Whamcloud - gitweb
LU-1247 obdfilter: fix invalid check of precrate objects
[fs/lustre-release.git] / lustre / obdfilter / filter.c
index afbab0b..75fd01b 100644 (file)
@@ -28,9 +28,8 @@
 /*
  * Copyright (c) 2001, 2010, Oracle and/or its affiliates. All rights reserved.
  * Use is subject to license terms.
- */
-/*
- * Copyright (c) 2011 Whamcloud, Inc.
+ *
+ * Copyright (c) 2011, 2012, Whamcloud, Inc.
  */
 /*
  * This file is part of Lustre, http://www.lustre.org/
@@ -138,12 +137,12 @@ int filter_finish_transno(struct obd_export *exp, struct inode *inode,
         if (!exp->exp_obd->obd_replayable || oti == NULL)
                 RETURN(rc);
 
-        cfs_mutex_down(&ted->ted_lcd_lock);
+        cfs_mutex_lock(&ted->ted_lcd_lock);
         lcd = ted->ted_lcd;
         /* if the export has already been disconnected, we have no last_rcvd slot,
          * update server data with latest transno then */
         if (lcd == NULL) {
-                cfs_mutex_up(&ted->ted_lcd_lock);
+                cfs_mutex_unlock(&ted->ted_lcd_lock);
                 CWARN("commit transaction for disconnected client %s: rc %d\n",
                       exp->exp_client_uuid.uuid, rc);
                 err = filter_update_server_data(exp->exp_obd);
@@ -168,7 +167,7 @@ int filter_finish_transno(struct obd_export *exp, struct inode *inode,
                         exp->exp_vbr_failed = 1;
                         cfs_spin_unlock(&exp->exp_lock);
                         cfs_spin_unlock(&obt->obt_lut->lut_translock);
-                        cfs_mutex_up(&ted->ted_lcd_lock);
+                        cfs_mutex_unlock(&ted->ted_lcd_lock);
                         RETURN(-EOVERFLOW);
                 }
         }
@@ -210,7 +209,7 @@ int filter_finish_transno(struct obd_export *exp, struct inode *inode,
 
         CDEBUG(log_pri, "wrote trans "LPU64" for client %s at #%d: err = %d\n",
                last_rcvd, lcd->lcd_uuid, ted->ted_lr_idx, err);
-        cfs_mutex_up(&ted->ted_lcd_lock);
+        cfs_mutex_unlock(&ted->ted_lcd_lock);
         RETURN(rc);
 }
 
@@ -355,7 +354,7 @@ static int filter_client_add(struct obd_device *obd, struct obd_export *exp,
         ted->ted_lr_idx = cl_idx;
         ted->ted_lr_off = le32_to_cpu(lsd->lsd_client_start) +
                           cl_idx * le16_to_cpu(lsd->lsd_client_size);
-        cfs_init_mutex(&ted->ted_lcd_lock);
+        cfs_mutex_init(&ted->ted_lcd_lock);
         LASSERTF(ted->ted_lr_off > 0, "ted_lr_off = %llu\n", ted->ted_lr_off);
 
         CDEBUG(D_INFO, "client at index %d (%llu) with UUID '%s' added\n",
@@ -456,12 +455,12 @@ static int filter_client_del(struct obd_export *exp)
          * be in server data or in client data in case of failure */
         filter_update_server_data(exp->exp_obd);
 
-        cfs_mutex_down(&ted->ted_lcd_lock);
+        cfs_mutex_lock(&ted->ted_lcd_lock);
         memset(ted->ted_lcd->lcd_uuid, 0, sizeof ted->ted_lcd->lcd_uuid);
         rc = fsfilt_write_record(exp->exp_obd, obt->obt_rcvd_filp,
                                  ted->ted_lcd,
                                  sizeof(*ted->ted_lcd), &off, 0);
-        cfs_mutex_up(&ted->ted_lcd_lock);
+        cfs_mutex_unlock(&ted->ted_lcd_lock);
         pop_ctxt(&saved, &exp->exp_obd->obd_lvfs_ctxt, NULL);
 
         CDEBUG(rc == 0 ? D_INFO : D_ERROR,
@@ -1199,14 +1198,14 @@ static int filter_read_groups(struct obd_device *obd, int last_group,
         struct filter_obd *filter = &obd->u.filter;
         int old_count, group, rc = 0;
 
