Never pass a NULL vfs_lookup to vfs_create() or iop->create by creating
temporary vfs_lookup objects where needed. Remove the obsolete NULL
checks.
Signed-off-by: Andreas Gruenbacher <[email protected]>
---
fs/9p/vfs_inode.c | 4 ++--
fs/cifs/dir.c | 5 ++---
fs/nfsd/vfs.c | 36 +++++++++++++++++++++---------------
ipc/mqueue.c | 23 ++++++++++++++---------
4 files changed, 39 insertions(+), 29 deletions(-)
--- a/fs/9p/vfs_inode.c
+++ b/fs/9p/vfs_inode.c
@@ -478,7 +478,7 @@ v9fs_vfs_create(struct inode *dir, struc
fid = NULL;
v9ses = v9fs_inode2v9ses(dir);
perm = unixmode2p9mode(v9ses, mode);
- if (lookup && lookup->flags & LOOKUP_OPEN)
+ if (lookup->flags & LOOKUP_OPEN)
flags = lookup->intent.open.flags - 1;
else
flags = O_RDWR;
@@ -492,7 +492,7 @@ v9fs_vfs_create(struct inode *dir, struc
}
/* if we are opening a file, assign the open fid to the file */
- if (lookup && lookup->flags & LOOKUP_OPEN) {
+ if (lookup->flags & LOOKUP_OPEN) {
filp = lookup_instantiate_filp(lookup, dentry,
v9fs_open_created);
if (IS_ERR(filp)) {
--- a/fs/cifs/dir.c
+++ b/fs/cifs/dir.c
@@ -153,7 +153,7 @@ cifs_create(struct inode *inode, struct
return -ENOMEM;
}
- if (lookup && (lookup->flags & LOOKUP_OPEN)) {
+ if (lookup->flags & LOOKUP_OPEN) {
int oflags = lookup->intent.open.flags;
desiredAccess = 0;
@@ -263,8 +263,7 @@ cifs_create(struct inode *inode, struct
direntry->d_op = &cifs_dentry_ops;
d_instantiate(direntry, newinode);
}
- if ((!lookup /* nfsd case - nfs srv does not set lookup */) ||
- ((lookup->flags & LOOKUP_OPEN) == FALSE)) {
+ if (!(lookup->flags & LOOKUP_OPEN)) {
/* mknod case - do not leave file open */
CIFSSMBClose(xid, pTcon, fileHandle);
} else if (newinode) {
--- a/fs/nfsd/vfs.c
+++ b/fs/nfsd/vfs.c
@@ -1139,7 +1139,8 @@ nfsd_create(struct svc_rqst *rqstp, stru
char *fname, int flen, struct iattr *iap,
int type, dev_t rdev, struct svc_fh *resfhp)
{
- struct dentry *dentry, *dchild = NULL;
+ struct vfs_lookup lookup;
+ struct dentry *dchild = NULL;
struct inode *dirp;
__be32 err;
int host_err;
@@ -1155,8 +1156,10 @@ nfsd_create(struct svc_rqst *rqstp, stru
if (err)
goto out;
- dentry = fhp->fh_dentry;
- dirp = dentry->d_inode;
+ lookup.path.dentry = fhp->fh_dentry;
+ lookup.path.mnt = fhp->fh_export->ex_mnt;
+ lookup.flags = 0;
+ dirp = lookup.path.dentry->d_inode;
err = nfserr_notdir;
if(!dirp->i_op || !dirp->i_op->lookup)
@@ -1168,7 +1171,7 @@ nfsd_create(struct svc_rqst *rqstp, stru
if (!resfhp->fh_dentry) {
/* called from nfsd_proc_mkdir, or possibly nfsd3_proc_create */
fh_lock_nested(fhp, I_MUTEX_PARENT);
- dchild = lookup_one_len(fname, dentry, flen);
+ dchild = lookup_one_len(fname, lookup.path.dentry, flen);
host_err = PTR_ERR(dchild);
if (IS_ERR(dchild))
goto out_nfserr;
@@ -1182,8 +1185,8 @@ nfsd_create(struct svc_rqst *rqstp, stru
/* not actually possible */
printk(KERN_ERR
"nfsd_create: parent %s/%s not locked!\n",
- dentry->d_parent->d_name.name,
- dentry->d_name.name);
+ lookup.path.dentry->d_parent->d_name.name,
+ lookup.path.dentry->d_name.name);
err = nfserr_io;
goto out;
}
@@ -1194,7 +1197,7 @@ nfsd_create(struct svc_rqst *rqstp, stru
err = nfserr_exist;
if (dchild->d_inode) {
dprintk("nfsd_create: dentry %s/%s not negative!\n",
- dentry->d_name.name, dchild->d_name.name);
+ lookup.path.dentry->d_name.name, dchild->d_name.name);
goto out;
}
@@ -1208,7 +1211,7 @@ nfsd_create(struct svc_rqst *rqstp, stru
err = 0;
switch (type) {
case S_IFREG:
- host_err = vfs_create(dirp, dchild, iap->ia_mode, NULL);
+ host_err = vfs_create(dirp, dchild, iap->ia_mode, &lookup);
break;
case S_IFDIR:
host_err = vfs_mkdir(dirp, dchild, iap->ia_mode);
@@ -1227,7 +1230,7 @@ nfsd_create(struct svc_rqst *rqstp, stru
goto out_nfserr;
if (EX_ISSYNC(fhp->fh_export)) {
- err = nfserrno(nfsd_sync_dir(dentry));
+ err = nfserrno(nfsd_sync_dir(lookup.path.dentry));
write_inode_now(dchild->d_inode, 1);
}
@@ -1267,7 +1270,8 @@ nfsd_create_v3(struct svc_rqst *rqstp, s
struct svc_fh *resfhp, int createmode, u32 *verifier,
int *truncp, int *created)
{
- struct dentry *dentry, *dchild = NULL;
+ struct vfs_lookup lookup;
+ struct dentry *dchild = NULL;
struct inode *dirp;
__be32 err;
int host_err;
@@ -1285,8 +1289,10 @@ nfsd_create_v3(struct svc_rqst *rqstp, s
if (err)
goto out;
- dentry = fhp->fh_dentry;
- dirp = dentry->d_inode;
+ lookup.path.dentry = fhp->fh_dentry;
+ lookup.path.mnt = fhp->fh_export->ex_mnt;
+ lookup.flags = 0;
+ dirp = lookup.path.dentry->d_inode;
/* Get all the sanity checks out of the way before
* we lock the parent. */
@@ -1298,7 +1304,7 @@ nfsd_create_v3(struct svc_rqst *rqstp, s
/*
* Compose the response file handle.
