Skip to contentSkip to conversation

FAILED: patch "[PATCH] smb: client: fix uid/gid override in getattr with posix" failed to apply to 5.10-stable tree

PATCH New [email protected] 2 participants 3 messages Sep 16 – Sep 18

Subsystems (3)
Stable branch SMB client (CIFS) Kernel SMB3 server (ksmbd)
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 5.10-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

To reproduce the conflict and resubmit, you may use the following commands:

git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.10.y
git checkout FETCH_HEAD
git cherry-pick -x cf4d35896621b7298eef51b7a465e5c0cb22f670
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<[email protected]>' --in-reply-to '2026091636-plot-cobalt-bbd3@gregkh' --subject-prefix 'PATCH 5.10.y' 'HEAD^..'

Possible dependencies:

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From cf4d35896621b7298eef51b7a465e5c0cb22f670 Mon Sep 17 00:00:00 2001
From: Paulo Alcantara <[email protected]>
Date: Sun, 6 Sep 2026 14:39:44 -0300
Subject: [PATCH] smb: client: fix uid/gid override in getattr with posix
extensions

When mounting with 'multiuser,posix' options, cifs_getattr() overrides the server-provided uid/gid with the current process's fsuid/fsgid. This is because the condition only checks for unix extensions (tcon->unix_ext) but not posix extensions (tcon->posix_extensions).

With SMB3 POSIX extensions, the server provides real uid/gid values just like with unix extensions, so they should be preserved rather than replaced with the caller's credentials.

Add a tcon->posix_extensions check to the condition so that uid/gid from the server are properly reported in stat results.

Reported-by Arthur Lesuisse <[email protected]>
Closes https://lore.kernel.org/r/DB9P190MB2012266F6B8DECBE5D26A1798DB52@DB9P190MB2012.EURP190.PROD.OUTLOOK.COM open here
Suggested-by Arthur Lesuisse <[email protected]>
Reviewed-by Namjae Jeon <[email protected]>
Signed-off-by Paulo Alcantara <[email protected]>
Cc Ronnie Sahlberg <[email protected]>
Cc Shyam Prasad N <[email protected]>
Cc Tom Talpey <[email protected]>
Cc Bharath SM <[email protected]>
Cc [email protected]
fs/smb/client/inode.c +4 -4
index 12ed8db10e00..49f9993ad567 100644
@@ -2992,14 +2992,14 @@ int cifs_getattr(struct mnt_idmap *idmap, const struct path *path,
 		stat->attributes |= STATX_ATTR_ENCRYPTED;
 
 	/*
-	 * If on a multiuser mount without unix extensions or cifsacl being
-	 * enabled, and the admin hasn't overridden them, set the ownership
-	 * to the fsuid/fsgid of the current process.
+	 * If on a multiuser mount without unix extensions, posix extensions
+	 * or cifsacl being enabled, and the admin hasn't overridden them,
+	 * set the ownership to the fsuid/fsgid of the current process.
 	 */
 	sbflags = cifs_sb_flags(cifs_sb);
 	if ((sbflags & CIFS_MOUNT_MULTIUSER) &&
 	    !(sbflags & CIFS_MOUNT_CIFS_ACL) &&
-	    !tcon->unix_ext) {
+	    !tcon->unix_ext && !tcon->posix_extensions) {
 		if (!(sbflags & CIFS_MOUNT_OVERR_UID))
 			stat->uid = current_fsuid();
 		if (!(sbflags & CIFS_MOUNT_OVERR_GID))
maintainer
#1 lore raw

PATCH 1/2 smb: client: use atomic_t for mnt_cifs_flags

From: Paulo Alcantara <[email protected]>

[ Upstream commit 4fc3a433c13944ee5766ec5b9bf6f1eb4d29b880 ]

Use atomic_t for cifs_sb_info::mnt_cifs_flags as it's currently accessed locklessly and may be changed concurrently in mount/remount and reconnect paths.

Signed-off-by Paulo Alcantara (Red Hat) <[email protected]>
Reviewed-by David Howells <[email protected]>
Cc [email protected]
Signed-off-by Steve French <[email protected]>

Adapt the dependency to the 5.10 client under fs/cifs. Keep its existing smb_vol mount setup, CIFS_SB interface and cifs_remap implementation; omit upstream-only cached directory, reparse and shutdown changes. Convert all mount flag readers, including the older DFS and fscache paths, and use atomic bit operations for updates after mount setup. Initialize the flags once from the existing mount options.

