[Cluster-devel] [GFS2/DLM] Pre-pull patch posting

[Cluster-devel] [GFS2/DLM] Pre-pull patch posting

[swhiteho]

Hi,

Since it seems that another merge window will probably be opening shortly
this is a posting of the current content of the GFS2/DLM -nmw git tree.

There are no new features this time, its all fixes and cleanups. I have
a few patches that I'm holding back which I'm intending to start off
the new -nmw git tree with after this merge, those include:

 o Patch to use page_mkwrite() on page faults (and thus allow us to
   use the filemap_fault() function rather than having a special
   GFS2 version of this like we do at the moment)
 o Patch to clean up gfs2_internal_read() as requested by Christoph
   Hellwig
 o Patches to separate the GFS2 aops into three separate operations
   structures (for writeback, ordered & jdata) as per ext3 which is
   waiting for the merge of Nick Piggin's AOP changes so that I don't
   land up creating a merge conflict.

Anyhow, back to what is in this set of patches... the major area of
change is the journalling code where there has been a lot of clean
up. Other than that, its mostly fixes for individual bugs and clean
ups that other people have sent to me,

Steve.

[Cluster-devel] [PATCH 01/51] [GFS2] Fix two races relating to glock callbacks

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

One of the races relates to referencing a variable while not holding
its protecting spinlock. The patch simply moves the test inside the
spin lock. The other races occurs when a demote to unlocked request
occurs during the time a demote to shared request is already running.
This of course only happens in the case that the lock was in the
exclusive mode to start with. The patch adds a check to see if another
demote request has occurred in the mean time and if it has, then it
performs a second demote.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 3f0974e..6a3eeba 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -545,12 +545,14 @@ static int rq_demote(struct gfs2_glock *gl)
 		return 0;
 	}
 	set_bit(GLF_LOCK, &gl->gl_flags);
-	spin_unlock(&gl->gl_spin);
 	if (gl->gl_demote_state == LM_ST_UNLOCKED ||
-	    gl->gl_state != LM_ST_EXCLUSIVE)
+	    gl->gl_state != LM_ST_EXCLUSIVE) {
+		spin_unlock(&gl->gl_spin);
 		gfs2_glock_drop_th(gl);
-	else
+	} else {
+		spin_unlock(&gl->gl_spin);
 		gfs2_glock_xmote_th(gl, NULL);
+	}
 	spin_lock(&gl->gl_spin);
 
 	return 0;
@@ -760,10 +762,20 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret)
 
 	if (!gh) {
 		gl->gl_stamp = jiffies;
-		if (ret & LM_OUT_CANCELED)
+		if (ret & LM_OUT_CANCELED) {
 			op_done = 0;
-		else
+		} else {
+			spin_lock(&gl->gl_spin);
+			if (gl->gl_state != gl->gl_demote_state) {
+				gl->gl_req_bh = NULL;
+				spin_unlock(&gl->gl_spin);
+				gfs2_glock_drop_th(gl);
+				gfs2_glock_put(gl);
+				return;
+			}
 			gfs2_demote_wake(gl);
+			spin_unlock(&gl->gl_spin);
+		}
 	} else {
 		spin_lock(&gl->gl_spin);
 		list_del_init(&gh->gh_list);
@@ -817,7 +829,7 @@ out:
  *
  */
 
-void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh)
+static void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh)
 {
 	struct gfs2_sbd *sdp = gl->gl_sbd;
 	int flags = gh ? gh->gh_flags : 0;
-- 
1.5.1.2

[Cluster-devel] [PATCH 02/51] [GFS2] Fix calculation of demote state

[swhiteho]

From: Josef Whiter <jwhiter@redhat.com>

If a glock is in the exclusive state and a request for demote to
deferred has been received, then further requests for demote to
shared are being ignored. This patch fixes that by ensuring that
we demote to unlocked in that case.

Signed-off-by: Josef Whiter <jwhiter@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 6a3eeba..6b6ae45 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -697,8 +697,9 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state, int remot
 			}
 			return;
 		}
-	} else if (gl->gl_demote_state != LM_ST_UNLOCKED) {
-		gl->gl_demote_state = state;
+	} else if (gl->gl_demote_state != LM_ST_UNLOCKED &&
+			gl->gl_demote_state != state) {
+		gl->gl_demote_state = LM_ST_UNLOCKED;
 	}
 	spin_unlock(&gl->gl_spin);
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 03/51] [GFS2] Clean up duplicate includes in fs/gfs2/

[swhiteho]

From: Jesper Juhl <jesper.juhl@gmail.com>

This patch cleans up duplicate includes in
	fs/gfs2/

Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 6b6ae45..d403fd7 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -25,8 +25,6 @@
 #include <asm/uaccess.h>
 #include <linux/seq_file.h>
 #include <linux/debugfs.h>
-#include <linux/module.h>
-#include <linux/kallsyms.h>
 
 #include "gfs2.h"
 #include "incore.h"
diff --git a/fs/gfs2/locking/dlm/lock_dlm.h b/fs/gfs2/locking/dlm/lock_dlm.h
index 24d70f7..9e8265d 100644
--- a/fs/gfs2/locking/dlm/lock_dlm.h
+++ b/fs/gfs2/locking/dlm/lock_dlm.h
@@ -13,7 +13,6 @@
 #include <linux/module.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
-#include <linux/module.h>
 #include <linux/types.h>
 #include <linux/string.h>
 #include <linux/list.h>
diff --git a/fs/gfs2/locking/nolock/main.c b/fs/gfs2/locking/nolock/main.c
index 0d149c8..d3b8ce6 100644
--- a/fs/gfs2/locking/nolock/main.c
+++ b/fs/gfs2/locking/nolock/main.c
@@ -9,7 +9,6 @@
 
 #include <linux/module.h>
 #include <linux/slab.h>
-#include <linux/module.h>
 #include <linux/init.h>
 #include <linux/types.h>
 #include <linux/fs.h>
-- 
1.5.1.2

[Cluster-devel] [PATCH 04/51] [GFS2] GFS2 not checking pointer on create when running under nfsd

[swhiteho]

From: Steve French <smfrench@gmail.com>

When looking at an unrelated problem, I noticed that nfsd does not
set nameidata pointer on create (ie nd is NULL).  This should
cause an oops in some cases in which when NFSd is mounted over GFS2.

Signed-off-by: Steve French <sfrench@us.ibm.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c
index 911c115..5b8b994 100644
--- a/fs/gfs2/ops_inode.c
+++ b/fs/gfs2/ops_inode.c
@@ -69,7 +69,7 @@ static int gfs2_create(struct inode *dir, struct dentry *dentry,
 			mark_inode_dirty(inode);
 			break;
 		} else if (PTR_ERR(inode) != -EEXIST ||
-			   (nd->intent.open.flags & O_EXCL)) {
+			   (nd && (nd->intent.open.flags & O_EXCL))) {
 			gfs2_holder_uninit(ghs);
 			return PTR_ERR(inode);
 		}
-- 
1.5.1.2

[Cluster-devel] [PATCH 05/51] [GFS2] Fix an oops in glock dumping

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

This fixes an oops which was occurring during glock dumping due to the
seq file code not taking a reference to the glock. Also this fixes a
memory leak which occurred in certain cases, in turn preventing the
filesystem from unmounting.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index d403fd7..e4bc8ae 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -46,7 +46,6 @@ struct glock_iter {
 	int hash;                     /* hash bucket index         */
 	struct gfs2_sbd *sdp;         /* incore superblock         */
 	struct gfs2_glock *gl;        /* current glock struct      */
-	struct hlist_head *hb_list;   /* current hash bucket ptr   */
 	struct seq_file *seq;         /* sequence file for debugfs */
 	char string[512];             /* scratch space             */
 };
@@ -1990,47 +1989,38 @@ int __init gfs2_glock_init(void)
 
 static int gfs2_glock_iter_next(struct glock_iter *gi)
 {
+	struct gfs2_glock *gl;
+
 	read_lock(gl_lock_addr(gi->hash));
-	while (1) {
-		if (!gi->hb_list) {  /* If we don't have a hash bucket yet */
-			gi->hb_list = &gl_hash_table[gi->hash].hb_list;
-			if (hlist_empty(gi->hb_list)) {
-				read_unlock(gl_lock_addr(gi->hash));
-				gi->hash++;
-				read_lock(gl_lock_addr(gi->hash));
-				gi->hb_list = NULL;
-				if (gi->hash >= GFS2_GL_HASH_SIZE) {
-					read_unlock(gl_lock_addr(gi->hash));
-					return 1;
-				}
-				else
-					continue;
-			}
-			if (!hlist_empty(gi->hb_list)) {
-				gi->gl = list_entry(gi->hb_list->first,
-						    struct gfs2_glock,
-						    gl_list);
-			}
-		} else {
-			if (gi->gl->gl_list.next == NULL) {
-				read_unlock(gl_lock_addr(gi->hash));
-				gi->hash++;
-				read_lock(gl_lock_addr(gi->hash));
-				gi->hb_list = NULL;
-				continue;
-			}
-			gi->gl = list_entry(gi->gl->gl_list.next,
-					    struct gfs2_glock, gl_list);
-		}
+	gl = gi->gl;
+	if (gl) {
+		gi->gl = hlist_entry(gl->gl_list.next, struct gfs2_glock,
+				     gl_list);
 		if (gi->gl)
-			break;
+			gfs2_glock_hold(gi->gl);
 	}
 	read_unlock(gl_lock_addr(gi->hash));
+	if (gl)
+		gfs2_glock_put(gl);
+
+	while(gi->gl == NULL) {
+		gi->hash++;
+		if (gi->hash >= GFS2_GL_HASH_SIZE)
+			return 1;
+		read_lock(gl_lock_addr(gi->hash));
+		gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first,
+				     struct gfs2_glock, gl_list);
+		if (gi->gl)
+			gfs2_glock_hold(gi->gl);
+		read_unlock(gl_lock_addr(gi->hash));
+	}
 	return 0;
 }
 
 static void gfs2_glock_iter_free(struct glock_iter *gi)
 {
+	if (gi->gl)
+		gfs2_glock_put(gi->gl);
 	kfree(gi);
 }
 
@@ -2044,12 +2034,17 @@ static struct glock_iter *gfs2_glock_iter_init(struct gfs2_sbd *sdp)
 
 	gi->sdp = sdp;
 	gi->hash = 0;
-	gi->gl = NULL;
-	gi->hb_list = NULL;
 	gi->seq = NULL;
 	memset(gi->string, 0, sizeof(gi->string));
 
-	if (gfs2_glock_iter_next(gi)) {
+	read_lock(gl_lock_addr(gi->hash));
+	gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first,
+			     struct gfs2_glock, gl_list);
+	if (gi->gl)
+		gfs2_glock_hold(gi->gl);
+	read_unlock(gl_lock_addr(gi->hash));
+
+	if (!gi->gl && gfs2_glock_iter_next(gi)) {
 		gfs2_glock_iter_free(gi);
 		return NULL;
 	}
@@ -2066,7 +2061,7 @@ static void *gfs2_glock_seq_start(struct seq_file *file, loff_t *pos)
 	if (!gi)
 		return NULL;
 
-	while (n--) {
+	while(n--) {
 		if (gfs2_glock_iter_next(gi)) {
 			gfs2_glock_iter_free(gi);
 			return NULL;
@@ -2093,7 +2088,9 @@ static void *gfs2_glock_seq_next(struct seq_file *file, void *iter_ptr,
 
 static void gfs2_glock_seq_stop(struct seq_file *file, void *iter_ptr)
 {
-	/* nothing for now */
+	struct glock_iter *gi = iter_ptr;
+	if (gi)
+		gfs2_glock_iter_free(gi);
 }
 
 static int gfs2_glock_seq_show(struct seq_file *file, void *iter_ptr)
-- 
1.5.1.2

[Cluster-devel] [PATCH 06/51] [GFS2] Move some code inside the log lock

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is the first of five patches for bug #248176:

There were still some critical variables being manipulated outside
the log_lock spinlock.  That usually resulted in a hang.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 3b395c4..a0371f8 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -117,7 +117,7 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 	struct buffer_head *bh;
 	struct gfs2_log_descriptor *ld;
 	struct gfs2_bufdata *bd1 = NULL, *bd2;
-	unsigned int total = sdp->sd_log_num_buf;
+	unsigned int total;
 	unsigned int offset = BUF_OFFSET;
 	unsigned int limit;
 	unsigned int num;
@@ -127,12 +127,16 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 	limit = buf_limit(sdp);
 	/* for 4k blocks, limit = 503 */
 