-        cfs_down(&filter->fo_init_lock);
+        cfs_mutex_lock(&filter->fo_init_lock);
         old_count = filter->fo_group_count;
         for (group = old_count; group <= last_group; group++) {
                 rc = filter_read_group_internal(obd, group, create);
                 if (rc != 0)
                         break;
         }
-        cfs_up(&filter->fo_init_lock);
+        cfs_mutex_unlock(&filter->fo_init_lock);
         return rc;
 }
 
@@ -1778,7 +1777,7 @@ static int filter_intent_policy(struct ldlm_namespace *ns,
          * therefore, that res->lr_lvb_data cannot increase beyond the
          * end of already granted lock. As a result, it is safe to
          * check against "stale" reply_lvb->lvb_size value without
-         * res->lr_lvb_sem.
+         * res->lr_lvb_mutex.
          */
         arg.size = reply_lvb->lvb_size;
         arg.victim = &l;
@@ -1970,6 +1969,7 @@ int filter_common_setup(struct obd_device *obd, struct lustre_cfg* lcfg,
 {
         struct filter_obd *filter = &obd->u.filter;
         struct vfsmount *mnt;
+        struct file_system_type *type;
         struct lustre_mount_info *lmi;
         struct obd_uuid uuid;
         __u8 *uuid_ptr;
@@ -1994,9 +1994,14 @@ int filter_common_setup(struct obd_device *obd, struct lustre_cfg* lcfg,
         } else {
                 /* old path - used by lctl */
                 CERROR("Using old MDS mount method\n");
-                mnt = ll_kern_mount(lustre_cfg_string(lcfg, 2),
-                                    MS_NOATIME|MS_NODIRATIME,
-                                    lustre_cfg_string(lcfg, 1), option);
+                type = get_fs_type(lustre_cfg_string(lcfg, 2));
+                if (!type) {
+                        CERROR("get_fs_type failed\n");
+                        RETURN(-ENODEV);
+                }
+                mnt = vfs_kern_mount(type, MS_NOATIME|MS_NODIRATIME,
+                                     lustre_cfg_string(lcfg, 1), option);
+                cfs_module_put(type->owner);
                 if (IS_ERR(mnt)) {
                         rc = PTR_ERR(mnt);
                         LCONSOLE_ERROR_MSG(0x135, "Can't mount disk %s (%d)\n",
@@ -2039,7 +2044,9 @@ int filter_common_setup(struct obd_device *obd, struct lustre_cfg* lcfg,
         filter->fo_fstype = mnt->mnt_sb->s_type->name;
         CDEBUG(D_SUPER, "%s: mnt = %p\n", filter->fo_fstype, mnt);
 
-        fsfilt_setup(obd, obd->u.obt.obt_sb);
+        rc = fsfilt_setup(obd, obd->u.obt.obt_sb);
+        if (rc)
+                GOTO(err_ops, rc);
 
         OBD_SET_CTXT_MAGIC(&obd->obd_lvfs_ctxt);
         obd->obd_lvfs_ctxt.pwdmnt = mnt;
@@ -2047,15 +2054,15 @@ int filter_common_setup(struct obd_device *obd, struct lustre_cfg* lcfg,
         obd->obd_lvfs_ctxt.fs = get_ds();
         obd->obd_lvfs_ctxt.cb_ops = filter_lvfs_ops;
 
-        cfs_init_mutex(&filter->fo_init_lock);
+        cfs_mutex_init(&filter->fo_init_lock);
         filter->fo_committed_group = 0;
         filter->fo_destroys_in_progress = 0;
         for (i = 0; i < 32; i++)
-                cfs_sema_init(&filter->fo_create_locks[i], 1);
+                cfs_mutex_init(&filter->fo_create_locks[i]);
 
         cfs_spin_lock_init(&filter->fo_objidlock);
         CFS_INIT_LIST_HEAD(&filter->fo_export_list);
-        cfs_sema_init(&filter->fo_alloc_lock, 1);
+        cfs_mutex_init(&filter->fo_alloc_lock);
         init_brw_stats(&filter->fo_filter_stats);
         cfs_spin_lock_init(&filter->fo_flags_lock);
         filter->fo_read_cache = 1; /* enable read-only cache by default */
@@ -2403,20 +2410,20 @@ static int filter_llog_finish(struct obd_device *obd, int count)
                  * This is safe to do, as llog is already synchronized
                  * and its import may go.
                  */
-                cfs_mutex_down(&ctxt->loc_sem);
+                cfs_mutex_lock(&ctxt->loc_mutex);
                 if (ctxt->loc_imp) {
                         class_import_put(ctxt->loc_imp);
                         ctxt->loc_imp = NULL;
                 }
-                cfs_mutex_up(&ctxt->loc_sem);
+                cfs_mutex_unlock(&ctxt->loc_mutex);
                 llog_ctxt_put(ctxt);
         }
 
