Merge master.kernel.org:/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw

* master.kernel.org:/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw: (73 commits)
  [DLM] Clean up lowcomms
  [GFS2] Change gfs2_fsync() to use write_inode_now()
  [GFS2] Fix indent in recovery.c
  [GFS2] Don't flush everything on fdatasync
  [GFS2] Add a comment about reading the super block
  [GFS2] Mount problem with the GFS2 code
  [GFS2] Remove gfs2_check_acl()
  [DLM] fix format warnings in rcom.c and recoverd.c
  [GFS2] lock function parameter
  [DLM] don't accept replies to old recovery messages
  [DLM] fix size of STATUS_REPLY message
  [GFS2] fs/gfs2/log.c:log_bmap() fix printk format warning
  [DLM] fix add_requestqueue checking nodes list
  [GFS2] Fix recursive locking in gfs2_getattr
  [GFS2] Fix recursive locking in gfs2_permission
  [GFS2] Reduce number of arguments to meta_io.c:getbuf()
  [GFS2] Move gfs2_meta_syncfs() into log.c
  [GFS2] Fix journal flush problem
  [GFS2] mark_inode_dirty after write to stuffed file
  [GFS2] Fix glock ordering on inode creation
  ...

fs/dlm/Kconfig

 menu "Distributed Lock Manager"
-	depends on INET && IP_SCTP && EXPERIMENTAL
+	depends on EXPERIMENTAL && INET
 
 config DLM
 	tristate "Distributed Lock Manager (DLM)"
 	depends on IPV6 || IPV6=n
 	select CONFIGFS_FS
+	select IP_SCTP if DLM_SCTP
 	help
 	A general purpose distributed lock manager for kernel or userspace
 	applications.
 
+choice
+	prompt "Select DLM communications protocol"
+	depends on DLM
+	default DLM_TCP
+	help
+	The DLM Can use TCP or SCTP for it's network communications.
+	SCTP supports multi-homed operations whereas TCP doesn't.
+	However, SCTP seems to have stability problems at the moment.
+
+config DLM_TCP
+	bool "TCP/IP"
+
+config DLM_SCTP
+	bool "SCTP"
+
+endchoice
+
 config DLM_DEBUG
 	bool "DLM debugging"
 	depends on DLM

fs/dlm/Makefile

@@ -4,7 +4,6 @@ dlm-y :=			ast.o \
 				dir.o \
 				lock.o \
 				lockspace.o \
-				lowcomms.o \
 				main.o \
 				member.o \
 				memory.o \
@@ -17,3 +16,6 @@ dlm-y :=			ast.o \
 				util.o
 dlm-$(CONFIG_DLM_DEBUG) +=	debug_fs.o
 
+dlm-$(CONFIG_DLM_TCP)   += lowcomms-tcp.o
+
+dlm-$(CONFIG_DLM_SCTP)  += lowcomms-sctp.o
\ No newline at end of file

fs/dlm/dlm_internal.h

@@ -471,6 +471,7 @@ struct dlm_ls {
 	char			*ls_recover_buf;
 	int			ls_recover_nodeid; /* for debugging */
 	uint64_t		ls_rcom_seq;
+	spinlock_t		ls_rcom_spin;
 	struct list_head	ls_recover_list;
 	spinlock_t		ls_recover_list_lock;
 	int			ls_recover_list_count;
@@ -488,7 +489,8 @@ struct dlm_ls {
 #define LSFL_RUNNING		1
 #define LSFL_RECOVERY_STOP	2
 #define LSFL_RCOM_READY		3
-#define LSFL_UEVENT_WAIT	4
+#define LSFL_RCOM_WAIT		4
+#define LSFL_UEVENT_WAIT	5
 
 /* much of this is just saving user space pointers associated with the
    lock that we pass back to the user lib with an ast */

fs/dlm/lock.c

@@ -2372,6 +2372,7 @@ static int send_lookup_reply(struct dlm_ls *ls, struct dlm_message *ms_in,
 static void receive_flags(struct dlm_lkb *lkb, struct dlm_message *ms)
 {
 	lkb->lkb_exflags = ms->m_exflags;
+	lkb->lkb_sbflags = ms->m_sbflags;
 	lkb->lkb_flags = (lkb->lkb_flags & 0xFFFF0000) |
 		         (ms->m_flags & 0x0000FFFF);
 }
@@ -3028,10 +3029,17 @@ int dlm_receive_message(struct dlm_header *hd, int nodeid, int recovery)
 
 	while (1) {
 		if (dlm_locking_stopped(ls)) {
-			if (!recovery)
-				dlm_add_requestqueue(ls, nodeid, hd);
-			error = -EINTR;
-			goto out;
+			if (recovery) {
+				error = -EINTR;
+				goto out;
+			}
+			error = dlm_add_requestqueue(ls, nodeid, hd);
+			if (error == -EAGAIN)
+				continue;
+			else {
+				error = -EINTR;
+				goto out;
+			}
 		}
 
 		if (lock_recovery_try(ls))

fs/dlm/lockspace.c

@@ -22,6 +22,7 @@
 #include "memory.h"
 #include "lock.h"
 #include "recover.h"
+#include "requestqueue.h"
 
 #ifdef CONFIG_DLM_DEBUG
 int dlm_create_debug_file(struct dlm_ls *ls);
@@ -478,6 +479,8 @@ static int new_lockspace(char *name, int namelen, void **lockspace,
 	ls->ls_recoverd_task = NULL;
 	mutex_init(&ls->ls_recoverd_active);
 	spin_lock_init(&ls->ls_recover_lock);
+	spin_lock_init(&ls->ls_rcom_spin);
+	get_random_bytes(&ls->ls_rcom_seq, sizeof(uint64_t));
 	ls->ls_recover_status = 0;
 	ls->ls_recover_seq = 0;
 	ls->ls_recover_args = NULL;
@@ -684,6 +687,7 @@ static int release_lockspace(struct dlm_ls *ls, int force)
 	 * Free structures on any other lists
 	 */
 
+	dlm_purge_requestqueue(ls);
 	kfree(ls->ls_recover_args);
 	dlm_clear_free_entries(ls);
 	dlm_clear_members(ls);

fs/dlm/lowcomms.h

@@ -14,8 +14,6 @@
 #ifndef __LOWCOMMS_DOT_H__
 #define __LOWCOMMS_DOT_H__
 
-int dlm_lowcomms_init(void);
-void dlm_lowcomms_exit(void);
 int dlm_lowcomms_start(void);
 void dlm_lowcomms_stop(void);
 int dlm_lowcomms_close(int nodeid);

fs/dlm/main.c

@@ -16,7 +16,6 @@
 #include "lock.h"
 #include "user.h"
 #include "memory.h"
-#include "lowcomms.h"
 #include "config.h"
 
 #ifdef CONFIG_DLM_DEBUG
@@ -47,20 +46,14 @@ static int __init init_dlm(void)
 	if (error)
 		goto out_config;
 
-	error = dlm_lowcomms_init();
-	if (error)
-		goto out_debug;
-
 	error = dlm_user_init();
 	if (error)
-		goto out_lowcomms;
+		goto out_debug;
 
 	printk("DLM (built %s %s) installed\n", __DATE__, __TIME__);
 
 	return 0;
 
- out_lowcomms:
-	dlm_lowcomms_exit();
  out_debug:
 	dlm_unregister_debugfs();
  out_config:
@@ -76,7 +69,6 @@ static int __init init_dlm(void)
 static void __exit exit_dlm(void)
 {
 	dlm_user_exit();
-	dlm_lowcomms_exit();
 	dlm_config_exit();
 	dlm_memory_exit();
 	dlm_lockspace_exit();

fs/dlm/member.c

@@ -186,6 +186,14 @@ int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv, int *neg_out)
 	struct dlm_member *memb, *safe;
 	int i, error, found, pos = 0, neg = 0, low = -1;
 
+	/* previously removed members that we've not finished removing need to
+	   count as a negative change so the "neg" recovery steps will happen */
+
+	list_for_each_entry(memb, &ls->ls_nodes_gone, list) {
+		log_debug(ls, "prev removed member %d", memb->nodeid);
+		neg++;
+	}
+
 	/* move departed members from ls_nodes to ls_nodes_gone */
 
 	list_for_each_entry_safe(memb, safe, &ls->ls_nodes, list) {

fs/dlm/rcom.c

@@ -90,13 +90,28 @@ static int check_config(struct dlm_ls *ls, struct rcom_config *rf, int nodeid)
 	return 0;
 }
 
+static void allow_sync_reply(struct dlm_ls *ls, uint64_t *new_seq)
+{
+	spin_lock(&ls->ls_rcom_spin);
+	*new_seq = ++ls->ls_rcom_seq;
+	set_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
+	spin_unlock(&ls->ls_rcom_spin);
+}
+
+static void disallow_sync_reply(struct dlm_ls *ls)
+{
+	spin_lock(&ls->ls_rcom_spin);
+	clear_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
+	clear_bit(LSFL_RCOM_READY, &ls->ls_flags);
+	spin_unlock(&ls->ls_rcom_spin);
+}
+
 int dlm_rcom_status(struct dlm_ls *ls, int nodeid)
 {
 	struct dlm_rcom *rc;
 	struct dlm_mhandle *mh;
 	int error = 0;
 
-	memset(ls->ls_recover_buf, 0, dlm_config.buffer_size);
 	ls->ls_recover_nodeid = nodeid;
 
 	if (nodeid == dlm_our_nodeid()) {
@@ -108,12 +123,14 @@ int dlm_rcom_status(struct dlm_ls *ls, int nodeid)
 	error = create_rcom(ls, nodeid, DLM_RCOM_STATUS, 0, &rc, &mh);
 	if (error)
 		goto out;
-	rc->rc_id = ++ls->ls_rcom_seq;
+
+	allow_sync_reply(ls, &rc->rc_id);
+	memset(ls->ls_recover_buf, 0, dlm_config.buffer_size);
 
 	send_rcom(ls, mh, rc);
 
 	error = dlm_wait_function(ls, &rcom_response);
-	clear_bit(LSFL_RCOM_READY, &ls->ls_flags);
+	disallow_sync_reply(ls);
 	if (error)
 		goto out;
 
