos, preauthentication, sessions, connections,
and module teardown.
- Correct alternate-data-stream and named-stream handling, COPYCHUNK
behavior, sparse-file and compression attributes, allocated-range
queries, file trimming, duplicate extents, DOS attributes, snapshots,
normalized names, and partial information responses.
- Fix locking, lease, oplock, durable reconnect, async request, and
CHANGE_NOTIFY races, including deferred-lock rollback, parent directory
lease notifications, and connection teardown lifetime bugs.
- Fix SMB3 encryption handling for compressed requests, expired encrypted
sessions, interim responses, bound multichannel connections, and
decryption failures.
- Fix SMB3 multichannel session lookup and session state transitions so
changes are scoped to the correct bound connections and cannot revive
connections that are already shutting down.
- Fix DACL access checks so ACE walks are bounded by the declared DACL
size, preventing data beyond the DACL boundary from being interpreted
during access validation.
- Fix session accounting and lifetime issues, including session counter
updates during publication and removal, session leaks on registration
failure, and procfs creation diagnostics.
- Improve TCP connection reliability by enabling TCP keepalive for
accepted connections and preserving TCP timers for kernel sockets,
preventing silent peers from holding connections indefinitely.
- Fix smbdirect RDMA cleanup ordering for completion queues, QPs, child
sockets, and listener locking.
- Improve async response framing, multi-iovec signing, RPC pipe status
handling, and ksmbd procfs monitoring for server, share, connection,
session, and open-file state.
- Remove the obsolete DES crypto header and Kconfig dependency now that
NTLMv1 support has been removed.
- Update the ksmbd repository URL in MAINTAINERS and add an additional
KSMBD reviewer.
-----BEGIN PGP SIGNATURE-----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=uocO
-----END PGP SIGNATURE-----
Merge tag 'ksmbd-for-7.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/smb
Pull smb server updates from Namjae Jeon:
"This contains server updates focused on SMB2 command sequencing, SMB3
request replay and encryption, Apple Time Machine interoperability,
protocol-compatibility fixes validated with smbtorture, security
hardening, SMB Direct transport support, connection reliability, and
other correctness improvements.
New features:
- Implement the SMB2 command sequence window
Enforce the credit-based MessageId range for each connection,
rejecting out-of-window, duplicate, and wrapped sequence numbers.
This prevents invalid requests and same-channel replays from being
processed
- Add SMB3 request replay support
SMB3 clients may resend requests with SMB2_FLAGS_REPLAY_OPERATION
after a channel disconnect when the original response was lost.
Track the required channel and open state to safely handle durable
CREATE replays and make oplock, lease, and lock replays idempotent,
avoiding duplicate state changes and improving multichannel
reconnect reliability
- Add opt-in Apple Time Machine support
Implement the AAPL negotiation and related Finder, stream,
COPYCHUNK, sparse-file, CHANGE_NOTIFY, and RPC compatibility
required for Time Machine shares, allowing macOS backupd to use
ksmbd for backups
- Add per-share SMB3 encryption support
Allow individual shares to require SMB3 encryption by advertising
SMB2_SHAREFLAG_ENCRYPT_DATA in TREE_CONNECT responses and rejecting
unencrypted tree connects and plaintext requests for protected
shares
- Add SMB Direct RDMA encryption support
Extend SMB Direct to support SMB3 encrypted payloads over RDMA,
with transform negotiation and encryption/decryption for RDMA
READ/WRITE
Other changes:
- Parse and retain AppInstanceVersion contexts, enforce version
ordering, close older active handles for newer takeovers, and
reject invalid or unversioned opens according to the SMB2 semantics
- Accept durable reconnect requests that omit VolatileFileId when the
persistent ID and reconnect context identify the handle, while
continuing to reject explicit volatile-ID mismatches
- Fix SMB2/SMB3 protocol validation and security issues, including
request offsets, file and object IDs, IPC responses, output buffer
sizes, SMB3.1.1 binding validation, signing-required handling,
durable handles, ACLs, maximal access, and security information
- Fix heap out-of-bounds accesses, use-after-free bugs, memory leaks,
invalid pointer dereferences, and sensitive-data lifetime issues in
authentication, Kerberos, preauthentication, sessions, connections,
and module teardown
- Correct alternate-data-stream and named-stream handling, COPYCHUNK
behavior, sparse-file and compression attributes, allocated-range
queries, file trimming, duplicate extents, DOS attributes,
snapshots, normalized names, and partial information responses
- Fix locking, lease, oplock, durable reconnect, async request, and
CHANGE_NOTIFY races, including deferred-lock rollback, parent
directory lease notifications, and connection teardown lifetime
bugs
- Fix SMB3 encryption handling for compressed requests, expired
encrypted sessions, interim responses, bound multichannel
connections, and decryption failures
- Fix SMB3 multichannel session lookup and session state transitions
so changes are scoped to the correct bound connections and cannot
revive connections that are already shutting down
- Fix DACL access checks so ACE walks are bounded by the declared
DACL size, preventing data beyond the DACL boundary from being
interpreted during access validation
- Fix session accounting and lifetime issues, including session
counter updates during publication and removal, session leaks on
registration failure, and procfs creation diagnostics
- Improve TCP connection reliability by enabling TCP keepalive for
accepted connections and preserving TCP timers for kernel sockets,
preventing silent peers from holding connections indefinitely
- Fix smbdirect RDMA cleanup ordering for completion queues, QPs,
child sockets, and listener locking
- Improve async response framing, multi-iovec signing, RPC pipe
status handling, and ksmbd procfs monitoring for server, share,
connection, session, and open-file state
- Remove the obsolete DES crypto header and Kconfig dependency now
that NTLMv1 support has been removed
- Update the ksmbd repository URL in MAINTAINERS and add an
additional KSMBD reviewer"
* tag 'ksmbd-for-7.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/smb: (142 commits)
MAINTAINERS: update ksmbd repository URL
MAINTAINERS: add myself as KSMBD reviewer
smb: server: remove unused DES crypto header
smb: server: Remove obsolete "select CRYPTO_LIB_DES" from Kconfig file
ksmbd: keep TCP timers alive for kernel sockets
ksmbd: enable TCP keepalive for accepted connections
smb/server: fix session counter on session removal
smb/server: update session counter under sessions table lock
smb/server: fix session leak in ksmbd_session_register()
smb/server: warn if ksmbd_proc_create() fails
ksmbd: bound smb_check_perm_dacl() ACE walks by DACL size
ksmbd: make RDMA encryption diagnostics conditional
ksmbd: add SMB Direct RDMA encryption transform
ksmbd: handle encrypted compressed requests
ksmbd: decrypt requests from expired encrypted sessions
ksmbd: disconnect on SMB3 decryption failure
ksmbd: encrypt interim responses to encrypted requests
ksmbd: scope session state changes to bound connections
ksmbd: fix encrypted request lookup on bound channels
ksmbd: add per-share SMB3 encryption enforcement
...
