cifs and win2k8

Hello,

This is my first time posting to this list, and I hope I am not breaking any rules.

I have been attempting to get the cifs kernel modules to mount the win2k8 dfs at my workplace. I did finally manage to get
the dfs to mount after modifying the kernel module with this small patch

(quoted diff) +1 -0

@@ -3109,6 +3112,7 @@
 cifs_sb, false);
 if (!refrc) {
 referral_walks_count++;
+ rc=0;
 goto try_mount_again;
 }
 }

Although I dont 100% understand what I did, as I had a hard time following the goto statements, I have a feeling that on initial setup it would follow the DFS referral, however ultimately fail as it exited because rc was still set to -6.

So now For my second problem.

Within the dfs tree there is a referral to //server/share$/folder. My user has access to //server/share$/folder, however do not have access to //server/$share. While following the referral the cifs module does a QUERY_PATH_INFO on \\server\share$\folder, and then is followed up by a QUERY_PATH_INFO on \\server\share$\ which rightfully returns a STATUS_ACCESS_DENIED from the server.

Traversing the same path with smbclient \\server\share$ is never queried.Perhaps I am doing something wrong?

I have been attempting to connect to this dfs from linux for about 2 years,and this is the closest I have gotten, so any help would be appreciated.

terry

I have been attempting to get the cifs kernel modules to mount the
win2k8 dfs at my workplace. I did finally manage to get
the dfs to mount after modifying the kernel module with this small patch

Within the dfs tree there is a referral to //server/share$/folder. My user has access to //server/share$/folder, however do not have access to
//server/$share. While following the referral the cifs module does a QUERY_PATH_INFO on \\server\share$\folder, and then is followed up by a QUERY_PATH_INFO on \\server\share$\ which rightfully returns a STATUS_ACCESS_DENIED from the server.

Hello Terrence,

Quoted 3 lines · Terrence Shipclark

I have been attempting to get the cifs kernel modules to mount the
win2k8 dfs at my workplace. I did finally manage to get
the dfs to mount after modifying the kernel module with this small patch

I believe this should already be fixed (in a more correct fashion) in >= 3.0.

Quoted 7 lines

Within the dfs tree there is a referral to //server/share$/folder. My user has access to //server/share$/folder, however do not have access to
//server/$share. While following the referral the cifs module does a QUERY_PATH_INFO on \\server\share$\folder, and then is followed up by a QUERY_PATH_INFO on \\server\share$\ which rightfully returns a STATUS_ACCESS_DENIED from the server.

I've also noticed that the mounting process seems to query more than it should, but I've taken the easy way out and asked the local windows admin to add "traverse filesystem" and "read attributes" permissions on the parent folders, which should be enough. I suspect that there is a "real fix" to avoid this, but haven't looked too hard myself.

Quoted 17 lines · On Sun, 23 Oct 2011 12:30:03 +0200 sean finney…

Hello Terrence,

Quoted 3 lines · Terrence Shipclark

I have been attempting to get the cifs kernel modules to mount the win2k8 dfs at my workplace.   I did finally manage to get the dfs to mount after modifying the kernel module with this small patch

I believe this should already be fixed (in a more correct fashion) in >= 3.0.

Quoted 7 lines

Within the dfs tree there is a referral to //server/share$/folder. My user has access to //server/share$/folder, however do not have access to
//server/$share.  While following the referral the cifs module does a QUERY_PATH_INFO on \\server\share$\folder, and then is followed up by a QUERY_PATH_INFO on \\server\share$\ which rightfully returns a STATUS_ACCESS_DENIED from the server.

Quoted 7 lines

I've also noticed that the mounting process seems to query more than it should, but I've taken the easy way out and asked the local windows admin to add "traverse filesystem" and "read attributes" permissions on the parent folders, which should be enough.  I suspect that there is a "real fix" to avoid this, but haven't looked too hard myself.

I believe this is a known bug...

3.0 got a patchset to allow superblocks to be shared, but that regressed the above use case (mounting a directory that you have access to but not the stuff below it).

Pavel had some patches that he was going to propose to fix it, but discussion of them sort of fell off. Pavel, did you ever come up with something that worked here?

3.0 got a patchset to allow superblocks to be shared, but that regressed the above use case (mounting a directory that you have access to but not the stuff below it).

Also, if you are interested, I can send you the right version of the patch I've already posted.

Sorry for so long delay on this. Here is the fixed version of the patch:

[PATCH] CIFS: Fix NT_STATUS_ACCESS_DENIED for mounts with prefixpath option

but I found out one use case where it doesn't work right:

1) mount //server/share/a/b/ test
2) mount //server/share/ test2
3) stat test2/a/b - return wrong inode number (autodisable server inode number happens)

In the same time if I add "ls test2" before step #3 it works! May be VFS can't handle negative dentry with children in this case, but I am not sure.

Reorganize code and make it send qpath info request only for a full path (//server/share/prefixpath) rather than request for every path compoment. In this case we end up with negative dentries for all components except full one and we will instantiate them later if such a mount is requested.

Signed-off-by Pavel Shilovsky <[email protected]>

---
 fs/cifs/cifsfs.c   |  138 +++++++++++++++++++++++++++++++-------------------- fs/cifs/cifsfs.h   |    3 +-
 fs/cifs/cifsglob.h |    6 ++
 fs/cifs/dir.c      |    3 +-
 fs/cifs/inode.c    |    7 ++-
 fs/cifs/readdir.c  |   43 ++++++++++------

 6 files changed, 124 insertions(+), 76 deletions(-)

fs/cifs/cifsfs.c +4 -5

index 8f1fe32..a188be1 100644
@@ -90,14 +90,13 @@ extern mempool_t *cifs_sm_req_poolp;
  extern mempool_t *cifs_req_poolp;
  extern mempool_t *cifs_mid_poolp;
 
-static int
+static void
  cifs_read_super(struct super_block *sb)
  {
-       struct inode *inode;
-       struct cifs_sb_info *cifs_sb;
-       int rc = 0;
+       struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
 