@@ -150,14 +167,21 @@ static void receive_rcom_status(struct dlm_ls *ls, struct dlm_rcom *rc_in)
 
 static void receive_sync_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in)
 {
-	if (rc_in->rc_id != ls->ls_rcom_seq) {
-		log_debug(ls, "reject old reply %d got %llx wanted %llx",
-			  rc_in->rc_type, rc_in->rc_id, ls->ls_rcom_seq);
-		return;
+	spin_lock(&ls->ls_rcom_spin);
+	if (!test_bit(LSFL_RCOM_WAIT, &ls->ls_flags) ||
+	    rc_in->rc_id != ls->ls_rcom_seq) {
+		log_debug(ls, "reject reply %d from %d seq %llx expect %llx",
+			  rc_in->rc_type, rc_in->rc_header.h_nodeid,
+			  (unsigned long long)rc_in->rc_id,
+			  (unsigned long long)ls->ls_rcom_seq);
+		goto out;
 	}
 	memcpy(ls->ls_recover_buf, rc_in, rc_in->rc_header.h_length);
 	set_bit(LSFL_RCOM_READY, &ls->ls_flags);
+	clear_bit(LSFL_RCOM_WAIT, &ls->ls_flags);
 	wake_up(&ls->ls_wait_general);
+ out:
+	spin_unlock(&ls->ls_rcom_spin);
 }
 
 static void receive_rcom_status_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in)
@@ -171,7 +195,6 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len)
 	struct dlm_mhandle *mh;
 	int error = 0, len = sizeof(struct dlm_rcom);
 
-	memset(ls->ls_recover_buf, 0, dlm_config.buffer_size);
 	ls->ls_recover_nodeid = nodeid;
 
 	if (nodeid == dlm_our_nodeid()) {
@@ -185,12 +208,14 @@ int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name, int last_len)
 	if (error)
 		goto out;
 	memcpy(rc->rc_buf, last_name, last_len);
-	rc->rc_id = ++ls->ls_rcom_seq;
+
+	allow_sync_reply(ls, &rc->rc_id);
+	memset(ls->ls_recover_buf, 0, dlm_config.buffer_size);
 
 	send_rcom(ls, mh, rc);
 
 	error = dlm_wait_function(ls, &rcom_response);
-	clear_bit(LSFL_RCOM_READY, &ls->ls_flags);
+	disallow_sync_reply(ls);
  out:
 	return error;
 }
@@ -370,9 +395,10 @@ static void receive_rcom_lock_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in)
 static int send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in)
 {
 	struct dlm_rcom *rc;
+	struct rcom_config *rf;
 	struct dlm_mhandle *mh;
 	char *mb;
-	int mb_len = sizeof(struct dlm_rcom);
+	int mb_len = sizeof(struct dlm_rcom) + sizeof(struct rcom_config);
 
 	mh = dlm_lowcomms_get_buffer(nodeid, mb_len, GFP_KERNEL, &mb);
 	if (!mh)
@@ -391,6 +417,9 @@ static int send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in)
 	rc->rc_id = rc_in->rc_id;
 	rc->rc_result = -ESRCH;
 
+	rf = (struct rcom_config *) rc->rc_buf;
+	rf->rf_lvblen = -1;
+
 	dlm_rcom_out(rc);
 	dlm_lowcomms_commit_buffer(mh);
 
@@ -412,9 +441,10 @@ void dlm_receive_rcom(struct dlm_header *hd, int nodeid)
 
 	ls = dlm_find_lockspace_global(hd->h_lockspace);
 	if (!ls) {
-		log_print("lockspace %x from %d not found",
-			  hd->h_lockspace, nodeid);
-		send_ls_not_ready(nodeid, rc);
+		log_print("lockspace %x from %d type %x not found",
+			  hd->h_lockspace, nodeid, rc->rc_type);
+		if (rc->rc_type == DLM_RCOM_STATUS)
+			send_ls_not_ready(nodeid, rc);
 		return;
 	}
 

fs/dlm/recover.c

@@ -252,6 +252,7 @@ static void recover_list_clear(struct dlm_ls *ls)
 	spin_lock(&ls->ls_recover_list_lock);
 	list_for_each_entry_safe(r, s, &ls->ls_recover_list, res_recover_list) {
 		list_del_init(&r->res_recover_list);
+		r->res_recover_locks_count = 0;
 		dlm_put_rsb(r);
 		ls->ls_recover_list_count--;
 	}

fs/dlm/recoverd.c

@@ -45,7 +45,7 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 	unsigned long start;
 	int error, neg = 0;
 
-	log_debug(ls, "recover %llx", rv->seq);
+	log_debug(ls, "recover %llx", (unsigned long long)rv->seq);
 
 	mutex_lock(&ls->ls_recoverd_active);
 
@@ -93,14 +93,6 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 		goto fail;
 	}
 
-	/*
-	 * Purge directory-related requests that are saved in requestqueue.
-	 * All dir requests from before recovery are invalid now due to the dir
-	 * rebuild and will be resent by the requesting nodes.
-	 */
-
-	dlm_purge_requestqueue(ls);
-
 	/*
 	 * Wait for all nodes to complete directory rebuild.
 	 */
@@ -164,10 +156,31 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 		 */
 
 		dlm_recover_rsbs(ls);
+	} else {
+		/*
+		 * Other lockspace members may be going through the "neg" steps
+		 * while also adding us to the lockspace, in which case they'll
+		 * be doing the recover_locks (RS_LOCKS) barrier.
+		 */
+		dlm_set_recover_status(ls, DLM_RS_LOCKS);
+
+		error = dlm_recover_locks_wait(ls);
+		if (error) {
+			log_error(ls, "recover_locks_wait failed %d", error);
+			goto fail;
+		}
 	}
 
 	dlm_release_root_list(ls);
 
+	/*
+	 * Purge directory-related requests that are saved in requestqueue.
+	 * All dir requests from before recovery are invalid now due to the dir
+	 * rebuild and will be resent by the requesting nodes.
+	 */
+
+	dlm_purge_requestqueue(ls);
+
 	dlm_set_recover_status(ls, DLM_RS_DONE);
 	error = dlm_recover_done_wait(ls);
 	if (error) {
@@ -199,7 +212,8 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 
 	dlm_astd_wake();
 
-	log_debug(ls, "recover %llx done: %u ms", rv->seq,
+	log_debug(ls, "recover %llx done: %u ms",
+		  (unsigned long long)rv->seq,
 		  jiffies_to_msecs(jiffies - start));
 	mutex_unlock(&ls->ls_recoverd_active);
 
@@ -207,11 +221,16 @@ static int ls_recover(struct dlm_ls *ls, struct dlm_recover *rv)
 
  fail:
 	dlm_release_root_list(ls);
-	log_debug(ls, "recover %llx error %d", rv->seq, error);
+	log_debug(ls, "recover %llx error %d",
+		  (unsigned long long)rv->seq, error);
 	mutex_unlock(&ls->ls_recoverd_active);
 	return error;
 }
 
+/* The dlm_ls_start() that created the rv we take here may already have been
+   stopped via dlm_ls_stop(); in that case we need to leave the RECOVERY_STOP
+   flag set. */
+
 static void do_ls_recovery(struct dlm_ls *ls)
 {
 	struct dlm_recover *rv = NULL;
@@ -219,7 +238,8 @@ static void do_ls_recovery(struct dlm_ls *ls)
 	spin_lock(&ls->ls_recover_lock);
 	rv = ls->ls_recover_args;
 	ls->ls_recover_args = NULL;
-	clear_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
+	if (rv && ls->ls_recover_seq == rv->seq)
+		clear_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
 	spin_unlock(&ls->ls_recover_lock);
 
 	if (rv) {

fs/dlm/requestqueue.c

@@ -30,26 +30,36 @@ struct rq_entry {
  * lockspace is enabled on some while still suspended on others.
  */
 
-void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd)
+int dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd)
 {
 	struct rq_entry *e;
 	int length = hd->h_length;
-
-	if (dlm_is_removed(ls, nodeid))
-		return;
+	int rv = 0;
 
 	e = kmalloc(sizeof(struct rq_entry) + length, GFP_KERNEL);
 	if (!e) {
 		log_print("dlm_add_requestqueue: out of memory\n");
-		return;
+		return 0;
 	}
 
 	e->nodeid = nodeid;
 	memcpy(e->request, hd, length);
 
+	/* We need to check dlm_locking_stopped() after taking the mutex to
+	   avoid a race where dlm_recoverd enables locking and runs
+	   process_requestqueue between our earlier dlm_locking_stopped check
+	   and this addition to the requestqueue. */
+
 	mutex_lock(&ls->ls_requestqueue_mutex);
-	list_add_tail(&e->list, &ls->ls_requestqueue);
+	if (dlm_locking_stopped(ls))
+		list_add_tail(&e->list, &ls->ls_requestqueue);
+	else {
+		log_debug(ls, "dlm_add_requestqueue skip from %d", nodeid);
+		kfree(e);
+		rv = -EAGAIN;
+	}
 	mutex_unlock(&ls->ls_requestqueue_mutex);
+	return rv;
 }
 
 int dlm_process_requestqueue(struct dlm_ls *ls)
@@ -120,6 +130,10 @@ static int purge_request(struct dlm_ls *ls, struct dlm_message *ms, int nodeid)
 {
 	uint32_t type = ms->m_type;
 
+	/* the ls is being cleaned up and freed by release_lockspace */
+	if (!ls->ls_count)
+		return 1;
+
 	if (dlm_is_removed(ls, nodeid))
 		return 1;
 

fs/dlm/requestqueue.h

@@ -13,7 +13,7 @@
 #ifndef __REQUESTQUEUE_DOT_H__
 #define __REQUESTQUEUE_DOT_H__
 
-void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd);
+int dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd);
 int dlm_process_requestqueue(struct dlm_ls *ls);
 void dlm_wait_requestqueue(struct dlm_ls *ls);
 void dlm_purge_requestqueue(struct dlm_ls *ls);

fs/gfs2/Kconfig

@@ -2,6 +2,7 @@ config GFS2_FS
 	tristate "GFS2 file system support"
 	depends on EXPERIMENTAL
 	select FS_POSIX_ACL
+	select CRC32
 	help
 	A cluster filesystem.
 

