Whamcloud - gitweb
LU-190 opencreate without mode can crash llite
[fs/lustre-release.git] / lustre / llite / namei.c
index f06d1cc..f9f9bc7 100644 (file)
@@ -1,22 +1,37 @@
 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
  * vim:expandtab:shiftwidth=8:tabstop=8:
  *
- *  Copyright (c) 2002, 2003 Cluster File Systems, Inc.
+ * GPL HEADER START
  *
- *   This file is part of Lustre, http://www.lustre.org.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
- *   Lustre is free software; you can redistribute it and/or
- *   modify it under the terms of version 2 of the GNU General Public
- *   License as published by the Free Software Foundation.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 only,
+ * as published by the Free Software Foundation.
  *
- *   Lustre is distributed in the hope that it will be useful,
- *   but WITHOUT ANY WARRANTY; without even the implied warranty of
- *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- *   GNU General Public License for more details.
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License version 2 for more details (a copy is included
+ * in the LICENSE file that accompanied this code).
  *
- *   You should have received a copy of the GNU General Public License
- *   along with Lustre; if not, write to the Free Software
- *   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ * You should have received a copy of the GNU General Public License
+ * version 2 along with this program; If not, see
+ * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ *
+ * GPL HEADER END
+ */
+/*
+ * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
+ * Use is subject to license terms.
+ */
+/*
+ * This file is part of Lustre, http://www.lustre.org/
+ * Lustre is a trademark of Sun Microsystems, Inc.
  */
 
 #include <linux/fs.h>
 #define DEBUG_SUBSYSTEM S_LLITE
 
 #include <obd_support.h>
+#include <lustre_fid.h>
 #include <lustre_lite.h>
 #include <lustre_dlm.h>
-#include <linux/lustre_version.h>
+#include <lustre_ver.h>
+#include <lustre_mdc.h>
 #include "llite_internal.h"
 
-/* methods */
+/*
+ * Check if we have something mounted at the named dchild.
+ * In such a case there would always be dentry present.
+ */
+static int ll_d_mountpoint(struct dentry *dparent, struct dentry *dchild,
+                           struct qstr *name)
+{
+        int mounted = 0;
+
+        if (unlikely(dchild)) {
+                mounted = d_mountpoint(dchild);
+        } else if (dparent) {
+                dchild = d_lookup(dparent, name);
+                if (dchild) {
+                        mounted = d_mountpoint(dchild);
+                        dput(dchild);
+                }
+        }
+        return mounted;
+}
 
-/* called from iget{4,5_locked}->find_inode() under inode_lock spinlock */
-#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,5,0))
-static int ll_test_inode(struct inode *inode, unsigned long ino, void *opaque)
-#else
-static int ll_test_inode(struct inode *inode, void *opaque)
-#endif
+int ll_unlock(__u32 mode, struct lustre_handle *lockh)
 {
-        static int last_ino, last_gen, last_count;
-        struct lustre_md *md = opaque;
+        ENTRY;
 
-        if (!(md->body->valid & (OBD_MD_FLGENER | OBD_MD_FLID))) {
-                CERROR("MDS body missing inum or generation\n");
-                return 0;
-        }
+        ldlm_lock_decref(lockh, mode);
 
-        if (last_ino == md->body->ino && last_gen == md->body->generation &&
-            last_count < 500) {
-                last_count++;
-        } else {
-                if (last_count > 1)
-                        CDEBUG(D_VFSTRACE, "compared %u/%u %u times\n",
-                               last_ino, last_gen, last_count);
-                last_count = 0;
-                last_ino = md->body->ino;
-                last_gen = md->body->generation;
-                CDEBUG(D_VFSTRACE,
-                       "comparing inode %p ino %lu/%u to body "LPU64"/%u\n",
-                       inode, inode->i_ino, inode->i_generation,
-                       md->body->ino, md->body->generation);
-        }
+        RETURN(0);
+}
 
-#if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
-        if (inode->i_ino != md->body->ino)
-                return 0;
-#endif
-        if (inode->i_generation != md->body->generation) {
-                struct ll_sb_info *sbi = ll_i2sbi(inode);
-                struct ll_inode_info *lli = ll_i2info(inode);
-
-                if (inode->i_state & (I_FREEING | I_CLEAR))
-                        return 0;
-                if (inode->i_nlink == 0)
-                        return 0;
-
-                /* add "duplicate" inode into deathrow for destroy */
-                spin_lock(&sbi->ll_deathrow_lock);
-                if (list_empty(&lli->lli_dead_list)) {
-                        atomic_inc(&inode->i_count);
-                        list_add(&lli->lli_dead_list, &sbi->ll_deathrow);
-                }
-                spin_unlock(&sbi->ll_deathrow_lock);
 
+/* called from iget5_locked->find_inode() under inode_lock spinlock */
+static int ll_test_inode(struct inode *inode, void *opaque)
+{
+        struct ll_inode_info *lli = ll_i2info(inode);
+        struct lustre_md     *md = opaque;
+
+        if (unlikely(!(md->body->valid & OBD_MD_FLID))) {
+                CERROR("MDS body missing FID\n");
                 return 0;
         }
 
-        /* Apply the attributes in 'opaque' to this inode */
-        if (!(inode->i_state & (I_FREEING | I_CLEAR)))
-                ll_update_inode(inode, md);
+        if (!lu_fid_eq(&lli->lli_fid, &md->body->fid1))
+                return 0;
+
         return 1;
 }
 
-extern struct dentry_operations ll_d_ops;
-
-int ll_unlock(__u32 mode, struct lustre_handle *lockh)
+static int ll_set_inode(struct inode *inode, void *opaque)
 {
-        ENTRY;
+        struct ll_inode_info *lli = ll_i2info(inode);
+        struct mdt_body *body = ((struct lustre_md *)opaque)->body;
 
-        ldlm_lock_decref(lockh, mode);
+        if (unlikely(!(body->valid & OBD_MD_FLID))) {
+                CERROR("MDS body missing FID\n");
+                return -EINVAL;
+        }
 
-        RETURN(0);
+        lli->lli_fid = body->fid1;
+        return 0;
 }
 
-/* Get an inode by inode number (already instantiated by the intent lookup).
+
+/*
+ * Get an inode by inode number (already instantiated by the intent lookup).
  * Returns inode or NULL
  */
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,5,0))
-int ll_set_inode(struct inode *inode, void *opaque)
-{
-        ll_read_inode2(inode, opaque);
-        return 0;
-}
-
 struct inode *ll_iget(struct super_block *sb, ino_t hash,
                       struct lustre_md *md)
 {
-        struct inode *inode;
+        struct inode         *inode;
+        ENTRY;
 