-       cifs_sb = CIFS_SB(sb);
+       /* BB should we make this contingent on mount parm? */
+       sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
 
        if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIXACL)
                sb->s_flags |= MS_POSIXACL;

@@ -115,26 +114,6 @@ cifs_read_super(struct super_block *sb)
        sb->s_bdi = &cifs_sb->bdi;
        sb->s_blocksize = CIFS_MAX_MSGSIZE;

       sb->s_blocksize_bits = 14;      /* default 2**14 = CIFS_MAX_MSGSIZE */
-       inode = cifs_root_iget(sb);

fs/cifs/cifsfs.c +0 -19

-
-       if (IS_ERR(inode)) {
-               rc = PTR_ERR(inode);
-               inode = NULL;
-               goto out_no_root;
-       }
-
-       sb->s_root = d_alloc_root(inode);
-
-       if (!sb->s_root) {
-               rc = -ENOMEM;
-               goto out_no_root;
-       }
-
-       /* do that *after* d_alloc_root() - we want NULL ->d_op for root here */
-       if (cifs_sb_master_tcon(cifs_sb)->nocase)
-               sb->s_d_op = &cifs_ci_dentry_ops;
-       else
-               sb->s_d_op = &cifs_dentry_ops;

 #ifdef CONFIG_CIFS_NFSD_EXPORT
       if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) {

fs/cifs/cifsfs.c +12 -8

@@ -143,14 +122,7 @@ cifs_read_super(struct super_block *sb)
        }
  #endif /* CONFIG_CIFS_NFSD_EXPORT */
 
-       return 0;
-
-out_no_root:
-       cERROR(1, "cifs_read_super: get root inode failed");
-       if (inode)
-               iput(inode);
-
-       return rc;
+       sb->s_flags |= MS_ACTIVE;
  }
 
  static void cifs_kill_sb(struct super_block *sb)

@@ -528,6 +500,17 @@ static const struct super_operations cifs_super_ops = {
  #endif
  };
 
+static struct dentry *
+cifs_alloc_root(struct super_block *sb)
+{
+       struct qstr q;
+
+       q.name = "/";
+       q.len = 1;
+       q.hash = full_name_hash(q.name, q.len);
+       return d_alloc_pseudo(sb, &q);
+}
+
  /*
  * Get root dentry from superblock according to prefix path mount option.
  * Return dentry with refcount + 1 on success and NULL otherwise.

@@ -535,8 +518,10 @@ static const struct super_operations cifs_super_ops = {

 static struct dentry *
 cifs_get_root(struct smb_vol *vol, struct super_block *sb) {
-       struct dentry *dentry;
+       struct dentry *dentry, *found;
+       struct inode *inode;
       struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
+       struct qstr q;
       char *full_path = NULL;
       char *s, *p;
       char sep;

fs/cifs/cifsfs.c +65 -20

@@ -548,20 +533,30 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb)
 
        cFYI(1, "Get root dentry for %s", full_path);
 
+       if (!sb->s_root) {
+               sb->s_root = cifs_alloc_root(sb);
+               if (IS_ERR(sb->s_root)) {
+                       dentry = ERR_PTR(-ENOMEM);
+                       goto out;
+               }
+
+               /*
+                * do that *after* cifs_alloc_root() - we want NULL ->d_op for
+                * root here
+                */
+               if (cifs_sb_master_tcon(cifs_sb)->nocase)
+                       sb->s_d_op = &cifs_ci_dentry_ops;
+               else
+                       sb->s_d_op = &cifs_dentry_ops;
+       }
+
        sep = CIFS_DIR_SEP(cifs_sb);
        dentry = dget(sb->s_root);
        p = s = full_path;
 
        do {
-               struct inode *dir = dentry->d_inode;
                struct dentry *child;
 
-               if (!dir) {
-                       dput(dentry);
-                       dentry = ERR_PTR(-ENOENT);
-                       break;
-               }
-
                /* skip separators */
                while (*s == sep)
                        s++;

@@ -572,12 +567,55 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb)
                while (*s && *s != sep)
                        s++;
 
-               mutex_lock(&dir->i_mutex);
-               child = lookup_one_len(p, dentry, s - p);
-               mutex_unlock(&dir->i_mutex);
+               q.name = p;
+               q.len = s - p;
+               if (dentry->d_flags & DCACHE_OP_HASH)
+                       dentry->d_op->d_hash(dentry, dentry->d_inode, &q);
+               else
+                       q.hash = full_name_hash(q.name, q.len);
+
+               child = d_lookup(dentry, &q);
+               if (!child) {
+                       child = d_alloc(dentry, &q);
+                       if (IS_ERR(child)) {
+                               dput(dentry);
+                               dentry = ERR_CAST(child);
+                               break;
+                       }
+                       d_rehash(child);
+               }
                dput(dentry);
                dentry = child;
        } while (!IS_ERR(dentry));
+
+       if (IS_ERR(dentry))
+               goto out;
+
+       mutex_lock(&cifs_root_mutex);
+
+       if (dentry->d_parent->d_inode)
+               mutex_lock(&dentry->d_parent->d_inode->i_mutex);
+
+       if (!dentry->d_inode) {
+               inode = cifs_mntroot_iget(sb, full_path);
+               if (IS_ERR(inode)) {
+                       dput(dentry);
+                       dentry = ERR_CAST(inode);
+                       goto out;
+               }
+
+               found = d_instantiate_unique(dentry, inode);
+               if (found) {
+                       dput(dentry);
+                       dentry = found;
+               }
+       }
+
+       if (dentry->d_parent->d_inode)
+               mutex_unlock(&dentry->d_parent->d_inode->i_mutex);
+
+       mutex_unlock(&cifs_root_mutex);
+out:
        kfree(full_path);
        return dentry;
  }

@@ -644,16 +682,7 @@ cifs_do_mount(struct file_system_type *fs_type,
                cifs_umount(cifs_sb);
        } else {
                sb->s_flags = flags;
-               /* BB should we make this contingent on mount parm? */
-               sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
-
-               rc = cifs_read_super(sb);
-               if (rc) {
-                       root = ERR_PTR(rc);
-                       goto out_super;
-               }
-
-               sb->s_flags |= MS_ACTIVE;
+               cifs_read_super(sb);
        }
 
        root = cifs_get_root(volume_info, sb);