fs/gfs2/acl.c

@@ -74,11 +74,11 @@ int gfs2_acl_validate_remove(struct gfs2_inode *ip, int access)
 {
 	if (!GFS2_SB(&ip->i_inode)->sd_args.ar_posix_acl)
 		return -EOPNOTSUPP;
-	if (current->fsuid != ip->i_di.di_uid && !capable(CAP_FOWNER))
+	if (current->fsuid != ip->i_inode.i_uid && !capable(CAP_FOWNER))
 		return -EPERM;
-	if (S_ISLNK(ip->i_di.di_mode))
+	if (S_ISLNK(ip->i_inode.i_mode))
 		return -EOPNOTSUPP;
-	if (!access && !S_ISDIR(ip->i_di.di_mode))
+	if (!access && !S_ISDIR(ip->i_inode.i_mode))
 		return -EACCES;
 
 	return 0;
@@ -145,14 +145,14 @@ static int acl_get(struct gfs2_inode *ip, int access, struct posix_acl **acl,
 }
 
 /**
- * gfs2_check_acl_locked - Check an ACL to see if we're allowed to do something
+ * gfs2_check_acl - Check an ACL to see if we're allowed to do something
  * @inode: the file we want to do something to
  * @mask: what we want to do
  *
  * Returns: errno
  */
 
-int gfs2_check_acl_locked(struct inode *inode, int mask)
+int gfs2_check_acl(struct inode *inode, int mask)
 {
 	struct posix_acl *acl = NULL;
 	int error;
@@ -170,21 +170,6 @@ int gfs2_check_acl_locked(struct inode *inode, int mask)
 	return -EAGAIN;
 }
 
-int gfs2_check_acl(struct inode *inode, int mask)
-{
-	struct gfs2_inode *ip = GFS2_I(inode);
-	struct gfs2_holder i_gh;
-	int error;
-
-	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh);
-	if (!error) {
-		error = gfs2_check_acl_locked(inode, mask);
-		gfs2_glock_dq_uninit(&i_gh);
-	}
-
-	return error;
-}
-
 static int munge_mode(struct gfs2_inode *ip, mode_t mode)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
@@ -198,10 +183,10 @@ static int munge_mode(struct gfs2_inode *ip, mode_t mode)
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (!error) {
 		gfs2_assert_withdraw(sdp,
-				(ip->i_di.di_mode & S_IFMT) == (mode & S_IFMT));
-		ip->i_di.di_mode = mode;
+				(ip->i_inode.i_mode & S_IFMT) == (mode & S_IFMT));
+		ip->i_inode.i_mode = mode;
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 
@@ -215,12 +200,12 @@ int gfs2_acl_create(struct gfs2_inode *dip, struct gfs2_inode *ip)
 	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
 	struct posix_acl *acl = NULL, *clone;
 	struct gfs2_ea_request er;
-	mode_t mode = ip->i_di.di_mode;
+	mode_t mode = ip->i_inode.i_mode;
 	int error;
 
 	if (!sdp->sd_args.ar_posix_acl)
 		return 0;
-	if (S_ISLNK(ip->i_di.di_mode))
+	if (S_ISLNK(ip->i_inode.i_mode))
 		return 0;
 
 	memset(&er, 0, sizeof(struct gfs2_ea_request));
@@ -232,7 +217,7 @@ int gfs2_acl_create(struct gfs2_inode *dip, struct gfs2_inode *ip)
 		return error;
 	if (!acl) {
 		mode &= ~current->fs->umask;
-		if (mode != ip->i_di.di_mode)
+		if (mode != ip->i_inode.i_mode)
 			error = munge_mode(ip, mode);
 		return error;
 	}
@@ -244,7 +229,7 @@ int gfs2_acl_create(struct gfs2_inode *dip, struct gfs2_inode *ip)
 	posix_acl_release(acl);
 	acl = clone;
 
-	if (S_ISDIR(ip->i_di.di_mode)) {
+	if (S_ISDIR(ip->i_inode.i_mode)) {
 		er.er_name = GFS2_POSIX_ACL_DEFAULT;
 		er.er_name_len = GFS2_POSIX_ACL_DEFAULT_LEN;
 		error = gfs2_system_eaops.eo_set(ip, &er);

fs/gfs2/acl.h

@@ -31,7 +31,6 @@ int gfs2_acl_validate_set(struct gfs2_inode *ip, int access,
 			  struct gfs2_ea_request *er,
 			  int *remove, mode_t *mode);
 int gfs2_acl_validate_remove(struct gfs2_inode *ip, int access);
-int gfs2_check_acl_locked(struct inode *inode, int mask);
 int gfs2_check_acl(struct inode *inode, int mask);
 int gfs2_acl_create(struct gfs2_inode *dip, struct gfs2_inode *ip);
 int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr);

fs/gfs2/daemon.c

@@ -112,6 +112,7 @@ int gfs2_logd(void *data)
 	struct gfs2_sbd *sdp = data;
 	struct gfs2_holder ji_gh;
 	unsigned long t;
+	int need_flush;
 
 	while (!kthread_should_stop()) {
 		/* Advance the log tail */
@@ -120,8 +121,10 @@ int gfs2_logd(void *data)
 		    gfs2_tune_get(sdp, gt_log_flush_secs) * HZ;
 
 		gfs2_ail1_empty(sdp, DIO_ALL);
-
-		if (time_after_eq(jiffies, t)) {
+		gfs2_log_lock(sdp);
+		need_flush = sdp->sd_log_num_buf > gfs2_tune_get(sdp, gt_incore_log_blocks);
+		gfs2_log_unlock(sdp);
+		if (need_flush || time_after_eq(jiffies, t)) {
 			gfs2_log_flush(sdp, NULL);
 			sdp->sd_log_flush_time = jiffies;
 		}

fs/gfs2/dir.h

@@ -31,17 +31,17 @@ struct gfs2_inum;
 typedef int (*gfs2_filldir_t) (void *opaque,
 			      const char *name, unsigned int length,
 			      u64 offset,
-			      struct gfs2_inum *inum, unsigned int type);
+			      struct gfs2_inum_host *inum, unsigned int type);
 
 int gfs2_dir_search(struct inode *dir, const struct qstr *filename,
-		    struct gfs2_inum *inum, unsigned int *type);
+		    struct gfs2_inum_host *inum, unsigned int *type);
 int gfs2_dir_add(struct inode *inode, const struct qstr *filename,
-		 const struct gfs2_inum *inum, unsigned int type);
+		 const struct gfs2_inum_host *inum, unsigned int type);
 int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *filename);
 int gfs2_dir_read(struct inode *inode, u64 * offset, void *opaque,
 		  gfs2_filldir_t filldir);
 int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename,
-		   struct gfs2_inum *new_inum, unsigned int new_type);
+		   struct gfs2_inum_host *new_inum, unsigned int new_type);
 
 int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip);
 

fs/gfs2/eaops.c

@@ -120,7 +120,7 @@ static int system_eo_set(struct gfs2_inode *ip, struct gfs2_ea_request *er)
 
 	if (GFS2_ACL_IS_ACCESS(er->er_name, er->er_name_len)) {
 		if (!(er->er_flags & GFS2_ERF_MODE)) {
-			er->er_mode = ip->i_di.di_mode;
+			er->er_mode = ip->i_inode.i_mode;
 			er->er_flags |= GFS2_ERF_MODE;
 		}
 		error = gfs2_acl_validate_set(ip, 1, er,

fs/gfs2/eattr.c

@@ -112,7 +112,7 @@ static int ea_foreach_i(struct gfs2_inode *ip, struct buffer_head *bh,
 static int ea_foreach(struct gfs2_inode *ip, ea_call_t ea_call, void *data)
 {
 	struct buffer_head *bh, *eabh;
-	u64 *eablk, *end;
+	__be64 *eablk, *end;
 	int error;
 
 	error = gfs2_meta_read(ip->i_gl, ip->i_di.di_eattr, DIO_WAIT, &bh);
@@ -129,7 +129,7 @@ static int ea_foreach(struct gfs2_inode *ip, ea_call_t ea_call, void *data)
 		goto out;
 	}
 
-	eablk = (u64 *)(bh->b_data + sizeof(struct gfs2_meta_header));
+	eablk = (__be64 *)(bh->b_data + sizeof(struct gfs2_meta_header));
 	end = eablk + GFS2_SB(&ip->i_inode)->sd_inptrs;
 
 	for (; eablk < end; eablk++) {
@@ -224,7 +224,8 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh,
 	struct gfs2_rgrpd *rgd;
 	struct gfs2_holder rg_gh;
 	struct buffer_head *dibh;
-	u64 *dataptrs, bn = 0;
+	__be64 *dataptrs;
+	u64 bn = 0;
 	u64 bstart = 0;
 	unsigned int blen = 0;
 	unsigned int blks = 0;
@@ -280,6 +281,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh,
 		if (!ip->i_di.di_blocks)
 			gfs2_consist_inode(ip);
 		ip->i_di.di_blocks--;
+		gfs2_set_inode_blocks(&ip->i_inode);
 	}
 	if (bstart)
 		gfs2_free_meta(ip, bstart, blen);
@@ -299,9 +301,9 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh,
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (!error) {
-		ip->i_di.di_ctime = get_seconds();
+		ip->i_inode.i_ctime.tv_sec = get_seconds();
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 
@@ -444,7 +446,7 @@ static int ea_get_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea,
 	struct buffer_head **bh;
 	unsigned int amount = GFS2_EA_DATA_LEN(ea);
 	unsigned int nptrs = DIV_ROUND_UP(amount, sdp->sd_jbsize);
-	u64 *dataptrs = GFS2_EA2DATAPTRS(ea);
+	__be64 *dataptrs = GFS2_EA2DATAPTRS(ea);
 	unsigned int x;
 	int error = 0;
 
@@ -597,6 +599,7 @@ static int ea_alloc_blk(struct gfs2_inode *ip, struct buffer_head **bhp)
 	ea->ea_num_ptrs = 0;
 
 	ip->i_di.di_blocks++;
+	gfs2_set_inode_blocks(&ip->i_inode);
 
 	return 0;
 }
@@ -629,7 +632,7 @@ static int ea_write(struct gfs2_inode *ip, struct gfs2_ea_header *ea,
 		ea->ea_num_ptrs = 0;
 		memcpy(GFS2_EA2DATA(ea), er->er_data, er->er_data_len);
 	} else {
-		u64 *dataptr = GFS2_EA2DATAPTRS(ea);
+		__be64 *dataptr = GFS2_EA2DATAPTRS(ea);
 		const char *data = er->er_data;
 		unsigned int data_len = er->er_data_len;
 		unsigned int copy;
@@ -648,6 +651,7 @@ static int ea_write(struct gfs2_inode *ip, struct gfs2_ea_header *ea,
 			gfs2_metatype_set(bh, GFS2_METATYPE_ED, GFS2_FORMAT_ED);
 
 			ip->i_di.di_blocks++;
+			gfs2_set_inode_blocks(&ip->i_inode);
 
 			copy = data_len > sdp->sd_jbsize ? sdp->sd_jbsize :
 							   data_len;
@@ -686,7 +690,7 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er,
 	if (error)
 		goto out;
 