         LASSERT(hash != 0);
         inode = iget5_locked(sb, hash, ll_test_inode, ll_set_inode, md);
 
         if (inode) {
-                if (inode->i_state & I_NEW)
-                        unlock_new_inode(inode);
-                CDEBUG(D_VFSTRACE, "inode: %lu/%u(%p)\n", inode->i_ino,
-                       inode->i_generation, inode);
+                if (inode->i_state & I_NEW) {
+                        int rc;
+
+                        ll_read_inode2(inode, md);
+                        rc = cl_inode_init(inode, md);
+                        if (rc != 0) {
+                                md->lsm = NULL;
+                                make_bad_inode(inode);
+                                unlock_new_inode(inode);
+                                iput(inode);
+                                inode = ERR_PTR(rc);
+                        } else
+                                unlock_new_inode(inode);
+                } else if (!(inode->i_state & (I_FREEING | I_CLEAR)))
+                                ll_update_inode(inode, md);
+                CDEBUG(D_VFSTRACE, "got inode: %p for "DFID"\n",
+                       inode, PFID(&md->body->fid1));
         }
-
-        return inode;
+        RETURN(inode);
 }
-#else
-struct inode *ll_iget(struct super_block *sb, ino_t hash,
-                      struct lustre_md *md)
+
+static void ll_drop_negative_dentry(struct inode *dir)
 {
-        struct inode *inode;
-        LASSERT(hash != 0);
-        inode = iget4(sb, hash, ll_test_inode, md);
-        if (inode)
-                CDEBUG(D_VFSTRACE, "inode: %lu/%u(%p)\n", inode->i_ino,
-                       inode->i_generation, inode);
-        return inode;
+        struct dentry *dentry, *tmp_alias, *tmp_subdir;
+
+        cfs_spin_lock(&ll_lookup_lock);
+        spin_lock(&dcache_lock);
+restart:
+        list_for_each_entry_safe(dentry, tmp_alias,
+                                 &dir->i_dentry,d_alias) {
+                if (!list_empty(&dentry->d_subdirs)) {
+                        struct dentry *child;
+                        list_for_each_entry_safe(child, tmp_subdir,
+                                                 &dentry->d_subdirs,
+                                                 d_child) {
+                                /* XXX Print some debug here? */
+                                if (!child->d_inode)
+                                /* Negative dentry. If we were
+                                   dropping dcache lock, go
+                                   throught the list again */
+                                        if (ll_drop_dentry(child))
+                                                goto restart;
+                        }
+                }
+        }
+        spin_unlock(&dcache_lock);
+        cfs_spin_unlock(&ll_lookup_lock);
 }
-#endif
 
-int ll_mdc_blocking_ast(struct ldlm_lock *lock, struct ldlm_lock_desc *desc,
-                        void *data, int flag)
+
+int ll_md_blocking_ast(struct ldlm_lock *lock, struct ldlm_lock_desc *desc,
+                       void *data, int flag)
 {
         int rc;
         struct lustre_handle lockh;
@@ -168,16 +197,33 @@ int ll_mdc_blocking_ast(struct ldlm_lock *lock, struct ldlm_lock_desc *desc,
                 break;
         case LDLM_CB_CANCELING: {
                 struct inode *inode = ll_inode_from_lock(lock);
+                struct ll_inode_info *lli;
                 __u64 bits = lock->l_policy_data.l_inodebits.bits;
+                struct lu_fid *fid;
+                ldlm_mode_t mode = lock->l_req_mode;
 
                 /* Invalidate all dentries associated with this inode */
                 if (inode == NULL)
                         break;
 
-                if (lock->l_resource->lr_name.name[0] != inode->i_ino ||
-                    lock->l_resource->lr_name.name[1] != inode->i_generation) {
-                        LDLM_ERROR(lock, "data mismatch with ino %lu/%u (%p)",
-                                   inode->i_ino, inode->i_generation, inode);
+                LASSERT(lock->l_flags & LDLM_FL_CANCELING);
+                /* For OPEN locks we differentiate between lock modes - CR, CW. PR - bug 22891 */
+                if ((bits & MDS_INODELOCK_LOOKUP) &&
+                    ll_have_md_lock(inode, MDS_INODELOCK_LOOKUP, LCK_MINMODE))
+                        bits &= ~MDS_INODELOCK_LOOKUP;
+                if ((bits & MDS_INODELOCK_UPDATE) &&
+                    ll_have_md_lock(inode, MDS_INODELOCK_UPDATE, LCK_MINMODE))
+                        bits &= ~MDS_INODELOCK_UPDATE;
+                if ((bits & MDS_INODELOCK_OPEN) &&
+                    ll_have_md_lock(inode, MDS_INODELOCK_OPEN, mode))
+                        bits &= ~MDS_INODELOCK_OPEN;
+
+                fid = ll_inode2fid(inode);
+                if (lock->l_resource->lr_name.name[0] != fid_seq(fid) ||
+                    lock->l_resource->lr_name.name[1] != fid_oid(fid) ||
+                    lock->l_resource->lr_name.name[2] != fid_ver(fid)) {
+                        LDLM_ERROR(lock, "data mismatch with object "
+                                   DFID" (%p)", PFID(fid), inode);
                 }
 
                 if (bits & MDS_INODELOCK_OPEN) {
@@ -188,8 +234,6 @@ int ll_mdc_blocking_ast(struct ldlm_lock *lock, struct ldlm_lock_desc *desc,
                                 break;
                         case LCK_PR:
                                 flags = FMODE_EXEC;
-                                if (!FMODE_EXEC)
-                                        CERROR("open PR lock without FMODE_EXEC\n");
                                 break;
                         case LCK_CR:
                                 flags = FMODE_READ;
@@ -199,57 +243,19 @@ int ll_mdc_blocking_ast(struct ldlm_lock *lock, struct ldlm_lock_desc *desc,
                                        "%d, inode %ld\n", lock->l_req_mode,
                                        inode->i_ino);
                         }
-                        ll_mdc_real_close(inode, flags);
+                        ll_md_real_close(inode, flags);
                 }
 
+                lli = ll_i2info(inode);
                 if (bits & MDS_INODELOCK_UPDATE)
-                        clear_bit(LLI_F_HAVE_MDS_SIZE_LOCK,
-                                  &(ll_i2info(inode)->lli_flags));
+                        lli->lli_flags &= ~LLIF_MDS_SIZE_LOCK;
 
                 if (S_ISDIR(inode->i_mode) &&
                      (bits & MDS_INODELOCK_UPDATE)) {
-                        struct dentry *dentry, *tmp, *dir;
-                        struct list_head *list;
-
                         CDEBUG(D_INODE, "invalidating inode %lu\n",
                                inode->i_ino);
                         truncate_inode_pages(inode->i_mapping, 0);
-
-                        /* Drop possible cached negative dentries */
-                        list = &inode->i_dentry;
-                        dir = NULL;
-                        spin_lock(&dcache_lock);
-
-                        /* It is possible to have several dentries (with
-                           racer?) */
-                        while ((list = list->next) != &inode->i_dentry) {
-                                dir = list_entry(list, struct dentry, d_alias);
-#ifdef LUSTRE_KERNEL_VERSION
-                                if (!(dir->d_flags & DCACHE_LUSTRE_INVALID))
-#else
-                                if (!d_unhashed(dir))
-#endif
-                                        break;
-
-                                dir = NULL;
-                        }
-
-                        if (dir) {
-restart:
-                                list_for_each_entry_safe(dentry, tmp,
-                                                         &dir->d_subdirs,
-                                                         d_child)
-                                {
-                                        /* XXX Print some debug here? */
-                                        if (!dentry->d_inode)
-                                                /* Negative dentry. If we were
-                                                   dropping dcache lock, go
-                                                   throught the list again */
-                                                if (ll_drop_dentry(dentry))
-                                                        goto restart;
-                                }
-                        }
-                        spin_unlock(&dcache_lock);
+                        ll_drop_negative_dentry(inode);
                 }
 
                 if (inode->i_sb->s_root &&
@@ -266,16 +272,12 @@ restart:
         RETURN(0);
 }
 
-int ll_mdc_cancel_unused(struct lustre_handle *conn, struct inode *inode,
-                         int flags, void *opaque)
+__u32 ll_i2suppgid(struct inode *i)
 {
-        struct ldlm_res_id res_id =
-                { .name = {inode->i_ino, inode->i_generation} };
-        struct obd_device *obddev = class_conn2obd(conn);
-        ENTRY;
-
-        RETURN(ldlm_cli_cancel_unused(obddev->obd_namespace, &res_id, flags,
-                                      opaque));
+        if (cfs_curproc_is_in_groups(i->i_gid))
+                return (__u32)i->i_gid;
+        else
+                return (__u32)(-1);
 }
 