@@ -1112,6 +1141,7 @@ init_cifs(void)
        spin_lock_init(&cifs_tcp_ses_lock);
        spin_lock_init(&cifs_file_list_lock);
        spin_lock_init(&GlobalMid_Lock);
+       mutex_init(&cifs_root_mutex);
 
        if (cifs_max_pending < 2) {
                cifs_max_pending = 2;

fs/cifs/cifsfs.h +2 -1

index 30ff560..cf28150 100644
@@ -43,7 +43,8 @@ extern const struct address_space_operations cifs_addr_ops_smallbuf;
 
  /* Functions related to inodes */
  extern const struct inode_operations cifs_dir_inode_ops;
-extern struct inode *cifs_root_iget(struct super_block *);
+extern struct inode *cifs_mntroot_iget(struct super_block *sb,
+                                      const char *full_path);
  extern int cifs_create(struct inode *, struct dentry *, int,
                       struct nameidata *);
  extern struct dentry *cifs_lookup(struct inode *, struct dentry *,

fs/cifs/cifsglob.h +6 -0

index 8238aa1..67f6852 100644
@@ -958,6 +958,12 @@ GLOBAL_EXTERN spinlock_t           cifs_tcp_ses_lock;
  */
  GLOBAL_EXTERN spinlock_t       cifs_file_list_lock;
 
+/*
+ * This lock protects root dentry in cifs_get_root from being instantiated
+ * twice.
+ */
+GLOBAL_EXTERN struct mutex cifs_root_mutex;
+
  #ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* unused temporarily */
  /* Outstanding dir notify requests */
  GLOBAL_EXTERN struct list_head GlobalDnotifyReqList;

fs/cifs/dir.c +0 -0

index d7eeb9d..d6b3319 100644

fs/cifs/dir.c +0 -0

                d_add(direntry, newInode);
                if (posix_open) {
                        filp = lookup_instantiate_filp(nd, direntry,

fs/cifs/inode.c +2 -2

index 2c50bd2..7b7fccd 100644
@@ -878,7 +878,7 @@ retry_iget5_locked:
  }
 
  /* gets root inode */
-struct inode *cifs_root_iget(struct super_block *sb)
+struct inode *cifs_mntroot_iget(struct super_block *sb, const char *full_path)
  {
        int xid;
        struct cifs_sb_info *cifs_sb = CIFS_SB(sb);

@@ -888,9 +888,10 @@ struct inode *cifs_root_iget(struct super_block *sb)
 
        xid = GetXid();
        if (tcon->unix_ext)
-               rc = cifs_get_inode_info_unix(&inode, "", sb, xid);
+               rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid);

       else
-               rc = cifs_get_inode_info(&inode, "", NULL, sb, xid, NULL);
+               rc = cifs_get_inode_info(&inode, full_path, NULL, sb, xid,
+                                        NULL);

       if (!inode) {
               inode = ERR_PTR(rc);

fs/cifs/readdir.c +2 -4

index 5de03ec..b2f8851 100644
@@ -65,10 +65,8 @@ static inline void dump_cifs_file_struct(struct file *file, char *label)
  }
  #endif /* DEBUG2 */
 
-/*
- * Find the dentry that matches "name". If there isn't one, create one. If it's
- * a negative dentry or the uniqueid changed, then drop it and recreate it.
- */
+
+/* Find the dentry that matches "name". If there isn't one, create one. */

 static struct dentry *
 cifs_readdir_lookup(struct dentry *parent, struct qstr *name, struct cifs_fattr *fattr)

fs/cifs/readdir.c +0 -0

@@ -84,32 +82,38 @@ cifs_readdir_lookup(struct dentry *parent, struct qstr *name,

       else
               name->hash = full_name_hash(name->name, name->len);

+       mutex_lock(&cifs_root_mutex);
+
       dentry = d_lookup(parent, name);
       if (dentry) {
               /* FIXME: check for inode number changes? */ if (dentry->d_inode != NULL)
-                       return dentry;
-               d_drop(dentry);
-               dput(dentry);
+                       goto out;
+       } else {
+               dentry = d_alloc(parent, name);
+               if (dentry == NULL)
+                       goto out;
+               d_rehash(dentry);
       }

-       dentry = d_alloc(parent, name);
-       if (dentry == NULL)
-               return NULL;
-
       inode = cifs_iget(sb, fattr);
       if (!inode) {
               dput(dentry);
-               return NULL;
+               dentry = NULL;
+               goto out;
       }

-       alias = d_materialise_unique(dentry, inode);
+       alias = d_instantiate_unique(dentry, inode); if (alias != NULL) {
               dput(dentry);
-               if (IS_ERR(alias))
-                       return NULL;
+               if (IS_ERR(alias)) {
+                       dentry = NULL;
+                       goto out;
+               }
               dentry = alias;
       }
+out:
+       mutex_unlock(&cifs_root_mutex);

       return dentry;
 }

fs/cifs/readdir.c +1 -0

@@ -708,6 +712,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
        char *tmp_buf = NULL;
        char *end_of_smb;
        unsigned int max_len;
+       ino_t ino;
 
        xid = GetXid();
 

@@ -732,8 +737,12 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
                }
                file->f_pos++;

       case 1:
-               if (filldir(direntry, "..", 2, file->f_pos,
-                    parent_ino(file->f_path.dentry), DT_DIR) < 0) {
+               if (!file->f_path.dentry->d_parent->d_inode) {
+                       cFYI(1, "parent dir is negative, filling as current");
+                       ino = file->f_path.dentry->d_inode->i_ino;
+               } else
+                       ino = parent_ino(file->f_path.dentry);
+               if (filldir(direntry, "..", 2, file->f_pos, ino, DT_DIR) < 0) { cERROR(1, "Filldir for parent dir failed"); rc = -ENOMEM;
                       break;
--
1.7.1

2011/11/1 Pavel Shilovsky <[email protected]>:

Quoted 284 lines

Sorry for so long delay on this. Here is the fixed version of the patch:

[PATCH] CIFS: Fix NT_STATUS_ACCESS_DENIED for mounts with prefixpath option

but I found out one use case where it doesn't work right:

1) mount //server/share/a/b/ test
2) mount //server/share/ test2
3) stat test2/a/b - return wrong inode number (autodisable server inode number happens)

In the same time if I add "ls test2" before step #3 it works! May be VFS can't handle negative dentry with children in this case, but I am not sure.