Quite alot of buf fixes again:
- Assorted locking, bounds-checking, cleanup, and error-path fixes across
UCMA/CMA, bng_re, bnxt_re, cxgb4, EFA, ERDMA, HFI1, HNS, ionic, iRDMA,
mlx4/mlx5, RXE, SIW, SRP/SRPT, and iSER target.
- netlink report for max # of supported resources
- get_zeroed_page()/etc removal
- Robust udata for ionic
- Allow unique RDMA device names per network namespace
- Completion counters and v2 admit queue support for EFA
- UC QP support for MANA
- Completion timestamps for ionic
- Harden uverbs data validation and resource lifetime handling, fixing several core use-after-free conditions.
- bnxt_re toggle-page ownership and lifetime bug fixes
- dmabuf SRQ support for mlx5
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRRRCHOFoQz/8F5bUaFwuHvBreFYQUCaoeO2wAKCRCFwuHvBreF
YcyJAQCn8JeqTuayLchCARJX+9fazTaJES9zj41i3M8a7BfxCAEA2b266g0S660m
7eQ761NiIb1iklSl9rmt8vM22EMcngM=
=qcwp
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull RDMA updates from Jason Gunthorpe:
"About the normal size, still a lot of AI bug fixes and so on, but some
interesting new functionality too:
- Assorted locking, bounds-checking, cleanup, and error-path fixes
across UCMA/CMA, bng_re, bnxt_re, cxgb4, EFA, ERDMA, HFI1, HNS,
ionic, iRDMA, mlx4/mlx5, RXE, SIW, SRP/SRPT, and iSER target.
- netlink report for max # of supported resources
- get_zeroed_page()/etc removal
- Robust udata for ionic
- Allow unique RDMA device names per network namespace
- Completion counters and v2 admit queue support for EFA
- UC QP support for MANA
- Completion timestamps for ionic
- Harden uverbs data validation and resource lifetime handling,
fixing several core use-after-free conditions.
- bnxt_re toggle-page ownership and lifetime bug fixes
- dmabuf SRQ support for mlx5"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (160 commits)
RDMA/ucma: Allow path records to exactly fit the output buffer
RDMA/uverbs: Guard legacy bundles without method_elm
RDMA/efa: Add support for 128B admin v2 SQ entry
RDMA/efa: Generalize the admin SQ
RDMA/efa: Decouple admin command payload from admin header
RDMA/rxe: Fix OOB in free_rd_atomic_resources()
RDMA/cma: Fix WARNING in res_to_rt
RDMA/cxgb4: Free debugfs on registration failure
RDMA/cxgb4: Cancel reg_work before freeing device on remove
RDMA/ucma: Lock the handler in ucma_set_ib_path()
RDMA/ucma: Lock the handler in ucma_write_cm_event()
RDMA/erdma: restrict the driver to little-endian systems
RDMA/ionic: Embed counter driver data in rdma_counter allocation
RDMA/ionic: Cap eq_count to the eth driver's interrupt vector budget
RDMA/siw: Fix use-after-free in siw_accept()
IB/isert: post the full-feature receive buffers after session registration
IB/isert: delay the final Login Response until the session is registered
RDMA/srp: fix heap information leak on a truncated SRP_CRED_REQ
RDMA/erdma: Hold QP references for AE and CM processing
RDMA/erdma: Hold CQ references when processing EQ events
...
The DES crypto header is no longer used after the removal of NTLMv1
authentication. Remove it now that the server no longer selects
CRYPTO_LIB_DES.
Fixes: ce812992f2 ("ksmbd: remove NTLMv1 authentication")
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
The DES encryption in the smb server code has been removed in 2021
with the removal of the insecure NTLMv1 authentication code. Thus
we don't need this "select" statement here anymore.
Fixes: ce812992f2 ("ksmbd: remove NTLMv1 authentication")
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
ksmbd creates its listening socket with sock_create_kern(). Kernel
sockets do not hold a network namespace reference by default. Accepted
sockets inherit this state.