 /* Pack the required supplementary groups into the supplied groups array.
@@ -284,25 +286,21 @@ int ll_mdc_cancel_unused(struct lustre_handle *conn, struct inode *inode,
  * array in case it might be useful.  Not needed if doing an MDS-side upcall. */
 void ll_i2gids(__u32 *suppgids, struct inode *i1, struct inode *i2)
 {
+#if 0
         int i;
+#endif
 
         LASSERT(i1 != NULL);
         LASSERT(suppgids != NULL);
 
-        if (in_group_p(i1->i_gid))
-                suppgids[0] = i1->i_gid;
-        else
-                suppgids[0] = -1;
+        suppgids[0] = ll_i2suppgid(i1);
 
-        if (i2) {
-                if (in_group_p(i2->i_gid))
-                        suppgids[1] = i2->i_gid;
+        if (i2)
+                suppgids[1] = ll_i2suppgid(i2);
                 else
                         suppgids[1] = -1;
-        } else {
-                suppgids[1] = -1;
-        }
 
+#if 0
         for (i = 0; i < current_ngroups; i++) {
                 if (suppgids[0] == -1) {
                         if (current_groups[i] != suppgids[1])
@@ -316,30 +314,7 @@ void ll_i2gids(__u32 *suppgids, struct inode *i1, struct inode *i2)
                 }
                 break;
         }
-}
-
-int ll_prepare_mdc_op_data(struct mdc_op_data *data, struct inode *i1,
-                            struct inode *i2, const char *name, int namelen,
-                            int mode)
-{
-        LASSERT(i1);
-
-        if (namelen > ll_i2sbi(i1)->ll_namelen)
-                return -ENAMETOOLONG;
-        ll_i2gids(data->suppgids, i1, i2);
-        ll_inode2fid(&data->fid1, i1);
-
-        if (i2)
-                ll_inode2fid(&data->fid2, i2);
-        else
-                memset(&data->fid2, 0, sizeof(data->fid2));
-
-        data->name = name;
-        data->namelen = namelen;
-        data->create_mode = mode;
-        data->mod_time = CURRENT_SECONDS;
-
-        return 0;
+#endif
 }
 
 static void ll_d_add(struct dentry *de, struct inode *inode)
@@ -364,7 +339,7 @@ static void ll_d_add(struct dentry *de, struct inode *inode)
                        de->d_name.len, de->d_name.name, de, de->d_hash.next);
                 LBUG();
         }
-        __d_rehash(de, 0);
+        d_rehash_cond(de, 0);
 }
 
 /* Search "inode"'s alias list for a dentry that has the same name and parent
@@ -373,12 +348,13 @@ static void ll_d_add(struct dentry *de, struct inode *inode)
  * in ll_revalidate_it.  After revaliadate inode will be have hashed aliases
  * and it triggers BUG_ON in d_instantiate_unique (bug #10954).
  */
-struct dentry *ll_find_alias(struct inode *inode, struct dentry *de)
+static struct dentry *ll_find_alias(struct inode *inode, struct dentry *de)
 {
         struct list_head *tmp;
         struct dentry *dentry;
         struct dentry *last_discon = NULL;
 
+        cfs_spin_lock(&ll_lookup_lock);
         spin_lock(&dcache_lock);
         list_for_each(tmp, &inode->i_dentry) {
                 dentry = list_entry(tmp, struct dentry, d_alias);
@@ -411,39 +387,47 @@ struct dentry *ll_find_alias(struct inode *inode, struct dentry *de)
                 dget_locked(dentry);
                 lock_dentry(dentry);
                 __d_drop(dentry);
-#ifdef LUSTRE_KERNEL_VERSION
-                dentry->d_flags &= ~DCACHE_LUSTRE_INVALID;
-#endif
                 unlock_dentry(dentry);
-                __d_rehash(dentry, 0); /* avoid taking dcache_lock inside */
+                ll_dops_init(dentry, 0);
+                d_rehash_cond(dentry, 0); /* avoid taking dcache_lock inside */
                 spin_unlock(&dcache_lock);
+                cfs_spin_unlock(&ll_lookup_lock);
                 iput(inode);
                 CDEBUG(D_DENTRY, "alias dentry %.*s (%p) parent %p inode %p "
                        "refc %d\n", de->d_name.len, de->d_name.name, de,
                        de->d_parent, de->d_inode, atomic_read(&de->d_count));
                 return dentry;
         }
+
         if (last_discon) {
-                 CDEBUG(D_DENTRY, "Reuse disconnected dentry %p inode %p "
+                CDEBUG(D_DENTRY, "Reuse disconnected dentry %p inode %p "
                         "refc %d\n", last_discon, last_discon->d_inode,
                         atomic_read(&last_discon->d_count));
-                 dget_locked(last_discon);
-                 spin_unlock(&dcache_lock);
-                 d_rehash(de);
-                 d_move(last_discon, de);
-                 iput(inode);
-                 return last_discon;
+                dget_locked(last_discon);
+                lock_dentry(last_discon);
+                last_discon->d_flags |= DCACHE_LUSTRE_INVALID;
+                unlock_dentry(last_discon);
+                spin_unlock(&dcache_lock);
+                cfs_spin_unlock(&ll_lookup_lock);
+                ll_dops_init(last_discon, 1);
+                d_rehash(de);
+                d_move(last_discon, de);
+                iput(inode);
+                return last_discon;
         }
-
+        lock_dentry(de);
+        de->d_flags |= DCACHE_LUSTRE_INVALID;
+        unlock_dentry(de);
         ll_d_add(de, inode);
 
         spin_unlock(&dcache_lock);
+        cfs_spin_unlock(&ll_lookup_lock);
 
         return de;
 }
 
-static int lookup_it_finish(struct ptlrpc_request *request, int offset,
-                            struct lookup_intent *it, void *data)
+int ll_lookup_it_finish(struct ptlrpc_request *request,
+                        struct lookup_intent *it, void *data)
 {
         struct it_cb_data *icbd = data;
         struct dentry **de = icbd->icbd_childp;
@@ -451,20 +435,22 @@ static int lookup_it_finish(struct ptlrpc_request *request, int offset,
         struct ll_sb_info *sbi = ll_i2sbi(parent);
         struct inode *inode = NULL;
         int rc;
+        ENTRY;
 
         /* NB 1 request reference will be taken away by ll_intent_lock()
          * when I return */
         if (!it_disposition(it, DISP_LOOKUP_NEG)) {
-                ENTRY;
+                struct dentry *save = *de;
+                __u32 bits;
 
-                rc = ll_prep_inode(sbi->ll_osc_exp, &inode, request, offset,
-                                   (*de)->d_sb);
+                rc = ll_prep_inode(&inode, request, (*de)->d_sb);
                 if (rc)
                         RETURN(rc);
 
                 CDEBUG(D_DLMTRACE, "setting l_data to inode %p (%lu/%u)\n",
                        inode, inode->i_ino, inode->i_generation);
-                mdc_set_lock_data(&it->d.lustre.it_lock_handle, inode);
+                md_set_lock_data(sbi->ll_md_exp,
+                                 &it->d.lustre.it_lock_handle, inode, &bits);
 