+	gfs2_log_lock(sdp);
+	total = sdp->sd_log_num_buf;
 	bd1 = bd2 = list_prepare_entry(bd1, &sdp->sd_log_le_buf, bd_le.le_list);
 	while(total) {
 		num = total;
 		if (total > limit)
 			num = limit;
+		gfs2_log_unlock(sdp);
 		bh = gfs2_log_get_buf(sdp);
+		gfs2_log_lock(sdp);
 		ld = (struct gfs2_log_descriptor *)bh->b_data;
 		ptr = (__be64 *)(bh->b_data + offset);
 		ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
@@ -152,21 +156,27 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 				break;
 		}
 
+		gfs2_log_unlock(sdp);
 		set_buffer_dirty(bh);
 		ll_rw_block(WRITE, 1, &bh);
+		gfs2_log_lock(sdp);
 
 		n = 0;
 		list_for_each_entry_continue(bd2, &sdp->sd_log_le_buf,
 					     bd_le.le_list) {
+			gfs2_log_unlock(sdp);
 			bh = gfs2_log_fake_buf(sdp, bd2->bd_bh);
 			set_buffer_dirty(bh);
 			ll_rw_block(WRITE, 1, &bh);
+			gfs2_log_lock(sdp);
 			if (++n >= num)
 				break;
 		}
 
+		BUG_ON(total < num);
 		total -= num;
 	}
+	gfs2_log_unlock(sdp);
 }
 
 static void buf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
@@ -524,7 +534,7 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 	struct gfs2_log_descriptor *ld;
 	unsigned int limit;
 	unsigned int total_dbuf;
-	unsigned int total_jdata = sdp->sd_log_num_jdata;
+	unsigned int total_jdata;
 	unsigned int num, n;
 	__be64 *ptr = NULL;
 
@@ -536,6 +546,7 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 	 */
 	gfs2_log_lock(sdp);
 	total_dbuf = sdp->sd_log_num_databuf;
+	total_jdata = sdp->sd_log_num_jdata;
 	bd2 = bd1 = list_prepare_entry(bd1, &sdp->sd_log_le_databuf,
 				       bd_le.le_list);
 	while(total_dbuf) {
@@ -621,10 +632,10 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 		}
 		gfs2_log_unlock(sdp);
 		if (bh) {
-			set_buffer_mapped(bh);
 			set_buffer_dirty(bh);
 			ll_rw_block(WRITE, 1, &bh);
 			bh = NULL;
+			ptr = NULL;
 		}
 		n = 0;
 		gfs2_log_lock(sdp);
-- 
1.5.1.2

[Cluster-devel] [PATCH 07/51] [GFS2] Revert part of earlier log.c changes

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is patch 2 of 5 for bug #248176.

The list_move code previously concocted in log.c for bug #238162
(see https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=238162#c23)
never runs as bh can now never be NULL at this point.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index f49a12e..f7c0608 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -83,11 +83,6 @@ static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 
 			gfs2_assert(sdp, bd->bd_ail == ai);
 
-			if (!bh){
-				list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list);
-                                continue;
-                        }
-
 			if (!buffer_busy(bh)) {
 				if (!buffer_uptodate(bh)) {
 					gfs2_log_unlock(sdp);
@@ -130,11 +125,6 @@ static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai, int fl
 					 bd_ail_st_list) {
 		bh = bd->bd_bh;
 
-		if (!bh){
-			list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list);
-			continue;
-		}
-
 		gfs2_assert(sdp, bd->bd_ail == ai);
 
 		if (buffer_busy(bh)) {
@@ -155,13 +145,14 @@ static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai, int fl
 
 static void gfs2_ail1_start(struct gfs2_sbd *sdp, int flags)
 {
-	struct list_head *head = &sdp->sd_ail1_list;
+	struct list_head *head;
 	u64 sync_gen;
 	struct list_head *first;
 	struct gfs2_ail *first_ai, *ai, *tmp;
 	int done = 0;
 
 	gfs2_log_lock(sdp);
+	head = &sdp->sd_ail1_list;
 	if (list_empty(head)) {
 		gfs2_log_unlock(sdp);
 		return;
-- 
1.5.1.2

[Cluster-devel] [PATCH 08/51] [GFS2] Prevent infinite loop in try_rgrp_unlink()

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is patch three of five for bug #248176.

The try_rgrp_unlink code in rgrp.c had an infinite loop.  This was
caused because the bitmap function rgblk_search can return a block
less than the "goal" block, in which case it was looping.  The fix is
to make it always march forward as needed.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index ce48c45..b93ac45 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -31,6 +31,7 @@
 #include "inode.h"
 
 #define BFITNOENT ((u32)~0)
+#define NO_BLOCK ((u64)~0)
 
 /*
  * These routines are used by the resource group routines (rgrp.c)
@@ -116,8 +117,7 @@ static unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, unsigned char *buffer,
  * @buffer: the buffer that holds the bitmaps
  * @buflen: the length (in bytes) of the buffer
  * @goal: start search at this block's bit-pair (within @buffer)
- * @old_state: GFS2_BLKST_XXX the state of the block we're looking for;
- *       bit 0 = alloc(1)/free(0), bit 1 = meta(1)/data(0)
+ * @old_state: GFS2_BLKST_XXX the state of the block we're looking for.
  *
  * Scope of @goal and returned block number is only within this bitmap buffer,
  * not entire rgrp or filesystem.  @buffer will be offset from the actual
@@ -137,9 +137,13 @@ static u32 gfs2_bitfit(struct gfs2_rgrpd *rgd, unsigned char *buffer,
 	byte = buffer + (goal / GFS2_NBBY);
 	bit = (goal % GFS2_NBBY) * GFS2_BIT_SIZE;
 	end = buffer + buflen;
-	alloc = (old_state & 1) ? 0 : 0x55;
+	alloc = (old_state == GFS2_BLKST_FREE) ? 0x55 : 0;
 
 	while (byte < end) {
+		/* If we're looking for a free block we can eliminate all
+		   bitmap settings with 0x55, which represents four data
+		   blocks in a row.  If we're looking for a data block, we can
+		   eliminate 0x00 which corresponds to four free blocks. */
 		if ((*byte & 0x55) == alloc) {
 			blk += (8 - bit) >> 1;
 
@@ -859,19 +863,21 @@ static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al)
 static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked)
 {
 	struct inode *inode;
-	u32 goal = 0;
+	u32 goal = 0, block;
 	u64 no_addr;
 
 	for(;;) {
 		if (goal >= rgd->rd_data)
 			break;
-		goal = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
-				    GFS2_BLKST_UNLINKED);
-		if (goal == BFITNOENT)
+		block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
+				     GFS2_BLKST_UNLINKED);
+		if (block == BFITNOENT)
 			break;
-		no_addr = goal + rgd->rd_data0;
+		/* rgblk_search can return a block < goal, so we need to
+		   keep it marching forward. */
+		no_addr = block + rgd->rd_data0;
 		goal++;
-		if (no_addr < *last_unlinked)
+		if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked)
 			continue;
 		*last_unlinked = no_addr;
 		inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN,
@@ -1152,7 +1158,7 @@ int gfs2_inplace_reserve_i(struct gfs2_inode *ip, char *file, unsigned int line)
 	struct gfs2_alloc *al = &ip->i_alloc;
 	struct inode *inode;
 	int error = 0;
-	u64 last_unlinked = 0;
+	u64 last_unlinked = NO_BLOCK;
 
 	if (gfs2_assert_warn(sdp, al->al_requested))
 		return -EINVAL;
@@ -1305,9 +1311,7 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
 		goal = 0;
 	}
 
-	if (old_state != new_state) {
-		gfs2_assert_withdraw(rgd->rd_sbd, blk != BFITNOENT);
-
+	if (blk != BFITNOENT && old_state != new_state) {
 		gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1);
 		gfs2_setbit(rgd, bi->bi_bh->b_data + bi->bi_offset,
 			    bi->bi_len, blk, new_state);
-- 
1.5.1.2

[Cluster-devel] [PATCH 09/51] [GFS2] use an temp variable to reduce a spin_unlock

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

this is more clear.

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/locking/dlm/plock.c b/fs/gfs2/locking/dlm/plock.c
index fba1f1d..1f7b038 100644
--- a/fs/gfs2/locking/dlm/plock.c
+++ b/fs/gfs2/locking/dlm/plock.c
@@ -346,15 +346,16 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count,
 
 static unsigned int dev_poll(struct file *file, poll_table *wait)
 {
+	unsigned int mask = 0;
+
 	poll_wait(file, &send_wq, wait);
 
 	spin_lock(&ops_lock);
-	if (!list_empty(&send_list)) {
-		spin_unlock(&ops_lock);
-		return POLLIN | POLLRDNORM;
-	}
+	if (!list_empty(&send_list))
+		mask = POLLIN | POLLRDNORM;
 	spin_unlock(&ops_lock);
-	return 0;
+
+	return mask;
 }
 
 static const struct file_operations dev_fops = {
-- 
1.5.1.2

[Cluster-devel] [PATCH 10/51] [GFS2] Detach buf data during in-place writeback

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is patch 5 of 5 for bug #248176

Metadata corruption was occurring because page references weren't
being removed in all cases.  I previously added a function called
detach_bufdata, but I discovered there already WAS a function out
there to do the job.  It's called gfs2_meta_cache_flush.  So I added
a call to that to remove the page references.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index f7c0608..00ab6c0 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -219,6 +219,7 @@ static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 {
 	struct list_head *head = &ai->ai_ail2_list;
 	struct gfs2_bufdata *bd;
+	struct gfs2_inode *bh_ip;
 
 	while (!list_empty(head)) {
 		bd = list_entry(head->prev, struct gfs2_bufdata,
@@ -228,6 +229,8 @@ static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 		list_del(&bd->bd_ail_st_list);
 		list_del(&bd->bd_ail_gl_list);
 		atomic_dec(&bd->bd_gl->gl_ail_count);
+		bh_ip = GFS2_I(bd->bd_bh->b_page->mapping->host);
+		gfs2_meta_cache_flush(bh_ip);
 		brelse(bd->bd_bh);
 	}
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 11/51] [GFS2] mark struct *_operations const

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

these struct *_operations are all method tables, thus should be const.

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/eaops.c b/fs/gfs2/eaops.c
index 1ab3e9d..aa8dbf3 100644
--- a/fs/gfs2/eaops.c
+++ b/fs/gfs2/eaops.c
@@ -200,28 +200,28 @@ static int security_eo_remove(struct gfs2_inode *ip, struct gfs2_ea_request *er)
 	return gfs2_ea_remove_i(ip, er);
 }
 
-static struct gfs2_eattr_operations gfs2_user_eaops = {
+static const struct gfs2_eattr_operations gfs2_user_eaops = {
 	.eo_get = user_eo_get,
 	.eo_set = user_eo_set,
 	.eo_remove = user_eo_remove,
 	.eo_name = "user",
 };
 
-struct gfs2_eattr_operations gfs2_system_eaops = {
+const struct gfs2_eattr_operations gfs2_system_eaops = {
 	.eo_get = system_eo_get,
 	.eo_set = system_eo_set,
 	.eo_remove = system_eo_remove,
 	.eo_name = "system",
 };
 
-static struct gfs2_eattr_operations gfs2_security_eaops = {
+static const struct gfs2_eattr_operations gfs2_security_eaops = {
 	.eo_get = security_eo_get,
 	.eo_set = security_eo_set,
 	.eo_remove = security_eo_remove,
 	.eo_name = "security",
 };
 
-struct gfs2_eattr_operations *gfs2_ea_ops[] = {
+const struct gfs2_eattr_operations *gfs2_ea_ops[] = {
 	NULL,
 	&gfs2_user_eaops,
 	&gfs2_system_eaops,
diff --git a/fs/gfs2/eaops.h b/fs/gfs2/eaops.h
index 508b4f7..da2f7fb 100644
--- a/fs/gfs2/eaops.h
+++ b/fs/gfs2/eaops.h
@@ -22,9 +22,9 @@ struct gfs2_eattr_operations {
 
 unsigned int gfs2_ea_name2type(const char *name, const char **truncated_name);
 
-extern struct gfs2_eattr_operations gfs2_system_eaops;
+extern const struct gfs2_eattr_operations gfs2_system_eaops;
 
-extern struct gfs2_eattr_operations *gfs2_ea_ops[];
+extern const struct gfs2_eattr_operations *gfs2_ea_ops[];
 
 #endif /* __EAOPS_DOT_H__ */
 
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index e4bc8ae..0054b7d 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -2103,7 +2103,7 @@ static int gfs2_glock_seq_show(struct seq_file *file, void *iter_ptr)
 	return 0;
 }
 