When an accepted socket is released, tcp_close() clears its pending TCP
timers for a kernel socket after the socket enters an orphaned state. If
the peer is unreachable while ksmbd sends a FIN, this can leave a
FIN-WAIT-1 orphan without a retransmission timer.
Upgrade the listening socket's network namespace reference before
kernel_listen(). Accepted sockets inherit the reference, so the TCP
stack can keep the retransmission timer active and apply its normal
orphan retry policy.
Preserve the existing graceful shutdown behavior.
Link: https://github.com/openwrt/openwrt/issues/24744
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
A client that disappears without sending a FIN or RST can leave its
ksmbd connection in ESTABLISHED indefinitely. ksmbd sets a socket
receive timeout, but the connection receive loop retries timeout errors
without a limit, so the connection remains in conn_list and consumes the
per-IP connection quota.
Enable SO_KEEPALIVE on accepted TCP sockets so the TCP stack can detect a
silent peer failure. The keepalive idle time, interval, and probe count
remain controlled by the existing TCP sysctl settings.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Reproducer:
1. server: systemctl start ksmbd
2. client: mount with `posix` option
mount -t cifs -o posix //${server_ip}/export /mnt
3. client: touch /mnt/file1 /mnt/file2
4. client: C program: int fd = open("/mnt/file2", O_RDONLY);
5. client: C program: rename("/mnt/file1", "/mnt/file2");
6. client: C program: struct stat stbuf; fstat(fd, &stbuf);
stbuf.st_nlink is 1, should be 0
This patch fixes xfstests generic/035 when mounted with `posix` option.
Signed-off-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Commit 110fee6b9b ("smb: client: fix missing timestamp updates with
O_TRUNC") introduced cifs_file_set_size(), which calls netfs_resize_file()
and cifs_setsize() on success. cifs_setsize() calls i_size_write(),
updating i_size to the new value. The subsequent blocks in both
cifs_setattr_unix() and cifs_setattr_nounix():
if ((attrs->ia_valid & ATTR_SIZE) &&
attrs->ia_size != i_size_read(inode)) {
truncate_setsize(inode, attrs->ia_size);
netfs_resize_file(&cifsInode->netfs, attrs->ia_size, true);
fscache_resize_cookie(cifs_inode_cookie(inode), attrs->ia_size);
}
are therefore unreachable on the success path: attrs->ia_size ==
i_size_read(inode) always holds after cifs_file_set_size() succeeds.
On the failure path, execution jumps to out/cifs_setattr_exit before
reaching these blocks.
truncate_setsize() and netfs_resize_file() are redundant with what
cifs_file_set_size() already did; fscache_resize_cookie() was moved there
by commit fa724e235c ("cifs: add fscache_resize_cookie() to
cifs_setsize()"). Remove both dead blocks.
Fixes: 110fee6b9b ("smb: client: fix missing timestamp updates with O_TRUNC")
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: Huiwen He <hehuiwen@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
cifs_remap_file_range() acquires i_rwsem on both inodes via
lock_two_nondirectories() before calling smb2_duplicate_extents().
cifs_setsize() (called inside smb2_duplicate_extents() when the clone
extends the file) therefore already runs under the lock, meaning the
fscache_resize_cookie() added to cifs_setsize() by commit fa724e235c
("cifs: add fscache_resize_cookie() to cifs_setsize()") is correctly
serialised for this path without further changes.
That same commit made the caller-side block:
if (rc == 0 && new_size > i_size) {
truncate_setsize(target_inode, new_size);
fscache_resize_cookie(cifs_inode_cookie(target_inode), new_size);
}
redundant: smb2_duplicate_extents() already performs the full size update
via cifs_setsize() when the operation extends the file. Remove the
now-dead block.
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Reviewed-by: Huiwen He <hehuiwen@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
cifs_setsize() calls fscache_resize_cookie() without holding i_rwsem.
When the fscache cookie is active (FSCACHE_COOKIE_IS_CACHING is set),
fscache_resize_cookie() performs a real resize that requires i_rwsem
held exclusively. If another file descriptor has the same inode open,
fscache_use_cookie() was already called from that cifs_open(), making
the cookie active. In that case, calling cifs_setsize() from
cifs_do_truncate() (invoked from cifs_open() without i_rwsem) races
against concurrent fscache I/O.
Strip fscache_resize_cookie() from cifs_setsize(), making it a pure
size/page-cache helper. Add cifs_resize_file_locked() for callers
that already hold i_rwsem: it calls netfs_resize_file() and
cifs_setsize(), then temporarily activates the cookie with
fscache_use_cookie() to perform the resize under the lock, then
deactivates it with cifs_fscache_unuse_inode_cookie(). Using
fscache_use_cookie() before the resize ensures correctness whether or
not another fd already holds the cookie active.
Switch cifs_file_set_size(), smb2_duplicate_extents(), and both size-
extension branches of smb3_simple_falloc() to the new wrapper; those
paths already hold i_rwsem via VFS setattr, lock_two_nondirectories(),
or cifs_fallocate() respectively. cifs_do_truncate() continues to
call cifs_setsize() followed by cifs_invalidate_cache(), since it runs
without i_rwsem.
Fixes: fa724e235c ("cifs: add fscache_resize_cookie() to cifs_setsize()")
Cc: stable@vger.kernel.org
Cc: David Howells <dhowells@redhat.com>
Cc: Paulo Alcantara <pc@manguebit.com>
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
cifs_do_truncate() is invoked from cifs_open() without i_rwsem, so it
cannot use cifs_resize_file_locked() to perform a proper fscache cookie
resize. Instead, add cifs_invalidate_cache() after cifs_setsize().
cifs_invalidate_cache() calls fscache_invalidate(), which works without
holding i_rwsem: it unconditionally increments inval_counter and sets
FSCACHE_COOKIE_NO_DATA_TO_READ, ensuring that stale cached data is not
served once the cookie is later activated by fscache_use_cookie().