                 /* We used to query real size from OSTs here, but actually
                    this is not needed. For stat() calls size would be updated
@@ -472,19 +458,40 @@ static int lookup_it_finish(struct ptlrpc_request *request, int offset,
                    2.4 and
                    vfs_getattr_it->ll_getattr()->ll_inode_revalidate_it() in 2.6
                    Everybody else who needs correct file size would call
-                   ll_glimpse_size or some equivalent themselves anyway.
+                   cl_glimpse_size or some equivalent themselves anyway.
                    Also see bug 7198. */
+                ll_dops_init(*de, 1);
                 *de = ll_find_alias(inode, *de);
+                if (*de != save) {
+                        struct ll_dentry_data *lld = ll_d2d(*de);
+
+                        /* just make sure the ll_dentry_data is ready */
+                        if (unlikely(lld == NULL))
+                                ll_dops_init(*de, 1);
+                }
+                /* we have lookup look - unhide dentry */
+                if (bits & MDS_INODELOCK_LOOKUP) {
+                        lock_dentry(*de);
+                        (*de)->d_flags &= ~DCACHE_LUSTRE_INVALID;
+                        unlock_dentry(*de);
+                }
         } else {
-                ENTRY;
+                ll_dops_init(*de, 1);
                 /* Check that parent has UPDATE lock. If there is none, we
                    cannot afford to hash this dentry (done by ll_d_add) as it
                    might get picked up later when UPDATE lock will appear */
-                if (ll_have_md_lock(parent, MDS_INODELOCK_UPDATE)) {
+                if (ll_have_md_lock(parent, MDS_INODELOCK_UPDATE, LCK_MINMODE)) {
                         spin_lock(&dcache_lock);
-                        ll_d_add(*de, inode);
+                        ll_d_add(*de, NULL);
                         spin_unlock(&dcache_lock);
                 } else {
+                        /* negative lookup - and don't have update lock to
+                         * parent */
+                        lock_dentry(*de);
+                        (*de)->d_flags |= DCACHE_LUSTRE_INVALID;
+                        unlock_dentry(*de);
+
+                        (*de)->d_inode = NULL;
                         /* We do not want to hash the dentry if don`t have a
                          * lock, but if this dentry is later used in d_move,
                          * we'd hit uninitialised list head d_hash, so we just
@@ -495,23 +502,24 @@ static int lookup_it_finish(struct ptlrpc_request *request, int offset,
                 }
         }
 
-        ll_set_dd(*de);
-        (*de)->d_op = &ll_d_ops;
-
         RETURN(0);
 }
 
 static struct dentry *ll_lookup_it(struct inode *parent, struct dentry *dentry,
                                    struct lookup_intent *it, int lookup_flags)
 {
+        struct lookup_intent lookup_it = { .it_op = IT_LOOKUP };
         struct dentry *save = dentry, *retval;
-        struct mdc_op_data op_data;
-        struct it_cb_data icbd;
         struct ptlrpc_request *req = NULL;
-        struct lookup_intent lookup_it = { .it_op = IT_LOOKUP };
-        int rc;
+        struct md_op_data *op_data;
+        struct it_cb_data icbd;
+        __u32 opc;
+        int rc, first = 0;
         ENTRY;
 
+        if (dentry->d_name.len > ll_i2sbi(parent)->ll_namelen)
+                RETURN(ERR_PTR(-ENAMETOOLONG));
+
         CDEBUG(D_VFSTRACE, "VFS Op:name=%.*s,dir=%lu/%u(%p),intent=%s\n",
                dentry->d_name.len, dentry->d_name.name, parent->i_ino,
                parent->i_generation, parent, LL_IT2STR(it));
@@ -525,33 +533,53 @@ static struct dentry *ll_lookup_it(struct inode *parent, struct dentry *dentry,
          * not valid, revalidate it here. */
         if (parent->i_sb->s_root && (parent->i_sb->s_root->d_inode == parent) &&
             (it->it_op & (IT_OPEN | IT_CREAT))) {
-                rc = ll_inode_revalidate_it(parent->i_sb->s_root, it);
+                rc = ll_inode_revalidate_it(parent->i_sb->s_root, it,
+                                            MDS_INODELOCK_LOOKUP);
                 if (rc)
                         RETURN(ERR_PTR(rc));
         }
 
+        if (it->it_op == IT_GETATTR) {
+                first = ll_statahead_enter(parent, &dentry, 1);
+                if (first >= 0) {
+                        ll_statahead_exit(parent, dentry, first);
+                        if (first == 1)
+                                RETURN(retval = dentry);
+                }
+        }
+
         icbd.icbd_childp = &dentry;
         icbd.icbd_parent = parent;
 
-        rc = ll_prepare_mdc_op_data(&op_data, parent, NULL, dentry->d_name.name,
-                                    dentry->d_name.len, lookup_flags);
-        if (rc)
-                RETURN(ERR_PTR(rc));
+        if (it->it_op & IT_CREAT ||
+            (it->it_op & IT_OPEN && it->it_create_mode & O_CREAT))
+                opc = LUSTRE_OPC_CREATE;
+        else
+                opc = LUSTRE_OPC_ANY;
 
-        it->it_create_mode &= ~current->fs->umask;
+        op_data = ll_prep_md_op_data(NULL, parent, NULL, dentry->d_name.name,
+                                     dentry->d_name.len, lookup_flags, opc,
+                                     NULL);
+        if (IS_ERR(op_data))
+                RETURN((void *)op_data);
 
-        rc = mdc_intent_lock(ll_i2mdcexp(parent), &op_data, NULL, 0, it,
-                             lookup_flags, &req, ll_mdc_blocking_ast, 0);
+        it->it_create_mode &= ~cfs_curproc_umask();
 
+        rc = md_intent_lock(ll_i2mdexp(parent), op_data, NULL, 0, it,
+                            lookup_flags, &req, ll_md_blocking_ast, 0);
+        ll_finish_md_op_data(op_data);
         if (rc < 0)
                 GOTO(out, retval = ERR_PTR(rc));
 
-        rc = lookup_it_finish(req, DLM_REPLY_REC_OFF, it, &icbd);
+        rc = ll_lookup_it_finish(req, it, &icbd);
         if (rc != 0) {
                 ll_intent_release(it);
                 GOTO(out, retval = ERR_PTR(rc));
         }
 
+        if (first == -EEXIST)
+                ll_statahead_mark(parent, dentry);
+
         if ((it->it_op & IT_OPEN) && dentry->d_inode &&
             !S_ISREG(dentry->d_inode->i_mode) &&
             !S_ISDIR(dentry->d_inode->i_mode)) {
@@ -569,22 +597,6 @@ static struct dentry *ll_lookup_it(struct inode *parent, struct dentry *dentry,
         return retval;
 }
 
-#if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
-#ifdef LUSTRE_KERNEL_VERSION
-static struct dentry *ll_lookup_nd(struct inode *parent, struct dentry *dentry,
-                                   struct nameidata *nd)
-{
-        struct dentry *de;
-        ENTRY;
-
-        if (nd && nd->flags & LOOKUP_LAST && !(nd->flags & LOOKUP_LINK_NOTLAST))
-                de = ll_lookup_it(parent, dentry, &nd->intent, nd->flags);
-        else
-                de = ll_lookup_it(parent, dentry, NULL, 0);
-
-        RETURN(de);
-}
-#else
 struct lookup_intent *ll_convert_intent(struct open_intent *oit,
                                         int lookup_flags)
 {
@@ -598,7 +610,7 @@ struct lookup_intent *ll_convert_intent(struct open_intent *oit,
                 it->it_op = IT_OPEN;
                 if (lookup_flags & LOOKUP_CREATE)
                         it->it_op |= IT_CREAT;
-                it->it_create_mode = oit->create_mode;
+                it->it_create_mode = (oit->create_mode & S_IALLUGO) | S_IFREG;
                 it->it_flags = oit->flags;
         } else {
                 it->it_op = IT_GETATTR;
@@ -631,11 +643,16 @@ static struct dentry *ll_lookup_nd(struct inode *parent, struct dentry *dentry,
                         if (IS_ERR(nd->intent.open.file))
                                 RETURN((struct dentry *)nd->intent.open.file);
 #endif
-
                 if (ll_d2d(dentry) && ll_d2d(dentry)->lld_it) {
                         it = ll_d2d(dentry)->lld_it;
                         ll_d2d(dentry)->lld_it = NULL;
                 } else {
+                        if ((nd->flags & LOOKUP_CREATE ) && !(nd->flags & LOOKUP_OPEN)) {
+                                /* We are sure this is new dentry, so we need to create
+                                   our private data and set the dentry ops */ 
+                                ll_dops_init(dentry, 1);
+                                RETURN(NULL);
+                        }
                         it = ll_convert_intent(&nd->intent.open, nd->flags);
                         if (IS_ERR(it))
                                 RETURN((struct dentry *)it);
@@ -655,19 +672,23 @@ static struct dentry *ll_lookup_nd(struct inode *parent, struct dentry *dentry,
                                                        (struct ptlrpc_request *)
                                                           it->d.lustre.it_data);
                                 } else {
-                                        struct file *filp;
-                                        nd->intent.open.file->private_data = it;
-                                        filp =lookup_instantiate_filp(nd,dentry,
-                                                                      NULL);
 #if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,17))
 /* 2.6.1[456] have a bug in open_namei() that forgets to check
  * nd->intent.open.file for error, so we need to return it as lookup's result
  * instead */
+                                        struct file *filp;
+                                        nd->intent.open.file->private_data = it;
+                                        filp =lookup_instantiate_filp(nd,dentry,
+                                                                      NULL);
                                         if (IS_ERR(filp)) {
                                                 if (de)
                                                         dput(de);
                                                 de = (struct dentry *) filp;
                                         }
+#else
+                                        nd->intent.open.file->private_data = it;
+                                        (void)lookup_instantiate_filp(nd,dentry,
+                                                                      NULL);
 #endif
 