Reorganize code and make it send qpath info request only for a full path (//server/share/prefixpath) rather than request for every path compoment. In this case we end up with negative dentries for all components except full one and we will instantiate them later if such a mount is requested.

Signed-off-by Pavel Shilovsky <[email protected]>

---
 fs/cifs/cifsfs.c   |  138 +++++++++++++++++++++++++++++++-------------------- fs/cifs/cifsfs.h   |    3 +-
 fs/cifs/cifsglob.h |    6 ++
 fs/cifs/dir.c      |    3 +-
 fs/cifs/inode.c    |    7 ++-
 fs/cifs/readdir.c  |   43 ++++++++++------

 6 files changed, 124 insertions(+), 76 deletions(-)

fs/cifs/cifsfs.c +4 -5

index 8f1fe32..a188be1 100644
@@ -90,14 +90,13 @@ extern mempool_t *cifs_sm_req_poolp;
  extern mempool_t *cifs_req_poolp;
  extern mempool_t *cifs_mid_poolp;
 
-static int
+static void
  cifs_read_super(struct super_block *sb)
  {
-       struct inode *inode;
-       struct cifs_sb_info *cifs_sb;
-       int rc = 0;
+       struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
 
-       cifs_sb = CIFS_SB(sb);
+       /* BB should we make this contingent on mount parm? */
+       sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
 
        if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIXACL)
                sb->s_flags |= MS_POSIXACL;

@@ -115,26 +114,6 @@ cifs_read_super(struct super_block *sb)
        sb->s_bdi = &cifs_sb->bdi;
        sb->s_blocksize = CIFS_MAX_MSGSIZE;

       sb->s_blocksize_bits = 14;      /* default 2**14 = CIFS_MAX_MSGSIZE */
-       inode = cifs_root_iget(sb);

fs/cifs/cifsfs.c +0 -19

-
-       if (IS_ERR(inode)) {
-               rc = PTR_ERR(inode);
-               inode = NULL;
-               goto out_no_root;
-       }
-
-       sb->s_root = d_alloc_root(inode);
-
-       if (!sb->s_root) {
-               rc = -ENOMEM;
-               goto out_no_root;
-       }
-
-       /* do that *after* d_alloc_root() - we want NULL ->d_op for root here */
-       if (cifs_sb_master_tcon(cifs_sb)->nocase)
-               sb->s_d_op = &cifs_ci_dentry_ops;
-       else
-               sb->s_d_op = &cifs_dentry_ops;

 #ifdef CONFIG_CIFS_NFSD_EXPORT
       if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) {

fs/cifs/cifsfs.c +12 -8

@@ -143,14 +122,7 @@ cifs_read_super(struct super_block *sb)
        }
  #endif /* CONFIG_CIFS_NFSD_EXPORT */
 
-       return 0;
-
-out_no_root:
-       cERROR(1, "cifs_read_super: get root inode failed");
-       if (inode)
-               iput(inode);
-
-       return rc;
+       sb->s_flags |= MS_ACTIVE;
  }
 
  static void cifs_kill_sb(struct super_block *sb)

@@ -528,6 +500,17 @@ static const struct super_operations cifs_super_ops = {
  #endif
  };
 
+static struct dentry *
+cifs_alloc_root(struct super_block *sb)
+{
+       struct qstr q;
+
+       q.name = "/";
+       q.len = 1;
+       q.hash = full_name_hash(q.name, q.len);
+       return d_alloc_pseudo(sb, &q);
+}
+
  /*
  * Get root dentry from superblock according to prefix path mount option.
  * Return dentry with refcount + 1 on success and NULL otherwise.

@@ -535,8 +518,10 @@ static const struct super_operations cifs_super_ops = {

 static struct dentry *
 cifs_get_root(struct smb_vol *vol, struct super_block *sb) {
-       struct dentry *dentry;
+       struct dentry *dentry, *found;
+       struct inode *inode;
       struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
+       struct qstr q;
       char *full_path = NULL;
       char *s, *p;
       char sep;

fs/cifs/cifsfs.c +65 -20

@@ -548,20 +533,30 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb)
 
        cFYI(1, "Get root dentry for %s", full_path);
 
+       if (!sb->s_root) {
+               sb->s_root = cifs_alloc_root(sb);
+               if (IS_ERR(sb->s_root)) {
+                       dentry = ERR_PTR(-ENOMEM);
+                       goto out;
+               }
+
+               /*
+                * do that *after* cifs_alloc_root() - we want NULL ->d_op for
+                * root here
+                */
+               if (cifs_sb_master_tcon(cifs_sb)->nocase)
+                       sb->s_d_op = &cifs_ci_dentry_ops;
+               else
+                       sb->s_d_op = &cifs_dentry_ops;
+       }
+
        sep = CIFS_DIR_SEP(cifs_sb);
        dentry = dget(sb->s_root);
        p = s = full_path;
 
        do {
-               struct inode *dir = dentry->d_inode;
                struct dentry *child;
 
-               if (!dir) {
-                       dput(dentry);
-                       dentry = ERR_PTR(-ENOENT);
-                       break;
-               }
-
                /* skip separators */
                while (*s == sep)
                        s++;

@@ -572,12 +567,55 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb)
                while (*s && *s != sep)
                        s++;
 