-static struct seq_operations gfs2_glock_seq_ops = {
+static const struct seq_operations gfs2_glock_seq_ops = {
 	.start = gfs2_glock_seq_start,
 	.next  = gfs2_glock_seq_next,
 	.stop  = gfs2_glock_seq_stop,
-- 
1.5.1.2

[Cluster-devel] [PATCH 12/51] [GFS2] use the declaration of gfs2_dops in the header file instead

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index cf5aa50..9a5e840 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -28,6 +28,7 @@
 #include "lm.h"
 #include "mount.h"
 #include "ops_fstype.h"
+#include "ops_dentry.h"
 #include "ops_super.h"
 #include "recovery.h"
 #include "rgrp.h"
@@ -38,8 +39,6 @@
 #define DO 0
 #define UNDO 1
 
-extern struct dentry_operations gfs2_dops;
-
 static struct gfs2_sbd *init_sbd(struct super_block *sb)
 {
 	struct gfs2_sbd *sdp;
-- 
1.5.1.2

[Cluster-devel] [PATCH 13/51] [GFS2] Reduce number of gfs2_scand processes to one

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

We only need a single gfs2_scand process rather than the one
per filesystem which we had previously. As a result the parameter
determining the frequency of gfs2_scand runs becomes a module
parameter rather than a mount parameter as it was before.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/daemon.c b/fs/gfs2/daemon.c
index 3548d9f..3731ab0 100644
--- a/fs/gfs2/daemon.c
+++ b/fs/gfs2/daemon.c
@@ -35,30 +35,6 @@
    The kthread functions used to start these daemons block and flush signals. */
 
 /**
- * gfs2_scand - Look for cached glocks and inodes to toss from memory
- * @sdp: Pointer to GFS2 superblock
- *
- * One of these daemons runs, finding candidates to add to sd_reclaim_list.
- * See gfs2_glockd()
- */
-
-int gfs2_scand(void *data)
-{
-	struct gfs2_sbd *sdp = data;
-	unsigned long t;
-
-	while (!kthread_should_stop()) {
-		gfs2_scand_internal(sdp);
-		t = gfs2_tune_get(sdp, gt_scand_secs) * HZ;
-		if (freezing(current))
-			refrigerator();
-		schedule_timeout_interruptible(t);
-	}
-
-	return 0;
-}
-
-/**
  * gfs2_glockd - Reclaim unused glock structures
  * @sdp: Pointer to GFS2 superblock
  *
diff --git a/fs/gfs2/daemon.h b/fs/gfs2/daemon.h
index 8010071..0de9b35 100644
--- a/fs/gfs2/daemon.h
+++ b/fs/gfs2/daemon.h
@@ -10,7 +10,6 @@
 #ifndef __DAEMON_DOT_H__
 #define __DAEMON_DOT_H__
 
-int gfs2_scand(void *data);
 int gfs2_glockd(void *data);
 int gfs2_recoverd(void *data);
 int gfs2_logd(void *data);
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 0054b7d..559937c 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -25,6 +25,8 @@
 #include <asm/uaccess.h>
 #include <linux/seq_file.h>
 #include <linux/debugfs.h>
+#include <linux/kthread.h>
+#include <linux/freezer.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -58,6 +60,8 @@ static void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh);
 static void gfs2_glock_drop_th(struct gfs2_glock *gl);
 static DECLARE_RWSEM(gfs2_umount_flush_sem);
 static struct dentry *gfs2_root;
+static struct task_struct *scand_process;
+static unsigned int scand_secs = 5;
 
 #define GFS2_GL_HASH_SHIFT      15
 #define GFS2_GL_HASH_SIZE       (1 << GFS2_GL_HASH_SHIFT)
@@ -1627,7 +1631,7 @@ static int examine_bucket(glock_examiner examiner, struct gfs2_sbd *sdp,
 		goto out;
 	gl = list_entry(head->first, struct gfs2_glock, gl_list);
 	while(1) {
-		if (gl->gl_sbd == sdp) {
+		if (!sdp || gl->gl_sbd == sdp) {
 			gfs2_glock_hold(gl);
 			read_unlock(gl_lock_addr(hash));
 			if (prev)
@@ -1645,6 +1649,7 @@ out:
 	read_unlock(gl_lock_addr(hash));
 	if (prev)
 		gfs2_glock_put(prev);
+	cond_resched();
 	return has_entries;
 }
 
@@ -1673,20 +1678,6 @@ out_schedule:
 }
 
 /**
- * gfs2_scand_internal - Look for glocks and inodes to toss from memory
- * @sdp: the filesystem
- *
- */
-
-void gfs2_scand_internal(struct gfs2_sbd *sdp)
-{
-	unsigned int x;
-
-	for (x = 0; x < GFS2_GL_HASH_SIZE; x++)
-		examine_bucket(scan_glock, sdp, x);
-}
-
-/**
  * clear_glock - look@a glock and see if we can free it from glock cache
  * @gl: the glock to look at
  *
@@ -1973,6 +1964,35 @@ static int gfs2_dump_lockstate(struct gfs2_sbd *sdp)
 	return error;
 }
 
+/**
+ * gfs2_scand - Look for cached glocks and inodes to toss from memory
+ * @sdp: Pointer to GFS2 superblock
+ *
+ * One of these daemons runs, finding candidates to add to sd_reclaim_list.
+ * See gfs2_glockd()
+ */
+
+static int gfs2_scand(void *data)
+{
+	unsigned x;
+	unsigned delay;
+
+	while (!kthread_should_stop()) {
+		for (x = 0; x < GFS2_GL_HASH_SIZE; x++)
+			examine_bucket(scan_glock, NULL, x);
+		if (freezing(current))
+			refrigerator();
+		delay = scand_secs;
+		if (delay < 1)
+			delay = 1;
+		schedule_timeout_interruptible(delay * HZ);
+	}
+
+	return 0;
+}
+
+
+
 int __init gfs2_glock_init(void)
 {
 	unsigned i;
@@ -1984,9 +2004,22 @@ int __init gfs2_glock_init(void)
 		rwlock_init(&gl_hash_locks[i]);
 	}
 #endif
+
+	scand_process = kthread_run(gfs2_scand, NULL, "gfs2_scand");
+	if (IS_ERR(scand_process))
+		return PTR_ERR(scand_process);
+
 	return 0;
 }
 
+void gfs2_glock_exit(void)
+{
+	kthread_stop(scand_process);
+}
+
+module_param(scand_secs, uint, S_IRUGO|S_IWUSR);
+MODULE_PARM_DESC(scand_secs, "The number of seconds between scand runs");
+
 static int gfs2_glock_iter_next(struct glock_iter *gi)
 {
 	struct gfs2_glock *gl;
diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h
index 7721ca3..f7a8e62 100644
--- a/fs/gfs2/glock.h
+++ b/fs/gfs2/glock.h
@@ -132,11 +132,11 @@ void gfs2_glock_cb(void *cb_data, unsigned int type, void *data);
 
 void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl);
 void gfs2_reclaim_glock(struct gfs2_sbd *sdp);
-
-void gfs2_scand_internal(struct gfs2_sbd *sdp);
 void gfs2_gl_hash_clear(struct gfs2_sbd *sdp, int wait);
 
 int __init gfs2_glock_init(void);
+void gfs2_glock_exit(void);
+
 int gfs2_create_debugfs_file(struct gfs2_sbd *sdp);
 void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp);
 int gfs2_register_debugfs(void);
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 170ba93..1390b30 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -429,7 +429,6 @@ struct gfs2_tune {
 	unsigned int gt_log_flush_secs;
 	unsigned int gt_jindex_refresh_secs; /* Check for new journal index */
 
-	unsigned int gt_scand_secs;
 	unsigned int gt_recoverd_secs;
 	unsigned int gt_logd_secs;
 	unsigned int gt_quotad_secs;
@@ -574,7 +573,6 @@ struct gfs2_sbd {
 
 	/* Daemon stuff */
 
-	struct task_struct *sd_scand_process;
 	struct task_struct *sd_recoverd_process;
 	struct task_struct *sd_logd_process;
 	struct task_struct *sd_quotad_process;
diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c
index d5d4e68..79c91fd 100644
--- a/fs/gfs2/main.c
+++ b/fs/gfs2/main.c
@@ -107,6 +107,8 @@ static int __init init_gfs2_fs(void)
 fail_unregister:
 	unregister_filesystem(&gfs2_fs_type);
 fail:
+	gfs2_glock_exit();
+
 	if (gfs2_bufdata_cachep)
 		kmem_cache_destroy(gfs2_bufdata_cachep);
 
@@ -127,6 +129,7 @@ fail:
 
 static void __exit exit_gfs2_fs(void)
 {
+	gfs2_glock_exit();
 	gfs2_unregister_debugfs();
 	unregister_filesystem(&gfs2_fs_type);
 	unregister_filesystem(&gfs2meta_fs_type);
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 9a5e840..58c730b 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -160,14 +160,6 @@ static int init_locking(struct gfs2_sbd *sdp, struct gfs2_holder *mount_gh,
 	if (undo)
 		goto fail_trans;
 
-	p = kthread_run(gfs2_scand, sdp, "gfs2_scand");
-	error = IS_ERR(p);
-	if (error) {
-		fs_err(sdp, "can't start scand thread: %d\n", error);
-		return error;
-	}
-	sdp->sd_scand_process = p;
-
 	for (sdp->sd_glockd_num = 0;
 	     sdp->sd_glockd_num < sdp->sd_args.ar_num_glockd;
 	     sdp->sd_glockd_num++) {
@@ -228,7 +220,6 @@ fail:
 	while (sdp->sd_glockd_num--)
 		kthread_stop(sdp->sd_glockd_process[sdp->sd_glockd_num]);
 
-	kthread_stop(sdp->sd_scand_process);
 	return error;
 }
 
diff --git a/fs/gfs2/ops_super.c b/fs/gfs2/ops_super.c
index 603d940..4316690 100644
--- a/fs/gfs2/ops_super.c
+++ b/fs/gfs2/ops_super.c
@@ -92,7 +92,6 @@ static void gfs2_put_super(struct super_block *sb)
 	kthread_stop(sdp->sd_recoverd_process);
 	while (sdp->sd_glockd_num--)
 		kthread_stop(sdp->sd_glockd_process[sdp->sd_glockd_num]);
-	kthread_stop(sdp->sd_scand_process);
 
 	if (!(sb->s_flags & MS_RDONLY)) {
 		error = gfs2_make_fs_ro(sdp);
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index f916b97..5589802 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -58,7 +58,6 @@ void gfs2_tune_init(struct gfs2_tune *gt)
 	gt->gt_incore_log_blocks = 1024;
 	gt->gt_log_flush_secs = 60;
 	gt->gt_jindex_refresh_secs = 60;
-	gt->gt_scand_secs = 15;
 	gt->gt_recoverd_secs = 60;
 	gt->gt_logd_secs = 1;
 	gt->gt_quotad_secs = 5;
diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c
index c26c21b..ba3a172 100644
--- a/fs/gfs2/sys.c
+++ b/fs/gfs2/sys.c
@@ -442,7 +442,6 @@ TUNE_ATTR(quota_simul_sync, 1);
 TUNE_ATTR(quota_cache_secs, 1);
 TUNE_ATTR(stall_secs, 1);
 TUNE_ATTR(statfs_quantum, 1);
-TUNE_ATTR_DAEMON(scand_secs, scand_process);
 TUNE_ATTR_DAEMON(recoverd_secs, recoverd_process);
 TUNE_ATTR_DAEMON(logd_secs, logd_process);
 TUNE_ATTR_DAEMON(quotad_secs, quotad_process);
@@ -464,7 +463,6 @@ static struct attribute *tune_attrs[] = {
 	&tune_attr_quota_cache_secs.attr,
 	&tune_attr_stall_secs.attr,
 	&tune_attr_statfs_quantum.attr,
-	&tune_attr_scand_secs.attr,
 	&tune_attr_recoverd_secs.attr,
 	&tune_attr_logd_secs.attr,
 	&tune_attr_quotad_secs.attr,
-- 
1.5.1.2

[Cluster-devel] [PATCH 14/51] [GFS2] invalid metadata block - REVISED

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is for bugzilla bug #248176: GFS2: invalid metadata block

Patches 1 thru 3 were accepted upstream, but there were problems
with 4 and 5.  Those issues have been resolved and now the recovery
tests are passing without errors.  This code has gone through
41 * 3 successful gfs2 recovery tests before it hit an
unrelated (openais) problem.

This is a complete rewrite of patch 4 for bug #248176.