                                 }
@@ -697,8 +718,6 @@ static struct dentry *ll_lookup_nd(struct inode *parent, struct dentry *dentry,
 
         RETURN(de);
 }
-#endif
-#endif
 
 /* We depend on "mode" being set with the proper file type/umask by now */
 static struct inode *ll_create_node(struct inode *dir, const char *name,
@@ -717,8 +736,7 @@ static struct inode *ll_create_node(struct inode *dir, const char *name,
         LASSERT(it_disposition(it, DISP_ENQ_CREATE_REF));
         request = it->d.lustre.it_data;
         it_clear_disposition(it, DISP_ENQ_CREATE_REF);
-        rc = ll_prep_inode(sbi->ll_osc_exp, &inode, request, DLM_REPLY_REC_OFF,
-                           dir->i_sb);
+        rc = ll_prep_inode(&inode, request, dir->i_sb);
         if (rc)
                 GOTO(out, inode = ERR_PTR(rc));
 
@@ -729,7 +747,8 @@ static struct inode *ll_create_node(struct inode *dir, const char *name,
          * stuff it in the lock. */
         CDEBUG(D_DLMTRACE, "setting l_ast_data to inode %p (%lu/%u)\n",
                inode, inode->i_ino, inode->i_generation);
-        mdc_set_lock_data(&it->d.lustre.it_lock_handle, inode);
+        md_set_lock_data(sbi->ll_md_exp,
+                         &it->d.lustre.it_lock_handle, inode, NULL);
         EXIT;
  out:
         ptlrpc_req_finished(request);
@@ -754,7 +773,6 @@ static int ll_create_it(struct inode *dir, struct dentry *dentry, int mode,
                         struct lookup_intent *it)
 {
         struct inode *inode;
-        struct ptlrpc_request *request = it->d.lustre.it_data;
         int rc = 0;
         ENTRY;
 
@@ -766,13 +784,10 @@ static int ll_create_it(struct inode *dir, struct dentry *dentry, int mode,
         if (rc)
                 RETURN(rc);
 
-        mdc_store_inode_generation(request, DLM_INTENT_REC_OFF,
-                                   DLM_REPLY_REC_OFF);
         inode = ll_create_node(dir, dentry->d_name.name, dentry->d_name.len,
                                NULL, 0, mode, 0, it);
-        if (IS_ERR(inode)) {
+        if (IS_ERR(inode))
                 RETURN(PTR_ERR(inode));
-        }
 
         d_instantiate(dentry, inode);
         /* Negative dentry may be unhashed if parent does not have UPDATE lock,
@@ -780,80 +795,73 @@ static int ll_create_it(struct inode *dir, struct dentry *dentry, int mode,
          * successful create. Hash it here. */
         spin_lock(&dcache_lock);
         if (d_unhashed(dentry))
-                __d_rehash(dentry, 0);
+                d_rehash_cond(dentry, 0);
         spin_unlock(&dcache_lock);
         RETURN(0);
 }
 
-static void ll_update_times(struct ptlrpc_request *request, int offset,
+static void ll_update_times(struct ptlrpc_request *request,
                             struct inode *inode)
 {
-        struct mds_body *body = lustre_msg_buf(request->rq_repmsg, offset,
-                                               sizeof(*body));
-        LASSERT(body);
+        struct mdt_body *body = req_capsule_server_get(&request->rq_pill,
+                                                       &RMF_MDT_BODY);
 
-        /* mtime is always updated with ctime, but can be set in past.
-           As write and utime(2) may happen within 1 second, and utime's
-           mtime has a priority over write's one, so take mtime from mds
-           for the same ctimes. */
+        LASSERT(body);
+        if (body->valid & OBD_MD_FLMTIME &&
+            body->mtime > LTIME_S(inode->i_mtime)) {
+                CDEBUG(D_INODE, "setting ino %lu mtime from %lu to "LPU64"\n",
+                       inode->i_ino, LTIME_S(inode->i_mtime), body->mtime);
+                LTIME_S(inode->i_mtime) = body->mtime;
+        }
         if (body->valid & OBD_MD_FLCTIME &&
-            body->ctime >= LTIME_S(inode->i_ctime)) {
+            body->ctime > LTIME_S(inode->i_ctime))
                 LTIME_S(inode->i_ctime) = body->ctime;
-
-                if (body->valid & OBD_MD_FLMTIME) {
-                        CDEBUG(D_INODE, "setting ino %lu mtime from %lu "
-                               "to "LPU64"\n", inode->i_ino,
-                               LTIME_S(inode->i_mtime), body->mtime);
-                        LTIME_S(inode->i_mtime) = body->mtime;
-                }
-        }
 }
 
 static int ll_new_node(struct inode *dir, struct qstr *name,
-                       const char *tgt, int mode,
-                       int rdev, struct dentry *dchild)
+                       const char *tgt, int mode, int rdev,
+                       struct dentry *dchild, __u32 opc)
 {
         struct ptlrpc_request *request = NULL;
+        struct md_op_data *op_data;
         struct inode *inode = NULL;
         struct ll_sb_info *sbi = ll_i2sbi(dir);
-        struct mdc_op_data op_data;
         int tgt_len = 0;
         int err;
 
         ENTRY;
         if (unlikely(tgt != NULL))
-                tgt_len = strlen(tgt)+1;
+                tgt_len = strlen(tgt) + 1;
 
-        err = ll_prepare_mdc_op_data(&op_data, dir, NULL, name->name,
-                                     name->len, 0);
-        if (err)
-                GOTO(err_exit, err);
+        op_data = ll_prep_md_op_data(NULL, dir, NULL, name->name,
+                                     name->len, 0, opc, NULL);
+        if (IS_ERR(op_data))
+                GOTO(err_exit, err = PTR_ERR(op_data));
 
-        err = mdc_create(sbi->ll_mdc_exp, &op_data, tgt, tgt_len,
-                         mode, current->fsuid, current->fsgid,
-                         current->cap_effective, rdev, &request);
+        err = md_create(sbi->ll_md_exp, op_data, tgt, tgt_len, mode,
+                        cfs_curproc_fsuid(), cfs_curproc_fsgid(),
+                        cfs_curproc_cap_pack(), rdev, &request);
+        ll_finish_md_op_data(op_data);
         if (err)
                 GOTO(err_exit, err);
 