-               mutex_lock(&dir->i_mutex);
-               child = lookup_one_len(p, dentry, s - p);
-               mutex_unlock(&dir->i_mutex);
+               q.name = p;
+               q.len = s - p;
+               if (dentry->d_flags & DCACHE_OP_HASH)
+                       dentry->d_op->d_hash(dentry, dentry->d_inode, &q);
+               else
+                       q.hash = full_name_hash(q.name, q.len);
+
+               child = d_lookup(dentry, &q);
+               if (!child) {
+                       child = d_alloc(dentry, &q);
+                       if (IS_ERR(child)) {
+                               dput(dentry);
+                               dentry = ERR_CAST(child);
+                               break;
+                       }
+                       d_rehash(child);
+               }
                dput(dentry);
                dentry = child;
        } while (!IS_ERR(dentry));
+
+       if (IS_ERR(dentry))
+               goto out;
+
+       mutex_lock(&cifs_root_mutex);
+
+       if (dentry->d_parent->d_inode)
+               mutex_lock(&dentry->d_parent->d_inode->i_mutex);
+
+       if (!dentry->d_inode) {
+               inode = cifs_mntroot_iget(sb, full_path);
+               if (IS_ERR(inode)) {
+                       dput(dentry);
+                       dentry = ERR_CAST(inode);
+                       goto out;
+               }
+
+               found = d_instantiate_unique(dentry, inode);
+               if (found) {
+                       dput(dentry);
+                       dentry = found;
+               }
+       }
+
+       if (dentry->d_parent->d_inode)
+               mutex_unlock(&dentry->d_parent->d_inode->i_mutex);
+
+       mutex_unlock(&cifs_root_mutex);
+out:
        kfree(full_path);
        return dentry;
  }

@@ -644,16 +682,7 @@ cifs_do_mount(struct file_system_type *fs_type,
                cifs_umount(cifs_sb);
        } else {
                sb->s_flags = flags;
-               /* BB should we make this contingent on mount parm? */
-               sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
-
-               rc = cifs_read_super(sb);
-               if (rc) {
-                       root = ERR_PTR(rc);
-                       goto out_super;
-               }
-
-               sb->s_flags |= MS_ACTIVE;
+               cifs_read_super(sb);
        }
 
        root = cifs_get_root(volume_info, sb);

@@ -1112,6 +1141,7 @@ init_cifs(void)
        spin_lock_init(&cifs_tcp_ses_lock);
        spin_lock_init(&cifs_file_list_lock);
        spin_lock_init(&GlobalMid_Lock);
+       mutex_init(&cifs_root_mutex);
 
        if (cifs_max_pending < 2) {
                cifs_max_pending = 2;

fs/cifs/cifsfs.h +2 -1

index 30ff560..cf28150 100644
@@ -43,7 +43,8 @@ extern const struct address_space_operations cifs_addr_ops_smallbuf;
 
  /* Functions related to inodes */
  extern const struct inode_operations cifs_dir_inode_ops;
-extern struct inode *cifs_root_iget(struct super_block *);
+extern struct inode *cifs_mntroot_iget(struct super_block *sb,
+                                      const char *full_path);
  extern int cifs_create(struct inode *, struct dentry *, int,
                       struct nameidata *);
  extern struct dentry *cifs_lookup(struct inode *, struct dentry *,

fs/cifs/cifsglob.h +6 -0

index 8238aa1..67f6852 100644
@@ -958,6 +958,12 @@ GLOBAL_EXTERN spinlock_t           cifs_tcp_ses_lock;
  */
  GLOBAL_EXTERN spinlock_t       cifs_file_list_lock;
 
+/*
+ * This lock protects root dentry in cifs_get_root from being instantiated
+ * twice.
+ */
+GLOBAL_EXTERN struct mutex cifs_root_mutex;
+
  #ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* unused temporarily */
  /* Outstanding dir notify requests */
  GLOBAL_EXTERN struct list_head GlobalDnotifyReqList;

fs/cifs/dir.c +0 -0

index d7eeb9d..d6b3319 100644

fs/cifs/dir.c cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, · line 556

Quoted 4 lines

fs/cifs/dir.c +1 -0

@@ -553,6 +553,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry,
 
        if (direntry->d_inode != NULL) {
                cFYI(1, "non-NULL inode in lookup");
+               newInode = direntry->d_inode;

this and ...

fs/cifs/dir.c cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, · line 600

Quoted 9 lines

fs/cifs/dir.c +1 -1

        } else {
                cFYI(1, "NULL inode in lookup");
        }

@@ -596,7 +597,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry,
                rc = cifs_get_inode_info(&newInode, full_path, NULL,
                                parent_dir_inode->i_sb, xid, NULL);
 
-       if ((rc == 0) && (newInode != NULL)) {
+       if ((rc == 0) && (newInode != NULL) && !direntry->d_inode) {

... this doesn't make sense - I made it for testing purpose only and forgot to remove. Since this patch isn't going to be a candidate to merge, seems no need to respin it.

Quoted 115 lines

fs/cifs/dir.c +0 -0

                d_add(direntry, newInode);
                if (posix_open) {
                        filp = lookup_instantiate_filp(nd, direntry,

fs/cifs/inode.c +2 -2

index 2c50bd2..7b7fccd 100644
@@ -878,7 +878,7 @@ retry_iget5_locked:
  }
 
  /* gets root inode */
-struct inode *cifs_root_iget(struct super_block *sb)
+struct inode *cifs_mntroot_iget(struct super_block *sb, const char *full_path)
  {
        int xid;
        struct cifs_sb_info *cifs_sb = CIFS_SB(sb);

@@ -888,9 +888,10 @@ struct inode *cifs_root_iget(struct super_block *sb)
 
        xid = GetXid();
        if (tcon->unix_ext)
-               rc = cifs_get_inode_info_unix(&inode, "", sb, xid);
+               rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid);

       else
-               rc = cifs_get_inode_info(&inode, "", NULL, sb, xid, NULL);
+               rc = cifs_get_inode_info(&inode, full_path, NULL, sb, xid,
+                                        NULL);

       if (!inode) {
               inode = ERR_PTR(rc);

fs/cifs/readdir.c +2 -4

index 5de03ec..b2f8851 100644
@@ -65,10 +65,8 @@ static inline void dump_cifs_file_struct(struct file *file, char *label)
  }
  #endif /* DEBUG2 */
 