-	error = gfs2_quota_check(ip, ip->i_di.di_uid, ip->i_di.di_gid);
+	error = gfs2_quota_check(ip, ip->i_inode.i_uid, ip->i_inode.i_gid);
 	if (error)
 		goto out_gunlock_q;
 
@@ -710,13 +714,13 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er,
 	if (!error) {
 		if (er->er_flags & GFS2_ERF_MODE) {
 			gfs2_assert_withdraw(GFS2_SB(&ip->i_inode),
-					    (ip->i_di.di_mode & S_IFMT) ==
+					    (ip->i_inode.i_mode & S_IFMT) ==
 					    (er->er_mode & S_IFMT));
-			ip->i_di.di_mode = er->er_mode;
+			ip->i_inode.i_mode = er->er_mode;
 		}
-		ip->i_di.di_ctime = get_seconds();
+		ip->i_inode.i_ctime.tv_sec = get_seconds();
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 
@@ -846,12 +850,12 @@ static int ea_set_simple_noalloc(struct gfs2_inode *ip, struct buffer_head *bh,
 
 	if (er->er_flags & GFS2_ERF_MODE) {
 		gfs2_assert_withdraw(GFS2_SB(&ip->i_inode),
-			(ip->i_di.di_mode & S_IFMT) == (er->er_mode & S_IFMT));
-		ip->i_di.di_mode = er->er_mode;
+			(ip->i_inode.i_mode & S_IFMT) == (er->er_mode & S_IFMT));
+		ip->i_inode.i_mode = er->er_mode;
 	}
-	ip->i_di.di_ctime = get_seconds();
+	ip->i_inode.i_ctime.tv_sec = get_seconds();
 	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-	gfs2_dinode_out(&ip->i_di, dibh->b_data);
+	gfs2_dinode_out(ip, dibh->b_data);
 	brelse(dibh);
 out:
 	gfs2_trans_end(GFS2_SB(&ip->i_inode));
@@ -931,12 +935,12 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er,
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 	struct buffer_head *indbh, *newbh;
-	u64 *eablk;
+	__be64 *eablk;
 	int error;
 	int mh_size = sizeof(struct gfs2_meta_header);
 
 	if (ip->i_di.di_flags & GFS2_DIF_EA_INDIRECT) {
-		u64 *end;
+		__be64 *end;
 
 		error = gfs2_meta_read(ip->i_gl, ip->i_di.di_eattr, DIO_WAIT,
 				       &indbh);
@@ -948,7 +952,7 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er,
 			goto out;
 		}
 
-		eablk = (u64 *)(indbh->b_data + mh_size);
+		eablk = (__be64 *)(indbh->b_data + mh_size);
 		end = eablk + sdp->sd_inptrs;
 
 		for (; eablk < end; eablk++)
@@ -971,11 +975,12 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er,
 		gfs2_metatype_set(indbh, GFS2_METATYPE_IN, GFS2_FORMAT_IN);
 		gfs2_buffer_clear_tail(indbh, mh_size);
 
-		eablk = (u64 *)(indbh->b_data + mh_size);
+		eablk = (__be64 *)(indbh->b_data + mh_size);
 		*eablk = cpu_to_be64(ip->i_di.di_eattr);
 		ip->i_di.di_eattr = blk;
 		ip->i_di.di_flags |= GFS2_DIF_EA_INDIRECT;
 		ip->i_di.di_blocks++;
+		gfs2_set_inode_blocks(&ip->i_inode);
 
 		eablk++;
 	}
@@ -1129,9 +1134,9 @@ static int ea_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_location *el)
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (!error) {
-		ip->i_di.di_ctime = get_seconds();
+		ip->i_inode.i_ctime.tv_sec = get_seconds();
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 
@@ -1202,7 +1207,7 @@ static int ea_acl_chmod_unstuffed(struct gfs2_inode *ip,
 	struct buffer_head **bh;
 	unsigned int amount = GFS2_EA_DATA_LEN(ea);
 	unsigned int nptrs = DIV_ROUND_UP(amount, sdp->sd_jbsize);
-	u64 *dataptrs = GFS2_EA2DATAPTRS(ea);
+	__be64 *dataptrs = GFS2_EA2DATAPTRS(ea);
 	unsigned int x;
 	int error;
 
@@ -1284,9 +1289,8 @@ int gfs2_ea_acl_chmod(struct gfs2_inode *ip, struct gfs2_ea_location *el,
 	if (!error) {
 		error = inode_setattr(&ip->i_inode, attr);
 		gfs2_assert_warn(GFS2_SB(&ip->i_inode), !error);
-		gfs2_inode_attr_out(ip);
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 
@@ -1300,7 +1304,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip)
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 	struct gfs2_rgrp_list rlist;
 	struct buffer_head *indbh, *dibh;
-	u64 *eablk, *end;
+	__be64 *eablk, *end;
 	unsigned int rg_blocks = 0;
 	u64 bstart = 0;
 	unsigned int blen = 0;
@@ -1319,7 +1323,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip)
 		goto out;
 	}
 
-	eablk = (u64 *)(indbh->b_data + sizeof(struct gfs2_meta_header));
+	eablk = (__be64 *)(indbh->b_data + sizeof(struct gfs2_meta_header));
 	end = eablk + sdp->sd_inptrs;
 
 	for (; eablk < end; eablk++) {
@@ -1363,7 +1367,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip)
 
 	gfs2_trans_add_bh(ip->i_gl, indbh, 1);
 
-	eablk = (u64 *)(indbh->b_data + sizeof(struct gfs2_meta_header));
+	eablk = (__be64 *)(indbh->b_data + sizeof(struct gfs2_meta_header));
 	bstart = 0;
 	blen = 0;
 
@@ -1387,6 +1391,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip)
 		if (!ip->i_di.di_blocks)
 			gfs2_consist_inode(ip);
 		ip->i_di.di_blocks--;
+		gfs2_set_inode_blocks(&ip->i_inode);
 	}
 	if (bstart)
 		gfs2_free_meta(ip, bstart, blen);
@@ -1396,7 +1401,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip)
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (!error) {
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 
@@ -1441,11 +1446,12 @@ static int ea_dealloc_block(struct gfs2_inode *ip)
 	if (!ip->i_di.di_blocks)
 		gfs2_consist_inode(ip);
 	ip->i_di.di_blocks--;
+	gfs2_set_inode_blocks(&ip->i_inode);
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (!error) {
 		gfs2_trans_add_bh(ip->i_gl, dibh, 1);
-		gfs2_dinode_out(&ip->i_di, dibh->b_data);
+		gfs2_dinode_out(ip, dibh->b_data);
 		brelse(dibh);
 	}
 

fs/gfs2/eattr.h

@@ -19,7 +19,7 @@ struct iattr;
 #define GFS2_EA_SIZE(ea) \
 ALIGN(sizeof(struct gfs2_ea_header) + (ea)->ea_name_len + \
       ((GFS2_EA_IS_STUFFED(ea)) ? GFS2_EA_DATA_LEN(ea) : \
-                                  (sizeof(u64) * (ea)->ea_num_ptrs)), 8)
+                                  (sizeof(__be64) * (ea)->ea_num_ptrs)), 8)
 
 #define GFS2_EA_IS_STUFFED(ea) (!(ea)->ea_num_ptrs)
 #define GFS2_EA_IS_LAST(ea) ((ea)->ea_flags & GFS2_EAFLAG_LAST)
@@ -29,13 +29,13 @@ ALIGN(sizeof(struct gfs2_ea_header) + (er)->er_name_len + (er)->er_data_len, 8)
 
 #define GFS2_EAREQ_SIZE_UNSTUFFED(sdp, er) \
 ALIGN(sizeof(struct gfs2_ea_header) + (er)->er_name_len + \
-      sizeof(u64) * DIV_ROUND_UP((er)->er_data_len, (sdp)->sd_jbsize), 8)
+      sizeof(__be64) * DIV_ROUND_UP((er)->er_data_len, (sdp)->sd_jbsize), 8)
 
 #define GFS2_EA2NAME(ea) ((char *)((struct gfs2_ea_header *)(ea) + 1))
 #define GFS2_EA2DATA(ea) (GFS2_EA2NAME(ea) + (ea)->ea_name_len)
 
 #define GFS2_EA2DATAPTRS(ea) \
-((u64 *)(GFS2_EA2NAME(ea) + ALIGN((ea)->ea_name_len, 8)))
+((__be64 *)(GFS2_EA2NAME(ea) + ALIGN((ea)->ea_name_len, 8)))
 
 #define GFS2_EA2NEXT(ea) \
 ((struct gfs2_ea_header *)((char *)(ea) + GFS2_EA_REC_LEN(ea)))

fs/gfs2/glock.c

@@ -96,7 +96,7 @@ static inline rwlock_t *gl_lock_addr(unsigned int x)
 	return &gl_hash_locks[x & (GL_HASH_LOCK_SZ-1)];
 }
 #else /* not SMP, so no spinlocks required */
-static inline rwlock_t *gl_lock_addr(x)
+static inline rwlock_t *gl_lock_addr(unsigned int x)
 {
 	return NULL;
 }
@@ -769,7 +769,7 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state)
 	} else {
 		spin_unlock(&gl->gl_spin);
 
-		new_gh = gfs2_holder_get(gl, state, LM_FLAG_TRY, GFP_KERNEL);
+		new_gh = gfs2_holder_get(gl, state, LM_FLAG_TRY, GFP_NOFS);
 		if (!new_gh)
 			return;
 		set_bit(HIF_DEMOTE, &new_gh->gh_iflags);
@@ -785,21 +785,6 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state)
 		gfs2_holder_put(new_gh);
 }
 