Part of the problem was that inodes were being recycled
before their buffers were flushed to the journal logs.
Another problem was that the clone bitmaps were being
searched for deleted inodes to recycle, but only the
"real" bitmaps should be searched for that purpose.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index b93ac45..2d7f7ea 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -865,12 +865,15 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked)
 	struct inode *inode;
 	u32 goal = 0, block;
 	u64 no_addr;
+	struct gfs2_sbd *sdp = rgd->rd_sbd;
 
 	for(;;) {
 		if (goal >= rgd->rd_data)
 			break;
+		down_write(&sdp->sd_log_flush_lock);
 		block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
 				     GFS2_BLKST_UNLINKED);
+		up_write(&sdp->sd_log_flush_lock);
 		if (block == BFITNOENT)
 			break;
 		/* rgblk_search can return a block < goal, so we need to
@@ -1295,7 +1298,9 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal,
 	   allocatable block anywhere else, we want to be able wrap around and
 	   search in the first part of our first-searched bit block.  */
 	for (x = 0; x <= length; x++) {
-		if (bi->bi_clone)
+		/* The GFS2_BLKST_UNLINKED state doesn't apply to the clone
+		   bitmaps, so we must search the originals for that. */
+		if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone)
 			blk = gfs2_bitfit(rgd, bi->bi_clone + bi->bi_offset,
 					  bi->bi_len, goal, old_state);
 		else
-- 
1.5.1.2

[Cluster-devel] [PATCH 15/51] [GFS2] Ensure journal file cache is flushed after recovery

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is for bugzilla bug #248176: GFS2: invalid metadata block

Patches 1 thru 3 were accepted upstream, but there were problems
with 4 and 5.  Those issues have been resolved and now the recovery
tests are passing without errors.  This code has gone through
41 * 3 successful gfs2 recovery tests before it hit an
unrelated (openais) problem.  I'm continuing to test it.

This is a complete rewrite of patch 5 for bug #248176, written by
Steve Whitehouse.  This is referred to in the bugzilla record as
"new 6" and "a different solution".

The problem was that the journal inodes, although protected by
a glock, were not synched with the other nodes because they don't
use the inode glock synch operations (i.e. no "glops" were defined).
Therefore, journal recovery on a journal-recovering node were causing
the blocks to get out of sync with the node that was actually trying
to use that journal as it comes back up from a reboot.

There are two possible solutions: (1) To make the journals use the
normal inode glock sync operations, or (2) To make the journal
operations take effect immediately (i.e. no caching).  Although
option 1 works, it turns out to be a lot more code.  Steve opted
for option 2, which is much simpler and therefore less prone to
regression errors.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

--

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 58c730b..f0bcaa2 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -358,7 +358,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 
 		ip = GFS2_I(sdp->sd_jdesc->jd_inode);
 		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED,
-					   LM_FLAG_NOEXP | GL_EXACT,
+					   LM_FLAG_NOEXP | GL_EXACT | GL_NOCACHE,
 					   &sdp->sd_jinode_gh);
 		if (error) {
 			fs_err(sdp, "can't acquire journal inode glock: %d\n",
diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c
index 5ada38c..beb6c7a 100644
--- a/fs/gfs2/recovery.c
+++ b/fs/gfs2/recovery.c
@@ -469,7 +469,7 @@ int gfs2_recover_journal(struct gfs2_jdesc *jd)
 		};
 
 		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED,
-					   LM_FLAG_NOEXP, &ji_gh);
+					   LM_FLAG_NOEXP | GL_NOCACHE, &ji_gh);
 		if (error)
 			goto fail_gunlock_j;
 	} else {
-- 
1.5.1.2

[Cluster-devel] [PATCH 16/51] [GFS2] use list_for_each_entry instead

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index f0bcaa2..32b2859 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -808,7 +808,6 @@ static struct super_block* get_gfs2_sb(const char *dev_name)
 	struct nameidata nd;
 	struct file_system_type *fstype;
 	struct super_block *sb = NULL, *s;
-	struct list_head *l;
 	int error;
 
 	error = path_lookup(dev_name, LOOKUP_FOLLOW, &nd);
@@ -820,8 +819,7 @@ static struct super_block* get_gfs2_sb(const char *dev_name)
 	error = vfs_getattr(nd.mnt, nd.dentry, &stat);
 
 	fstype = get_fs_type("gfs2");
-	list_for_each(l, &fstype->fs_supers) {
-		s = list_entry(l, struct super_block, s_instances);
+	list_for_each_entry(s, &fstype->fs_supers, s_instances) {
 		if ((S_ISBLK(stat.mode) && s->s_dev == stat.rdev) ||
 		    (S_ISDIR(stat.mode) && s == nd.dentry->d_inode->i_sb)) {
 			sb = s;
-- 
1.5.1.2

[Cluster-devel] [PATCH 17/51] [GFS2] unneeded typecast

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

sb->s_fs_info is a void pointer, thus the type cast is not needed.

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 32b2859..25cfab9 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -849,7 +849,7 @@ static int gfs2_get_sb_meta(struct file_system_type *fs_type, int flags,
 		error = -ENOENT;
 		goto error;
 	}
-	sdp = (struct gfs2_sbd*) sb->s_fs_info;
+	sdp = sb->s_fs_info;
 	if (sdp->sd_vfs_meta) {
 		printk(KERN_WARNING "GFS2: gfs2meta mount already exists\n");
 		error = -EBUSY;
-- 
1.5.1.2

[Cluster-devel] [PATCH 18/51] [GFS2] better code for translating characters

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

the original code could work, but I think this code could work better.

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 25cfab9..6c820cb 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -144,7 +144,8 @@ static int init_names(struct gfs2_sbd *sdp, int silent)
 	snprintf(sdp->sd_proto_name, GFS2_FSNAME_LEN, "%s", proto);
 	snprintf(sdp->sd_table_name, GFS2_FSNAME_LEN, "%s", table);
 
-	while ((table = strchr(sdp->sd_table_name, '/')))
+	table = sdp->sd_table_name;
+	while ((table = strchr(table, '/')))
 		*table = '_';
 
 out:
-- 
1.5.1.2

[Cluster-devel] [PATCH 19/51] [GFS2] Force unstuff of hidden quota inode

[swhiteho]

From: Abhijith Das <adas@redhat.com>

This patch forcibly unstuffs (if stuffed) the hidden quota inode at the
first availble opportunity. In any practical scenario the quota inode
won't be stuffed, so this is ok to do. Unstuffing the quota inode allows
us to ignore the case of a stuffed quota inode in gfs2_adjust_quota().

Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index 6e546ee..5dfa465 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -614,6 +614,16 @@ static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc,
 	s64 value;
 	int err = -EIO;
 
+	if (gfs2_is_stuffed(ip)) {
+		struct gfs2_alloc *al = NULL;
+		al = gfs2_alloc_get(ip);
+		/* just request 1 blk */
+		al->al_requested = 1;
+		gfs2_inplace_reserve(ip);
+		gfs2_unstuff_dinode(ip, NULL);
+		gfs2_inplace_release(ip);
+		gfs2_alloc_put(ip);
+	}
 	page = grab_cache_page(mapping, index);
 	if (!page)
 		return -ENOMEM;
-- 
1.5.1.2

[Cluster-devel] [PATCH 20/51] [GFS2] fixed a NULL pointer assignment BUG

[swhiteho]

From: Denis Cheng <crquan@gmail.com>

Signed-off-by: Denis Cheng <crquan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 6c820cb..c1c6672 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -292,8 +292,9 @@ static int init_sb(struct gfs2_sbd *sdp, int silent, int undo)
 		fs_err(sdp, "can't get root dentry\n");
 		error = -ENOMEM;
 		iput(inode);
-	}
-	sb->s_root->d_op = &gfs2_dops;
+	} else
+		sb->s_root->d_op = &gfs2_dops;
+	
 out:
 	gfs2_glock_dq_uninit(&sb_gh);
 	return error;
-- 
1.5.1.2

[Cluster-devel] [PATCH 21/51] [GFS2] Fix quota do_list operation hang

[swhiteho]

From: Abhijith Das <adas@redhat.com>

This is the filesystem part of the patches to fix this bz. There are
additional userland patches (gfs2_quota, libgfs2) for the complete
solution. This patch adds a new field qu_ll_next to the gfs2_quota
structure. This field allows us to create linked lists of quotas in the
ondisk quota inode. Instead of scanning through the entire sparse quota
file for valid quotas, we can now simply walk through the user and group
quota linked lists to perform the do_list operation.

Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index 5dfa465..addb51e 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -70,6 +70,7 @@ struct gfs2_quota_host {
 	u64 qu_limit;
 	u64 qu_warn;
 	s64 qu_value;
+	u32 qu_ll_next;
 };
 
 struct gfs2_quota_change_host {
@@ -580,6 +581,7 @@ static void gfs2_quota_in(struct gfs2_quota_host *qu, const void *buf)
 	qu->qu_limit = be64_to_cpu(str->qu_limit);
 	qu->qu_warn = be64_to_cpu(str->qu_warn);
 	qu->qu_value = be64_to_cpu(str->qu_value);
+	qu->qu_ll_next = be32_to_cpu(str->qu_ll_next);
 }
 
 static void gfs2_quota_out(const struct gfs2_quota_host *qu, void *buf)
@@ -589,6 +591,7 @@ static void gfs2_quota_out(const struct gfs2_quota_host *qu, void *buf)
 	str->qu_limit = cpu_to_be64(qu->qu_limit);
 	str->qu_warn = cpu_to_be64(qu->qu_warn);
 	str->qu_value = cpu_to_be64(qu->qu_value);
+	str->qu_ll_next = cpu_to_be32(qu->qu_ll_next);
 	memset(&str->qu_reserved, 0, sizeof(str->qu_reserved));
 }
 
diff --git a/include/linux/gfs2_ondisk.h b/include/linux/gfs2_ondisk.h
index a44a6a0..c3c19f9 100644
--- a/include/linux/gfs2_ondisk.h
+++ b/include/linux/gfs2_ondisk.h
@@ -170,6 +170,33 @@ struct gfs2_rgrp {
 };
 
 /*
+ * quota linked list: user quotas and group quotas form two separate 
+ * singly linked lists. ll_next stores uids or gids of next quotas in the 
+ * linked list.
+
+Given the uid/gid, how to calculate the quota file offsets for the corresponding
+gfs2_quota structures on disk:
+
+for user quotas, given uid,
+offset = uid * sizeof(struct gfs2_quota);
+
+for group quotas, given gid,
+offset = (gid * sizeof(struct gfs2_quota)) + sizeof(struct gfs2_quota);
+
+
+  uid:0   gid:0       uid:12   gid:12      uid:17   gid:17     uid:5142 gid:5142
++-------+-------+    +-------+-------+    +-------+- - - -+    +- - - -+-------+
+| valid | valid | :: | valid | valid | :: | valid | inval | :: | inval | valid |
++-------+-------+    +-------+-------+    +-------+- - - -+    +- - - -+-------+
+next:12   next:12    next:17 next:5142    next:NULL                    next:NULL
+    |       |            |       |            |<-- user quota list         |
+     \______|___________/ \______|___________/         group quota list -->|
+            |                    |                                         |
+             \__________________/ \_______________________________________/
+
+*/
+
+/*
  * quota structure
  */
 
@@ -177,7 +204,8 @@ struct gfs2_quota {
 	__be64 qu_limit;
 	__be64 qu_warn;
 	__be64 qu_value;
-	__u8 qu_reserved[64];
+	__be32 qu_ll_next; /* location of next quota in list */
+	__u8 qu_reserved[60];
 };
 
 /*
-- 
1.5.1.2

[Cluster-devel] [PATCH 22/51] [GFS2] Clean up invalidatepage/releasepage

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

This patch fixes some bugs relating to journaled data files by cleaning
up the gfs2_invalidatepage() and gfs2_releasepage() functions. We now
never block during gfs2_releasepage(), instead we always either release
or refuse to release depending on the status of the buffers.