-/*
- * Find the dentry that matches "name". If there isn't one, create one. If it's
- * a negative dentry or the uniqueid changed, then drop it and recreate it.
- */
+
+/* Find the dentry that matches "name". If there isn't one, create one. */

 static struct dentry *
 cifs_readdir_lookup(struct dentry *parent, struct qstr *name, struct cifs_fattr *fattr)

fs/cifs/readdir.c +0 -0

@@ -84,32 +82,38 @@ cifs_readdir_lookup(struct dentry *parent, struct qstr *name,

       else
               name->hash = full_name_hash(name->name, name->len);

+       mutex_lock(&cifs_root_mutex);
+
       dentry = d_lookup(parent, name);
       if (dentry) {
               /* FIXME: check for inode number changes? */ if (dentry->d_inode != NULL)
-                       return dentry;
-               d_drop(dentry);
-               dput(dentry);
+                       goto out;
+       } else {
+               dentry = d_alloc(parent, name);
+               if (dentry == NULL)
+                       goto out;
+               d_rehash(dentry);
       }

-       dentry = d_alloc(parent, name);
-       if (dentry == NULL)
-               return NULL;
-
       inode = cifs_iget(sb, fattr);
       if (!inode) {
               dput(dentry);
-               return NULL;
+               dentry = NULL;
+               goto out;
       }

-       alias = d_materialise_unique(dentry, inode);
+       alias = d_instantiate_unique(dentry, inode); if (alias != NULL) {
               dput(dentry);
-               if (IS_ERR(alias))
-                       return NULL;
+               if (IS_ERR(alias)) {
+                       dentry = NULL;
+                       goto out;
+               }
               dentry = alias;
       }
+out:
+       mutex_unlock(&cifs_root_mutex);

       return dentry;
 }

fs/cifs/readdir.c +1 -0

@@ -708,6 +712,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
        char *tmp_buf = NULL;
        char *end_of_smb;
        unsigned int max_len;
+       ino_t ino;
 
        xid = GetXid();
 

@@ -732,8 +737,12 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
                }
                file->f_pos++;

       case 1:
-               if (filldir(direntry, "..", 2, file->f_pos,
-                    parent_ino(file->f_path.dentry), DT_DIR) < 0) {
+               if (!file->f_path.dentry->d_parent->d_inode) {
+                       cFYI(1, "parent dir is negative, filling as current");
+                       ino = file->f_path.dentry->d_inode->i_ino;
+               } else
+                       ino = parent_ino(file->f_path.dentry);

Quoted 8 lines

+               if (filldir(direntry, "..", 2, file->f_pos, ino, DT_DIR) < 0) { cERROR(1, "Filldir for parent dir failed"); rc = -ENOMEM;
                       break;
--
1.7.1

Quoted 421 lines · On Tue, 1 Nov 2011 22:56:18 +0300 Pavel Shilovsky <[email protected]> wrote:

2011/11/1 Pavel Shilovsky <[email protected]>:

Quoted 284 lines

Sorry for so long delay on this. Here is the fixed version of the patch:

[PATCH] CIFS: Fix NT_STATUS_ACCESS_DENIED for mounts with prefixpath option

but I found out one use case where it doesn't work right:

1) mount //server/share/a/b/ test
2) mount //server/share/ test2
3) stat test2/a/b - return wrong inode number (autodisable server inode number happens)

In the same time if I add "ls test2" before step #3 it works! May be VFS can't handle negative dentry with children in this case, but I am not sure.

Reorganize code and make it send qpath info request only for a full path (//server/share/prefixpath) rather than request for every path compoment. In this case we end up with negative dentries for all components except full one and we will instantiate them later if such a mount is requested.

Signed-off-by Pavel Shilovsky <[email protected]>

---
 fs/cifs/cifsfs.c   |  138 +++++++++++++++++++++++++++++++-------------------- fs/cifs/cifsfs.h   |    3 +-
 fs/cifs/cifsglob.h |    6 ++
 fs/cifs/dir.c      |    3 +-
 fs/cifs/inode.c    |    7 ++-
 fs/cifs/readdir.c  |   43 ++++++++++------

 6 files changed, 124 insertions(+), 76 deletions(-)

fs/cifs/cifsfs.c +4 -5

index 8f1fe32..a188be1 100644
@@ -90,14 +90,13 @@ extern mempool_t *cifs_sm_req_poolp;
  extern mempool_t *cifs_req_poolp;
  extern mempool_t *cifs_mid_poolp;
 
-static int
+static void
  cifs_read_super(struct super_block *sb)
  {
-       struct inode *inode;
-       struct cifs_sb_info *cifs_sb;
-       int rc = 0;
+       struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
 
-       cifs_sb = CIFS_SB(sb);
+       /* BB should we make this contingent on mount parm? */
+       sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
 
        if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIXACL)
                sb->s_flags |= MS_POSIXACL;

@@ -115,26 +114,6 @@ cifs_read_super(struct super_block *sb)
        sb->s_bdi = &cifs_sb->bdi;
        sb->s_blocksize = CIFS_MAX_MSGSIZE;

       sb->s_blocksize_bits = 14;      /* default 2**14 = CIFS_MAX_MSGSIZE */
-       inode = cifs_root_iget(sb);

fs/cifs/cifsfs.c +0 -19

-
-       if (IS_ERR(inode)) {
-               rc = PTR_ERR(inode);
-               inode = NULL;
-               goto out_no_root;
-       }
-
-       sb->s_root = d_alloc_root(inode);
-
-       if (!sb->s_root) {
-               rc = -ENOMEM;
-               goto out_no_root;
-       }
-
-       /* do that *after* d_alloc_root() - we want NULL ->d_op for root here */
-       if (cifs_sb_master_tcon(cifs_sb)->nocase)
-               sb->s_d_op = &cifs_ci_dentry_ops;
-       else
-               sb->s_d_op = &cifs_dentry_ops;

 #ifdef CONFIG_CIFS_NFSD_EXPORT
       if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) {

fs/cifs/cifsfs.c +12 -8

@@ -143,14 +122,7 @@ cifs_read_super(struct super_block *sb)
        }
  #endif /* CONFIG_CIFS_NFSD_EXPORT */
 
-       return 0;
-
-out_no_root:
-       cERROR(1, "cifs_read_super: get root inode failed");
-       if (inode)
-               iput(inode);
-
-       return rc;
+       sb->s_flags |= MS_ACTIVE;
  }
 
  static void cifs_kill_sb(struct super_block *sb)

@@ -528,6 +500,17 @@ static const struct super_operations cifs_super_ops = {
  #endif
  };
 
+static struct dentry *
+cifs_alloc_root(struct super_block *sb)
+{
+       struct qstr q;
+
+       q.name = "/";
+       q.len = 1;
+       q.hash = full_name_hash(q.name, q.len);
+       return d_alloc_pseudo(sb, &q);
+}
+
  /*
  * Get root dentry from superblock according to prefix path mount option.
  * Return dentry with refcount + 1 on success and NULL otherwise.