Truncation to zero leaves no valid cached data, making invalidation the
correct semantic here.
Fixes: fa724e235c ("cifs: add fscache_resize_cookie() to cifs_setsize()")
Cc: stable@vger.kernel.org
Cc: David Howells <dhowells@redhat.com>
Cc: Paulo Alcantara <pc@manguebit.com>
Signed-off-by: Frank Sorenson <sorenson@redhat.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
When free_tgts() frees all structures in ce->tlist, ce->tgthint
is left pointing to one of the freed cache_dfs_tgt structures.
If ce->tgthint is not reset before it is used later, it results
in a use-after-free.
Set ce->tgthint to NULL in free_tgts() after the elements are
freed to reflect that no elements remain.
Fixes: 54be1f6c1c ("cifs: Add DFS cache routines")
Cc: stable@vger.kernel.org # depends on: smb: client: harden DFS cache against invalid target hints
Signed-off-by: Fredric Cover <fredric.cover.lkernel@gmail.com>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Currently, get_tgt_name() returns ERR_PTR(-ENOENT) when ce->tgthint is
NULL, and dfs_cache_noreq_update_tgthint() assumes ce->tgthint is always
valid.
In preparation for clearing ce->tgthint in free_tgts(), harden callers
of get_tgt_name() against ERR_PTR results and harden
dfs_cache_noreq_update_tgthint() against NULL pointer dereferences.
Cc: stable@vger.kernel.org
Signed-off-by: Fredric Cover <fredric.cover.lkernel@gmail.com>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
When a file has the setuid or setgid bit set and is written to, the VFS
strips those bits and issues a setattr with ATTR_KILL_SUID/ATTR_KILL_SGID
together with an ATTR_MODE carrying the already-cleared mode.
Both cifs_setattr_unix() and cifs_setattr_nounix() unconditionally dropped
ATTR_MODE in that case:
/* skip mode change if it's just for clearing setuid/setgid */
if (attrs->ia_valid & (ATTR_KILL_SUID|ATTR_KILL_SGID))
attrs->ia_valid &= ~ATTR_MODE;
This is fine for the default mount, where the mode is only emulated via
the DOS read-only attribute and cannot represent the setuid/setgid bits
anyway. However, with the "cifsacl" or "modefromsid" mount options the
mode is stored on the server through an ACL (id_mode_to_cifs_acl()), with
the SMB3.1.1 POSIX extensions the mode is sent to the server directly,
and with the SMB1 Unix extensions (cifs_setattr_unix) the mode is sent
via CIFSSMBUnixSetPathInfo(). In all those cases dropping ATTR_MODE means
the cleared mode is never pushed to the server, so the setuid/setgid bit
survives the write.
This is a security issue: on local filesystems the setuid bit is stripped
when a file is written, but over these cifs.ko mounts the bit persists on
the server, potentially allowing an unexpected privilege escalation on
subsequent execution.
Fix this in two places:
1. cifs_setattr_nounix(): only take the "skip mode change" shortcut
when the mode is emulated via the DOS read-only attribute (i.e.
neither cifsacl/modefromsid nor the SMB3.1.1 POSIX extensions are
in effect), so that the cleared mode is propagated to the server
in the ACL / POSIX cases.
2. cifs_setattr_unix(): this function is only called when Unix
extensions are in effect, so the mode is always stored on the
server. Remove the shortcut entirely so that the cleared mode is
always pushed.
Fixes: d32c4f2626 ("CIFS: ignore mode change if it's just for clearing setuid/setgid bits")
Cc: stable@vger.kernel.org
Signed-off-by: Jiangshan Yi <yijiangshan@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
smb2_async_readv() and smb2_async_writev() end with the same
send-error block: if the error is replayable and smb2_should_replay()
agrees, tell netfs to retry the subrequest. The write path also sets
wdata->replay. The read path does not set rdata->replay.
smb2_should_replay() is not a pure predicate. It consumes the retry
budget and computes the exponential back-off, doubling cur_sleep up to
CIFS_MAX_SLEEP. That back-off is only applied where the replay flag is
tested at the top of the reissued request:
if (rdata->replay) {
/* Back-off before retry */
if (rdata->cur_sleep)
msleep(rdata->cur_sleep);
smb2_set_replay(server, &rqst);
}
So on the read path the back-off is recomputed on every send-error
retry and then discarded, and SMB2_FLAGS_REPLAY_OPERATION is not set
on the reissued request.
netfs does not pace the retry either. netfs_reissue_read() calls
->issue_read() directly, and fs/netfs/read_retry.c contains no delay
of its own, so read send-error retries reissue immediately while the
equivalent write retries back off.
The read response callback already sets rdata->replay under the same
conditions, so the read path does use the replay mechanism. Only this
send-error path omits it.
Where the back-off belongs was settled while the commit below was
under review. David Howells asked whether netfslib should be doing the
back-off [1], and objected to sleeping inside the response callback
because that runs in the cifsd thread and would stall the socket [2].
The sleep was therefore taken out of smb2_should_replay() and moved to
just before the replay in smb2_async_readv() and smb2_async_writev()
[3]. Setting the flag here preserves that arrangement: the sleep still
happens at the top of the reissued request, not in a callback.