-        ll_update_times(request, REPLY_REC_OFF, dir);
+        ll_update_times(request, dir);
 
         if (dchild) {
-                err = ll_prep_inode(sbi->ll_osc_exp, &inode, request,
-                                    REPLY_REC_OFF, dchild->d_sb);
+                err = ll_prep_inode(&inode, request, dchild->d_sb);
                 if (err)
                      GOTO(err_exit, err);
 
                 d_drop(dchild);
                 d_instantiate(dchild, inode);
-                EXIT;
         }
+        EXIT;
 err_exit:
         ptlrpc_req_finished(request);
 
         return err;
 }
 
-
 static int ll_mknod_generic(struct inode *dir, struct qstr *name, int mode,
                             unsigned rdev, struct dentry *dchild)
 {
@@ -864,7 +872,7 @@ static int ll_mknod_generic(struct inode *dir, struct qstr *name, int mode,
                name->len, name->name, dir->i_ino, dir->i_generation, dir,
                mode, rdev);
 
-        mode &= ~current->fs->umask;
+        mode &= ~cfs_curproc_umask();
 
         switch (mode & S_IFMT) {
         case 0:
@@ -874,7 +882,8 @@ static int ll_mknod_generic(struct inode *dir, struct qstr *name, int mode,
         case S_IFBLK:
         case S_IFIFO:
         case S_IFSOCK:
-                err = ll_new_node(dir, name, NULL, mode, rdev, dchild);
+                err = ll_new_node(dir, name, NULL, mode, rdev, dchild,
+                                  LUSTRE_OPC_MKNOD);
                 break;
         case S_IFDIR:
                 err = -EPERM;
@@ -885,9 +894,8 @@ static int ll_mknod_generic(struct inode *dir, struct qstr *name, int mode,
         RETURN(err);
 }
 
-#if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
-#ifndef LUSTRE_KERNEL_VERSION
-static int ll_create_nd(struct inode *dir, struct dentry *dentry, int mode, struct nameidata *nd)
+static int ll_create_nd(struct inode *dir, struct dentry *dentry,
+                        int mode, struct nameidata *nd)
 {
         struct lookup_intent *it = ll_d2d(dentry)->lld_it;
         int rc;
@@ -919,18 +927,6 @@ out:
 
         return rc;
 }
-#else
-static int ll_create_nd(struct inode *dir, struct dentry *dentry, int mode, struct nameidata *nd)
-{
-
-        if (!nd || !nd->intent.d.lustre.it_disposition)
-                /* No saved request? Just mknod the file */
-                return ll_mknod_generic(dir, &dentry->d_name, mode, 0, dentry);
-
-        return ll_create_it(dir, dentry, mode, &nd->intent);
-}
-#endif
-#endif
 
 static int ll_symlink_generic(struct inode *dir, struct qstr *name,
                               const char *tgt, struct dentry *dchild)
@@ -943,17 +939,17 @@ static int ll_symlink_generic(struct inode *dir, struct qstr *name,
                dir, 3000, tgt);
 
         err = ll_new_node(dir, name, (char *)tgt, S_IFLNK | S_IRWXUGO,
-                          0, dchild);
+                          0, dchild, LUSTRE_OPC_SYMLINK);
         RETURN(err);
 }
 
 static int ll_link_generic(struct inode *src,  struct inode *dir,
                            struct qstr *name, struct dentry *dchild)
 {
+        struct ll_sb_info *sbi = ll_i2sbi(dir);
         struct ptlrpc_request *request = NULL;
-        struct mdc_op_data op_data;
+        struct md_op_data *op_data;
         int err;
-        struct ll_sb_info *sbi = ll_i2sbi(dir);
 
         ENTRY;
         CDEBUG(D_VFSTRACE,
@@ -961,92 +957,98 @@ static int ll_link_generic(struct inode *src,  struct inode *dir,
                src->i_ino, src->i_generation, src, dir->i_ino,
                dir->i_generation, dir, name->len, name->name);
 
-        err = ll_prepare_mdc_op_data(&op_data, src, dir, name->name,
-                                     name->len, 0);
+        op_data = ll_prep_md_op_data(NULL, src, dir, name->name, name->len,
+                                     0, LUSTRE_OPC_ANY, NULL);
+        if (IS_ERR(op_data))
+                RETURN(PTR_ERR(op_data));
+
+        err = md_link(sbi->ll_md_exp, op_data, &request);
+        ll_finish_md_op_data(op_data);
         if (err)
                 GOTO(out, err);
-        err = mdc_link(sbi->ll_mdc_exp, &op_data, &request);
-        if (err)
-               GOTO(out, err);
-
-        if (dchild) {
+        if (dchild)
                 d_drop(dchild);
-        }
-        ll_update_times(request, REPLY_REC_OFF, dir);
 
+        ll_update_times(request, dir);
         EXIT;
 out:
         ptlrpc_req_finished(request);
         RETURN(err);
 }
 
-static int ll_mkdir_generic(struct inode *dir, struct qstr *name, int mode,
-                            struct dentry *dchild)
+static int ll_mkdir_generic(struct inode *dir, struct qstr *name,
+                            int mode, struct dentry *dchild)
 
 {
         int err;
-
         ENTRY;
+
         CDEBUG(D_VFSTRACE, "VFS Op:name=%.*s,dir=%lu/%u(%p)\n",
                name->len, name->name, dir->i_ino, dir->i_generation, dir);
 
-        mode = (mode & (S_IRWXUGO|S_ISVTX) & ~current->fs->umask) | S_IFDIR;
-        err = ll_new_node(dir, name, NULL, mode, 0, dchild);
+        mode = (mode & (S_IRWXUGO|S_ISVTX) & ~cfs_curproc_umask()) | S_IFDIR;
+        err = ll_new_node(dir, name, NULL, mode, 0, dchild, LUSTRE_OPC_MKDIR);
 
         RETURN(err);
 }
 
+/* Try to find the child dentry by its name.
+   If found, put the result fid into @fid. */
+static void ll_get_child_fid(struct inode * dir, struct qstr *name,
+                             struct lu_fid *fid)
+{
+        struct dentry *parent, *child;
+
+        parent = list_entry(dir->i_dentry.next, struct dentry, d_alias);
+        child = d_lookup(parent, name);
+        if (child) {
+                if (child->d_inode)
+                        *fid = *ll_inode2fid(child->d_inode);
+                dput(child);
+        }
+}
+
 static int ll_rmdir_generic(struct inode *dir, struct dentry *dparent,
-                            struct qstr *name)
+                            struct dentry *dchild, struct qstr *name)
 {
         struct ptlrpc_request *request = NULL;
-        struct mdc_op_data op_data;
-        struct dentry *dentry;
+        struct md_op_data *op_data;
         int rc;
         ENTRY;
+
         CDEBUG(D_VFSTRACE, "VFS Op:name=%.*s,dir=%lu/%u(%p)\n",
                name->len, name->name, dir->i_ino, dir->i_generation, dir);
 
-        /* Check if we have something mounted at the dir we are going to delete
-         * In such a case there would always be dentry present. */
-        if (dparent) {
-                dentry = d_lookup(dparent, name);
-                if (dentry) {
-                        int mounted = d_mountpoint(dentry);
-                        dput(dentry);
-                        if (mounted)
-                                GOTO(out, rc = -EBUSY);
-                }
-        }
+        if (unlikely(ll_d_mountpoint(dparent, dchild, name)))
+                RETURN(-EBUSY);
 
-        rc = ll_prepare_mdc_op_data(&op_data, dir, NULL, name->name,
-                                    name->len, S_IFDIR);
-        if (rc)
-                GOTO(out, rc);
-        rc = mdc_unlink(ll_i2sbi(dir)->ll_mdc_exp, &op_data, &request);
-        if (rc)
-                GOTO(out, rc);
-        ll_update_times(request, REPLY_REC_OFF, dir);
+        op_data = ll_prep_md_op_data(NULL, dir, NULL, name->name, name->len,
+                                     S_IFDIR, LUSTRE_OPC_ANY, NULL);
+        if (IS_ERR(op_data))
+                RETURN(PTR_ERR(op_data));
 