Provide cifs_sb_flags as a single-evaluation macro so this backport adds no functions. Retain the cifs_getattr flag snapshot needed for cf4d35896621b7298eef51b7a465e5c0cb22f670 to cherry-pick cleanly, while leaving its POSIX ownership fix to that target commit.

[ sashal: Reduced backport -- upstream 4fc3a433c1394 touches 27 file(s), this
backport carries 19. Not backported here:
fs/smb/client/cached_dir.c
fs/smb/client/cifsacl.c
fs/smb/client/cifsfs.c
fs/smb/client/cifs_fs_sb.h
fs/smb/client/cifsglob.h
fs/smb/client/cifs_ioctl.h
fs/smb/client/cifs_unicode.c
fs/smb/client/cifs_unicode.h
... and 19 more
This note is generated from the file lists only; see the resolution record for the reasoning. ]

Stable-dep-of: cf4d35896621 ("smb: client: fix uid/gid override in getattr with posix extensions")

Signed-off-by Sasha Levin <[email protected]>
---
 fs/cifs/cifs_dfs_ref.c |  2 +-
 fs/cifs/cifs_fs_sb.h   |  6 ++-
 fs/cifs/cifs_unicode.c |  4 +-
 fs/cifs/cifsacl.c      |  6 +--
 fs/cifs/cifsfs.c       | 68 ++++++++++++++---------------
 fs/cifs/cifsglob.h     | 10 +++--
 fs/cifs/connect.c      | 97 +++++++++++++++++++++---------------------
 fs/cifs/dir.c          | 16 +++----
 fs/cifs/file.c         | 38 ++++++++---------
 fs/cifs/fscache.c      |  2 +-
 fs/cifs/inode.c        | 90 ++++++++++++++++++++-------------------
 fs/cifs/link.c         |  4 +-
 fs/cifs/misc.c         | 10 ++---
 fs/cifs/readdir.c      | 18 ++++----
 fs/cifs/smb1ops.c      |  4 +-
 fs/cifs/smb2misc.c     |  4 +-
 fs/cifs/smb2ops.c      |  2 +-
 fs/cifs/smb2pdu.c      |  4 +-
 fs/cifs/xattr.c        |  6 +--
 19 files changed, 201 insertions(+), 190 deletions(-)
fs/cifs/cifs_dfs_ref.c +1 -1
index 42125601ebb10..b2bbf54551094 100644
@@ -317,7 +317,7 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt)
 	mnt = ERR_PTR(-ENOMEM);
 
 	cifs_sb = CIFS_SB(mntpt->d_sb);
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_DFS) {
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NO_DFS) {
 		mnt = ERR_PTR(-EREMOTE);
 		goto cdda_exit;
 	}
fs/cifs/cifs_fs_sb.h +5 -1
index 6e7c4427369d8..f1fe2d5a7589b 100644
@@ -20,6 +20,7 @@
 #ifndef _CIFS_FS_SB_H
 #define _CIFS_FS_SB_H
 
+#include <linux/atomic.h>
 #include <linux/backing-dev.h>
 
 #define CIFS_MOUNT_NO_PERM      1 /* do not do client vfs_perm check */

@@ -56,6 +57,9 @@
 #define CIFS_MOUNT_RO_CACHE	0x20000000  /* assumes share will not change */
 #define CIFS_MOUNT_RW_CACHE	0x40000000  /* assumes only client accessing */
 
+/* Mount and reconnect paths may update these flags concurrently. */
+#define cifs_sb_flags(cifs_sb) ((unsigned int)atomic_read(&(cifs_sb)->mnt_cifs_flags))
+
 struct cifs_sb_info {
 	struct rb_root tlink_tree;
 	spinlock_t tlink_tree_lock;

@@ -72,7 +76,7 @@ struct cifs_sb_info {
 	kgid_t	mnt_backupgid;
 	umode_t	mnt_file_mode;
 	umode_t	mnt_dir_mode;
-	unsigned int mnt_cifs_flags;
+	atomic_t mnt_cifs_flags;
 	char   *mountdata; /* options received at mount time or via DFS refs */
 	struct delayed_work prune_tlinks;
 	struct rcu_head rcu;
fs/cifs/cifs_unicode.c +2 -2
index 171ad8b42107e..91f84bd6299a6 100644
@@ -18,9 +18,9 @@ int cifs_remap(struct cifs_sb_info *cifs_sb)
 {
 	int map_type;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SFM_CHR)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MAP_SFM_CHR)
 		map_type = SFM_MAP_UNI_RSVD;
-	else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR)
+	else if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MAP_SPECIAL_CHR)
 		map_type = SFU_MAP_UNI_RSVD;
 	else
 		map_type = NO_MAP_UNI_RSVD;
fs/cifs/cifsacl.c +3 -3
index c8a0a2edc6500..6f5c4b51dd265 100644
@@ -359,7 +359,7 @@ sid_to_id(struct cifs_sb_info *cifs_sb, struct cifs_sid *psid,
 		return -EIO;
 	}
 