-void gfs2_glock_inode_squish(struct inode *inode)
-{
-	struct gfs2_holder gh;
-	struct gfs2_glock *gl = GFS2_I(inode)->i_gl;
-	gfs2_holder_init(gl, LM_ST_UNLOCKED, 0, &gh);
-	set_bit(HIF_DEMOTE, &gh.gh_iflags);
-	spin_lock(&gl->gl_spin);
-	gfs2_assert(inode->i_sb->s_fs_info, list_empty(&gl->gl_holders));
-	list_add_tail(&gh.gh_list, &gl->gl_waiters2);
-	run_queue(gl);
-	spin_unlock(&gl->gl_spin);
-	wait_for_completion(&gh.gh_wait);
-	gfs2_holder_uninit(&gh);
-}
-
 /**
  * state_change - record that the glock is now in a different state
  * @gl: the glock
@@ -847,12 +832,12 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret)
 
 	if (prev_state != LM_ST_UNLOCKED && !(ret & LM_OUT_CACHEABLE)) {
 		if (glops->go_inval)
-			glops->go_inval(gl, DIO_METADATA | DIO_DATA);
+			glops->go_inval(gl, DIO_METADATA);
 	} else if (gl->gl_state == LM_ST_DEFERRED) {
 		/* We might not want to do this here.
 		   Look at moving to the inode glops. */
 		if (glops->go_inval)
-			glops->go_inval(gl, DIO_DATA);
+			glops->go_inval(gl, 0);
 	}
 
 	/*  Deal with each possible exit condition  */
@@ -954,7 +939,7 @@ void gfs2_glock_xmote_th(struct gfs2_glock *gl, unsigned int state, int flags)
 	gfs2_assert_warn(sdp, state != gl->gl_state);
 
 	if (gl->gl_state == LM_ST_EXCLUSIVE && glops->go_sync)
-		glops->go_sync(gl, DIO_METADATA | DIO_DATA | DIO_RELEASE);
+		glops->go_sync(gl);
 
 	gfs2_glock_hold(gl);
 	gl->gl_req_bh = xmote_bh;
@@ -995,7 +980,7 @@ static void drop_bh(struct gfs2_glock *gl, unsigned int ret)
 	state_change(gl, LM_ST_UNLOCKED);
 
 	if (glops->go_inval)
-		glops->go_inval(gl, DIO_METADATA | DIO_DATA);
+		glops->go_inval(gl, DIO_METADATA);
 
 	if (gh) {
 		spin_lock(&gl->gl_spin);
@@ -1041,7 +1026,7 @@ void gfs2_glock_drop_th(struct gfs2_glock *gl)
 	gfs2_assert_warn(sdp, gl->gl_state != LM_ST_UNLOCKED);
 
 	if (gl->gl_state == LM_ST_EXCLUSIVE && glops->go_sync)
-		glops->go_sync(gl, DIO_METADATA | DIO_DATA | DIO_RELEASE);
+		glops->go_sync(gl);
 
 	gfs2_glock_hold(gl);
 	gl->gl_req_bh = drop_bh;
@@ -1244,9 +1229,6 @@ int gfs2_glock_nq(struct gfs2_holder *gh)
 
 	clear_bit(GLF_PREFETCH, &gl->gl_flags);
 
-	if (error == GLR_TRYFAILED && (gh->gh_flags & GL_DUMP))
-		dump_glock(gl);
-
 	return error;
 }
 
@@ -1923,7 +1905,7 @@ static int examine_bucket(glock_examiner examiner, struct gfs2_sbd *sdp,
 
 static void scan_glock(struct gfs2_glock *gl)
 {
-	if (gl->gl_ops == &gfs2_inode_glops)
+	if (gl->gl_ops == &gfs2_inode_glops && gl->gl_object)
 		return;
 
 	if (gfs2_glmutex_trylock(gl)) {
@@ -2078,7 +2060,7 @@ static int dump_inode(struct gfs2_inode *ip)
 	printk(KERN_INFO "    num = %llu %llu\n",
 		    (unsigned long long)ip->i_num.no_formal_ino,
 		    (unsigned long long)ip->i_num.no_addr);
-	printk(KERN_INFO "    type = %u\n", IF2DT(ip->i_di.di_mode));
+	printk(KERN_INFO "    type = %u\n", IF2DT(ip->i_inode.i_mode));
 	printk(KERN_INFO "    i_flags =");
 	for (x = 0; x < 32; x++)
 		if (test_bit(x, &ip->i_flags))

fs/gfs2/glock.h

@@ -27,8 +27,6 @@
 #define GL_ATIME		0x00000200
 #define GL_NOCACHE		0x00000400
 #define GL_NOCANCEL		0x00001000
-#define GL_AOP			0x00004000
-#define GL_DUMP			0x00008000
 
 #define GLR_TRYFAILED		13
 #define GLR_CANCELED		14
@@ -108,7 +106,6 @@ void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs);
 void gfs2_glock_prefetch_num(struct gfs2_sbd *sdp, u64 number,
 			     const struct gfs2_glock_operations *glops,
 			     unsigned int state, int flags);
-void gfs2_glock_inode_squish(struct inode *inode);
 
 /**
  * gfs2_glock_nq_init - intialize a holder and enqueue it on a glock

fs/gfs2/glops.c

@@ -92,7 +92,7 @@ static void gfs2_pte_inval(struct gfs2_glock *gl)
 
 	ip = gl->gl_object;
 	inode = &ip->i_inode;
-	if (!ip || !S_ISREG(ip->i_di.di_mode))
+	if (!ip || !S_ISREG(inode->i_mode))
 		return;
 
 	if (!test_bit(GIF_PAGED, &ip->i_flags))
@@ -106,90 +106,21 @@ static void gfs2_pte_inval(struct gfs2_glock *gl)
 	clear_bit(GIF_SW_PAGED, &ip->i_flags);
 }
 
-/**
- * gfs2_page_inval - Invalidate all pages associated with a glock
- * @gl: the glock
- *
- */
-
-static void gfs2_page_inval(struct gfs2_glock *gl)
-{
-	struct gfs2_inode *ip;
-	struct inode *inode;
-
-	ip = gl->gl_object;
-	inode = &ip->i_inode;
-	if (!ip || !S_ISREG(ip->i_di.di_mode))
-		return;
-
-	truncate_inode_pages(inode->i_mapping, 0);
-	gfs2_assert_withdraw(GFS2_SB(&ip->i_inode), !inode->i_mapping->nrpages);
-	clear_bit(GIF_PAGED, &ip->i_flags);
-}
-
-/**
- * gfs2_page_wait - Wait for writeback of data
- * @gl: the glock
- *
- * Syncs data (not metadata) for a regular file.
- * No-op for all other types.
- */
-
-static void gfs2_page_wait(struct gfs2_glock *gl)
-{
-	struct gfs2_inode *ip = gl->gl_object;
-	struct inode *inode = &ip->i_inode;
-	struct address_space *mapping = inode->i_mapping;
-	int error;
-
-	if (!S_ISREG(ip->i_di.di_mode))
-		return;
-
-	error = filemap_fdatawait(mapping);
-
-	/* Put back any errors cleared by filemap_fdatawait()
-	   so they can be caught by someone who can pass them
-	   up to user space. */
-
-	if (error == -ENOSPC)
-		set_bit(AS_ENOSPC, &mapping->flags);
-	else if (error)
-		set_bit(AS_EIO, &mapping->flags);
-
-}
-
-static void gfs2_page_writeback(struct gfs2_glock *gl)
-{
-	struct gfs2_inode *ip = gl->gl_object;
-	struct inode *inode = &ip->i_inode;
-	struct address_space *mapping = inode->i_mapping;
-
-	if (!S_ISREG(ip->i_di.di_mode))
-		return;
-
-	filemap_fdatawrite(mapping);
-}
-
 /**
  * meta_go_sync - sync out the metadata for this glock
  * @gl: the glock
- * @flags: DIO_*
  *
  * Called when demoting or unlocking an EX glock.  We must flush
  * to disk all dirty buffers/pages relating to this glock, and must not
  * not return to caller to demote/unlock the glock until I/O is complete.
  */
 
-static void meta_go_sync(struct gfs2_glock *gl, int flags)
+static void meta_go_sync(struct gfs2_glock *gl)
 {
-	if (!(flags & DIO_METADATA))
-		return;
-
 	if (test_and_clear_bit(GLF_DIRTY, &gl->gl_flags)) {
 		gfs2_log_flush(gl->gl_sbd, gl);
 		gfs2_meta_sync(gl);
-		if (flags & DIO_RELEASE)
-			gfs2_ail_empty_gl(gl);
+		gfs2_ail_empty_gl(gl);
 	}
 
 }
@@ -264,31 +195,31 @@ static void inode_go_drop_th(struct gfs2_glock *gl)
 /**
  * inode_go_sync - Sync the dirty data and/or metadata for an inode glock
  * @gl: the glock protecting the inode
- * @flags:
  *
  */
 
-static void inode_go_sync(struct gfs2_glock *gl, int flags)
+static void inode_go_sync(struct gfs2_glock *gl)
 {
-	int meta = (flags & DIO_METADATA);
-	int data = (flags & DIO_DATA);
+	struct gfs2_inode *ip = gl->gl_object;
+
+	if (ip && !S_ISREG(ip->i_inode.i_mode))
+		ip = NULL;
 
 	if (test_bit(GLF_DIRTY, &gl->gl_flags)) {
-		if (meta && data) {
-			gfs2_page_writeback(gl);
-			gfs2_log_flush(gl->gl_sbd, gl);
-			gfs2_meta_sync(gl);
-			gfs2_page_wait(gl);
-			clear_bit(GLF_DIRTY, &gl->gl_flags);
-		} else if (meta) {
-			gfs2_log_flush(gl->gl_sbd, gl);
-			gfs2_meta_sync(gl);
-		} else if (data) {
-			gfs2_page_writeback(gl);
-			gfs2_page_wait(gl);
+		gfs2_log_flush(gl->gl_sbd, gl);
+		if (ip)
+			filemap_fdatawrite(ip->i_inode.i_mapping);
+		gfs2_meta_sync(gl);
+		if (ip) {
+			struct address_space *mapping = ip->i_inode.i_mapping;
+			int error = filemap_fdatawait(mapping);
+			if (error == -ENOSPC)
+				set_bit(AS_ENOSPC, &mapping->flags);
+			else if (error)
+				set_bit(AS_EIO, &mapping->flags);
 		}
-		if (flags & DIO_RELEASE)
-			gfs2_ail_empty_gl(gl);
+		clear_bit(GLF_DIRTY, &gl->gl_flags);
+		gfs2_ail_empty_gl(gl);
 	}
 }
 