         if (filter->fo_lcm) {
-                cfs_mutex_down(&ctxt->loc_sem);
+                cfs_mutex_lock(&ctxt->loc_mutex);
                 llog_recov_thread_fini(filter->fo_lcm, obd->obd_force);
                 filter->fo_lcm = NULL;
-                cfs_mutex_up(&ctxt->loc_sem);
+                cfs_mutex_unlock(&ctxt->loc_mutex);
         }
         RETURN(filter_olg_fini(&obd->obd_olg));
 }
@@ -3159,6 +3166,7 @@ static int filter_getattr(struct obd_export *exp, struct obd_info *oinfo)
 {
         struct dentry *dentry = NULL;
         struct obd_device *obd;
+        __u64 curr_version;
         int rc = 0;
         ENTRY;
 
@@ -3179,7 +3187,14 @@ static int filter_getattr(struct obd_export *exp, struct obd_info *oinfo)
 
         /* Limit the valid bits in the return data to what we actually use */
         oinfo->oi_oa->o_valid = OBD_MD_FLID;
-        obdo_from_inode(oinfo->oi_oa, dentry->d_inode, NULL, FILTER_VALID_FLAGS);
+        obdo_from_inode(oinfo->oi_oa, dentry->d_inode, FILTER_VALID_FLAGS);
+
+        /* Store inode version in reply */
+        curr_version = fsfilt_get_version(exp->exp_obd, dentry->d_inode);
+        if ((__s64)curr_version != -EOPNOTSUPP) {
+                oinfo->oi_oa->o_valid |= OBD_MD_FLDATAVERSION;
+                oinfo->oi_oa->o_data_version = curr_version;
+        }
 
         f_dput(dentry);
         RETURN(rc);
@@ -3496,7 +3511,7 @@ int filter_setattr(struct obd_export *exp, struct obd_info *oinfo,
         oa->o_valid = OBD_MD_FLID;
 
         /* Quota release need uid/gid info */
-        obdo_from_inode(oa, dentry->d_inode, NULL,
+        obdo_from_inode(oa, dentry->d_inode,
                         FILTER_VALID_FLAGS | OBD_MD_FLUID | OBD_MD_FLGID);
 
         EXIT;
@@ -3572,7 +3587,7 @@ static int filter_destroy_precreated(struct obd_export *exp, struct obdo *oa,
         int skip_orphan;
         ENTRY;
 
-        LASSERT(down_trylock(&filter->fo_create_locks[oa->o_seq]) != 0);
+        LASSERT_MUTEX_LOCKED(&filter->fo_create_locks[oa->o_seq]);
 
         memset(&doa, 0, sizeof(doa));
 
@@ -3654,18 +3669,18 @@ static int filter_handle_precreate(struct obd_export *exp, struct obdo *oa,
                 }
                 /* This causes inflight precreates to abort and drop lock */
                 cfs_set_bit(group, &filter->fo_destroys_in_progress);
-                cfs_down(&filter->fo_create_locks[group]);
+                cfs_mutex_lock(&filter->fo_create_locks[group]);
                 if (!cfs_test_bit(group, &filter->fo_destroys_in_progress)) {
                         CERROR("%s:["LPU64"] destroys_in_progress already cleared\n",
                                exp->exp_obd->obd_name, group);
-                        cfs_up(&filter->fo_create_locks[group]);
+                        cfs_mutex_unlock(&filter->fo_create_locks[group]);
                         RETURN(0);
                 }
                 diff = oa->o_id - last;
                 CDEBUG(D_HA, "filter_last_id() = "LPU64" -> diff = %d\n",
                        last, diff);
 