-	if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UID_FROM_ACL) ||
+	if ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_UID_FROM_ACL) ||
 	    (cifs_sb_master_tcon(cifs_sb)->posix_extensions)) {
 		uint32_t unix_id;
 		bool is_group;

@@ -1332,12 +1332,12 @@ id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64 nmode,
 		return -ENOMEM;
 	}
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MODE_FROM_SID)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MODE_FROM_SID)
 		mode_from_sid = true;
 	else
 		mode_from_sid = false;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UID_FROM_ACL)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_UID_FROM_ACL)
 		id_from_sid = true;
 	else
 		id_from_sid = false;
fs/cifs/cifsfs.c +34 -34

156 lines not on this page · raw message

fs/cifs/cifsglob.h +7 -3
index d0905aa0b04cd..79335daa86efc 100644
@@ -1452,9 +1452,13 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file);
 #define CIFS_CACHE_RW_FLG	(CIFS_CACHE_READ_FLG | CIFS_CACHE_WRITE_FLG)
 #define CIFS_CACHE_RHW_FLG	(CIFS_CACHE_RW_FLG | CIFS_CACHE_HANDLE_FLG)
 
-#define CIFS_CACHE_READ(cinode) ((cinode->oplock & CIFS_CACHE_READ_FLG) || (CIFS_SB(cinode->vfs_inode.i_sb)->mnt_cifs_flags & CIFS_MOUNT_RO_CACHE))
+#define CIFS_CACHE_READ(cinode) \
+	(((cinode)->oplock & CIFS_CACHE_READ_FLG) || \
+	 (cifs_sb_flags(CIFS_SB((cinode)->vfs_inode.i_sb)) & CIFS_MOUNT_RO_CACHE))
 #define CIFS_CACHE_HANDLE(cinode) (cinode->oplock & CIFS_CACHE_HANDLE_FLG)
-#define CIFS_CACHE_WRITE(cinode) ((cinode->oplock & CIFS_CACHE_WRITE_FLG) || (CIFS_SB(cinode->vfs_inode.i_sb)->mnt_cifs_flags & CIFS_MOUNT_RW_CACHE))
+#define CIFS_CACHE_WRITE(cinode) \
+	(((cinode)->oplock & CIFS_CACHE_WRITE_FLG) || \
+	 (cifs_sb_flags(CIFS_SB((cinode)->vfs_inode.i_sb)) & CIFS_MOUNT_RW_CACHE))
 
 /*
  * One of these for each file inode

@@ -1515,7 +1519,7 @@ CIFS_FILE_SB(struct file *file)
 
 static inline char CIFS_DIR_SEP(const struct cifs_sb_info *cifs_sb)
 {
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_POSIX_PATHS)
 		return '/';
 	else
 		return '\\';
fs/cifs/connect.c +49 -48

233 lines not on this page · raw message

fs/cifs/dir.c +8 -8
index 73d43d8911aa5..e1493768d2f2d 100644
@@ -107,7 +107,7 @@ build_path_from_dentry_optional_prefix(struct dentry *direntry, bool prefix)
 	else
 		dfsplen = 0;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_USE_PREFIX_PATH)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_USE_PREFIX_PATH)
 		pplen = cifs_sb->prepath ? strlen(cifs_sb->prepath) + 1 : 0;
 
 cifs_bp_rename_retry:

@@ -181,7 +181,7 @@ build_path_from_dentry_optional_prefix(struct dentry *direntry, bool prefix)
 
 	if (dfsplen) {
 		strncpy(full_path, tcon->treeName, dfsplen);
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) {
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_POSIX_PATHS) {
 			int i;
 			for (i = 0; i < dfsplen; i++) {
 				if (full_path[i] == '\\')