-        EXIT;
-out:
+        ll_get_child_fid(dir, name, &op_data->op_fid3);
+        rc = md_unlink(ll_i2sbi(dir)->ll_md_exp, op_data, &request);
+        ll_finish_md_op_data(op_data);
+        if (rc == 0)
+                ll_update_times(request, dir);
         ptlrpc_req_finished(request);
-        return(rc);
+        RETURN(rc);
 }
 
 int ll_objects_destroy(struct ptlrpc_request *request, struct inode *dir)
 {
-        struct mds_body *body;
+        struct mdt_body *body;
         struct lov_mds_md *eadata;
         struct lov_stripe_md *lsm = NULL;
         struct obd_trans_info oti = { 0 };
         struct obdo *oa;
+        struct obd_capa *oc = NULL;
         int rc;
         ENTRY;
 
         /* req is swabbed so this is safe */
-        body = lustre_msg_buf(request->rq_repmsg, REPLY_REC_OFF, sizeof(*body));
-
+        body = req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY);
         if (!(body->valid & OBD_MD_FLEASIZE))
                 RETURN(0);
 
@@ -1059,152 +1061,137 @@ int ll_objects_destroy(struct ptlrpc_request *request, struct inode *dir)
          * to this file. Use this EA to unlink the objects on the OST.
          * It's opaque so we don't swab here; we leave it to obd_unpackmd() to
          * check it is complete and sensible. */
-        eadata = lustre_swab_repbuf(request, REPLY_REC_OFF + 1,
-                                    body->eadatasize, NULL);
+        eadata = req_capsule_server_sized_get(&request->rq_pill, &RMF_MDT_MD,
+                                              body->eadatasize);
         LASSERT(eadata != NULL);
-        if (eadata == NULL) {
-                CERROR("Can't unpack MDS EA data\n");
-                GOTO(out, rc = -EPROTO);
-        }
 
-        rc = obd_unpackmd(ll_i2obdexp(dir), &lsm, eadata, body->eadatasize);
+        rc = obd_unpackmd(ll_i2dtexp(dir), &lsm, eadata, body->eadatasize);
         if (rc < 0) {
                 CERROR("obd_unpackmd: %d\n", rc);
                 GOTO(out, rc);
         }
         LASSERT(rc >= sizeof(*lsm));
 
-        rc = obd_checkmd(ll_i2obdexp(dir), ll_i2mdcexp(dir), lsm);
-        if (rc)
-                GOTO(out_free_memmd, rc);
-
-        oa = obdo_alloc();
+        OBDO_ALLOC(oa);
         if (oa == NULL)
                 GOTO(out_free_memmd, rc = -ENOMEM);
 
         oa->o_id = lsm->lsm_object_id;
+        oa->o_seq = lsm->lsm_object_seq;
         oa->o_mode = body->mode & S_IFMT;
-        oa->o_valid = OBD_MD_FLID | OBD_MD_FLTYPE;
+        oa->o_valid = OBD_MD_FLID | OBD_MD_FLTYPE | OBD_MD_FLGROUP;
 
         if (body->valid & OBD_MD_FLCOOKIE) {
                 oa->o_valid |= OBD_MD_FLCOOKIE;
                 oti.oti_logcookies =
-                        lustre_msg_buf(request->rq_repmsg, REPLY_REC_OFF + 2,
-                                       sizeof(struct llog_cookie) *
-                                       lsm->lsm_stripe_count);
+                        req_capsule_server_sized_get(&request->rq_pill,
+                                                     &RMF_LOGCOOKIES,
+                                                   sizeof(struct llog_cookie) *
+                                                     lsm->lsm_stripe_count);
                 if (oti.oti_logcookies == NULL) {
                         oa->o_valid &= ~OBD_MD_FLCOOKIE;
                         body->valid &= ~OBD_MD_FLCOOKIE;
                 }
         }
 
-        rc = obd_destroy(ll_i2obdexp(dir), oa, lsm, &oti, ll_i2mdcexp(dir));
-        obdo_free(oa);
+        if (body->valid & OBD_MD_FLOSSCAPA) {
+                rc = md_unpack_capa(ll_i2mdexp(dir), request, &RMF_CAPA2, &oc);
+                if (rc)
+                        GOTO(out_free_memmd, rc);
+        }
+
+        rc = obd_destroy(ll_i2dtexp(dir), oa, lsm, &oti, ll_i2mdexp(dir), oc);
+        capa_put(oc);
+        OBDO_FREE(oa);
         if (rc)
                 CERROR("obd destroy objid "LPX64" error %d\n",
                        lsm->lsm_object_id, rc);
  out_free_memmd:
-        obd_free_memmd(ll_i2obdexp(dir), &lsm);
+        obd_free_memmd(ll_i2dtexp(dir), &lsm);
  out:
         return rc;
 }
 
-static int ll_unlink_generic(struct inode * dir, struct qstr *name)
+/* ll_unlink_generic() doesn't update the inode with the new link count.
+ * Instead, ll_ddelete() and ll_d_iput() will update it based upon if there
+ * is any lock existing. They will recycle dentries and inodes based upon locks
+ * too. b=20433 */
+static int ll_unlink_generic(struct inode *dir, struct dentry *dparent,
+                             struct dentry *dchild, struct qstr *name)
 {
         struct ptlrpc_request *request = NULL;
-        struct mdc_op_data op_data;
+        struct md_op_data *op_data;
         int rc;
         ENTRY;
-
         CDEBUG(D_VFSTRACE, "VFS Op:name=%.*s,dir=%lu/%u(%p)\n",
                name->len, name->name, dir->i_ino, dir->i_generation, dir);
 
-        rc = ll_prepare_mdc_op_data(&op_data, dir, NULL, name->name,
-                                    name->len, 0);
-        if (rc)
-                GOTO(out, rc);
-        rc = mdc_unlink(ll_i2sbi(dir)->ll_mdc_exp, &op_data, &request);
+        /*
+         * XXX: unlink bind mountpoint maybe call to here,
+         * just check it as vfs_unlink does.
+         */
+        if (unlikely(ll_d_mountpoint(dparent, dchild, name)))
+                RETURN(-EBUSY);
+
+        op_data = ll_prep_md_op_data(NULL, dir, NULL, name->name,
+                                     name->len, 0, LUSTRE_OPC_ANY, NULL);
+        if (IS_ERR(op_data))
+                RETURN(PTR_ERR(op_data));
+
+        ll_get_child_fid(dir, name, &op_data->op_fid3);
+        rc = md_unlink(ll_i2sbi(dir)->ll_md_exp, op_data, &request);
+        ll_finish_md_op_data(op_data);
         if (rc)
                 GOTO(out, rc);
 
-        ll_update_times(request, REPLY_REC_OFF, dir);
+        ll_update_times(request, dir);
 
         rc = ll_objects_destroy(request, dir);
-        if (rc)
-                GOTO(out, rc);
-        EXIT;
  out:
         ptlrpc_req_finished(request);
-        return(rc);
+        RETURN(rc);
 }
 
-static int ll_rename_generic(struct inode *src, struct qstr *src_name,
-                             struct inode *tgt, struct qstr *tgt_name)
+static int ll_rename_generic(struct inode *src, struct dentry *src_dparent,
+                             struct dentry *src_dchild, struct qstr *src_name,
+                             struct inode *tgt, struct dentry *tgt_dparent,
+                             struct dentry *tgt_dchild, struct qstr *tgt_name)
 {
         struct ptlrpc_request *request = NULL;
         struct ll_sb_info *sbi = ll_i2sbi(src);
-        struct mdc_op_data op_data;
+        struct md_op_data *op_data;
         int err;
-
         ENTRY;
         CDEBUG(D_VFSTRACE,"VFS Op:oldname=%.*s,src_dir=%lu/%u(%p),newname=%.*s,"
                "tgt_dir=%lu/%u(%p)\n", src_name->len, src_name->name,
                src->i_ino, src->i_generation, src, tgt_name->len,
                tgt_name->name, tgt->i_ino, tgt->i_generation, tgt);
 