Set rdata->replay here, matching smb2_async_writev().
Fixes: 2c1238a747 ("cifs: make retry logic in read/write path consistent with other paths")
Link: https://lore.kernel.org/all/1652858.1769038134@warthog.procyon.org.uk/ [1]
Link: https://lore.kernel.org/all/1653031.1769038583@warthog.procyon.org.uk/ [2]
Link: https://lore.kernel.org/all/CANT5p=pXP3+CywpmK-on2uTvxO3S=31_B85_UDR7RoK1dQVtMA@mail.gmail.com/ [3]
Assisted-by: Codex:gpt-5.5
Assisted-by: Claude:claude-opus-5
Signed-off-by: Christopher Lusk <clusk@northecho.dev>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
smb2_new_read_req() allocates the request buffer with
smb2_plain_req_init() but only publishes it to the caller with
*buf = req at the very end of the function. Two error returns sit in
between:
rc = smb2_plain_req_init(SMB2_READ, io_parms->tcon, server,
(void **) &req, total_len);
if (rc)
return rc;
if (server == NULL)
return -ECONNABORTED;
[...]
rdata->mr = smbd_register_mr(server->smbd_conn,
&rdata->subreq.io_iter,
true, need_invalidate);
if (!rdata->mr)
return -EAGAIN;
On either of them the buffer is neither released nor handed back, so
it is leaked. The caller cannot clean up after it: smb2_async_readv()
does 'goto out' on a non-zero return, which skips the
cifs_small_buf_release(buf) at async_readv_out, and buf has not been
assigned at that point in any case.
The write path has never had this problem. smb2_async_writev()
registers the memory region inline and jumps to its release label
instead of returning:
wdata->mr = smbd_register_mr(...);
if (!wdata->mr) {
rc = -EAGAIN;
goto async_writev_out;
}
Commit b797209219 ("cifs: smbd: Retry on memory registration
failure") changed both sides from -ENOBUFS to -EAGAIN in a single
patch, which puts the two shapes next to each other.
Only the -EAGAIN return is reachable in practice, because
smb2_plain_req_init() calls smb2_reconnect() first and that already
fails with -EIO when server is NULL, before anything is allocated.
Both returns are given the same treatment here rather than leaving
one of them correct only by accident.
Because -EAGAIN is a replayable error, the failure also reaches the
retry block at the end of smb2_async_readv(), which marks the
subrequest NETFS_SREQ_NEED_RETRY, so a failing registration can be
retried rather than ending the I/O, and every attempt that reaches it
leaks another buffer. smb2_should_replay() short-circuits on
tcon->retry, so on a hard mount the attempt count is not bounded by
the retrans setting.
Only the asynchronous read path is affected. The synchronous
SMB2_read() caller passes rdata == NULL and the memory registration
block is guarded on rdata.
The memory registration failure path was pointed out by the Sashiko
AI reviewer while it was reviewing an unrelated patch to
smb2_async_readv().
Fixes: bd3dcc6a22 ("CIFS: SMBD: Upper layer performs SMB read via RDMA write through memory registration")
Link: https://sashiko.dev/#/patchset/20260729192002.876156-1-clusk%40northecho.dev
Link: https://lore.kernel.org/all/20260729192002.876156-1-clusk@northecho.dev/
Assisted-by: Claude:claude-opus-5
Signed-off-by: Christopher Lusk <clusk@northecho.dev>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Some SMB servers return STATUS_OBJECT_NAME_NOT_FOUND or
STATUS_DELETE_PENDING when creating an O_TMPFILE.
The SMB client maps them to ENOENT.
The ENOENT error is supposed to be returned when the path
(a target directory in this case) does not exist, while a
lack of support of O_TMPFILE by the target file system should
be indicated by EOPNOTSUPP.
Reported-by: Mikhail Stefantsev <mikhail@mstefan99.com>
Fixes: 3e7d63037a ("smb: client: add support for O_TMPFILE")
Closes: https://lore.kernel.org/linux-cifs/75fb5359-b268-492d-8d4a-504d1af60f2a@app.fastmail.com/
Signed-off-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Paulo Alcantara <pc@manguebit.org>
Add library APIs for most AES encryption modes that are used in the
kernel (ECB, CBC, CBC-CTS, CTR, XCTR, XTS, GCM, CCM).
These AES modes have many in-kernel users that are currently using the
crypto_skcipher or crypto_aead APIs. These existing APIs are difficult
to use and inefficient. Until now, the lack of proper library support
for these has been the main gap in the crypto library.
This set of changes is the next stage of addressing it:
- Implement the new APIs on top of the existing support for
single-block AES in the library.
- Fully document the new APIs.
- Migrate the only user of the old AES-GCM library API to the new,
more flexible API; then remove the old API and its implementation.
- Wire up the new APIs to the traditional crypto API by adding
crypto_skcipher and crypto_aead algorithms.
This makes the new APIs be covered by the traditional crypto API's
self-tests. It also makes them be already used for real on systems
that don't have architecture-optimized code for these modes.
But most importantly, this is a prerequisite for migrating the
architecture-optimized code for these AES modes (i.e.
arch/*/crypto/aes*) into the library, which as usual will
eliminate a lot of redundant "glue" code.
Note that unlike some of the other algorithms that have been migrated
to the library, e.g. SHA-512, for these AES modes there was too much
to get done in one cycle. Nor did it make sense to handle these modes
one at a time, because they tend to be coupled together or depend on
each other, especially in the architecture-optimized AES code.