@@ -208,7 +208,7 @@ check_name(struct dentry *direntry, struct cifs_tcon *tcon)
 		     le32_to_cpu(tcon->fsAttrInfo.MaxPathNameComponentLength)))
 		return -ENAMETOOLONG;
 
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS)) {
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_POSIX_PATHS)) {
 		for (i = 0; i < direntry->d_name.len; i++) {
 			if (direntry->d_name.name[i] == '\\') {
 				cifs_dbg(FYI, "Invalid file name\n");

@@ -384,7 +384,7 @@ cifs_do_create(struct inode *inode, struct dentry *direntry, unsigned int xid,
 				.device	= 0,
 		};
 
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) {
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SET_UID) {
 			args.uid = current_fsuid();
 			if (inode->i_mode & S_ISGID)
 				args.gid = inode->i_gid;

@@ -418,10 +418,10 @@ cifs_do_create(struct inode *inode, struct dentry *direntry, unsigned int xid,
 		if (newinode) {
 			if (server->ops->set_lease_key)
 				server->ops->set_lease_key(newinode, fid);
-			if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM)
+			if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_DYNPERM)
 				newinode->i_mode = mode;
 			if ((*oplock & CIFS_CREATE_ACTION) &&
-			    (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID)) {
+			    (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SET_UID)) {
 				newinode->i_uid = current_fsuid();
 				if (inode->i_mode & S_ISGID)
 					newinode->i_gid = inode->i_gid;

@@ -545,8 +545,8 @@ cifs_atomic_open(struct inode *inode, struct dentry *direntry,
 	}
 
 	if (file->f_flags & O_DIRECT &&
-	    CIFS_SB(inode->i_sb)->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) {
-		if (CIFS_SB(inode->i_sb)->mnt_cifs_flags & CIFS_MOUNT_NO_BRL)
+	    cifs_sb_flags(CIFS_SB(inode->i_sb)) & CIFS_MOUNT_STRICT_IO) {
+		if (cifs_sb_flags(CIFS_SB(inode->i_sb)) & CIFS_MOUNT_NO_BRL)
 			file->f_op = &cifs_file_direct_nobrl_ops;
 		else
 			file->f_op = &cifs_file_direct_ops;
fs/cifs/file.c +19 -19
index 0da9c11cae278..0ec0cf7edf7b3 100644
@@ -473,7 +473,7 @@ void _cifsFileInfo_put(struct cifsFileInfo *cifs_file,
 		 * close  because it may cause a error when we open this file
 		 * again and get at least level II oplock.
 		 */
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO)
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_STRICT_IO)
 			set_bit(CIFS_INO_INVALID_MAPPING, &cifsi->flags);
 		cifs_set_oplock_level(cifsi, 0);
 	}

@@ -544,8 +544,8 @@ int cifs_open(struct inode *inode, struct file *file)
 		 inode, file->f_flags, full_path);
 
 	if (file->f_flags & O_DIRECT &&
-	    cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) {
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL)
+	    cifs_sb_flags(cifs_sb) & CIFS_MOUNT_STRICT_IO) {
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NO_BRL)
 			file->f_op = &cifs_file_direct_nobrl_ops;
 		else
 			file->f_op = &cifs_file_direct_ops;

@@ -657,7 +657,7 @@ cifs_relock_file(struct cifsFileInfo *cfile)
 
 	if (cap_unix(tcon->ses) &&
 	    (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) &&
-	    ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0))
+	    ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOPOSIXBRL) == 0))
 		rc = cifs_push_posix_locks(cfile);
 	else
 		rc = tcon->ses->server->ops->push_mand_locks(cfile);

@@ -1361,7 +1361,7 @@ cifs_push_locks(struct cifsFileInfo *cfile)
 
 	if (cap_unix(tcon->ses) &&
 	    (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) &&
-	    ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0))
+	    ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOPOSIXBRL) == 0))
 		rc = cifs_push_posix_locks(cfile);
 	else
 		rc = tcon->ses->server->ops->push_mand_locks(cfile);

@@ -1741,7 +1741,7 @@ int cifs_flock(struct file *file, int cmd, struct file_lock *fl)
 
 	if (cap_unix(tcon->ses) &&
 	    (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) &&
-	    ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0))
+	    ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOPOSIXBRL) == 0))
 		posix_lck = true;
 
 	if (!lock && !unlock) {

@@ -1789,7 +1789,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *flock)
 
 	if (cap_unix(tcon->ses) &&
 	    (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) &&
-	    ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0))
+	    ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOPOSIXBRL) == 0))
 		posix_lck = true;
 	/*
 	 * BB add code here to normalize offset and length to account for