-        err = ll_prepare_mdc_op_data(&op_data, src, tgt, NULL, 0, 0);
-        if (err)
-                GOTO(out, err);
-        err = mdc_rename(sbi->ll_mdc_exp, &op_data,
-                         src_name->name, src_name->len,
-                         tgt_name->name, tgt_name->len, &request);
-        if (err)
-                GOTO(out, err);
-        ll_update_times(request, REPLY_REC_OFF, src);
-        ll_update_times(request, REPLY_REC_OFF, tgt);
-        err = ll_objects_destroy(request, src);
-        if (err)
-                GOTO(out, err);
+        if (unlikely(ll_d_mountpoint(src_dparent, src_dchild, src_name) ||
+            ll_d_mountpoint(tgt_dparent, tgt_dchild, tgt_name)))
+                RETURN(-EBUSY);
+
+        op_data = ll_prep_md_op_data(NULL, src, tgt, NULL, 0, 0,
+                                     LUSTRE_OPC_ANY, NULL);
+        if (IS_ERR(op_data))
+                RETURN(PTR_ERR(op_data));
+
+        ll_get_child_fid(src, src_name, &op_data->op_fid3);
+        ll_get_child_fid(tgt, tgt_name, &op_data->op_fid4);
+        err = md_rename(sbi->ll_md_exp, op_data,
+                        src_name->name, src_name->len,
+                        tgt_name->name, tgt_name->len, &request);
+        ll_finish_md_op_data(op_data);
+        if (!err) {
+                ll_update_times(request, src);
+                ll_update_times(request, tgt);
+                err = ll_objects_destroy(request, src);
+        }
 
-        EXIT;
-out:
         ptlrpc_req_finished(request);
 
-        return(err);
-}
-
-#ifdef LUSTRE_KERNEL_VERSION
-static int ll_mknod_raw(struct nameidata *nd, int mode, dev_t rdev)
-{
-        return ll_mknod_generic(nd->dentry->d_inode, &nd->last, mode,rdev,NULL);
-}
-static int ll_rename_raw(struct nameidata *srcnd, struct nameidata *tgtnd)
-{
-        return ll_rename_generic(srcnd->dentry->d_inode, &srcnd->last,
-                                 tgtnd->dentry->d_inode, &tgtnd->last);
-}
-static int ll_link_raw(struct nameidata *srcnd, struct nameidata *tgtnd)
-{
-        return ll_link_generic(srcnd->dentry->d_inode, tgtnd->dentry->d_inode,
-                               &tgtnd->last, NULL);
-}
-static int ll_symlink_raw(struct nameidata *nd, const char *tgt)
-{
-        return ll_symlink_generic(nd->dentry->d_inode, &nd->last, tgt, NULL);
-}
-static int ll_rmdir_raw(struct nameidata *nd)
-{
-        return ll_rmdir_generic(nd->dentry->d_inode, nd->dentry, &nd->last);
-}
-static int ll_mkdir_raw(struct nameidata *nd, int mode)
-{
-        return ll_mkdir_generic(nd->dentry->d_inode, &nd->last, mode, NULL);
-}
-static int ll_unlink_raw(struct nameidata *nd)
-{
-        return ll_unlink_generic(nd->dentry->d_inode, &nd->last);
+        RETURN(err);
 }
-#endif
 
 static int ll_mknod(struct inode *dir, struct dentry *dchild, int mode,
                     ll_dev_t rdev)
@@ -1213,10 +1200,9 @@ static int ll_mknod(struct inode *dir, struct dentry *dchild, int mode,
                                 old_encode_dev(rdev), dchild);
 }
 
-#if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
 static int ll_unlink(struct inode * dir, struct dentry *dentry)
 {
-        return ll_unlink_generic(dir, &dentry->d_name);
+        return ll_unlink_generic(dir, NULL, dentry, &dentry->d_name);
 }
 static int ll_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 {
@@ -1224,7 +1210,7 @@ static int ll_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 }
 static int ll_rmdir(struct inode *dir, struct dentry *dentry)
 {
-        return ll_rmdir_generic(dir, NULL, &dentry->d_name);
+        return ll_rmdir_generic(dir, NULL, dentry, &dentry->d_name);
 }
 static int ll_symlink(struct inode *dir, struct dentry *dentry,
                       const char *oldname)
@@ -1234,35 +1220,28 @@ static int ll_symlink(struct inode *dir, struct dentry *dentry,
 static int ll_link(struct dentry *old_dentry, struct inode *dir,
                    struct dentry *new_dentry)
 {
-        return ll_link_generic(old_dentry->d_inode, dir,
-                               &new_dentry->d_name, new_dentry);
+        return ll_link_generic(old_dentry->d_inode, dir, &new_dentry->d_name,
+                               new_dentry);
 }
 static int ll_rename(struct inode *old_dir, struct dentry *old_dentry,
                      struct inode *new_dir, struct dentry *new_dentry)
 {
-        return ll_rename_generic(old_dir, &old_dentry->d_name, new_dir,
+        int err;
+        err = ll_rename_generic(old_dir, NULL,
+                                 old_dentry, &old_dentry->d_name,
+                                 new_dir, NULL, new_dentry,
                                  &new_dentry->d_name);
-}
+        if (!err) {
+#ifndef HAVE_FS_RENAME_DOES_D_MOVE
+                if (!S_ISDIR(old_dentry->d_inode->i_mode))
 #endif
+                        d_move(old_dentry, new_dentry);
+        }
+        return err;
+}
 
 struct inode_operations ll_dir_inode_operations = {
-#ifdef LUSTRE_KERNEL_VERSION
-        .link_raw           = ll_link_raw,
-        .unlink_raw         = ll_unlink_raw,
-        .symlink_raw        = ll_symlink_raw,
-        .mkdir_raw          = ll_mkdir_raw,
-        .rmdir_raw          = ll_rmdir_raw,
-        .mknod_raw          = ll_mknod_raw,
-        .rename_raw         = ll_rename_raw,
-        .setattr            = ll_setattr,
-        .setattr_raw        = ll_setattr_raw,
-#endif
         .mknod              = ll_mknod,
-#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,5,0))
-        .create_it          = ll_create_it,
-        .lookup_it          = ll_lookup_it,
-        .revalidate_it      = ll_inode_revalidate_it,
-#else
         .lookup             = ll_lookup_nd,
         .create             = ll_create_nd,
         /* We need all these non-raw things for NFSD, to not patch it. */
@@ -1274,7 +1253,6 @@ struct inode_operations ll_dir_inode_operations = {
         .rename             = ll_rename,
         .setattr            = ll_setattr,
         .getattr            = ll_getattr,
-#endif
         .permission         = ll_inode_permission,
         .setxattr           = ll_setxattr,
         .getxattr           = ll_getxattr,
@@ -1283,15 +1261,8 @@ struct inode_operations ll_dir_inode_operations = {
 };
 
 struct inode_operations ll_special_inode_operations = {
-#ifdef LUSTRE_KERNEL_VERSION
-        .setattr_raw    = ll_setattr_raw,
-#endif
         .setattr        = ll_setattr,
-#if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
         .getattr        = ll_getattr,
-#else
-        .revalidate_it  = ll_inode_revalidate_it,
-#endif
         .permission     = ll_inode_permission,
         .setxattr       = ll_setxattr,
         .getxattr       = ll_getxattr,