[PATCH RFC 07/24] vfs: make vfs_create break delegations on parent directory
From: Jeff Layton
Date: Fri Mar 15 2024 - 12:57:45 EST
In order to add directory delegation support, we need to break
delegations on the parent whenever there is going to be a change in the
directory.
Add a new delegated_inode parameter to vfs_create. Most callers will
set that to NULL, but do_mknodat can use that to synchronously wait
for the delegation break to complete.
Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
---
fs/ecryptfs/inode.c | 2 +-
fs/namei.c | 15 +++++++++++++--
fs/nfsd/nfs3proc.c | 2 +-
fs/nfsd/vfs.c | 2 +-
fs/open.c | 2 +-
fs/overlayfs/overlayfs.h | 2 +-
fs/smb/server/vfs.c | 2 +-
include/linux/fs.h | 2 +-
8 files changed, 20 insertions(+), 9 deletions(-)
diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c
index 3d0cddbf037c..a99b1e264c46 100644
--- a/fs/ecryptfs/inode.c
+++ b/fs/ecryptfs/inode.c
@@ -189,7 +189,7 @@ ecryptfs_do_create(struct inode *directory_inode,
rc = lock_parent(ecryptfs_dentry, &lower_dentry, &lower_dir);
if (!rc)
rc = vfs_create(&nop_mnt_idmap, lower_dir,
- lower_dentry, mode, true);
+ lower_dentry, mode, true, NULL);
if (rc) {
printk(KERN_ERR "%s: Failure to create dentry in lower fs; "
"rc = [%d]\n", __func__, rc);
diff --git a/fs/namei.c b/fs/namei.c
index 88598a62ec64..01e04cf155eb 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -3174,6 +3174,7 @@ static inline umode_t vfs_prepare_mode(struct mnt_idmap *idmap,
* @dentry: pointer to dentry of the base directory
* @mode: mode of the new file
* @want_excl: whether the file must not yet exist
+ * @delegated_inode: return pointer for delegated_inode
*
* Create a new file.
*
@@ -3184,7 +3185,8 @@ static inline umode_t vfs_prepare_mode(struct mnt_idmap *idmap,
* raw inode simply pass @nop_mnt_idmap.
*/
int vfs_create(struct mnt_idmap *idmap, struct inode *dir,
- struct dentry *dentry, umode_t mode, bool want_excl)
+ struct dentry *dentry, umode_t mode, bool want_excl,
+ struct inode **delegated_inode)
{
int error;
@@ -3197,6 +3199,9 @@ int vfs_create(struct mnt_idmap *idmap, struct inode *dir,
mode = vfs_prepare_mode(idmap, dir, mode, S_IALLUGO, S_IFREG);
error = security_inode_create(dir, dentry, mode);
+ if (error)
+ return error;
+ error = try_break_deleg(dir, delegated_inode);
if (error)
return error;
error = dir->i_op->create(idmap, dir, dentry, mode, want_excl);
@@ -4047,6 +4052,7 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
struct path path;
int error;
unsigned int lookup_flags = 0;
+ struct inode *delegated_inode = NULL;
error = may_mknod(mode);
if (error)
@@ -4066,7 +4072,7 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
switch (mode & S_IFMT) {
case 0: case S_IFREG:
error = vfs_create(idmap, path.dentry->d_inode,
- dentry, mode, true);
+ dentry, mode, true, &delegated_inode);
if (!error)
ima_post_path_mknod(idmap, dentry);
break;
@@ -4081,6 +4087,11 @@ static int do_mknodat(int dfd, struct filename *name, umode_t mode,
}
out2:
done_path_create(&path, dentry);
+ if (delegated_inode) {
+ error = break_deleg_wait(&delegated_inode);
+ if (!error)
+ goto retry;
+ }
if (retry_estale(error, lookup_flags)) {
lookup_flags |= LOOKUP_REVAL;
goto retry;
diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c
index dfcc957e460d..e920a6291f2d 100644
--- a/fs/nfsd/nfs3proc.c
+++ b/fs/nfsd/nfs3proc.c
@@ -313,7 +313,7 @@ nfsd3_create_file(struct svc_rqst *rqstp, struct svc_fh *fhp,
status = fh_fill_pre_attrs(fhp);
if (status != nfs_ok)
goto out;
- host_err = vfs_create(&nop_mnt_idmap, inode, child, iap->ia_mode, true);
+ host_err = vfs_create(&nop_mnt_idmap, inode, child, iap->ia_mode, true, NULL);
if (host_err < 0) {
status = nfserrno(host_err);
goto out;
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
index 34cc2d1a4944..47b8ab1d4b17 100644
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1491,7 +1491,7 @@ nfsd_create_locked(struct svc_rqst *rqstp, struct svc_fh *fhp,
switch (type) {
case S_IFREG:
host_err = vfs_create(&nop_mnt_idmap, dirp, dchild,
- iap->ia_mode, true);
+ iap->ia_mode, true, NULL);
if (!host_err)
nfsd_check_ignore_resizing(iap);
break;
diff --git a/fs/open.c b/fs/open.c
index 0a73afe04d34..0b50ea7e8aec 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -1137,7 +1137,7 @@ struct file *dentry_create(const struct path *path, int flags, umode_t mode,
error = vfs_create(mnt_idmap(path->mnt),
d_inode(path->dentry->d_parent),
- path->dentry, mode, true);
+ path->dentry, mode, true, NULL);
if (!error)
error = vfs_open(path, f);
diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h
index 5b1f56294c4d..be2518e6da95 100644
--- a/fs/overlayfs/overlayfs.h
+++ b/fs/overlayfs/overlayfs.h
@@ -232,7 +232,7 @@ static inline int ovl_do_create(struct ovl_fs *ofs,
struct inode *dir, struct dentry *dentry,
umode_t mode)
{
- int err = vfs_create(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, true);
+ int err = vfs_create(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, true, NULL);
pr_debug("create(%pd2, 0%o) = %i\n", dentry, mode, err);
return err;
diff --git a/fs/smb/server/vfs.c b/fs/smb/server/vfs.c
index 5b4e5876c2ac..b313eb5a1d28 100644
--- a/fs/smb/server/vfs.c
+++ b/fs/smb/server/vfs.c
@@ -187,7 +187,7 @@ int ksmbd_vfs_create(struct ksmbd_work *work, const char *name, umode_t mode)
mode |= S_IFREG;
err = vfs_create(mnt_idmap(path.mnt), d_inode(path.dentry),
- dentry, mode, true);
+ dentry, mode, true, NULL);
if (!err) {
ksmbd_vfs_inherit_owner(work, d_inode(path.dentry),
d_inode(dentry));
diff --git a/include/linux/fs.h b/include/linux/fs.h
index e72c825476de..8fb4101fea49 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1837,7 +1837,7 @@ bool inode_owner_or_capable(struct mnt_idmap *idmap,
* VFS helper functions..
*/
int vfs_create(struct mnt_idmap *, struct inode *,
- struct dentry *, umode_t, bool);
+ struct dentry *, umode_t, bool, struct inode **);
int vfs_mkdir(struct mnt_idmap *, struct inode *,
struct dentry *, umode_t, struct inode **);
int vfs_mknod(struct mnt_idmap *, struct inode *, struct dentry *,
--
2.44.0