@@ -1919,7 +1919,7 @@ struct cifsFileInfo *find_readable_file(struct cifsInodeInfo *cifs_inode,
 	struct cifs_sb_info *cifs_sb = CIFS_SB(cifs_inode->vfs_inode.i_sb);
 
 	/* only filter by fsuid on multiuser mounts */
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER))
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MULTIUSER))
 		fsuid_only = false;
 
 	spin_lock(&cifs_inode->open_file_lock);

@@ -1975,7 +1975,7 @@ cifs_get_writable_file(struct cifsInodeInfo *cifs_inode, int flags,
 	cifs_sb = CIFS_SB(cifs_inode->vfs_inode.i_sb);
 
 	/* only filter by fsuid on multiuser mounts */
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER))
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MULTIUSER))
 		fsuid_only = false;
 
 	spin_lock(&cifs_inode->open_file_lock);

@@ -2526,7 +2526,7 @@ static int cifs_write_end(struct file *file, struct address_space *mapping,
 	struct cifs_sb_info *cifs_sb = CIFS_SB(cfile->dentry->d_sb);
 	__u32 pid;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_RWPIDFORWARD)
 		pid = cfile->pid;
 	else
 		pid = current->tgid;

@@ -2608,7 +2608,7 @@ int cifs_strict_fsync(struct file *file, loff_t start, loff_t end,
 	}
 
 	tcon = tlink_tcon(smbfile->tlink);
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC)) {
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOSSYNC)) {
 		server = tcon->ses->server;
 		if (server->ops->flush == NULL) {
 			rc = -ENOSYS;

@@ -2653,7 +2653,7 @@ int cifs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 		 file, datasync);
 
 	tcon = tlink_tcon(smbfile->tlink);
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC)) {
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOSSYNC)) {
 		server = tcon->ses->server;
 		if (server->ops->flush == NULL) {
 			rc = -ENOSYS;

@@ -2900,7 +2900,7 @@ cifs_write_from_iter(loff_t offset, size_t len, struct iov_iter *from,
 	size_t start;
 	unsigned int xid;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_RWPIDFORWARD)
 		pid = open_file->pid;
 	else
 		pid = current->tgid;

@@ -3301,7 +3301,7 @@ cifs_strict_writev(struct kiocb *iocb, struct iov_iter *from)
 	if (CIFS_CACHE_WRITE(cinode)) {
 		if (cap_unix(tcon->ses) &&
 		(CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability))
-		  && ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0)) {
+		  && ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOPOSIXBRL) == 0)) {
 			written = generic_file_write_iter(iocb, from);
 			goto out;
 		}

@@ -3638,7 +3638,7 @@ cifs_send_async_read(loff_t offset, size_t len, struct cifsFileInfo *open_file,
 
 	server = cifs_pick_channel(tlink_tcon(open_file->tlink)->ses);
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_RWPIDFORWARD)
 		pid = open_file->pid;
 	else
 		pid = current->tgid;

@@ -4012,7 +4012,7 @@ cifs_strict_readv(struct kiocb *iocb, struct iov_iter *to)
 
 	if (cap_unix(tcon->ses) &&
 	    (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) &&
-	    ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0))
+	    ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NOPOSIXBRL) == 0))
 		return generic_file_read_iter(iocb, to);
 
 	/*

@@ -4066,7 +4066,7 @@ cifs_read(struct file *file, char *read_data, size_t read_size, loff_t *offset)
 		return -ENOSYS;
 	}
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_RWPIDFORWARD)
 		pid = open_file->pid;
 	else
 		pid = current->tgid;

@@ -4397,7 +4397,7 @@ static int cifs_readpages(struct file *file, struct address_space *mapping,
 		return rc;
 	}
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_RWPIDFORWARD)
 		pid = open_file->pid;
 	else
 		pid = current->tgid;

@@ -4629,7 +4629,7 @@ bool is_size_safe_to_change(struct cifsInodeInfo *cifsInode, __u64 end_of_file)
 		struct cifs_sb_info *cifs_sb;
 
 		cifs_sb = CIFS_SB(cifsInode->vfs_inode.i_sb);
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO) {
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_DIRECT_IO) {
 			/* since no page cache to corrupt on directio
 			we can change size safely */
 			return true;
fs/cifs/fscache.c +1 -1
index da688185403c3..79074a104890c 100644
@@ -164,7 +164,7 @@ static void cifs_fscache_enable_inode_cookie(struct inode *inode)
 	if (cifsi->fscache)
 		return;
 