This fixes Red Hat bugzillas #248969 and #252392.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Bob Peterson <rpeterso@redhat.com>

diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 777ca46..88342e0 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -156,9 +156,11 @@ static void inode_go_sync(struct gfs2_glock *gl)
 		ip = NULL;
 
 	if (test_bit(GLF_DIRTY, &gl->gl_flags)) {
-		if (ip)
+		if (ip && !gfs2_is_jdata(ip))
 			filemap_fdatawrite(ip->i_inode.i_mapping);
 		gfs2_log_flush(gl->gl_sbd, gl);
+		if (ip && gfs2_is_jdata(ip))
+			filemap_fdatawrite(ip->i_inode.i_mapping);
 		gfs2_meta_sync(gl);
 		if (ip) {
 			struct address_space *mapping = ip->i_inode.i_mapping;
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 00ab6c0..d0e6b42 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -229,8 +229,10 @@ static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 		list_del(&bd->bd_ail_st_list);
 		list_del(&bd->bd_ail_gl_list);
 		atomic_dec(&bd->bd_gl->gl_ail_count);
-		bh_ip = GFS2_I(bd->bd_bh->b_page->mapping->host);
-		gfs2_meta_cache_flush(bh_ip);
+		if (bd->bd_bh->b_page->mapping) {
+			bh_ip = GFS2_I(bd->bd_bh->b_page->mapping->host);
+			gfs2_meta_cache_flush(bh_ip);
+		}
 		brelse(bd->bd_bh);
 	}
 }
diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c
index 42a5f58..8407d1d 100644
--- a/fs/gfs2/ops_address.c
+++ b/fs/gfs2/ops_address.c
@@ -616,58 +616,13 @@ static sector_t gfs2_bmap(struct address_space *mapping, sector_t lblock)
 	return dblock;
 }
 
-static void discard_buffer(struct gfs2_sbd *sdp, struct buffer_head *bh)
-{
-	struct gfs2_bufdata *bd;
-
-	gfs2_log_lock(sdp);
-	bd = bh->b_private;
-	if (bd) {
-		bd->bd_bh = NULL;
-		bh->b_private = NULL;
-		if (!bd->bd_ail && list_empty(&bd->bd_le.le_list))
-			kmem_cache_free(gfs2_bufdata_cachep, bd);
-	}
-	gfs2_log_unlock(sdp);
-
-	lock_buffer(bh);
-	clear_buffer_dirty(bh);
-	bh->b_bdev = NULL;
-	clear_buffer_mapped(bh);
-	clear_buffer_req(bh);
-	clear_buffer_new(bh);
-	clear_buffer_delay(bh);
-	unlock_buffer(bh);
-}
-
 static void gfs2_invalidatepage(struct page *page, unsigned long offset)
 {
-	struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host);
-	struct buffer_head *head, *bh, *next;
-	unsigned int curr_off = 0;
-
 	BUG_ON(!PageLocked(page));
 	if (offset == 0)
 		ClearPageChecked(page);
-	if (!page_has_buffers(page))
-		return;
 
-	bh = head = page_buffers(page);
-	do {
-		unsigned int next_off = curr_off + bh->b_size;
-		next = bh->b_this_page;
-
-		if (offset <= curr_off)
-			discard_buffer(sdp, bh);
-
-		curr_off = next_off;
-		bh = next;
-	} while (bh != head);
-
-	if (!offset)
-		try_to_release_page(page, 0);
-
-	return;
+	block_invalidatepage(page, offset);
 }
 
 /**
@@ -736,59 +691,6 @@ out:
 }
 
 /**
- * stuck_releasepage - We're stuck in gfs2_releasepage().  Print stuff out.
- * @bh: the buffer we're stuck on
- *
- */
-
-static void stuck_releasepage(struct buffer_head *bh)
-{
-	struct inode *inode = bh->b_page->mapping->host;
-	struct gfs2_sbd *sdp = inode->i_sb->s_fs_info;
-	struct gfs2_bufdata *bd = bh->b_private;
-	struct gfs2_glock *gl;
-static unsigned limit = 0;
-
-	if (limit > 3)
-		return;
-	limit++;
-
-	fs_warn(sdp, "stuck in gfs2_releasepage() %p\n", inode);
-	fs_warn(sdp, "blkno = %llu, bh->b_count = %d\n",
-		(unsigned long long)bh->b_blocknr, atomic_read(&bh->b_count));
-	fs_warn(sdp, "pinned = %u\n", buffer_pinned(bh));
-	fs_warn(sdp, "bh->b_private = %s\n", (bd) ? "!NULL" : "NULL");
-
-	if (!bd)
-		return;
-
-	gl = bd->bd_gl;
-
-	fs_warn(sdp, "gl = (%u, %llu)\n",
-		gl->gl_name.ln_type, (unsigned long long)gl->gl_name.ln_number);
-
-	fs_warn(sdp, "bd_list_tr = %s, bd_le.le_list = %s\n",
-		(list_empty(&bd->bd_list_tr)) ? "no" : "yes",
-		(list_empty(&bd->bd_le.le_list)) ? "no" : "yes");
-
-	if (gl->gl_ops == &gfs2_inode_glops) {
-		struct gfs2_inode *ip = gl->gl_object;
-		unsigned int x;
-
-		if (!ip)
-			return;
-
-		fs_warn(sdp, "ip = %llu %llu\n",
-			(unsigned long long)ip->i_no_formal_ino,
-			(unsigned long long)ip->i_no_addr);
-
-		for (x = 0; x < GFS2_MAX_META_HEIGHT; x++)
-			fs_warn(sdp, "ip->i_cache[%u] = %s\n",
-				x, (ip->i_cache[x]) ? "!NULL" : "NULL");
-	}
-}
-
-/**
  * gfs2_releasepage - free the metadata associated with a page
  * @page: the page that's being released
  * @gfp_mask: passed from Linux VFS, ignored by us
@@ -805,38 +707,31 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 	struct gfs2_sbd *sdp = aspace->i_sb->s_fs_info;
 	struct buffer_head *bh, *head;
 	struct gfs2_bufdata *bd;
-	unsigned long t = jiffies + gfs2_tune_get(sdp, gt_stall_secs) * HZ;
 
 	if (!page_has_buffers(page))
 		goto out;
 
+	gfs2_log_lock(sdp);
 	head = bh = page_buffers(page);
 	do {
-		while (atomic_read(&bh->b_count)) {
-			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? */
-				return 0;
-			}
-
-			yield();
-		}
-
+		if (atomic_read(&bh->b_count))
+			goto cannot_release;
+		bd = bh->b_private;
+		if (bd && bd->bd_ail)
+			goto cannot_release;
 		gfs2_assert_warn(sdp, !buffer_pinned(bh));
 		gfs2_assert_warn(sdp, !buffer_dirty(bh));
+		bh = bh->b_this_page;
+	} while(bh != head);
+	gfs2_log_unlock(sdp);
 
+	head = bh = page_buffers(page);
+	do {
 		gfs2_log_lock(sdp);
 		bd = bh->b_private;
 		if (bd) {
 			gfs2_assert_warn(sdp, bd->bd_bh == bh);
 			gfs2_assert_warn(sdp, list_empty(&bd->bd_list_tr));
-			gfs2_assert_warn(sdp, !bd->bd_ail);
 			bd->bd_bh = NULL;
 			if (!list_empty(&bd->bd_le.le_list))
 				bd = NULL;
@@ -851,6 +746,9 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 
 out:
 	return try_to_free_buffers(page);
+cannot_release:
+	gfs2_log_unlock(sdp);
+	return 0;
 }
 
 const struct address_space_operations gfs2_file_aops = {
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index c1c6672..9e0e9be 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -35,6 +35,7 @@
 #include "super.h"
 #include "sys.h"
 #include "util.h"
+#include "log.h"
 
 #define DO 0
 #define UNDO 1
@@ -887,6 +888,7 @@ error:
 static void gfs2_kill_sb(struct super_block *sb)
 {
 	gfs2_delete_debugfs_file(sb->s_fs_info);
+	gfs2_meta_syncfs(sb->s_fs_info);
 	kill_block_super(sb);
 }
 
-- 
1.5.1.2

[Cluster-devel] [PATCH 23/51] [GFS2] Add a missing gfs2_trans_add_bh()

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

This was missing from the dir_split_leaf() function although in
most cases its not a problem due to other functions having
already previously called gfs2_trans_add_bh. This makes certain
that it is correct.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Wendy Cheng <wcheng@redhat.com>

diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c
index 2beb2f4..08c6dd0 100644
--- a/fs/gfs2/dir.c
+++ b/fs/gfs2/dir.c
@@ -1043,6 +1043,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name)
 
 	error = gfs2_meta_inode_buffer(dip, &dibh);
 	if (!gfs2_assert_withdraw(GFS2_SB(&dip->i_inode), !error)) {
+		gfs2_trans_add_bh(dip->i_gl, dibh, 1);
 		dip->i_di.di_blocks++;
 		gfs2_set_inode_blocks(&dip->i_inode);
 		gfs2_dinode_out(dip, dibh->b_data);
-- 
1.5.1.2

[Cluster-devel] [PATCH 24/51] [GFS2] Add NULL entry to token table

[swhiteho]

From: Benjamin Marzinski <bmarzins@redhat.com>

match_token() was returning garbage data instead of a fail value. This data
happened to match a valid option id for an option that required an argument (in
this case, lockproto=%s) For match_token() to correctly fail if the option
doesn't match any of the tokens, the token table must end with a NULL entry.
This patch adds the NULL entry.

Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/mount.c b/fs/gfs2/mount.c
index 4864659..b941f9f 100644
--- a/fs/gfs2/mount.c
+++ b/fs/gfs2/mount.c
@@ -42,6 +42,7 @@ enum {
 	Opt_nosuiddir,
 	Opt_data_writeback,
 	Opt_data_ordered,
+	Opt_err,
 };
 
 static match_table_t tokens = {
@@ -64,7 +65,8 @@ static match_table_t tokens = {
 	{Opt_suiddir, "suiddir"},
 	{Opt_nosuiddir, "nosuiddir"},
 	{Opt_data_writeback, "data=writeback"},
-	{Opt_data_ordered, "data=ordered"}
+	{Opt_data_ordered, "data=ordered"},
+	{Opt_err, NULL}
 };
 
 /**
@@ -237,6 +239,7 @@ int gfs2_mount_args(struct gfs2_sbd *sdp, char *data_arg, int remount)
 		case Opt_data_ordered:
 			args->ar_data = GFS2_DATA_ORDERED;
 			break;
+		case Opt_err:
 		default:
 			fs_info(sdp, "unknown option: %s\n", o);
 			error = -EINVAL;
-- 
1.5.1.2

[Cluster-devel] [PATCH 25/51] [GFS2] Reduce truncate IO traffic

[swhiteho]

From: Wendy Cheng <wcheng@redhat.com>

Current GFS2 setattr call unconditionally invokes do_shrink even the
requested size and actual file size are equal. This has generated large
amount of extra IOs found during NFS benchmark runs. This patch moves
the relevant logic out of shrink code path. Since setattr is a system
call, the time stamps update is still required.

Signed-off-by: S. Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index cd805a6..9b89904 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -1085,6 +1085,33 @@ static int do_shrink(struct gfs2_inode *ip, u64 size)
 	return error;
 }
 
+static int do_touch(struct gfs2_inode *ip, u64 size)
+{
+	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
+	struct buffer_head *dibh;
+	int error;
+
+	error = gfs2_trans_begin(sdp, RES_DINODE, 0);
+	if (error)
+		return error;
+
+	down_write(&ip->i_rw_mutex);
+
+	error = gfs2_meta_inode_buffer(ip, &dibh);
+	if (error)
+		goto do_touch_out;
+
+	ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME;
+	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
+	gfs2_dinode_out(ip, dibh->b_data);
+	brelse(dibh);
+
+do_touch_out:
+	up_write(&ip->i_rw_mutex);
+	gfs2_trans_end(sdp);
+	return error;
+}
+
 /**
  * gfs2_truncatei - make a file a given size
  * @ip: the inode
@@ -1105,8 +1132,11 @@ int gfs2_truncatei(struct gfs2_inode *ip, u64 size)
 
 	if (size > ip->i_di.di_size)
 		error = do_grow(ip, size);
-	else
+	else if (size < ip->i_di.di_size)
 		error = do_shrink(ip, size);
+	else
+		/* update time stamps */
+		error = do_touch(ip, size);
 
 	return error;
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 26/51] [DLM] Fix lowcomms socket closing

[swhiteho]

From: Patrick Caulfield <pcaulfie@redhat.com>

This patch fixes the slight mess made in lowcomms closing by previous patches
and fixes all sorts of DLM hangs.

Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 9e9d2e8..62a8a6c 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -334,18 +334,8 @@ static void close_connection(struct connection *con, bool and_other)
 		con->rx_page = NULL;
 	}
 
-	/* If we are an 'othercon' then NULL the pointer to us
-	   from the parent and tidy ourself up */
-	if (test_bit(CF_IS_OTHERCON, &con->flags)) {
-		struct connection *parent = __nodeid2con(con->nodeid, 0);
-		parent->othercon = NULL;
-		kmem_cache_free(con_cache, con);
-	}
-	else {
-		/* Parent connections get reused */
-		con->retries = 0;
-		mutex_unlock(&con->sock_mutex);
-	}
+	con->retries = 0;
+	mutex_unlock(&con->sock_mutex);
 }
 
 /* We only send shutdown messages to nodes that are not part of the cluster */
@@ -731,6 +721,8 @@ static int tcp_accept_from_sock(struct connection *con)
 			INIT_WORK(&othercon->swork, process_send_sockets);
 			INIT_WORK(&othercon->rwork, process_recv_sockets);
 			set_bit(CF_IS_OTHERCON, &othercon->flags);
+		}
+		if (!othercon->sock) {
 			newcon->othercon = othercon;
 			othercon->sock = newsock;
 			newsock->sk->sk_user_data = othercon;
-- 
1.5.1.2

[Cluster-devel] [PATCH 27/51] [GFS2] Wendy's dump lockname in hex & fix glock dump

[swhiteho]

From: Abhijith Das <adas@redhat.com>

With this patch, gfs2 glockdump through the debugfs filesystem will only
dump glocks for the specified filesystem instead of all glocks. Also, to
aid debugging, the glock number is dumped in hex instead of decimal.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: S. Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Abhijith Das <adas@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 559937c..3d94918 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -1844,7 +1844,7 @@ static int dump_glock(struct glock_iter *gi, struct gfs2_glock *gl)
 
 	spin_lock(&gl->gl_spin);
 
-	print_dbg(gi, "Glock 0x%p (%u, %llu)\n", gl, gl->gl_name.ln_type,
+	print_dbg(gi, "Glock 0x%p (%u, 0x%llx)\n", gl, gl->gl_name.ln_type,
 		   (unsigned long long)gl->gl_name.ln_number);
 	print_dbg(gi, "  gl_flags =");
 	for (x = 0; x < 32; x++) {
@@ -2024,20 +2024,21 @@ static int gfs2_glock_iter_next(struct glock_iter *gi)
 {
 	struct gfs2_glock *gl;
 
+restart:
 	read_lock(gl_lock_addr(gi->hash));
 	gl = gi->gl;
 	if (gl) {
-		gi->gl = hlist_entry(gl->gl_list.next, struct gfs2_glock,
-				     gl_list);
+		gi->gl = hlist_entry(gl->gl_list.next,
+				     struct gfs2_glock, gl_list);
 		if (gi->gl)
 			gfs2_glock_hold(gi->gl);
 	}
 	read_unlock(gl_lock_addr(gi->hash));
 	if (gl)
 		gfs2_glock_put(gl);
-
-	while(gi->gl == NULL) {
+	if (gl && gi->gl == NULL)
 		gi->hash++;
+	while(gi->gl == NULL) {
 		if (gi->hash >= GFS2_GL_HASH_SIZE)
 			return 1;
 		read_lock(gl_lock_addr(gi->hash));
@@ -2046,7 +2047,12 @@ static int gfs2_glock_iter_next(struct glock_iter *gi)
 		if (gi->gl)
 			gfs2_glock_hold(gi->gl);
 		read_unlock(gl_lock_addr(gi->hash));
+		gi->hash++;
 	}
+
+	if (gi->sdp != gi->gl->gl_sbd)
+		goto restart;
+
 	return 0;
 }
 
@@ -2068,16 +2074,10 @@ static struct glock_iter *gfs2_glock_iter_init(struct gfs2_sbd *sdp)
 	gi->sdp = sdp;
 	gi->hash = 0;
 	gi->seq = NULL;
+	gi->gl = NULL;
 	memset(gi->string, 0, sizeof(gi->string));
 
-	read_lock(gl_lock_addr(gi->hash));
-	gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first,
-			     struct gfs2_glock, gl_list);
-	if (gi->gl)
-		gfs2_glock_hold(gi->gl);
-	read_unlock(gl_lock_addr(gi->hash));
-
-	if (!gi->gl && gfs2_glock_iter_next(gi)) {
+	if (gfs2_glock_iter_next(gi)) {
 		gfs2_glock_iter_free(gi);
 		return NULL;
 	}
-- 
1.5.1.2

[Cluster-devel] [PATCH 28/51] [GFS2] Patch to protect sd_log_num_jdata

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

This is a patch to GFS2 to protect sd_log_num_jdata with the
gfs2_log_lock.  Without this patch, there is a timing window
where you can get hit the following assert from function
gfs2_log_flush():

gfs2_assert_withdraw(sdp,
			sdp->sd_log_num_buf + sdp->sd_log_num_jdata ==
			sdp->sd_log_commited_buf +
			sdp->sd_log_commited_databuf);

I've tested it on my roth cluster and it fixes the problem.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index a0371f8..7ef3356 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -492,11 +492,12 @@ static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 
 	gfs2_trans_add_gl(bd->bd_gl);
 	if (gfs2_is_jdata(ip)) {
-		sdp->sd_log_num_jdata++;
 		gfs2_pin(sdp, bd->bd_bh);
 		tr->tr_num_databuf_new++;
 	}
 	gfs2_log_lock(sdp);
+	if (gfs2_is_jdata(ip))
+		sdp->sd_log_num_jdata++;
 	sdp->sd_log_num_databuf++;
 	list_add(&le->le_list, &sdp->sd_log_le_databuf);
 	gfs2_log_unlock(sdp);
-- 
1.5.1.2

[Cluster-devel] [PATCH 29/51] [GFS2] panic after can't parse mount arguments

[swhiteho]

From: Abhijith Das <adas@redhat.com>

When you try to mount gfs2 with -o garbage, the mount fails and the gfs2
superblock is deallocated and becomes NULL. The vfs comes around later
on and calls gfs2_kill_sb. At this point the hidden gfs2 superblock
pointer (sb->s_fs_info) is NULL and dereferencing it through
gfs2_meta_syncfs causes the panic. (the other function call to
gfs2_delete_debugfs_file() succeeds because this function already checks
for a NULL pointer)

Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 9e0e9be..314c113 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -887,8 +887,10 @@ error:
 
 static void gfs2_kill_sb(struct super_block *sb)
 {
-	gfs2_delete_debugfs_file(sb->s_fs_info);
-	gfs2_meta_syncfs(sb->s_fs_info);
+	if (sb->s_fs_info) {
+		gfs2_delete_debugfs_file(sb->s_fs_info);
+		gfs2_meta_syncfs(sb->s_fs_info);
+	}
 	kill_block_super(sb);
 }
 
-- 
1.5.1.2

[Cluster-devel] [PATCH 30/51] [GFS2] delay glock demote for a minimum hold time

[swhiteho]

From: Benjamin Marzinski <bmarzins@redhat.com>

When a lot of IO, with some distributed mmap IO, is run on a GFS2 filesystem in
a cluster, it will deadlock. The reason is that do_no_page() will repeatedly
call gfs2_sharewrite_nopage(), because each node keeps giving up the glock
too early, and is forced to call unmap_mapping_range(). This bumps the
mapping->truncate_count sequence count, forcing do_no_page() to retry. This
patch institutes a minimum glock hold time a tenth a second.  This insures
that even in heavy contention cases, the node has enough time to get some
useful work done before it gives up the glock.

A second issue is that when gfs2_glock_dq() is called from within a page fault
to demote a lock, and the associated page needs to be written out, it will
try to acqire a lock on it, but it has already been locked at a higher level.
This patch puts makes gfs2_glock_dq() use the work queue as well, to avoid this
issue. This is the same patch as Steve Whitehouse originally proposed to fix
this issue, execpt that gfs2_glock_dq() now grabs a reference to the glock
before it queues up the work on it.

Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 3d94918..931368a 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -27,6 +27,8 @@
 #include <linux/debugfs.h>
 #include <linux/kthread.h>
 #include <linux/freezer.h>
+#include <linux/workqueue.h>
+#include <linux/jiffies.h>
 
 #include "gfs2.h"
 #include "incore.h"
@@ -58,10 +60,13 @@ static int gfs2_dump_lockstate(struct gfs2_sbd *sdp);
 static int dump_glock(struct glock_iter *gi, struct gfs2_glock *gl);
 static void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh);
 static void gfs2_glock_drop_th(struct gfs2_glock *gl);
+static void run_queue(struct gfs2_glock *gl);
+
 static DECLARE_RWSEM(gfs2_umount_flush_sem);
 static struct dentry *gfs2_root;
 static struct task_struct *scand_process;
 static unsigned int scand_secs = 5;
+static struct workqueue_struct *glock_workqueue;
 
 #define GFS2_GL_HASH_SHIFT      15
 #define GFS2_GL_HASH_SIZE       (1 << GFS2_GL_HASH_SHIFT)
@@ -277,6 +282,18 @@ static struct gfs2_glock *gfs2_glock_find(const struct gfs2_sbd *sdp,
 	return gl;
 }
 
+static void glock_work_func(struct work_struct *work)
+{
+	struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_work.work);
+
+	spin_lock(&gl->gl_spin);
+	if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags))
+		set_bit(GLF_DEMOTE, &gl->gl_flags);
+	run_queue(gl);
+	spin_unlock(&gl->gl_spin);
+	gfs2_glock_put(gl);
+}
+
 /**
  * gfs2_glock_get() - Get a glock, or create one if one doesn't exist
  * @sdp: The GFS2 superblock
@@ -316,6 +333,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	gl->gl_name = name;
 	atomic_set(&gl->gl_ref, 1);
 	gl->gl_state = LM_ST_UNLOCKED;
+	gl->gl_demote_state = LM_ST_EXCLUSIVE;
 	gl->gl_hash = hash;
 	gl->gl_owner_pid = 0;
 	gl->gl_ip = 0;
@@ -324,10 +342,12 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	gl->gl_req_bh = NULL;
 	gl->gl_vn = 0;
 	gl->gl_stamp = jiffies;
+	gl->gl_tchange = jiffies;
 	gl->gl_object = NULL;
 	gl->gl_sbd = sdp;
 	gl->gl_aspace = NULL;
 	lops_init_le(&gl->gl_le, &gfs2_glock_lops);
+	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
 
 	/* If this glock protects actual on-disk data or metadata blocks,
 	   create a VFS inode to manage the pages/buffers holding them. */
@@ -441,6 +461,8 @@ static void wait_on_holder(struct gfs2_holder *gh)
 
 static void gfs2_demote_wake(struct gfs2_glock *gl)
 {
+	BUG_ON(!spin_is_locked(&gl->gl_spin));
+	gl->gl_demote_state = LM_ST_EXCLUSIVE;
         clear_bit(GLF_DEMOTE, &gl->gl_flags);
         smp_mb__after_clear_bit();
         wake_up_bit(&gl->gl_flags, GLF_DEMOTE);
@@ -682,10 +704,14 @@ static void gfs2_glmutex_unlock(struct gfs2_glock *gl)
  * practise: LM_ST_SHARED and LM_ST_UNLOCKED
  */
 
-static void handle_callback(struct gfs2_glock *gl, unsigned int state, int remote)
+static void handle_callback(struct gfs2_glock *gl, unsigned int state,
+			    int remote, unsigned long delay)
 {
+	int bit = delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE;
+
 	spin_lock(&gl->gl_spin);
-	if (test_and_set_bit(GLF_DEMOTE, &gl->gl_flags) == 0) {
+	set_bit(bit, &gl->gl_flags);
+	if (gl->gl_demote_state == LM_ST_EXCLUSIVE) {
 		gl->gl_demote_state = state;
 		gl->gl_demote_time = jiffies;
 		if (remote && gl->gl_ops->go_type == LM_TYPE_IOPEN &&
@@ -727,6 +753,7 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state)
 	}
 
 	gl->gl_state = new_state;
+	gl->gl_tchange = jiffies;
 }
 
 /**
@@ -813,7 +840,6 @@ out:
 		gl->gl_req_gh = NULL;
 		gl->gl_req_bh = NULL;
 		clear_bit(GLF_LOCK, &gl->gl_flags);
-		run_queue(gl);
 		spin_unlock(&gl->gl_spin);
 	}
 
@@ -885,7 +911,6 @@ static void drop_bh(struct gfs2_glock *gl, unsigned int ret)
 	gfs2_assert_warn(sdp, !ret);
 
 	state_change(gl, LM_ST_UNLOCKED);
-	gfs2_demote_wake(gl);
 
 	if (glops->go_inval)
 		glops->go_inval(gl, DIO_METADATA);
@@ -898,10 +923,10 @@ static void drop_bh(struct gfs2_glock *gl, unsigned int ret)
 	}
 
 	spin_lock(&gl->gl_spin);
+	gfs2_demote_wake(gl);
 	gl->gl_req_gh = NULL;
 	gl->gl_req_bh = NULL;
 	clear_bit(GLF_LOCK, &gl->gl_flags);
-	run_queue(gl);
 	spin_unlock(&gl->gl_spin);
 
 	gfs2_glock_put(gl);
@@ -1209,9 +1234,10 @@ void gfs2_glock_dq(struct gfs2_holder *gh)
 {
 	struct gfs2_glock *gl = gh->gh_gl;
 	const struct gfs2_glock_operations *glops = gl->gl_ops;
+	unsigned delay = 0;
 
 	if (gh->gh_flags & GL_NOCACHE)
-		handle_callback(gl, LM_ST_UNLOCKED, 0);
+		handle_callback(gl, LM_ST_UNLOCKED, 0, 0);
 
 	gfs2_glmutex_lock(gl);
 