@@ -535,8 +518,10 @@ static const struct super_operations cifs_super_ops = {

 static struct dentry *
 cifs_get_root(struct smb_vol *vol, struct super_block *sb) {
-       struct dentry *dentry;
+       struct dentry *dentry, *found;
+       struct inode *inode;
       struct cifs_sb_info *cifs_sb = CIFS_SB(sb);
+       struct qstr q;
       char *full_path = NULL;
       char *s, *p;
       char sep;

fs/cifs/cifsfs.c +65 -20

@@ -548,20 +533,30 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb)
 
        cFYI(1, "Get root dentry for %s", full_path);
 
+       if (!sb->s_root) {
+               sb->s_root = cifs_alloc_root(sb);
+               if (IS_ERR(sb->s_root)) {
+                       dentry = ERR_PTR(-ENOMEM);
+                       goto out;
+               }
+
+               /*
+                * do that *after* cifs_alloc_root() - we want NULL ->d_op for
+                * root here
+                */
+               if (cifs_sb_master_tcon(cifs_sb)->nocase)
+                       sb->s_d_op = &cifs_ci_dentry_ops;
+               else
+                       sb->s_d_op = &cifs_dentry_ops;
+       }
+
        sep = CIFS_DIR_SEP(cifs_sb);
        dentry = dget(sb->s_root);
        p = s = full_path;
 
        do {
-               struct inode *dir = dentry->d_inode;
                struct dentry *child;
 
-               if (!dir) {
-                       dput(dentry);
-                       dentry = ERR_PTR(-ENOENT);
-                       break;
-               }
-
                /* skip separators */
                while (*s == sep)
                        s++;

@@ -572,12 +567,55 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb)
                while (*s && *s != sep)
                        s++;
 
-               mutex_lock(&dir->i_mutex);
-               child = lookup_one_len(p, dentry, s - p);
-               mutex_unlock(&dir->i_mutex);
+               q.name = p;
+               q.len = s - p;
+               if (dentry->d_flags & DCACHE_OP_HASH)
+                       dentry->d_op->d_hash(dentry, dentry->d_inode, &q);
+               else
+                       q.hash = full_name_hash(q.name, q.len);
+
+               child = d_lookup(dentry, &q);
+               if (!child) {
+                       child = d_alloc(dentry, &q);
+                       if (IS_ERR(child)) {
+                               dput(dentry);
+                               dentry = ERR_CAST(child);
+                               break;
+                       }
+                       d_rehash(child);
+               }
                dput(dentry);
                dentry = child;
        } while (!IS_ERR(dentry));
+
+       if (IS_ERR(dentry))
+               goto out;
+
+       mutex_lock(&cifs_root_mutex);
+
+       if (dentry->d_parent->d_inode)
+               mutex_lock(&dentry->d_parent->d_inode->i_mutex);
+
+       if (!dentry->d_inode) {
+               inode = cifs_mntroot_iget(sb, full_path);
+               if (IS_ERR(inode)) {
+                       dput(dentry);
+                       dentry = ERR_CAST(inode);
+                       goto out;
+               }
+
+               found = d_instantiate_unique(dentry, inode);
+               if (found) {
+                       dput(dentry);
+                       dentry = found;
+               }
+       }
+
+       if (dentry->d_parent->d_inode)
+               mutex_unlock(&dentry->d_parent->d_inode->i_mutex);
+
+       mutex_unlock(&cifs_root_mutex);
+out:
        kfree(full_path);
        return dentry;
  }

@@ -644,16 +682,7 @@ cifs_do_mount(struct file_system_type *fs_type,
                cifs_umount(cifs_sb);
        } else {
                sb->s_flags = flags;
-               /* BB should we make this contingent on mount parm? */
-               sb->s_flags |= MS_NODIRATIME | MS_NOATIME;
-
-               rc = cifs_read_super(sb);
-               if (rc) {
-                       root = ERR_PTR(rc);
-                       goto out_super;
-               }
-
-               sb->s_flags |= MS_ACTIVE;
+               cifs_read_super(sb);
        }
 
        root = cifs_get_root(volume_info, sb);

@@ -1112,6 +1141,7 @@ init_cifs(void)
        spin_lock_init(&cifs_tcp_ses_lock);
        spin_lock_init(&cifs_file_list_lock);
        spin_lock_init(&GlobalMid_Lock);
+       mutex_init(&cifs_root_mutex);
 
        if (cifs_max_pending < 2) {
                cifs_max_pending = 2;

fs/cifs/cifsfs.h +2 -1

index 30ff560..cf28150 100644
@@ -43,7 +43,8 @@ extern const struct address_space_operations cifs_addr_ops_smallbuf;
 
  /* Functions related to inodes */
  extern const struct inode_operations cifs_dir_inode_ops;
-extern struct inode *cifs_root_iget(struct super_block *);
+extern struct inode *cifs_mntroot_iget(struct super_block *sb,
+                                      const char *full_path);
  extern int cifs_create(struct inode *, struct dentry *, int,
                       struct nameidata *);
  extern struct dentry *cifs_lookup(struct inode *, struct dentry *,

fs/cifs/cifsglob.h +6 -0

index 8238aa1..67f6852 100644
@@ -958,6 +958,12 @@ GLOBAL_EXTERN spinlock_t           cifs_tcp_ses_lock;
  */
  GLOBAL_EXTERN spinlock_t       cifs_file_list_lock;
 