Thus, most of the benefits (reductions in lines of code, performance
improvements, etc.) will follow in later cycles when
architecture-optimized code is migrated into the library and users of
crypto_skcipher and crypto_aead are updated to use the new APIs.
The design of the new APIs was informed by writing proof-of-concept
patches for many kernel subsystems currently accessing these same
algorithms via crypto_skcipher or crypto_aead (patches 18-33 of
https://lore.kernel.org/r/20260707053503.209874-1-ebiggers@kernel.org/).
While those patches will be resent for real later, the total diffstat
for them was negative 1905 lines. So clearly the new APIs are quite a
bit easier to use and align better with what users actually need.
Besides the new AES encryption APIs, there are also a few changes for
improved AES-CMAC key and context zeroization.
-----BEGIN PGP SIGNATURE-----
iIoEABYIADIWIQSacvsUNc7UX4ntmEPzXCl4vpKOKwUCaoEqhRQcZWJpZ2dlcnNA
a2VybmVsLm9yZwAKCRDzXCl4vpKOKy8YAP4o44z09KdugtigFIFX7qQnLk0uPAw+
sTPqmu+FA4jOGAD/Vy23hzAVpWmSBFhyo7HMK9W0ABK7cqj6x23I/N6JRgM=
=SLPT
-----END PGP SIGNATURE-----
Merge tag 'libcrypto-updates-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux
Pull crypto library updates from Eric Biggers:
"Add library APIs for most AES encryption modes that are used in the
kernel (ECB, CBC, CBC-CTS, CTR, XCTR, XTS, GCM, CCM).
These AES modes have many in-kernel users that are currently using the
crypto_skcipher or crypto_aead APIs. These existing APIs are difficult
to use and inefficient. Until now, the lack of proper library support
for these has been the main gap in the crypto library.
This set of changes is the next stage of addressing it:
- Implement the new APIs on top of the existing support for
single-block AES in the library.
- Fully document the new APIs.
- Migrate the only user of the old AES-GCM library API to the new,
more flexible API; then remove the old API and its implementation.
- Wire up the new APIs to the traditional crypto API by adding
crypto_skcipher and crypto_aead algorithms.
This makes the new APIs be covered by the traditional crypto API's
self-tests. It also makes them be already used for real on systems
that don't have architecture-optimized code for these modes.
But most importantly, this is a prerequisite for migrating the
architecture-optimized code for these AES modes (i.e.
arch/*/crypto/aes*) into the library, which as usual will eliminate
a lot of redundant "glue" code.
Note that unlike some of the other algorithms that have been migrated
to the library, e.g. SHA-512, for these AES modes there was too much
to get done in one cycle. Nor did it make sense to handle these modes
one at a time, because they tend to be coupled together or depend on
each other, especially in the architecture-optimized AES code.
Thus, most of the benefits (reductions in lines of code, performance
improvements, etc.) will follow in later cycles when
architecture-optimized code is migrated into the library and users of
crypto_skcipher and crypto_aead are updated to use the new APIs.
The design of the new APIs was informed by writing proof-of-concept
patches for many kernel subsystems currently accessing these same
algorithms via crypto_skcipher or crypto_aead (patches 18-33 of
https://lore.kernel.org/r/20260707053503.209874-1-ebiggers@kernel.org/).
While those patches will be resent for real later, the total diffstat
for them was negative 1905 lines. So clearly the new APIs are quite a
bit easier to use and align better with what users actually need.
Besides the new AES encryption APIs, there are also a few changes for
improved AES-CMAC key and context zeroization"
* tag 'libcrypto-updates-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiggers/linux:
mac80211: fils_aead: Use __cleanup() instead of memzero_explicit()
Bluetooth: SMP: clear the aes_cmac_key when done
smb: clear the aes_cmac_key and aes_cmac_ctx when done
lib/crypto: aes-cmac: Add zeroization functions
lib/crypto: aesgcm: Remove old AES-GCM library
x86/sev: Remove obsolete virtual address check
x86/sev: Use new AES-GCM library
crypto: aes - Add CCM support using library
crypto: aes - Add GCM support using library
crypto: aes - Add XTS support using library
crypto: aes - Add CTR and XCTR support using library
crypto: aes - Add CBC and CBC-CTS support using library
crypto: aes - Add ECB support using library
lib/crypto: aes: Add CCM support
lib/crypto: aes: Add GCM support
lib/crypto: aes: Add XTS support
lib/crypto: aes: Add CTR and XCTR support
lib/crypto: aes: Add CBC and CBC-CTS support
lib/crypto: aes: Add ECB support
crypto: xts - Split out __xts_verify_key() helper
Please consider pulling these changes from the signed vfs-7.3-rc1.misc tag.
Thanks!
Christian
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCan7RJQAKCRCRxhvAZXjc
otI6AQCIwshv02UI9fmbTC9pFr8b+v/9IWQwdVULaH61HXgICQD7BE/CktL8nfRs
zCzGS6kBdWgiOA6bdVpgGGLqiMRnQgo=
=Kc+G
-----END PGP SIGNATURE-----
Merge tag 'vfs-7.3-rc1.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull misc vfs updates from Christian Brauner:
"Bigger cleanups:
- The lockref dead-count handling is tidied up.
The open-coded check for a count below zero as the dead marker
relies on information the caller should not have.
- make put_mnt_ns() leave mounts connected. Destroying a mount
namespace disconnected its mounts from their mount points. So a
file descriptor still open on the parent of a mount point could be
used to peek under it.