*/
- dchild = lookup_one_len(fname, dentry, flen);
+ dchild = lookup_one_len(fname, lookup.path.dentry, flen);
host_err = PTR_ERR(dchild);
if (IS_ERR(dchild))
goto out_nfserr;
@@ -1352,14 +1358,14 @@ nfsd_create_v3(struct svc_rqst *rqstp, s
goto out;
}
- host_err = vfs_create(dirp, dchild, iap->ia_mode, NULL);
+ host_err = vfs_create(dirp, dchild, iap->ia_mode, &lookup);
if (host_err < 0)
goto out_nfserr;
if (created)
*created = 1;
if (EX_ISSYNC(fhp->fh_export)) {
- err = nfserrno(nfsd_sync_dir(dentry));
+ err = nfserrno(nfsd_sync_dir(lookup.path.dentry));
/* setattr will sync the child (or not) */
}
--- a/ipc/mqueue.c
+++ b/ipc/mqueue.c
@@ -599,9 +599,11 @@ static int mq_attr_ok(struct mq_attr *at
/*
* Invoked when creating a new queue via sys_mq_open
*/
-static struct file *do_create(struct dentry *dir, struct dentry *dentry,
- int oflag, mode_t mode, struct mq_attr __user *u_attr)
+static struct file *do_create(struct vfs_lookup *lookup, int oflag, mode_t mode,
+ struct mq_attr __user *u_attr)
{
+ struct dentry *dir = mqueue_mnt->mnt_root;
+ struct dentry *dentry = lookup->path.dentry;
struct mq_attr attr;
int ret;
@@ -617,7 +619,7 @@ static struct file *do_create(struct den
}
mode &= ~current->fs->umask;
- ret = vfs_create(dir->d_inode, dentry, mode, NULL);
+ ret = vfs_create(dir->d_inode, dentry, mode, lookup);
dentry->d_fsdata = NULL;
if (ret)
goto out;
@@ -631,10 +633,11 @@ out:
}
/* Opens existing queue */
-static struct file *do_open(struct dentry *dentry, int oflag)
+static struct file *do_open(struct vfs_lookup *lookup, int oflag)
{
static int oflag2acc[O_ACCMODE] = { MAY_READ, MAY_WRITE,
MAY_READ | MAY_WRITE };
+ struct dentry *dentry = lookup->path.dentry;
if ((oflag & O_ACCMODE) == (O_RDWR | O_WRONLY)) {
dput(dentry);
@@ -654,6 +657,7 @@ static int oflag2acc[O_ACCMODE] = { MAY_
asmlinkage long sys_mq_open(const char __user *u_name, int oflag, mode_t mode,
struct mq_attr __user *u_attr)
{
+ struct vfs_lookup lookup;
struct dentry *dentry;
struct file *filp;
char *name;
@@ -676,7 +680,9 @@ asmlinkage long sys_mq_open(const char _
error = PTR_ERR(dentry);
goto out_err;
}
- mntget(mqueue_mnt);
+ lookup.path.mnt = mntget(mqueue_mnt);
+ lookup.path.dentry = dentry;
+ lookup.flags = 0;
if (oflag & O_CREAT) {
if (dentry->d_inode) { /* entry already exists */
@@ -684,17 +690,16 @@ asmlinkage long sys_mq_open(const char _
error = -EEXIST;
if (oflag & O_EXCL)
goto out;
- filp = do_open(dentry, oflag);
+ filp = do_open(&lookup, oflag);
} else {
- filp = do_create(mqueue_mnt->mnt_root, dentry,
- oflag, mode, u_attr);
+ filp = do_create(&lookup, oflag, mode, u_attr);
}
} else {
error = -ENOENT;
if (!dentry->d_inode)
goto out;
audit_inode(name, dentry->d_inode);
- filp = do_open(dentry, oflag);
+ filp = do_open(&lookup, oflag);
}
if (IS_ERR(filp)) {
On Wed, Aug 08, 2007 at 07:16:30PM +0200, Andreas Gruenbacher wrote:
> Never pass a NULL vfs_lookup to vfs_create() or iop->create by creating
> temporary vfs_lookup objects where needed. Remove the obsolete NULL
> checks.
ipc/mqueue.c should be rewritten to use vfs_path_lookup instead.
nfsd is special - people need to deal with the !nameidata or !intent
case anyway because the fs/exportfs/expfs.c code uses lookup_one_len
all over when the filesystem is mared nfs exportable, but fortunately
the filesystem that are nfs exportable are usually not those using
intents. We might eventually fix this one day by having some kind
open by handle intent, but that should come as part of a large
patch series where we qualify the type of lookup for every single
lookup.