@@ -301,15 +232,20 @@ static void inode_go_sync(struct gfs2_glock *gl, int flags)
 
 static void inode_go_inval(struct gfs2_glock *gl, int flags)
 {
+	struct gfs2_inode *ip = gl->gl_object;
 	int meta = (flags & DIO_METADATA);
-	int data = (flags & DIO_DATA);
 
 	if (meta) {
 		gfs2_meta_inval(gl);
-		gl->gl_vn++;
+		if (ip)
+			set_bit(GIF_INVALID, &ip->i_flags);
+	}
+
+	if (ip && S_ISREG(ip->i_inode.i_mode)) {
+		truncate_inode_pages(ip->i_inode.i_mapping, 0);
+		gfs2_assert_withdraw(GFS2_SB(&ip->i_inode), !ip->i_inode.i_mapping->nrpages);
+		clear_bit(GIF_PAGED, &ip->i_flags);
 	}
-	if (data)
-		gfs2_page_inval(gl);
 }
 
 /**
@@ -351,11 +287,10 @@ static int inode_go_lock(struct gfs2_holder *gh)
 	if (!ip)
 		return 0;
 
-	if (ip->i_vn != gl->gl_vn) {
+	if (test_bit(GIF_INVALID, &ip->i_flags)) {
 		error = gfs2_inode_refresh(ip);
 		if (error)
 			return error;
-		gfs2_inode_attr_in(ip);
 	}
 
 	if ((ip->i_di.di_flags & GFS2_DIF_TRUNC_IN_PROG) &&
@@ -379,11 +314,8 @@ static void inode_go_unlock(struct gfs2_holder *gh)
 	struct gfs2_glock *gl = gh->gh_gl;
 	struct gfs2_inode *ip = gl->gl_object;
 
-	if (ip == NULL)
-		return;
-	if (test_bit(GLF_DIRTY, &gl->gl_flags))
-		gfs2_inode_attr_in(ip);
-	gfs2_meta_cache_flush(ip);
+	if (ip)
+		gfs2_meta_cache_flush(ip);
 }
 
 /**
@@ -491,13 +423,13 @@ static void trans_go_xmote_bh(struct gfs2_glock *gl)
 	struct gfs2_sbd *sdp = gl->gl_sbd;
 	struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode);
 	struct gfs2_glock *j_gl = ip->i_gl;
-	struct gfs2_log_header head;
+	struct gfs2_log_header_host head;
 	int error;
 
 	if (gl->gl_state != LM_ST_UNLOCKED &&
 	    test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
 		gfs2_meta_cache_flush(GFS2_I(sdp->sd_jdesc->jd_inode));
-		j_gl->gl_ops->go_inval(j_gl, DIO_METADATA | DIO_DATA);
+		j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);
 
 		error = gfs2_find_jhead(sdp->sd_jdesc, &head);
 		if (error)

fs/gfs2/incore.h

@@ -14,8 +14,6 @@
 
 #define DIO_WAIT	0x00000010
 #define DIO_METADATA	0x00000020
-#define DIO_DATA	0x00000040
-#define DIO_RELEASE	0x00000080
 #define DIO_ALL		0x00000100
 
 struct gfs2_log_operations;
@@ -41,7 +39,7 @@ struct gfs2_log_operations {
 	void (*lo_before_commit) (struct gfs2_sbd *sdp);
 	void (*lo_after_commit) (struct gfs2_sbd *sdp, struct gfs2_ail *ai);
 	void (*lo_before_scan) (struct gfs2_jdesc *jd,
-				struct gfs2_log_header *head, int pass);
+				struct gfs2_log_header_host *head, int pass);
 	int (*lo_scan_elements) (struct gfs2_jdesc *jd, unsigned int start,
 				 struct gfs2_log_descriptor *ld, __be64 *ptr,
 				 int pass);
@@ -67,8 +65,8 @@ struct gfs2_rgrpd {
 	struct list_head rd_list_mru;
 	struct list_head rd_recent;	/* Recently used rgrps */
 	struct gfs2_glock *rd_gl;	/* Glock for this rgrp */
-	struct gfs2_rindex rd_ri;
-	struct gfs2_rgrp rd_rg;
+	struct gfs2_rindex_host rd_ri;
+	struct gfs2_rgrp_host rd_rg;
 	u64 rd_rg_vn;
 	struct gfs2_bitmap *rd_bits;
 	unsigned int rd_bh_count;
@@ -103,18 +101,17 @@ struct gfs2_bufdata {
 };
 
 struct gfs2_glock_operations {
-	void (*go_xmote_th) (struct gfs2_glock * gl, unsigned int state,
-			     int flags);
-	void (*go_xmote_bh) (struct gfs2_glock * gl);
-	void (*go_drop_th) (struct gfs2_glock * gl);
-	void (*go_drop_bh) (struct gfs2_glock * gl);
-	void (*go_sync) (struct gfs2_glock * gl, int flags);
-	void (*go_inval) (struct gfs2_glock * gl, int flags);
-	int (*go_demote_ok) (struct gfs2_glock * gl);
-	int (*go_lock) (struct gfs2_holder * gh);
-	void (*go_unlock) (struct gfs2_holder * gh);
-	void (*go_callback) (struct gfs2_glock * gl, unsigned int state);
-	void (*go_greedy) (struct gfs2_glock * gl);
+	void (*go_xmote_th) (struct gfs2_glock *gl, unsigned int state, int flags);
+	void (*go_xmote_bh) (struct gfs2_glock *gl);
+	void (*go_drop_th) (struct gfs2_glock *gl);
+	void (*go_drop_bh) (struct gfs2_glock *gl);
+	void (*go_sync) (struct gfs2_glock *gl);
+	void (*go_inval) (struct gfs2_glock *gl, int flags);
+	int (*go_demote_ok) (struct gfs2_glock *gl);
+	int (*go_lock) (struct gfs2_holder *gh);
+	void (*go_unlock) (struct gfs2_holder *gh);
+	void (*go_callback) (struct gfs2_glock *gl, unsigned int state);
+	void (*go_greedy) (struct gfs2_glock *gl);
 	const int go_type;
 };
 
@@ -217,6 +214,7 @@ struct gfs2_alloc {
 };
 
 enum {
+	GIF_INVALID		= 0,
 	GIF_QD_LOCKED		= 1,
 	GIF_PAGED		= 2,
 	GIF_SW_PAGED		= 3,
@@ -224,12 +222,11 @@ enum {
 
 struct gfs2_inode {
 	struct inode i_inode;
-	struct gfs2_inum i_num;
+	struct gfs2_inum_host i_num;
 
 	unsigned long i_flags;		/* GIF_... */
 
-	u64 i_vn;
-	struct gfs2_dinode i_di; /* To be replaced by ref to block */
+	struct gfs2_dinode_host i_di; /* To be replaced by ref to block */
 
 	struct gfs2_glock *i_gl; /* Move into i_gh? */
 	struct gfs2_holder i_iopen_gh;
@@ -450,7 +447,7 @@ struct gfs2_sbd {
 	struct super_block *sd_vfs_meta;
 	struct kobject sd_kobj;
 	unsigned long sd_flags;	/* SDF_... */
-	struct gfs2_sb sd_sb;
+	struct gfs2_sb_host sd_sb;
 
 	/* Constants computed on mount */
 
@@ -503,8 +500,8 @@ struct gfs2_sbd {
 
 	spinlock_t sd_statfs_spin;
 	struct mutex sd_statfs_mutex;
-	struct gfs2_statfs_change sd_statfs_master;
-	struct gfs2_statfs_change sd_statfs_local;
+	struct gfs2_statfs_change_host sd_statfs_master;
+	struct gfs2_statfs_change_host sd_statfs_local;
 	unsigned long sd_statfs_sync_time;
 
 	/* Resource group stuff */

fs/gfs2/inode.h

@@ -22,13 +22,19 @@ static inline int gfs2_is_jdata(struct gfs2_inode *ip)
 
 static inline int gfs2_is_dir(struct gfs2_inode *ip)
 {
-	return S_ISDIR(ip->i_di.di_mode);
+	return S_ISDIR(ip->i_inode.i_mode);
+}
+
+static inline void gfs2_set_inode_blocks(struct inode *inode)
+{
+	struct gfs2_inode *ip = GFS2_I(inode);
+	inode->i_blocks = ip->i_di.di_blocks <<
+		(GFS2_SB(inode)->sd_sb.sb_bsize_shift - GFS2_BASIC_BLOCK_SHIFT);
 }
 
 void gfs2_inode_attr_in(struct gfs2_inode *ip);
-void gfs2_inode_attr_out(struct gfs2_inode *ip);
-struct inode *gfs2_inode_lookup(struct super_block *sb, struct gfs2_inum *inum, unsigned type);
-struct inode *gfs2_ilookup(struct super_block *sb, struct gfs2_inum *inum);
+struct inode *gfs2_inode_lookup(struct super_block *sb, struct gfs2_inum_host *inum, unsigned type);
+struct inode *gfs2_ilookup(struct super_block *sb, struct gfs2_inum_host *inum);
 
 int gfs2_inode_refresh(struct gfs2_inode *ip);
 
@@ -37,19 +43,15 @@ int gfs2_change_nlink(struct gfs2_inode *ip, int diff);
 struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name,
 			   int is_root, struct nameidata *nd);
 struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name,
-			   unsigned int mode);
+			   unsigned int mode, dev_t dev);
 int gfs2_rmdiri(struct gfs2_inode *dip, const struct qstr *name,
 		struct gfs2_inode *ip);
 int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name,
 		   struct gfs2_inode *ip);
 int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to);
 int gfs2_readlinki(struct gfs2_inode *ip, char **buf, unsigned int *len);
-
 int gfs2_glock_nq_atime(struct gfs2_holder *gh);
-int gfs2_glock_nq_m_atime(unsigned int num_gh, struct gfs2_holder *ghs);
-
 int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr);
-
 struct inode *gfs2_lookup_simple(struct inode *dip, const char *name);
 
 #endif /* __INODE_DOT_H__ */

fs/gfs2/log.c

@@ -15,6 +15,7 @@
 #include <linux/gfs2_ondisk.h>
 #include <linux/crc32.h>
 #include <linux/lm_interface.h>
+#include <linux/delay.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -142,7 +143,7 @@ static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai, int fl
 	return list_empty(&ai->ai_ail1_list);
 }
 
-void gfs2_ail1_start(struct gfs2_sbd *sdp, int flags)
+static void gfs2_ail1_start(struct gfs2_sbd *sdp, int flags)
 {
 	struct list_head *head = &sdp->sd_ail1_list;
 	u64 sync_gen;
@@ -261,6 +262,12 @@ static void ail2_empty(struct gfs2_sbd *sdp, unsigned int new_tail)
  * @sdp: The GFS2 superblock
  * @blks: The number of blocks to reserve
  *
+ * Note that we never give out the last 6 blocks of the journal. Thats
+ * due to the fact that there is are a small number of header blocks
+ * associated with each log flush. The exact number can't be known until
+ * flush time, so we ensure that we have just enough free blocks at all
+ * times to avoid running out during a log flush.
+ *
  * Returns: errno
  */
 
