ceph: add ino32 mount option

The ino32 mount option forces the ceph fs to report 32 bit
ino values.  This is useful for 64 bit kernels with 32 bit userspace.

Signed-off-by: Yehuda Sadeh <yehuda@hq.newdream.net>

4 files changed, 65 insertions, 25 deletions

diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index ebafa65a29b..cbe875d3a52 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -161,7 +161,7 @@ more:
         filp->f_pos = di->offset;
         err = filldir(dirent, dentry->d_name.name,
                       dentry->d_name.len, di->offset,
-                      dentry->d_inode->i_ino,
+                      ceph_translate_ino(dentry->d_sb, dentry->d_inode->i_ino),
                       dentry->d_inode->i_mode >> 12);
 
         if (last) {
@@ -245,15 +245,17 @@ static int ceph_readdir(struct file *filp, void *dirent, filldir_t filldir)
 
                 dout("readdir off 0 -> '.'\n");
                 if (filldir(dirent, ".", 1, ceph_make_fpos(0, 0),
-                            inode->i_ino, inode->i_mode >> 12) < 0)
+                            ceph_translate_ino(inode->i_sb, inode->i_ino),
+                            inode->i_mode >> 12) < 0)
                         return 0;
                 filp->f_pos = 1;
                 off = 1;
         }
         if (filp->f_pos == 1) {
+                ino_t ino = filp->f_dentry->d_parent->d_inode->i_ino;
                 dout("readdir off 1 -> '..'\n");
                 if (filldir(dirent, "..", 2, ceph_make_fpos(0, 1),
-                            filp->f_dentry->d_parent->d_inode->i_ino,
+                            ceph_translate_ino(inode->i_sb, ino),
                             inode->i_mode >> 12) < 0)
                         return 0;
                 filp->f_pos = 2;
@@ -377,7 +379,8 @@ more:
                 if (filldir(dirent,
                             rinfo->dir_dname[off - fi->offset],
                             rinfo->dir_dname_len[off - fi->offset],
-                            pos, ino, ftype) < 0) {
+                            pos,
+                            ceph_translate_ino(inode->i_sb, ino), ftype) < 0) {
                         dout("filldir stopping us...\n");
                         return 0;
                 }
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 60456361e07..b54c97da1c4 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -36,6 +36,13 @@ static void ceph_vmtruncate_work(struct work_struct *work);
 /*
  * find or create an inode, given the ceph ino number
  */
+static int ceph_set_ino_cb(struct inode *inode, void *data)
+{
+        ceph_inode(inode)->i_vino = *(struct ceph_vino *)data;
+        inode->i_ino = ceph_vino_to_ino(*(struct ceph_vino *)data);
+        return 0;
+}
+
 struct inode *ceph_get_inode(struct super_block *sb, struct ceph_vino vino)
 {
         struct inode *inode;
@@ -1809,7 +1816,7 @@ int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry,
         err = ceph_do_getattr(inode, CEPH_STAT_CAP_INODE_ALL);
         if (!err) {
                 generic_fillattr(inode, stat);
-                stat->ino = inode->i_ino;
+                stat->ino = ceph_translate_ino(inode->i_sb, inode->i_ino);
                 if (ceph_snap(inode) != CEPH_NOSNAP)
                         stat->dev = ceph_snap(inode);
                 else
diff --git a/fs/ceph/super.c b/fs/ceph/super.c
index 9c5085465a6..e39ea78c489 100644
--- a/fs/ceph/super.c
+++ b/fs/ceph/super.c
@@ -131,6 +131,7 @@ enum {
         Opt_rbytes,
         Opt_norbytes,
         Opt_noasyncreaddir,
+        Opt_ino32,
 };
 
 static match_table_t fsopt_tokens = {
@@ -150,6 +151,7 @@ static match_table_t fsopt_tokens = {
         {Opt_rbytes, "rbytes"},
         {Opt_norbytes, "norbytes"},
         {Opt_noasyncreaddir, "noasyncreaddir"},
+        {Opt_ino32, "ino32"},
         {-1, NULL}
 };
 