Locked mounts were already kept connected to prevent exactly that.
But a mount is only locked when its tree is copied across a user
namespace boundary. So a mount namespace set up by a privileged
component had no locked mounts and its mounts were disconnected.
Passing UMOUNT_CONNECTED keeps every mount connected and prevents
that bug.
- vfs_prepare_mode() passes S_IFDIR for directories. I meant to fix
that ago but didn't get to it. So now someone finally did it.
This kills the exception where the mode could be 0 when a directory
was created whereas every other creation operation passed it
explicitly already.
- move long delayed work for ufs, jffs2, hfsplus, hfs and affs from
the per-cpu system_long_wq to the new unbound system_dfl_long_wq.
None of that work relies on per-cpu state and the work item is
enqueued with queue_delayed_work() whose timer is global anyway. So
it may as well benefit from scheduler task placement.
Smaller fixes and cleanups:
- unlock_buffer() and journal_end_buffer_io_sync() use
clear_and_wake_up_bit()
- the pipe page pools are unified into a single per-pipe pool and the
extra wake_up(rd_wait) is limited to EPOLLET consumers
- eventpoll now computes its timer slack lazily in ep_poll()
- shrink_dcache_for_umount() keeps making progress on busy roots
- excess xarray nodes are freed in clear_inode()
- romfs detects hard link cycles
- the user path of nested backing files is fixed
- pidfd holds exec_update_lock around the namespace ioctl
- non-memcg-aware nr_cached_objects is skipped during memcg slab
shrink
- iomap_write_iter() always returns status
- mangle_path() is renamed to seq_mangle_path()
- inode timestamp accessors are annotated
- new regression test for pipe->poll_usage.
- a few documentation, kernel-doc and selftest fixes"
* tag 'vfs-7.3-rc1.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (67 commits)
selftests/namespaces: Fix racy pipe handshake in timens and pidns_separate
selftests/epoll: add a regression test for pipe->poll_usage
pipe: only enable the extra wake_up(rd_wait) for EPOLLET consumers
pidfd: hold exec_update_lock around namespace ioctl
fs: fix user path of nested backing files
fs: remove stale inode_insert5() kernel-doc parameter
fs: fix switch/case indentation in sysfs() syscall
fs: document semantics of kstat::{uid,gid} fields
dcache: keep shrink_dcache_for_umount() making progress on busy roots
seq_file: rename mangle_path to seq_mangle_path
nstree: add/fix struct ns_id_req kernel-doc member fields
dcache: use lockref routines for dead count checks
lockref: tidy up dead count handling
initramfs: fix typo in reserve_initrd_mem comment
fs/pipe: unify the page pools into a single per-pipe pool
fs: annotate inode timestamp accessors
eventpoll: compute timer slack lazily in ep_poll()
selftests/filesystems: add mntns cleanup test
put_mnt_ns(): leave mounts connected
affs: Move long delayed work on system_dfl_long_wq
...
Please consider pulling these changes from the signed vfs-7.3-rc1.lookup tag.
Thanks!
Christian
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCan7RJQAKCRCRxhvAZXjc
ouj6AQCnBHstCt5RRimywLYU+eTmdKe203zRQYxXpwlYOAzm7wEAm1mIC7/Of6bD
jkiJilcPDcyYcJsqwBw/Vcz5HR+v8wo=
=TX6K
-----END PGP SIGNATURE-----
Merge tag 'vfs-7.3-rc1.lookup' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs lookup updates from Christian Brauner:
"This refactors lookup_open() and adds vfs_lookup_open() for nfsd.
mnt_want_write() and parent locking are moved into lookup_open()
itself.
audit_inode_child() is also now called in lookup_open() on failure.
That is the calling convention in vfs_create() and vfs_mkdir(), but
lookup_open() made no such call when atomic_open() should have created
a file and did not. And neither did the regular ->create() path fwiw.
This also contains work to remove the unneeded excl argument from the
->create() inode op"
* tag 'vfs-7.3-rc1.lookup' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
fs/namei.c: fix coding style in atomic_open() and lookup_open()
fs/namei.c: fix kerneldoc of atomic_open() and vfs_lookup_open()
fs/namei.c: update stale comments in lookup_open()
Remove excl arg to ->create inode_operation
fs/namei.c: update kerneldoc of atomic_open()
vfs: call audit_inode_child() in lookup_open() on failure
vfs: move create error && negative dentry case in lookup_open() up
VFS: add vfs_lookup_open() for nfsd
VFS: move delegated_inode retry loop into lookup_open()
VFS: move mnt_want_write() and locking into lookup_open()
Please consider pulling these changes from the signed vfs-7.3-rc1.kthread tag.
Thanks!
Christian
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCan7RJQAKCRCRxhvAZXjc
ohHuAQCFeGhxqsb44KhuJaWXewiXbefr6eg31lXmxgSfg0DILAD/Tw/WmpiUbX5M
H+/jh6NmLKnQ/ADjlWFRy3R1K1Il3w4=
=jwPP
-----END PGP SIGNATURE-----
Merge tag 'vfs-7.3-rc1.kthread' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull kthread vfs updates from Christian Brauner:
"This stops kernel threads from sharing filesystem state with
userspace. This work is about 3 cycles old and has been in -next
for about that time.
When the kernel boots init_task creates PID 1 and then kthreadd. From
that point every kthread and PID 1 share the same fs_struct. That is
why pivot_root() has to rewrite the fs_struct of all kthreads. The
rewriting exists so that kthreads can use init's filesystem state when
they want to. It also means userspace can move the ground out from
under the kernel.