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_FSCACHE))
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_FSCACHE))
 		return;
 
 	cifs_fscache_acquire_inode_cookie(cifsi, tcon);
fs/cifs/inode.c +46 -44

299 lines not on this page · raw message

fs/cifs/link.c +2 -2
index 4fc6eb8e786d4..84545eea2d181 100644
@@ -638,7 +638,7 @@ cifs_get_link(struct dentry *direntry, struct inode *inode,
 	 * First try Minshall+French Symlinks, if configured
 	 * and fallback to UNIX Extensions Symlinks.
 	 */
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MF_SYMLINKS)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MF_SYMLINKS)
 		rc = query_mf_symlink(xid, tcon, cifs_sb, full_path,
 				      &target_path);
 

@@ -694,7 +694,7 @@ cifs_symlink(struct inode *inode, struct dentry *direntry, const char *symname)
 	cifs_dbg(FYI, "symname is %s\n", symname);
 
 	/* BB what if DFS and this volume is on different share? BB */
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MF_SYMLINKS)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MF_SYMLINKS)
 		rc = create_mf_symlink(xid, pTcon, cifs_sb, full_path, symname);
 	else if (pTcon->unix_ext)
 		rc = CIFSUnixCreateSymLink(xid, pTcon, full_path, symname,
fs/cifs/misc.c +5 -5
index b4199b4085db3..10b054f05f390 100644
@@ -537,13 +537,13 @@ dump_smb(void *buf, int smb_buf_length)
 void
 cifs_autodisable_serverino(struct cifs_sb_info *cifs_sb)
 {
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) {
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SERVER_INUM) {
 		struct cifs_tcon *tcon = NULL;
 
 		if (cifs_sb->master_tlink)
 			tcon = cifs_sb_master_tcon(cifs_sb);
 
-		cifs_sb->mnt_cifs_flags &= ~CIFS_MOUNT_SERVER_INUM;
+		atomic_andnot(CIFS_MOUNT_SERVER_INUM, &cifs_sb->mnt_cifs_flags);
 		cifs_sb->mnt_cifs_serverino_autodisabled = true;
 		cifs_dbg(VFS, "Autodisabling the use of server inode numbers on %s\n",
 			 tcon ? tcon->treeName : "new server");

@@ -643,11 +643,11 @@ void cifs_done_oplock_break(struct cifsInodeInfo *cinode)
 bool
 backup_cred(struct cifs_sb_info *cifs_sb)
 {
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_BACKUPUID) {
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_CIFS_BACKUPUID) {
 		if (uid_eq(cifs_sb->mnt_backupuid, current_fsuid()))
 			return true;
 	}
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_BACKUPGID) {
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_CIFS_BACKUPGID) {
 		if (in_group_p(cifs_sb->mnt_backupgid))
 			return true;
 	}

@@ -1206,7 +1206,7 @@ int update_super_prepath(struct cifs_tcon *tcon, char *prefix)
 	} else
 		cifs_sb->prepath = NULL;
 
-	cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_USE_PREFIX_PATH;
+	atomic_or(CIFS_MOUNT_USE_PREFIX_PATH, &cifs_sb->mnt_cifs_flags);
 
 out:
 	cifs_put_tcon_super(sb);
fs/cifs/readdir.c +9 -9
index 9a7315aa24e0d..dab5d1b4bc1f0 100644
@@ -112,7 +112,7 @@ cifs_prime_dcache(struct dentry *parent, struct qstr *name,
 			 * want to clobber the existing one with the one that
 			 * the readdir code created.
 			 */
-			if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM))
+			if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SERVER_INUM))
 				fattr->cf_uniqueid = CIFS_I(inode)->uniqueid;
 
 			/* update inode in place

@@ -220,11 +220,11 @@ cifs_fill_common_info(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb)
 	 * may look wrong since the inodes may not have timed out by the time
 	 * "ls" does a stat() call on them.
 	 */
-	if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) ||
-	    (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MODE_FROM_SID))
+	if ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_CIFS_ACL) ||
+	    (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MODE_FROM_SID))
 		fattr->cf_flags |= CIFS_FATTR_NEED_REVAL;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL &&
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_UNX_EMUL &&
 	    fattr->cf_cifsattrs & ATTR_SYSTEM) {
 		if (fattr->cf_eof == 0)  {
 			fattr->cf_mode &= ~S_IFMT;