@@ -225,6 +227,9 @@ static int parse_fsopt_token(char *c, void *private)
         case Opt_noasyncreaddir:
                 fsopt->flags |= CEPH_MOUNT_OPT_NOASYNCREADDIR;
                 break;
+        case Opt_ino32:
+                fsopt->flags |= CEPH_MOUNT_OPT_INO32;
+                break;
         default:
                 BUG_ON(token);
         }
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 20b907d76ae..5405c903704 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -27,6 +27,7 @@
 #define CEPH_MOUNT_OPT_DIRSTAT         (1<<4) /* `cat dirname` for stats */
 #define CEPH_MOUNT_OPT_RBYTES          (1<<5) /* dir st_bytes = rbytes */
 #define CEPH_MOUNT_OPT_NOASYNCREADDIR  (1<<7) /* no dcache readdir */
+#define CEPH_MOUNT_OPT_INO32           (1<<8) /* 32 bit inos */
 
 #define CEPH_MOUNT_OPT_DEFAULT    (CEPH_MOUNT_OPT_RBYTES)
 
@@ -319,6 +320,16 @@ static inline struct ceph_inode_info *ceph_inode(struct inode *inode)
         return container_of(inode, struct ceph_inode_info, vfs_inode);
 }
 
+static inline struct ceph_fs_client *ceph_inode_to_client(struct inode *inode)
+{
+        return (struct ceph_fs_client *)inode->i_sb->s_fs_info;
+}
+
+static inline struct ceph_fs_client *ceph_sb_to_client(struct super_block *sb)
+{
+        return (struct ceph_fs_client *)sb->s_fs_info;
+}
+
 static inline struct ceph_vino ceph_vino(struct inode *inode)
 {
         return ceph_inode(inode)->i_vino;
@@ -327,19 +338,49 @@ static inline struct ceph_vino ceph_vino(struct inode *inode)
 /*
  * ino_t is <64 bits on many architectures, blech.
  *
- * don't include snap in ino hash, at least for now.
+ *               i_ino (kernel inode)   st_ino (userspace)
+ * i386          32                     32
+ * x86_64+ino32  64                     32
+ * x86_64        64                     64
+ */
+static inline u32 ceph_ino_to_ino32(ino_t ino)
+{
+        ino ^= ino >> (sizeof(ino) * 8 - 32);
+        if (!ino)
+                ino = 1;
+        return ino;
+}
+
+/*
+ * kernel i_ino value
  */
 static inline ino_t ceph_vino_to_ino(struct ceph_vino vino)
 {
         ino_t ino = (ino_t)vino.ino;  /* ^ (vino.snap << 20); */
 #if BITS_PER_LONG == 32
-        ino ^= vino.ino >> (sizeof(u64)-sizeof(ino_t)) * 8;
-        if (!ino)
-                ino = 1;
+        ino = ceph_ino_to_ino32(ino);
 #endif
         return ino;
 }
 
+/*
+ * user-visible ino (stat, filldir)
+ */
+#if BITS_PER_LONG == 32
+static inline ino_t ceph_translate_ino(struct super_block *sb, ino_t ino)
+{
+        return ino;
+}
+#else
+static inline ino_t ceph_translate_ino(struct super_block *sb, ino_t ino)
+{
+        if (ceph_test_mount_opt(ceph_sb_to_client(sb), INO32))
+                ino = ceph_ino_to_ino32(ino);
+        return ino;
+}
+#endif
+
+
 /* for printf-style formatting */
 #define ceph_vinop(i) ceph_inode(i)->i_vino.ino, ceph_inode(i)->i_vino.snap
 
@@ -428,13 +469,6 @@ static inline loff_t ceph_make_fpos(unsigned frag, unsigned off)
         return ((loff_t)frag << 32) | (loff_t)off;
 }
 
-static inline int ceph_set_ino_cb(struct inode *inode, void *data)
-{
-        ceph_inode(inode)->i_vino = *(struct ceph_vino *)data;
-        inode->i_ino = ceph_vino_to_ino(*(struct ceph_vino *)data);
-        return 0;
-}
-
 /*
  * caps helpers
  */
@@ -503,15 +537,6 @@ extern void ceph_reservation_status(struct ceph_fs_client *client,
                                     int *total, int *avail, int *used,
                                     int *reserved, int *min);
 
-static inline struct ceph_fs_client *ceph_inode_to_client(struct inode *inode)
-{
-        return (struct ceph_fs_client *)inode->i_sb->s_fs_info;
-}
-
-static inline struct ceph_fs_client *ceph_sb_to_client(struct super_block *sb)
-{
-        return (struct ceph_fs_client *)sb->s_fs_info;
-}
 
 
 /*