@@ -1229,8 +1255,14 @@ void gfs2_glock_dq(struct gfs2_holder *gh)
 	}
 
 	clear_bit(GLF_LOCK, &gl->gl_flags);
-	run_queue(gl);
 	spin_unlock(&gl->gl_spin);
+
+	gfs2_glock_hold(gl);
+	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
+	    !test_bit(GLF_DEMOTE, &gl->gl_flags))
+		delay = gl->gl_ops->go_min_hold_time;
+	if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
+		gfs2_glock_put(gl);
 }
 
 void gfs2_glock_dq_wait(struct gfs2_holder *gh)
@@ -1457,18 +1489,21 @@ static void blocking_cb(struct gfs2_sbd *sdp, struct lm_lockname *name,
 			unsigned int state)
 {
 	struct gfs2_glock *gl;
+	unsigned long delay = 0;
+	unsigned long holdtime;
+	unsigned long now = jiffies;
 
 	gl = gfs2_glock_find(sdp, name);
 	if (!gl)
 		return;
 
-	handle_callback(gl, state, 1);
-
-	spin_lock(&gl->gl_spin);
-	run_queue(gl);
-	spin_unlock(&gl->gl_spin);
+	holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time;
+	if (time_before(now, holdtime))
+		delay = holdtime - now;
 
-	gfs2_glock_put(gl);
+	handle_callback(gl, state, 1, delay);
+	if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
+		gfs2_glock_put(gl);
 }
 
 /**
@@ -1509,7 +1544,8 @@ void gfs2_glock_cb(void *cb_data, unsigned int type, void *data)
 			return;
 		if (!gfs2_assert_warn(sdp, gl->gl_req_bh))
 			gl->gl_req_bh(gl, async->lc_ret);
-		gfs2_glock_put(gl);
+		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
+			gfs2_glock_put(gl);
 		up_read(&gfs2_umount_flush_sem);
 		return;
 	}
@@ -1602,7 +1638,7 @@ void gfs2_reclaim_glock(struct gfs2_sbd *sdp)
 	if (gfs2_glmutex_trylock(gl)) {
 		if (list_empty(&gl->gl_holders) &&
 		    gl->gl_state != LM_ST_UNLOCKED && demote_ok(gl))
-			handle_callback(gl, LM_ST_UNLOCKED, 0);
+			handle_callback(gl, LM_ST_UNLOCKED, 0, 0);
 		gfs2_glmutex_unlock(gl);
 	}
 
@@ -1702,7 +1738,7 @@ static void clear_glock(struct gfs2_glock *gl)
 	if (gfs2_glmutex_trylock(gl)) {
 		if (list_empty(&gl->gl_holders) &&
 		    gl->gl_state != LM_ST_UNLOCKED)
-			handle_callback(gl, LM_ST_UNLOCKED, 0);
+			handle_callback(gl, LM_ST_UNLOCKED, 0, 0);
 		gfs2_glmutex_unlock(gl);
 	}
 }
@@ -2009,11 +2045,18 @@ int __init gfs2_glock_init(void)
 	if (IS_ERR(scand_process))
 		return PTR_ERR(scand_process);
 
+	glock_workqueue = create_workqueue("glock_workqueue");
+	if (IS_ERR(glock_workqueue)) {
+		kthread_stop(scand_process);
+		return PTR_ERR(glock_workqueue);
+	}
+
 	return 0;
 }
 
 void gfs2_glock_exit(void)
 {
+	destroy_workqueue(glock_workqueue);
 	kthread_stop(scand_process);
 }
 
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 88342e0..7ef6b23 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -454,6 +454,7 @@ const struct gfs2_glock_operations gfs2_inode_glops = {
 	.go_lock = inode_go_lock,
 	.go_unlock = inode_go_unlock,
 	.go_type = LM_TYPE_INODE,
+	.go_min_hold_time = HZ / 10,
 };
 
 const struct gfs2_glock_operations gfs2_rgrp_glops = {
@@ -464,6 +465,7 @@ const struct gfs2_glock_operations gfs2_rgrp_glops = {
 	.go_lock = rgrp_go_lock,
 	.go_unlock = rgrp_go_unlock,
 	.go_type = LM_TYPE_RGRP,
+	.go_min_hold_time = HZ / 10,
 };
 
 const struct gfs2_glock_operations gfs2_trans_glops = {
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 1390b30..23b611a 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -11,6 +11,7 @@
 #define __INCORE_DOT_H__
 
 #include <linux/fs.h>
+#include <linux/workqueue.h>
 
 #define DIO_WAIT	0x00000010
 #define DIO_METADATA	0x00000020
@@ -130,6 +131,7 @@ struct gfs2_glock_operations {
 	int (*go_lock) (struct gfs2_holder *gh);
 	void (*go_unlock) (struct gfs2_holder *gh);
 	const int go_type;
+	const unsigned long go_min_hold_time;
 };
 
 enum {
@@ -161,6 +163,7 @@ enum {
 	GLF_LOCK		= 1,
 	GLF_STICKY		= 2,
 	GLF_DEMOTE		= 3,
+	GLF_PENDING_DEMOTE	= 4,
 	GLF_DIRTY		= 5,
 };
 
@@ -193,6 +196,7 @@ struct gfs2_glock {
 
 	u64 gl_vn;
 	unsigned long gl_stamp;
+	unsigned long gl_tchange;
 	void *gl_object;
 
 	struct list_head gl_reclaim;
@@ -203,6 +207,7 @@ struct gfs2_glock {
 	struct gfs2_log_element gl_le;
 	struct list_head gl_ail_list;
 	atomic_t gl_ail_count;
+	struct delayed_work gl_work;
 };
 
 struct gfs2_alloc {
-- 
1.5.1.2

[Cluster-devel] [PATCH 31/51] [GFS2] fix inode meta data corruption

[swhiteho]

From: Wendy Cheng <wcheng@redhat.com>

Fix a nasty inode meta data corruption issue by keeping the buffer head in
icache array. This buffer needs to stay in memory until journal flush occurs
Otherwise, gfs2_meta_inode_buffer could do a disk read before the inode hits
disk. It ends up with meta data corruptions. The buffer will be released as
part of the existing journal flush logic.

Signed-off-by: S. Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 34f7bcd..013f00b 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -244,6 +244,11 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
 	return 0;
 }
 
+static void gfs2_inode_bh(struct gfs2_inode *ip, struct buffer_head *bh)
+{
+	ip->i_cache[0] = bh;
+}
+
 /**
  * gfs2_inode_refresh - Refresh the incore copy of the dinode
  * @ip: The GFS2 inode
@@ -688,7 +693,7 @@ out:
 static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl,
 			const struct gfs2_inum_host *inum, unsigned int mode,
 			unsigned int uid, unsigned int gid,
-			const u64 *generation, dev_t dev)
+			const u64 *generation, dev_t dev, struct buffer_head **bhp)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
 	struct gfs2_dinode *di;
@@ -743,13 +748,15 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl,
 	di->di_mtime_nsec = cpu_to_be32(tv.tv_nsec);
 	di->di_ctime_nsec = cpu_to_be32(tv.tv_nsec);
 	memset(&di->di_reserved, 0, sizeof(di->di_reserved));
+	
+	set_buffer_uptodate(dibh);
 
-	brelse(dibh);
+	*bhp = dibh;
 }
 
 static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl,
 		       unsigned int mode, const struct gfs2_inum_host *inum,
-		       const u64 *generation, dev_t dev)
+		       const u64 *generation, dev_t dev, struct buffer_head **bhp)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode);
 	unsigned int uid, gid;
@@ -770,7 +777,7 @@ static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl,
 	if (error)
 		goto out_quota;
 
-	init_dinode(dip, gl, inum, mode, uid, gid, generation, dev);
+	init_dinode(dip, gl, inum, mode, uid, gid, generation, dev, bhp);
 	gfs2_quota_change(dip, +1, uid, gid);
 	gfs2_trans_end(sdp);
 
@@ -909,6 +916,7 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name,
 	struct gfs2_inum_host inum = { .no_addr = 0, .no_formal_ino = 0 };
 	int error;
 	u64 generation;
+	struct buffer_head *bh=NULL;
 
 	if (!name->len || name->len > GFS2_FNAMESIZE)
 		return ERR_PTR(-ENAMETOOLONG);
@@ -935,7 +943,7 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name,
 	if (error)
 		goto fail_gunlock;
 
-	error = make_dinode(dip, ghs[1].gh_gl, mode, &inum, &generation, dev);
+	error = make_dinode(dip, ghs[1].gh_gl, mode, &inum, &generation, dev, &bh);
 	if (error)
 		goto fail_gunlock2;
 
@@ -945,6 +953,8 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name,
 	if (IS_ERR(inode))
 		goto fail_gunlock2;
 
+	gfs2_inode_bh(GFS2_I(inode), bh);
+
 	error = gfs2_inode_refresh(GFS2_I(inode));
 	if (error)
 		goto fail_gunlock2;
-- 
1.5.1.2

[Cluster-devel] [PATCH 32/51] [GFS2] Correct lock ordering in unlink

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

This patch corrects the lock ordering in unlink to be the same as
that in the rest of GFS2, i.e. parent -> child -> rgrp.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c
index 5b8b994..2cbe5a3 100644
--- a/fs/gfs2/ops_inode.c
+++ b/fs/gfs2/ops_inode.c
@@ -278,17 +278,25 @@ static int gfs2_unlink(struct inode *dir, struct dentry *dentry)
 	gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2);
 
 
-	error = gfs2_glock_nq_m(3, ghs);
+	error = gfs2_glock_nq(ghs); /* parent */
 	if (error)
-		goto out;
+		goto out_parent;
+
+	error = gfs2_glock_nq(ghs + 1); /* child */
+	if (error)
+		goto out_child;
+
+	error = gfs2_glock_nq(ghs + 2); /* rgrp */
+	if (error)
+		goto out_rgrp;
 
 	error = gfs2_unlink_ok(dip, &dentry->d_name, ip);
 	if (error)
-		goto out_gunlock;
+		goto out_rgrp;
 
 	error = gfs2_trans_begin(sdp, 2*RES_DINODE + RES_LEAF + RES_RG_BIT, 0);
 	if (error)
-		goto out_gunlock;
+		goto out_rgrp;
 
 	error = gfs2_dir_del(dip, &dentry->d_name);
         if (error)
@@ -298,12 +306,15 @@ static int gfs2_unlink(struct inode *dir, struct dentry *dentry)
 
 out_end_trans:
 	gfs2_trans_end(sdp);
-out_gunlock:
-	gfs2_glock_dq_m(3, ghs);
-out:
-	gfs2_holder_uninit(ghs);
-	gfs2_holder_uninit(ghs + 1);
+	gfs2_glock_dq(ghs + 2);
+out_rgrp:
 	gfs2_holder_uninit(ghs + 2);
+	gfs2_glock_dq(ghs + 1);
+out_child:
+	gfs2_holder_uninit(ghs + 1);
+	gfs2_glock_dq(ghs);
+out_parent:
+	gfs2_holder_uninit(ghs);
 	gfs2_glock_dq_uninit(&ri_gh);
 	return error;
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 33/51] [GFS2] Introduce gfs2_remove_from_ail

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

This collects together the operations required to remove a gfs2_bufdata
from the ail lists. Its only called from two places to start with, but
expect to see more of this function in future.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index 7ef6b23..b17346a 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -60,11 +60,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 		blkno = bh->b_blocknr;
 		gfs2_assert_withdraw(sdp, !buffer_busy(bh));
 
-		bd->bd_ail = NULL;
-		list_del(&bd->bd_ail_st_list);
-		list_del(&bd->bd_ail_gl_list);
-		atomic_dec(&gl->gl_ail_count);
-		brelse(bh);
+		gfs2_remove_from_ail(NULL, bd);
 		gfs2_log_unlock(sdp);
 
 		gfs2_trans_add_revoke(sdp, blkno);
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index d0e6b42..d8232ec 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -60,6 +60,26 @@ unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct,
 }
 