@@ -435,7 +435,7 @@ initiate_cifs_search(const unsigned int xid, struct file *file,
 	else if ((tcon->ses->capabilities &
 		  tcon->ses->server->vals->cap_nt_find) == 0) {
 		cifsFile->srch_inf.info_level = SMB_FIND_FILE_INFO_STANDARD;
-	} else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) {
+	} else if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SERVER_INUM) {
 		cifsFile->srch_inf.info_level = SMB_FIND_FILE_ID_FULL_DIR_INFO;
 	} else /* not srvinos - BB fixme add check for backlevel? */ {
 		cifsFile->srch_inf.info_level = SMB_FIND_FILE_DIRECTORY_INFO;

@@ -455,8 +455,8 @@ initiate_cifs_search(const unsigned int xid, struct file *file,
 	else if STATUS_STOPPED_ON_SYMLINK
 		call get_symlink_reparse_path and retry with new path */
 	else if ((rc == -EOPNOTSUPP) &&
-		(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM)) {
-		cifs_sb->mnt_cifs_flags &= ~CIFS_MOUNT_SERVER_INUM;
+		(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SERVER_INUM)) {
+		atomic_andnot(CIFS_MOUNT_SERVER_INUM, &cifs_sb->mnt_cifs_flags);
 		goto ffirst_retry;
 	}
 error_exit:

@@ -908,14 +908,14 @@ static int cifs_filldir(char *find_entry, struct file *file,
 		break;
 	}
 
-	if (de.ino && (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM)) {
+	if (de.ino && (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SERVER_INUM)) {
 		fattr.cf_uniqueid = de.ino;
 	} else {
 		fattr.cf_uniqueid = iunique(sb, ROOT_I);
 		cifs_autodisable_serverino(cifs_sb);
 	}
 
-	if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MF_SYMLINKS) &&
+	if ((cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MF_SYMLINKS) &&
 	    couldbe_mf_symlink(&fattr))
 		/*
 		 * trying to get the type and mode can be slow,
fs/cifs/smb1ops.c +2 -2
index 80287c26cfac0..5d22ded1f68f2 100644
@@ -1052,7 +1052,7 @@ cifs_make_node(unsigned int xid, struct inode *inode,
 			.mtime	= NO_CHANGE_64,
 			.device	= dev,
 		};
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) {
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_SET_UID) {
 			args.uid = current_fsuid();
 			args.gid = current_fsgid();
 		} else {

@@ -1077,7 +1077,7 @@ cifs_make_node(unsigned int xid, struct inode *inode,
 	 * SMB1 SFU emulation: should work with all servers, but only
 	 * support block and char device (no socket & fifo)
 	 */
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL))
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_UNX_EMUL))
 		goto out;
 
 	if (!S_ISCHR(mode) && !S_ISBLK(mode))
fs/cifs/smb2misc.c +2 -2
index 2013489431ac1..179c264469c0d 100644
@@ -477,9 +477,9 @@ cifs_convert_path_to_utf16(const char *from, struct cifs_sb_info *cifs_sb)
 	__le16 *to;
 	int map_type;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SFM_CHR)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MAP_SFM_CHR)
 		map_type = SFM_MAP_UNI_RSVD;
-	else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR)
+	else if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_MAP_SPECIAL_CHR)
 		map_type = SFU_MAP_UNI_RSVD;
 	else
 		map_type = NO_MAP_UNI_RSVD;
fs/cifs/smb2ops.c +1 -1
index 51f8ea0c89065..da6a8f43b72ef 100644
@@ -5108,7 +5108,7 @@ smb2_make_node(unsigned int xid, struct inode *inode,
 	 * supports block and char device (no socket & fifo),
 	 * and was used by default in earlier versions of Windows
 	 */
-	if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL))
+	if (!(cifs_sb_flags(cifs_sb) & CIFS_MOUNT_UNX_EMUL))
 		goto out;
 
 	/*
fs/cifs/smb2pdu.c +2 -2
index 19b4f426e9afa..a17e33a0ee7eb 100644
@@ -2805,7 +2805,7 @@ SMB2_open_init(struct cifs_tcon *tcon, struct TCP_Server_Info *server,
 		bool set_mode;
 		bool set_owner;
 
-		if ((oparms->cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MODE_FROM_SID) &&
+		if ((cifs_sb_flags(oparms->cifs_sb) & CIFS_MOUNT_MODE_FROM_SID) &&
 		    (oparms->mode != ACL_NO_MODE))
 			set_mode = true;
 		else {