PID 1 now gets a completely separate fs_struct. All kthreads are
anchored in a private SB_KERNMOUNT instance of nullfs that cannot be
mounted on and cannot be used to follow other mounts. Userspace init
can no longer affect kthread filesystem state and kthreads can no
longer affect userspace fs state without explicit opting in to that.
Path lookup from a kthread now fails by default. It makes it
deliberately hard to offload security sensitive operations into init's
filesystem state from a kthread.
Places that legitimately need to look something up there opt in
through the new scoped_with_init_fs() which temporarily overrides the
caller's fs_struct with init's. usermodehelpers remain the only kernel
tasks that genuinely share init's filesystem state, since they execute
random binaries in the root filesystem (excellent...).
The visible result is that /proc/2/root is a nullfs with an empty
mountinfo while /proc/1/root is the real root"
* tag 'vfs-7.3-rc1.kthread' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (26 commits)
initramfs_test: use test init/exit hooks to override init fs
fs: stop rewriting paths for PF_EXITING | PF_DUMPCORE
fs: stop rewriting kthread fs structs
fs: start all kthreads in nullfs
nullfs: make nullfs multi-instance
devtmpfs: create private mount namespace
fs: add umh argument to struct kernel_clone_args
fs: stop sharing fs_struct between init_task and pid 1
af_unix: use scoped_with_init_fs() for coredump socket lookup
initramfs: use scoped_with_init_fs() for rootfs unpacking
pnfs/blocklayout: use scoped_with_init_fs() for SCSI device lookup
ksmbd: use scoped_with_init_fs() for VFS path operations
ksmbd: use scoped_with_init_fs() for filesystem info path lookup
ksmbd: use scoped_with_init_fs() for share path resolution
fs: use scoped_with_init_fs() for kernel_read_file_from_path_initns()
coredump: use scoped_with_init_fs() for coredump path resolution
btrfs: use scoped_with_init_fs() for update_dev_time()
scsi: target: use scoped_with_init_fs() for APTPL metadata
scsi: target: use scoped_with_init_fs() for ALUA metadata
crypto: ccp: use scoped_with_init_fs() for SEV file access
...
See the procedure below:
smb2_sess_setup
ksmbd_smb2_session_create
__session_create
hash_add(sessions_table, &sess->hlist, sess->id)
ksmbd_counter_inc(KSMBD_COUNTER_SESSIONS)
ksmbd_conn_handler_loop
ksmbd_server_terminate_conn
ksmbd_sessions_deregister
hash_del(&sess->hlist)
// do not decrement KSMBD_COUNTER_SESSIONS
KSMBD_COUNTER_SESSIONS tracks sessions published in sessions_table, but
session removal does not decrement it. The value therefore keeps growing
after sessions are expired, rejected during registration, or removed on
the last channel disconnect.
Fixes: b38f99c121 ("ksmbd: add procfs interface for runtime monitoring and statistics")
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
KSMBD_COUNTER_SESSIONS tracks sessions published in sessions_table.
Increment it while holding sessions_table_lock so publishing a session and
updating the counter happen together.
Fixes: b38f99c121 ("ksmbd: add procfs interface for runtime monitoring and statistics")
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
See the procedure below:
smb2_sess_setup
ksmbd_smb2_session_create
__session_create
atomic_set(&sess->refcnt, 2)
hash_add(sessions_table, &sess->hlist, sess->id)
ksmbd_session_register
xa_store(&conn->sessions, sess->id, sess) // fail
ksmbd_user_session_put
atomic_dec(&sess->refcnt) // refcnt is 1, session is not freed
Remove the session from sessions_table and drop its table reference if
xa_store() fails.
Fixes: f5c779b7dd ("ksmbd: fix racy issue from session setup and logoff")
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Print a warning if the sessions procfs entry cannot be created.
Signed-off-by: Ze Tan <tanze@kylinos.cn>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
smb_check_perm_dacl() validates that the DACL fits inside the NT
security descriptor, but then bounds its two ACE walks by the
remaining NTSD length (acl_size) rather than the DACL's declared
size (pdacl_size).
When pdacl->size is smaller than the trailing NTSD buffer, bytes
after the declared DACL boundary - still inside the stored security
descriptor - are parsed as ACEs during access checks. A crafted
DACL can place an access-granting ACE beyond pdacl->size, and the
current code accepts it during SMB2_CREATE access validation, while
parse_dacl() and smb_inherit_dacl() stop at pdacl_size.
Bound both ACE walks by pdacl_size to match the DACL boundary
semantics used elsewhere in the server.
Validation:
- semantic KUnit harness shows the post-boundary ACE is selected
before the fix and rejected (EACCES) after it
- linux master (7.2-rc6), x86_64
Fixes: 8f0541186e ("ksmbd: fix heap-based overflow in set_ntacl_dacl()")
Signed-off-by: Hang Nan <2122295973@qq.com>
Reviewed-by: ChenXiaoSong <chenxiaosong@kylinos.cn>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
The temporary RDMA encryption diagnostics logged every SMB3 request
and successful payload operation with pr_err(), which made normal
traffic too noisy.
Keep only negotiation, RDMA READ preparation, RDMA WRITE transform
metadata, crypto completion, and final transfer completion messages as
KSMBD_DEBUG_RDMA diagnostics. Keep error reports for malformed metadata,
crypto, RDMA transfer, and file write failures at error level.
This preserves the diagnostics needed to verify RDMA transform
operation without flooding the kernel error log during normal I/O.
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