@@ -274,7 +281,7 @@ int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks)
 
 	mutex_lock(&sdp->sd_log_reserve_mutex);
 	gfs2_log_lock(sdp);
-	while(sdp->sd_log_blks_free <= blks) {
+	while(sdp->sd_log_blks_free <= (blks + 6)) {
 		gfs2_log_unlock(sdp);
 		gfs2_ail1_empty(sdp, 0);
 		gfs2_log_flush(sdp, NULL);
@@ -319,7 +326,8 @@ static u64 log_bmap(struct gfs2_sbd *sdp, unsigned int lbn)
 	bh_map.b_size = 1 << inode->i_blkbits;
 	error = gfs2_block_map(inode, lbn, 0, &bh_map);
 	if (error || !bh_map.b_blocknr)
-		printk(KERN_INFO "error=%d, dbn=%llu lbn=%u", error, bh_map.b_blocknr, lbn);
+		printk(KERN_INFO "error=%d, dbn=%llu lbn=%u", error,
+		       (unsigned long long)bh_map.b_blocknr, lbn);
 	gfs2_assert_withdraw(sdp, !error && bh_map.b_blocknr);
 
 	return bh_map.b_blocknr;
@@ -643,12 +651,9 @@ void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
 	up_read(&sdp->sd_log_flush_lock);
 
 	gfs2_log_lock(sdp);
-	if (sdp->sd_log_num_buf > gfs2_tune_get(sdp, gt_incore_log_blocks)) {
-		gfs2_log_unlock(sdp);
-		gfs2_log_flush(sdp, NULL);
-	} else {
-		gfs2_log_unlock(sdp);
-	}
+	if (sdp->sd_log_num_buf > gfs2_tune_get(sdp, gt_incore_log_blocks))
+		wake_up_process(sdp->sd_logd_process);
+	gfs2_log_unlock(sdp);
 }
 
 /**
@@ -686,3 +691,21 @@ void gfs2_log_shutdown(struct gfs2_sbd *sdp)
 	up_write(&sdp->sd_log_flush_lock);
 }
 
+
+/**
+ * gfs2_meta_syncfs - sync all the buffers in a filesystem
+ * @sdp: the filesystem
+ *
+ */
+
+void gfs2_meta_syncfs(struct gfs2_sbd *sdp)
+{
+	gfs2_log_flush(sdp, NULL);
+	for (;;) {
+		gfs2_ail1_start(sdp, DIO_ALL);
+		if (gfs2_ail1_empty(sdp, DIO_ALL))
+			break;
+		msleep(10);
+	}
+}
+

fs/gfs2/log.h

@@ -48,7 +48,6 @@ static inline void gfs2_log_pointers_init(struct gfs2_sbd *sdp,
 unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct,
 			    unsigned int ssize);
 
-void gfs2_ail1_start(struct gfs2_sbd *sdp, int flags);
 int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags);
 
 int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks);
@@ -61,5 +60,6 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl);
 void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *trans);
 
 void gfs2_log_shutdown(struct gfs2_sbd *sdp);
+void gfs2_meta_syncfs(struct gfs2_sbd *sdp);
 
 #endif /* __LOG_DOT_H__ */

fs/gfs2/lops.c

@@ -182,7 +182,7 @@ static void buf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 }
 
 static void buf_lo_before_scan(struct gfs2_jdesc *jd,
-			       struct gfs2_log_header *head, int pass)
+			       struct gfs2_log_header_host *head, int pass)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 
@@ -328,7 +328,7 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 }
 
 static void revoke_lo_before_scan(struct gfs2_jdesc *jd,
-				  struct gfs2_log_header *head, int pass)
+				  struct gfs2_log_header_host *head, int pass)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
 
@@ -509,7 +509,7 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 {
 	LIST_HEAD(started);
 	struct gfs2_bufdata *bd1 = NULL, *bd2, *bdt;
-	struct buffer_head *bh = NULL;
+	struct buffer_head *bh = NULL,*bh1 = NULL;
 	unsigned int offset = sizeof(struct gfs2_log_descriptor);
 	struct gfs2_log_descriptor *ld;
 	unsigned int limit;
@@ -537,8 +537,13 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 		list_for_each_entry_safe_continue(bd1, bdt,
 						  &sdp->sd_log_le_databuf,
 						  bd_le.le_list) {
+			/* store off the buffer head in a local ptr since
+			 * gfs2_bufdata might change when we drop the log lock
+			 */
+			bh1 = bd1->bd_bh;
+
 			/* An ordered write buffer */
-			if (bd1->bd_bh && !buffer_pinned(bd1->bd_bh)) {
+			if (bh1 && !buffer_pinned(bh1)) {
 				list_move(&bd1->bd_le.le_list, &started);
 				if (bd1 == bd2) {
 					bd2 = NULL;
@@ -547,20 +552,21 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 							bd_le.le_list);
 				}
 				total_dbuf--;
-				if (bd1->bd_bh) {
-					get_bh(bd1->bd_bh);
-					if (buffer_dirty(bd1->bd_bh)) {
+				if (bh1) {
+					if (buffer_dirty(bh1)) {
+						get_bh(bh1);
+
 						gfs2_log_unlock(sdp);
-						wait_on_buffer(bd1->bd_bh);
-						ll_rw_block(WRITE, 1,
-							    &bd1->bd_bh);
+
+						ll_rw_block(SWRITE, 1, &bh1);
+						brelse(bh1);
+
 						gfs2_log_lock(sdp);
 					}
-					brelse(bd1->bd_bh);
 					continue;
 				}
 				continue;
-			} else if (bd1->bd_bh) { /* A journaled buffer */
+			} else if (bh1) { /* A journaled buffer */
 				int magic;
 				gfs2_log_unlock(sdp);
 				if (!bh) {
@@ -582,16 +588,16 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 					ld->ld_data2 = cpu_to_be32(0);
 					memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
 				}
-				magic = gfs2_check_magic(bd1->bd_bh);
-				*ptr++ = cpu_to_be64(bd1->bd_bh->b_blocknr);
+				magic = gfs2_check_magic(bh1);
+				*ptr++ = cpu_to_be64(bh1->b_blocknr);
 				*ptr++ = cpu_to_be64((__u64)magic);
-				clear_buffer_escaped(bd1->bd_bh);
+				clear_buffer_escaped(bh1);
 				if (unlikely(magic != 0))
-					set_buffer_escaped(bd1->bd_bh);
+					set_buffer_escaped(bh1);
 				gfs2_log_lock(sdp);
 				if (n++ > num)
 					break;
-			} else if (!bd1->bd_bh) {
+			} else if (!bh1) {
 				total_dbuf--;
 				sdp->sd_log_num_databuf--;
 				list_del_init(&bd1->bd_le.le_list);

fs/gfs2/lops.h

@@ -60,7 +60,7 @@ static inline void lops_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 }
 
 static inline void lops_before_scan(struct gfs2_jdesc *jd,
-				    struct gfs2_log_header *head,
+				    struct gfs2_log_header_host *head,
 				    unsigned int pass)
 {
 	int x;

fs/gfs2/meta_io.c

@@ -127,17 +127,17 @@ void gfs2_meta_sync(struct gfs2_glock *gl)
 
 /**
  * getbuf - Get a buffer with a given address space
- * @sdp: the filesystem
- * @aspace: the address space
+ * @gl: the glock
  * @blkno: the block number (filesystem scope)
  * @create: 1 if the buffer should be created
  *
  * Returns: the buffer
  */
 
-static struct buffer_head *getbuf(struct gfs2_sbd *sdp, struct inode *aspace,
-				  u64 blkno, int create)
+static struct buffer_head *getbuf(struct gfs2_glock *gl, u64 blkno, int create)
 {
+	struct address_space *mapping = gl->gl_aspace->i_mapping;
+	struct gfs2_sbd *sdp = gl->gl_sbd;
 	struct page *page;
 	struct buffer_head *bh;
 	unsigned int shift;
@@ -150,13 +150,13 @@ static struct buffer_head *getbuf(struct gfs2_sbd *sdp, struct inode *aspace,
 
 	if (create) {
 		for (;;) {
-			page = grab_cache_page(aspace->i_mapping, index);
+			page = grab_cache_page(mapping, index);
 			if (page)
 				break;
 			yield();
 		}
 	} else {
-		page = find_lock_page(aspace->i_mapping, index);
+		page = find_lock_page(mapping, index);
 		if (!page)
 			return NULL;
 	}
@@ -202,7 +202,7 @@ static void meta_prep_new(struct buffer_head *bh)
 struct buffer_head *gfs2_meta_new(struct gfs2_glock *gl, u64 blkno)
 {
 	struct buffer_head *bh;
-	bh = getbuf(gl->gl_sbd, gl->gl_aspace, blkno, CREATE);
+	bh = getbuf(gl, blkno, CREATE);
 	meta_prep_new(bh);
 	return bh;
 }
@@ -220,7 +220,7 @@ struct buffer_head *gfs2_meta_new(struct gfs2_glock *gl, u64 blkno)
 int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
 		   struct buffer_head **bhp)
 {
-	*bhp = getbuf(gl->gl_sbd, gl->gl_aspace, blkno, CREATE);
+	*bhp = getbuf(gl, blkno, CREATE);
 	if (!buffer_uptodate(*bhp))
 		ll_rw_block(READ_META, 1, bhp);
 	if (flags & DIO_WAIT) {
@@ -379,11 +379,10 @@ void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh,
 void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-	struct inode *aspace = ip->i_gl->gl_aspace;
 	struct buffer_head *bh;
 
 	while (blen) {
-		bh = getbuf(sdp, aspace, bstart, NO_CREATE);
+		bh = getbuf(ip->i_gl, bstart, NO_CREATE);
 		if (bh) {
 			struct gfs2_bufdata *bd = bh->b_private;
 
@@ -472,6 +471,9 @@ int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num,
 	struct buffer_head *bh = NULL, **bh_slot = ip->i_cache + height;
 	int in_cache = 0;
 