+/*
+ * This lock protects root dentry in cifs_get_root from being instantiated
+ * twice.
+ */
+GLOBAL_EXTERN struct mutex cifs_root_mutex;
+
  #ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* unused temporarily */
  /* Outstanding dir notify requests */
  GLOBAL_EXTERN struct list_head GlobalDnotifyReqList;

fs/cifs/dir.c +0 -0

index d7eeb9d..d6b3319 100644

fs/cifs/dir.c cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, · line 556

Quoted 4 lines

fs/cifs/dir.c +1 -0

@@ -553,6 +553,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry,
 
        if (direntry->d_inode != NULL) {
                cFYI(1, "non-NULL inode in lookup");
+               newInode = direntry->d_inode;

this and ...

fs/cifs/dir.c cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, · line 600

Quoted 9 lines

fs/cifs/dir.c +1 -1

        } else {
                cFYI(1, "NULL inode in lookup");
        }

@@ -596,7 +597,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry,
                rc = cifs_get_inode_info(&newInode, full_path, NULL,
                                parent_dir_inode->i_sb, xid, NULL);
 
-       if ((rc == 0) && (newInode != NULL)) {
+       if ((rc == 0) && (newInode != NULL) && !direntry->d_inode) {

... this doesn't make sense - I made it for testing purpose only and forgot to remove. Since this patch isn't going to be a candidate to merge, seems no need to respin it.

Quoted 115 lines

fs/cifs/dir.c +0 -0

                d_add(direntry, newInode);
                if (posix_open) {
                        filp = lookup_instantiate_filp(nd, direntry,

fs/cifs/inode.c +2 -2

index 2c50bd2..7b7fccd 100644
@@ -878,7 +878,7 @@ retry_iget5_locked:
  }
 
  /* gets root inode */
-struct inode *cifs_root_iget(struct super_block *sb)
+struct inode *cifs_mntroot_iget(struct super_block *sb, const char *full_path)
  {
        int xid;
        struct cifs_sb_info *cifs_sb = CIFS_SB(sb);

@@ -888,9 +888,10 @@ struct inode *cifs_root_iget(struct super_block *sb)
 
        xid = GetXid();
        if (tcon->unix_ext)
-               rc = cifs_get_inode_info_unix(&inode, "", sb, xid);
+               rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid);

       else
-               rc = cifs_get_inode_info(&inode, "", NULL, sb, xid, NULL);
+               rc = cifs_get_inode_info(&inode, full_path, NULL, sb, xid,
+                                        NULL);

       if (!inode) {
               inode = ERR_PTR(rc);

fs/cifs/readdir.c +2 -4

index 5de03ec..b2f8851 100644
@@ -65,10 +65,8 @@ static inline void dump_cifs_file_struct(struct file *file, char *label)
  }
  #endif /* DEBUG2 */
 
-/*
- * Find the dentry that matches "name". If there isn't one, create one. If it's
- * a negative dentry or the uniqueid changed, then drop it and recreate it.
- */
+
+/* Find the dentry that matches "name". If there isn't one, create one. */

 static struct dentry *
 cifs_readdir_lookup(struct dentry *parent, struct qstr *name, struct cifs_fattr *fattr)

fs/cifs/readdir.c +0 -0

@@ -84,32 +82,38 @@ cifs_readdir_lookup(struct dentry *parent, struct qstr *name,

       else
               name->hash = full_name_hash(name->name, name->len);

+       mutex_lock(&cifs_root_mutex);
+
       dentry = d_lookup(parent, name);
       if (dentry) {
               /* FIXME: check for inode number changes? */ if (dentry->d_inode != NULL)
-                       return dentry;
-               d_drop(dentry);
-               dput(dentry);
+                       goto out;
+       } else {
+               dentry = d_alloc(parent, name);
+               if (dentry == NULL)
+                       goto out;
+               d_rehash(dentry);
       }

-       dentry = d_alloc(parent, name);
-       if (dentry == NULL)
-               return NULL;
-
       inode = cifs_iget(sb, fattr);
       if (!inode) {
               dput(dentry);
-               return NULL;
+               dentry = NULL;
+               goto out;
       }

-       alias = d_materialise_unique(dentry, inode);
+       alias = d_instantiate_unique(dentry, inode); if (alias != NULL) {
               dput(dentry);
-               if (IS_ERR(alias))
-                       return NULL;
+               if (IS_ERR(alias)) {
+                       dentry = NULL;
+                       goto out;
+               }
               dentry = alias;
       }
+out:
+       mutex_unlock(&cifs_root_mutex);

       return dentry;
 }

fs/cifs/readdir.c +1 -0

@@ -708,6 +712,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
        char *tmp_buf = NULL;
        char *end_of_smb;
        unsigned int max_len;
+       ino_t ino;
 
        xid = GetXid();
 

@@ -732,8 +737,12 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir)
                }
                file->f_pos++;

       case 1:
-               if (filldir(direntry, "..", 2, file->f_pos,
-                    parent_ino(file->f_path.dentry), DT_DIR) < 0) {
+               if (!file->f_path.dentry->d_parent->d_inode) {
+                       cFYI(1, "parent dir is negative, filling as current");
+                       ino = file->f_path.dentry->d_inode->i_ino;
+               } else
+                       ino = parent_ino(file->f_path.dentry);

Quoted 11 lines

Quoted 8 lines

+               if (filldir(direntry, "..", 2, file->f_pos, ino, DT_DIR) < 0) { cERROR(1, "Filldir for parent dir failed"); rc = -ENOMEM;
                       break;
--
1.7.1

(cc'ing fsdevel since I think we could use more eyes on this scheme)

I'm not a big fan of this patch. It seems to me that this assumes that it's ok to have a dentry with a negative dentry as a parent. That makes no sense though. Anegative dentry is the result of a lookup that returned the equivalent of -ENOENT. I'm very reticent to play that sort of game with the dcache code...

What may be better is to do something like this:

Go ahead and instantiate positive dentries from the root of the share to the root of the vfsmount. However, we would mark those inodes as "placeholders" and give them a bogus inode number of some sort (maybe even 0?).