@@ -2813,7 +2813,7 @@ SMB2_open_init(struct cifs_tcon *tcon, struct TCP_Server_Info *server,
 			oparms->mode = ACL_NO_MODE;
 		}
 
-		if (oparms->cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UID_FROM_ACL)
+		if (cifs_sb_flags(oparms->cifs_sb) & CIFS_MOUNT_UID_FROM_ACL)
 			set_owner = true;
 		else
 			set_owner = false;
fs/cifs/xattr.c +3 -3
index c6931eab3f97e..01cf57b64341c 100644
@@ -154,7 +154,7 @@ static int cifs_xattr_set(const struct xattr_handler *handler,
 			break;
 		}
 
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR)
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NO_XATTR)
 			goto out;
 
 		if (pTcon->ses->server->ops->set_EA) {

@@ -321,7 +321,7 @@ static int cifs_xattr_get(const struct xattr_handler *handler,
 			break;
 		}
 
-		if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR)
+		if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NO_XATTR)
 			goto out;
 
 		if (pTcon->ses->server->ops->query_all_EAs)

@@ -402,7 +402,7 @@ ssize_t cifs_listxattr(struct dentry *direntry, char *data, size_t buf_size)
 	struct cifs_tcon *pTcon;
 	char *full_path;
 
-	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR)
+	if (cifs_sb_flags(cifs_sb) & CIFS_MOUNT_NO_XATTR)
 		return -EOPNOTSUPP;
 
 	tlink = cifs_sb_tlink(cifs_sb);
Signature · 1 line
2.53.0
maintainer
#2 lore raw

PATCH 2/2 smb: client: fix uid/gid override in getattr with posix extensions

From: Paulo Alcantara <[email protected]>

[ Upstream commit cf4d35896621b7298eef51b7a465e5c0cb22f670 ]

When mounting with 'multiuser,posix' options, cifs_getattr() overrides the server-provided uid/gid with the current process's fsuid/fsgid. This is because the condition only checks for unix extensions (tcon->unix_ext) but not posix extensions (tcon->posix_extensions).

With SMB3 POSIX extensions, the server provides real uid/gid values just like with unix extensions, so they should be preserved rather than replaced with the caller's credentials.

Add a tcon->posix_extensions check to the condition so that uid/gid from the server are properly reported in stat results.

Reported-by Arthur Lesuisse <[email protected]>
Closes https://lore.kernel.org/r/DB9P190MB2012266F6B8DECBE5D26A1798DB52@DB9P190MB2012.EURP190.PROD.OUTLOOK.COM open here
Suggested-by Arthur Lesuisse <[email protected]>
Reviewed-by Namjae Jeon <[email protected]>
Signed-off-by Paulo Alcantara <[email protected]>
Cc Ronnie Sahlberg <[email protected]>
Cc Shyam Prasad N <[email protected]>
Cc Tom Talpey <[email protected]>
Cc Bharath SM <[email protected]>
Cc [email protected]
Signed-off-by Sasha Levin <[email protected]>
---
 fs/cifs/inode.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
fs/cifs/inode.c +4 -4
index 6f44f54ea60e2..2258ad7f43eb3 100644
@@ -2418,14 +2418,14 @@ int cifs_getattr(const struct path *path, struct kstat *stat,
 		stat->attributes |= STATX_ATTR_ENCRYPTED;
 
 	/*
-	 * If on a multiuser mount without unix extensions or cifsacl being
-	 * enabled, and the admin hasn't overridden them, set the ownership
-	 * to the fsuid/fsgid of the current process.
+	 * If on a multiuser mount without unix extensions, posix extensions
+	 * or cifsacl being enabled, and the admin hasn't overridden them,
+	 * set the ownership to the fsuid/fsgid of the current process.
 	 */
 	sbflags = cifs_sb_flags(cifs_sb);
 	if ((sbflags & CIFS_MOUNT_MULTIUSER) &&
 	    !(sbflags & CIFS_MOUNT_CIFS_ACL) &&
-	    !tcon->unix_ext) {
+	    !tcon->unix_ext && !tcon->posix_extensions) {
 		if (!(sbflags & CIFS_MOUNT_OVERR_UID))
 			stat->uid = current_fsuid();
 		if (!(sbflags & CIFS_MOUNT_OVERR_GID))
Signature · 1 line
2.53.0