+	BUG_ON(!gl);
+	BUG_ON(!sdp);
+
 	spin_lock(&ip->i_spin);
 	if (*bh_slot && (*bh_slot)->b_blocknr == num) {
 		bh = *bh_slot;
@@ -481,7 +483,7 @@ int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num,
 	spin_unlock(&ip->i_spin);
 
 	if (!bh)
-		bh = getbuf(gl->gl_sbd, gl->gl_aspace, num, CREATE);
+		bh = getbuf(gl, num, CREATE);
 
 	if (!bh)
 		return -ENOBUFS;
@@ -532,7 +534,6 @@ int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num,
 struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen)
 {
 	struct gfs2_sbd *sdp = gl->gl_sbd;
-	struct inode *aspace = gl->gl_aspace;
 	struct buffer_head *first_bh, *bh;
 	u32 max_ra = gfs2_tune_get(sdp, gt_max_readahead) >>
 			  sdp->sd_sb.sb_bsize_shift;
@@ -544,7 +545,7 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen)
 	if (extlen > max_ra)
 		extlen = max_ra;
 
-	first_bh = getbuf(sdp, aspace, dblock, CREATE);
+	first_bh = getbuf(gl, dblock, CREATE);
 
 	if (buffer_uptodate(first_bh))
 		goto out;
@@ -555,7 +556,7 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen)
 	extlen--;
 
 	while (extlen) {
-		bh = getbuf(sdp, aspace, dblock, CREATE);
+		bh = getbuf(gl, dblock, CREATE);
 
 		if (!buffer_uptodate(bh) && !buffer_locked(bh))
 			ll_rw_block(READA, 1, &bh);
@@ -571,20 +572,3 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen)
 	return first_bh;
 }
 
-/**
- * gfs2_meta_syncfs - sync all the buffers in a filesystem
- * @sdp: the filesystem
- *
- */
-
-void gfs2_meta_syncfs(struct gfs2_sbd *sdp)
-{
-	gfs2_log_flush(sdp, NULL);
-	for (;;) {
-		gfs2_ail1_start(sdp, DIO_ALL);
-		if (gfs2_ail1_empty(sdp, DIO_ALL))
-			break;
-		msleep(10);
-	}
-}
-

fs/gfs2/meta_io.h

@@ -67,7 +67,6 @@ static inline int gfs2_meta_inode_buffer(struct gfs2_inode *ip,
 }
 
 struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen);
-void gfs2_meta_syncfs(struct gfs2_sbd *sdp);
 
 #define buffer_busy(bh) \
 ((bh)->b_state & ((1ul << BH_Dirty) | (1ul << BH_Lock) | (1ul << BH_Pinned)))

fs/gfs2/ops_address.c

@@ -156,19 +156,6 @@ static int gfs2_writepage(struct page *page, struct writeback_control *wbc)
 	return 0;
 }
 
-static int zero_readpage(struct page *page)
-{
-	void *kaddr;
-
-	kaddr = kmap_atomic(page, KM_USER0);
-	memset(kaddr, 0, PAGE_CACHE_SIZE);
-	kunmap_atomic(kaddr, KM_USER0);
-
-	SetPageUptodate(page);
-
-	return 0;
-}
-
 /**
  * stuffed_readpage - Fill in a Linux page with stuffed file data
  * @ip: the inode
@@ -183,9 +170,7 @@ static int stuffed_readpage(struct gfs2_inode *ip, struct page *page)
 	void *kaddr;
 	int error;
 
-	/* Only the first page of a stuffed file might contain data */
-	if (unlikely(page->index))
-		return zero_readpage(page);
+	BUG_ON(page->index);
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (error)
@@ -230,9 +215,9 @@ static int gfs2_readpage(struct file *file, struct page *page)
 				/* gfs2_sharewrite_nopage has grabbed the ip->i_gl already */
 				goto skip_lock;
 		}
-		gfs2_holder_init(ip->i_gl, LM_ST_SHARED, GL_ATIME|GL_AOP, &gh);
+		gfs2_holder_init(ip->i_gl, LM_ST_SHARED, GL_ATIME|LM_FLAG_TRY_1CB, &gh);
 		do_unlock = 1;
-		error = gfs2_glock_nq_m_atime(1, &gh);
+		error = gfs2_glock_nq_atime(&gh);
 		if (unlikely(error))
 			goto out_unlock;
 	}
@@ -254,6 +239,8 @@ static int gfs2_readpage(struct file *file, struct page *page)
 out:
 	return error;
 out_unlock:
+	if (error == GLR_TRYFAILED)
+		error = AOP_TRUNCATED_PAGE;
 	unlock_page(page);
 	if (do_unlock)
 		gfs2_holder_uninit(&gh);
@@ -293,9 +280,9 @@ static int gfs2_readpages(struct file *file, struct address_space *mapping,
 				goto skip_lock;
 		}
 		gfs2_holder_init(ip->i_gl, LM_ST_SHARED,
-				 LM_FLAG_TRY_1CB|GL_ATIME|GL_AOP, &gh);
+				 LM_FLAG_TRY_1CB|GL_ATIME, &gh);
 		do_unlock = 1;
-		ret = gfs2_glock_nq_m_atime(1, &gh);
+		ret = gfs2_glock_nq_atime(&gh);
 		if (ret == GLR_TRYFAILED)
 			goto out_noerror;
 		if (unlikely(ret))
@@ -366,10 +353,13 @@ static int gfs2_prepare_write(struct file *file, struct page *page,
 	unsigned int write_len = to - from;
 
 
-	gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_ATIME|GL_AOP, &ip->i_gh);
-	error = gfs2_glock_nq_m_atime(1, &ip->i_gh);
-	if (error)
+	gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_ATIME|LM_FLAG_TRY_1CB, &ip->i_gh);
+	error = gfs2_glock_nq_atime(&ip->i_gh);
+	if (unlikely(error)) {
+		if (error == GLR_TRYFAILED)
+			error = AOP_TRUNCATED_PAGE;
 		goto out_uninit;
+	}
 
 	gfs2_write_calc_reserv(ip, write_len, &data_blocks, &ind_blocks);
 
@@ -386,7 +376,7 @@ static int gfs2_prepare_write(struct file *file, struct page *page,
 		if (error)
 			goto out_alloc_put;
 
-		error = gfs2_quota_check(ip, ip->i_di.di_uid, ip->i_di.di_gid);
+		error = gfs2_quota_check(ip, ip->i_inode.i_uid, ip->i_inode.i_gid);
 		if (error)
 			goto out_qunlock;
 
@@ -482,8 +472,10 @@ static int gfs2_commit_write(struct file *file, struct page *page,
 
 		SetPageUptodate(page);
 
-		if (inode->i_size < file_size)
+		if (inode->i_size < file_size) {
 			i_size_write(inode, file_size);
+			mark_inode_dirty(inode);
+		}
 	} else {
 		if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED ||
 		    gfs2_is_jdata(ip))
@@ -498,11 +490,6 @@ static int gfs2_commit_write(struct file *file, struct page *page,
 		di->di_size = cpu_to_be64(inode->i_size);
 	}
 
-	di->di_mode = cpu_to_be32(inode->i_mode);
-	di->di_atime = cpu_to_be64(inode->i_atime.tv_sec);
-	di->di_mtime = cpu_to_be64(inode->i_mtime.tv_sec);
-	di->di_ctime = cpu_to_be64(inode->i_ctime.tv_sec);
-
 	brelse(dibh);
 	gfs2_trans_end(sdp);
 	if (al->al_requested) {
@@ -624,7 +611,7 @@ static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb,
 	 * on this path. All we need change is atime.
 	 */
 	gfs2_holder_init(ip->i_gl, LM_ST_SHARED, GL_ATIME, &gh);
-	rv = gfs2_glock_nq_m_atime(1, &gh);
+	rv = gfs2_glock_nq_atime(&gh);
 	if (rv)
 		goto out;
 
@@ -737,6 +724,9 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 			if (!atomic_read(&aspace->i_writecount))
 				return 0;
 
+			if (!(gfp_mask & __GFP_WAIT))
+				return 0;
+
 			if (time_after_eq(jiffies, t)) {
 				stuck_releasepage(bh);
 				/* should we withdraw here? */

fs/gfs2/ops_dentry.c

@@ -43,7 +43,7 @@ static int gfs2_drevalidate(struct dentry *dentry, struct nameidata *nd)
 	struct inode *inode = dentry->d_inode;
 	struct gfs2_holder d_gh;
 	struct gfs2_inode *ip;
-	struct gfs2_inum inum;
+	struct gfs2_inum_host inum;
 	unsigned int type;
 	int error;
 
@@ -76,7 +76,7 @@ static int gfs2_drevalidate(struct dentry *dentry, struct nameidata *nd)
 	if (!gfs2_inum_equal(&ip->i_num, &inum))
 		goto invalid_gunlock;
 
-	if (IF2DT(ip->i_di.di_mode) != type) {
+	if (IF2DT(ip->i_inode.i_mode) != type) {
 		gfs2_consist_inode(dip);
 		goto fail_gunlock;
 	}

fs/gfs2/ops_export.h

@@ -15,7 +15,7 @@
 
 extern struct export_operations gfs2_export_ops;
 struct gfs2_fh_obj {
-	struct gfs2_inum this;
+	struct gfs2_inum_host this;
 	__u32            imode;
 };
 

fs/gfs2/ops_file.h

@@ -17,7 +17,7 @@ extern struct file gfs2_internal_file_sentinel;
 extern int gfs2_internal_read(struct gfs2_inode *ip,
 			      struct file_ra_state *ra_state,
 			      char *buf, loff_t *pos, unsigned size);
-
+extern void gfs2_set_inode_flags(struct inode *inode);
 extern const struct file_operations gfs2_file_fops;
 extern const struct file_operations gfs2_dir_fops;
 

fs/gfs2/ops_fstype.c

@@ -237,7 +237,7 @@ static int init_locking(struct gfs2_sbd *sdp, struct gfs2_holder *mount_gh,
 }
 
 static struct inode *gfs2_lookup_root(struct super_block *sb,
-				      struct gfs2_inum *inum)
+				      struct gfs2_inum_host *inum)
 {
 	return gfs2_inode_lookup(sb, inum, DT_DIR);
 }
@@ -246,7 +246,7 @@ static int init_sb(struct gfs2_sbd *sdp, int silent, int undo)
 {
 	struct super_block *sb = sdp->sd_vfs;
 	struct gfs2_holder sb_gh;
-	struct gfs2_inum *inum;
+	struct gfs2_inum_host *inum;
 	struct inode *inode;
 	int error = 0;