-                if (-diff > OST_MAX_PRECREATE) {
+                if (-diff > (OST_MAX_PRECREATE * 3) / 2) {
                         CERROR("%s: ignoring bogus orphan destroy request: "
                                "obdid "LPU64" last_id "LPU64"\n", obd->obd_name,
                                oa->o_id, last);
@@ -3683,7 +3698,7 @@ static int filter_handle_precreate(struct obd_export *exp, struct obdo *oa,
                         cfs_clear_bit(group, &filter->fo_destroys_in_progress);
                 }
         } else {
-                cfs_down(&filter->fo_create_locks[group]);
+                cfs_mutex_lock(&filter->fo_create_locks[group]);
                 if (oti->oti_conn_cnt < exp->exp_conn_cnt) {
                         CERROR("%s: dropping old precreate request\n",
                                obd->obd_name);
@@ -3712,7 +3727,7 @@ static int filter_handle_precreate(struct obd_export *exp, struct obdo *oa,
         /* else diff == 0 */
         GOTO(out, rc = 0);
 out:
-        cfs_up(&filter->fo_create_locks[group]);
+        cfs_mutex_unlock(&filter->fo_create_locks[group]);
         return rc;
 }
 
@@ -3831,7 +3846,7 @@ static int filter_precreate(struct obd_device *obd, struct obdo *oa,
 
         filter = &obd->u.filter;
 
-        LASSERT(down_trylock(&filter->fo_create_locks[group]) != 0);
+        LASSERT_MUTEX_LOCKED(&filter->fo_create_locks[group]);
 
         OBD_FAIL_TIMEOUT(OBD_FAIL_TGT_DELAY_PRECREATE, obd_timeout / 2);
 
@@ -4101,9 +4116,9 @@ int filter_create(struct obd_export *exp, struct obdo *oa,
                         rc = -EINVAL;
                 } else {
                         diff = 1;
-                        cfs_down(&filter->fo_create_locks[oa->o_seq]);
+                        cfs_mutex_lock(&filter->fo_create_locks[oa->o_seq]);
                         rc = filter_precreate(obd, oa, oa->o_seq, &diff);
-                        cfs_up(&filter->fo_create_locks[oa->o_seq]);
+                        cfs_mutex_unlock(&filter->fo_create_locks[oa->o_seq]);
                 }
         } else {
                 rc = filter_handle_precreate(exp, oa, oa->o_seq, oti);
@@ -4259,7 +4274,7 @@ int filter_destroy(struct obd_export *exp, struct obdo *oa,
         cleanup_phase = 4; /* fsfilt_commit */
 
         /* Quota release need uid/gid of inode */
-        obdo_from_inode(oa, dchild->d_inode, NULL, OBD_MD_FLUID|OBD_MD_FLGID);
+        obdo_from_inode(oa, dchild->d_inode, OBD_MD_FLUID | OBD_MD_FLGID);
 
         filter_fmd_drop(exp, oa->o_id, oa->o_seq);
 
@@ -4391,8 +4406,7 @@ static int filter_sync(struct obd_export *exp, struct obd_info *oinfo,
         UNLOCK_INODE_MUTEX(dentry->d_inode);
 
         oinfo->oi_oa->o_valid = OBD_MD_FLID;
-        obdo_from_inode(oinfo->oi_oa, dentry->d_inode, NULL,
-                        FILTER_VALID_FLAGS);
+        obdo_from_inode(oinfo->oi_oa, dentry->d_inode, FILTER_VALID_FLAGS);
 
         pop_ctxt(&saved, &exp->exp_obd->obd_lvfs_ctxt, NULL);
 
@@ -4633,8 +4647,8 @@ int filter_iocontrol(unsigned int cmd, struct obd_export *exp,
                 CDEBUG(D_HA, "syncing ost %s\n", obd->obd_name);
                 rc = fsfilt_sync(obd, obd->u.obt.obt_sb);
 
-                lvfs_set_rdonly(obd, obd->u.obt.obt_sb);
-                RETURN(0);
+                rc = lvfs_set_rdonly(obd, obd->u.obt.obt_sb);
+                RETURN(rc);
         }
 
         case OBD_IOC_CATLOGLIST: {
@@ -4755,7 +4769,6 @@ static struct obd_ops filter_obd_ops = {
         .o_create         = filter_create,
         .o_setattr        = filter_setattr,
         .o_destroy        = filter_destroy,
-        .o_brw            = filter_brw,
         .o_punch          = filter_truncate,
         .o_sync           = filter_sync,
         .o_preprw         = filter_preprw,