 /**
+ * gfs2_remove_from_ail - Remove an entry from the ail lists, updating counters
+ * @mapping: The associated mapping (maybe NULL)
+ * @bd: The gfs2_bufdata to remove
+ *
+ * The log lock _must_ be held when calling this function
+ *
+ */
+
+void gfs2_remove_from_ail(struct address_space *mapping, struct gfs2_bufdata *bd)
+{
+	bd->bd_ail = NULL;
+	list_del(&bd->bd_ail_st_list);
+	list_del(&bd->bd_ail_gl_list);
+	atomic_dec(&bd->bd_gl->gl_ail_count);
+	if (mapping)
+		gfs2_meta_cache_flush(GFS2_I(mapping->host));
+	brelse(bd->bd_bh);
+}
+
+/**
  * gfs2_ail1_start_one - Start I/O on a part of the AIL
  * @sdp: the filesystem
  * @tr: the part of the AIL
@@ -219,21 +239,12 @@ static void gfs2_ail2_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 {
 	struct list_head *head = &ai->ai_ail2_list;
 	struct gfs2_bufdata *bd;
-	struct gfs2_inode *bh_ip;
 
 	while (!list_empty(head)) {
 		bd = list_entry(head->prev, struct gfs2_bufdata,
 				bd_ail_st_list);
 		gfs2_assert(sdp, bd->bd_ail == ai);
-		bd->bd_ail = NULL;
-		list_del(&bd->bd_ail_st_list);
-		list_del(&bd->bd_ail_gl_list);
-		atomic_dec(&bd->bd_gl->gl_ail_count);
-		if (bd->bd_bh->b_page->mapping) {
-			bh_ip = GFS2_I(bd->bd_bh->b_page->mapping->host);
-			gfs2_meta_cache_flush(bh_ip);
-		}
-		brelse(bd->bd_bh);
+		gfs2_remove_from_ail(bd->bd_bh->b_page->mapping, bd);
 	}
 }
 
diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h
index 8e7aa0f..6394235 100644
--- a/fs/gfs2/log.h
+++ b/fs/gfs2/log.h
@@ -58,6 +58,7 @@ struct buffer_head *gfs2_log_fake_buf(struct gfs2_sbd *sdp,
 				      struct buffer_head *real);
 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_remove_from_ail(struct address_space *mapping, struct gfs2_bufdata *bd);
 
 void gfs2_log_shutdown(struct gfs2_sbd *sdp);
 void gfs2_meta_syncfs(struct gfs2_sbd *sdp);
-- 
1.5.1.2

[Cluster-devel] [PATCH 34/51] [GFS2] Don't mark jdata dirty in gfs2_unstuffer_page()

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

Journaled data is marked dirty by gfs2_unpin and should not be marked
dirty here.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index 9b89904..1e56f4d 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -95,7 +95,8 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh,
 	set_buffer_uptodate(bh);
 	if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED || gfs2_is_jdata(ip))
 		gfs2_trans_add_bh(ip->i_gl, bh, 0);
-	mark_buffer_dirty(bh);
+	if (!gfs2_is_jdata(ip))
+		mark_buffer_dirty(bh);
 
 	if (release) {
 		unlock_page(page);
-- 
1.5.1.2

[Cluster-devel] [PATCH 35/51] [GFS2] Move pin/unpin into lops.c, clean up locking

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

gfs2_pin and gfs2_unpin are only used in lops.c, despite being
defined in meta_io.c, so this patch moves them into lops.c and
makes them static. At the same time, its possible to clean up
the locking in the buf and databuf _lo_add() functions so that
we only need to grab the spinlock once. Also we have to move
lock_buffer() around the _lo_add() functions since we can't
do that in gfs2_pin() any more since we hold the spinlock
for the duration of that function.

As a result, the code shrinks by 12 lines and we do far fewer
operations when adding buffers to the log. It also makes the
code somewhat easier to read & understand.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 7ef3356..3ec5871 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -27,7 +27,71 @@
 #include "trans.h"
 #include "util.h"
 
-static void glock_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
+/**
+ * gfs2_pin - Pin a buffer in memory
+ * @sdp: The superblock
+ * @bh: The buffer to be pinned
+ *
+ * The log lock must be held when calling this function
+ */
+static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh)
+{
+	struct gfs2_bufdata *bd;
+
+	gfs2_assert_withdraw(sdp, test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags));
+
+	clear_buffer_dirty(bh);
+	if (test_set_buffer_pinned(bh))
+		gfs2_assert_withdraw(sdp, 0);
+	if (!buffer_uptodate(bh))
+		gfs2_io_error_bh(sdp, bh);
+	bd = bh->b_private;
+	/* If this buffer is in the AIL and it has already been written
+	 * to in-place disk block, remove it from the AIL.
+	 */
+	if (bd->bd_ail)
+		list_move(&bd->bd_ail_st_list, &bd->bd_ail->ai_ail2_list);
+	get_bh(bh);
+}
+
+/**
+ * gfs2_unpin - Unpin a buffer
+ * @sdp: the filesystem the buffer belongs to
+ * @bh: The buffer to unpin
+ * @ai:
+ *
+ */
+
+static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh,
+		       struct gfs2_ail *ai)
+{
+	struct gfs2_bufdata *bd = bh->b_private;
+
+	gfs2_assert_withdraw(sdp, buffer_uptodate(bh));
+
+	if (!buffer_pinned(bh))
+		gfs2_assert_withdraw(sdp, 0);
+
+	lock_buffer(bh);
+	mark_buffer_dirty(bh);
+	clear_buffer_pinned(bh);
+
+	gfs2_log_lock(sdp);
+	if (bd->bd_ail) {
+		list_del(&bd->bd_ail_st_list);
+		brelse(bh);
+	} else {
+		struct gfs2_glock *gl = bd->bd_gl;
+		list_add(&bd->bd_ail_gl_list, &gl->gl_ail_list);
+		atomic_inc(&gl->gl_ail_count);
+	}
+	bd->bd_ail = ai;
+	list_add(&bd->bd_ail_st_list, &ai->ai_ail1_list);
+	gfs2_log_unlock(sdp);
+	unlock_buffer(bh);
+}
+
+static void __glock_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 {
 	struct gfs2_glock *gl;
 	struct gfs2_trans *tr = current->journal_info;
@@ -38,15 +102,19 @@ static void glock_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 	if (gfs2_assert_withdraw(sdp, gfs2_glock_is_held_excl(gl)))
 		return;
 
-	gfs2_log_lock(sdp);
-	if (!list_empty(&le->le_list)){
-		gfs2_log_unlock(sdp);
+	if (!list_empty(&le->le_list))
 		return;
-	}
+
 	gfs2_glock_hold(gl);
 	set_bit(GLF_DIRTY, &gl->gl_flags);
 	sdp->sd_log_num_gl++;
 	list_add(&le->le_list, &sdp->sd_log_le_gl);
+}
+
+static void glock_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
+{
+	gfs2_log_lock(sdp);
+	__glock_lo_add(sdp, le);
 	gfs2_log_unlock(sdp);
 }
 
@@ -71,30 +139,25 @@ static void buf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 	struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le);
 	struct gfs2_trans *tr;
 
+	lock_buffer(bd->bd_bh);
 	gfs2_log_lock(sdp);
-	if (!list_empty(&bd->bd_list_tr)) {
-		gfs2_log_unlock(sdp);
-		return;
-	}
+	if (!list_empty(&bd->bd_list_tr))
+		goto out;
 	tr = current->journal_info;
 	tr->tr_touched = 1;
 	tr->tr_num_buf++;
 	list_add(&bd->bd_list_tr, &tr->tr_list_buf);
-	gfs2_log_unlock(sdp);
-
 	if (!list_empty(&le->le_list))
-		return;
-
-	gfs2_trans_add_gl(bd->bd_gl);
-
+		goto out;
+	__glock_lo_add(sdp, &bd->bd_gl->gl_le);
 	gfs2_meta_check(sdp, bd->bd_bh);
 	gfs2_pin(sdp, bd->bd_bh);
-	gfs2_log_lock(sdp);
 	sdp->sd_log_num_buf++;
 	list_add(&le->le_list, &sdp->sd_log_le_buf);
-	gfs2_log_unlock(sdp);
-
 	tr->tr_num_buf_new++;
+out:
+	gfs2_log_unlock(sdp);
+	unlock_buffer(bd->bd_bh);
 }
 
 static void buf_lo_incore_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
@@ -476,31 +539,29 @@ static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 	struct address_space *mapping = bd->bd_bh->b_page->mapping;
 	struct gfs2_inode *ip = GFS2_I(mapping->host);
 
+	lock_buffer(bd->bd_bh);
 	gfs2_log_lock(sdp);
-	if (!list_empty(&bd->bd_list_tr)) {
-		gfs2_log_unlock(sdp);
-		return;
-	}
+	if (!list_empty(&bd->bd_list_tr))
+		goto out;
 	tr->tr_touched = 1;
 	if (gfs2_is_jdata(ip)) {
 		tr->tr_num_buf++;
 		list_add(&bd->bd_list_tr, &tr->tr_list_buf);
 	}
-	gfs2_log_unlock(sdp);
 	if (!list_empty(&le->le_list))
-		return;
+		goto out;
 
-	gfs2_trans_add_gl(bd->bd_gl);
+	__glock_lo_add(sdp, &bd->bd_gl->gl_le);
 	if (gfs2_is_jdata(ip)) {
 		gfs2_pin(sdp, bd->bd_bh);
 		tr->tr_num_databuf_new++;
-	}
-	gfs2_log_lock(sdp);
-	if (gfs2_is_jdata(ip))
 		sdp->sd_log_num_jdata++;
+	}
 	sdp->sd_log_num_databuf++;
 	list_add(&le->le_list, &sdp->sd_log_le_databuf);
+out:
 	gfs2_log_unlock(sdp);
+	unlock_buffer(bd->bd_bh);
 }
 
 static int gfs2_check_magic(struct buffer_head *bh)
diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index 8da343b..d762e4f 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -298,76 +298,6 @@ void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh,
 }
 
 /**
- * gfs2_pin - Pin a buffer in memory
- * @sdp: the filesystem the buffer belongs to
- * @bh: The buffer to be pinned
- *
- */
-
-void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh)
-{
-	struct gfs2_bufdata *bd = bh->b_private;
-
-	gfs2_assert_withdraw(sdp, test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags));
-
-	if (test_set_buffer_pinned(bh))
-		gfs2_assert_withdraw(sdp, 0);
-
-	wait_on_buffer(bh);
-
-	/* If this buffer is in the AIL and it has already been written
-	   to in-place disk block, remove it from the AIL. */
-
-	gfs2_log_lock(sdp);
-	if (bd->bd_ail && !buffer_in_io(bh))
-		list_move(&bd->bd_ail_st_list, &bd->bd_ail->ai_ail2_list);
-	gfs2_log_unlock(sdp);
-
-	clear_buffer_dirty(bh);
-	wait_on_buffer(bh);
-
-	if (!buffer_uptodate(bh))
-		gfs2_io_error_bh(sdp, bh);
-
-	get_bh(bh);
-}
-
-/**
- * gfs2_unpin - Unpin a buffer
- * @sdp: the filesystem the buffer belongs to
- * @bh: The buffer to unpin
- * @ai:
- *
- */
-
-void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh,
-	        struct gfs2_ail *ai)
-{
-	struct gfs2_bufdata *bd = bh->b_private;
-
-	gfs2_assert_withdraw(sdp, buffer_uptodate(bh));
-
-	if (!buffer_pinned(bh))
-		gfs2_assert_withdraw(sdp, 0);
-
-	mark_buffer_dirty(bh);
-	clear_buffer_pinned(bh);
-
-	gfs2_log_lock(sdp);
-	if (bd->bd_ail) {
-		list_del(&bd->bd_ail_st_list);
-		brelse(bh);
-	} else {
-		struct gfs2_glock *gl = bd->bd_gl;
-		list_add(&bd->bd_ail_gl_list, &gl->gl_ail_list);
-		atomic_inc(&gl->gl_ail_count);
-	}
-	bd->bd_ail = ai;
-	list_add(&bd->bd_ail_st_list, &ai->ai_ail1_list);
-	gfs2_log_unlock(sdp);
-}
-
-/**
  * gfs2_meta_wipe - make inode's buffers so they aren't dirty/pinned anymore
  * @ip: the inode who owns the buffers
  * @bstart: the first buffer in the run
diff --git a/fs/gfs2/meta_io.h b/fs/gfs2/meta_io.h
index 527bf19..fd67f07 100644
--- a/fs/gfs2/meta_io.h
+++ b/fs/gfs2/meta_io.h
@@ -50,9 +50,6 @@ int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh);
 
 void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh,
 			 int meta);
-void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh);
-void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh,
-		struct gfs2_ail *ai);
 
 void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen);
 
-- 
1.5.1.2

[Cluster-devel] [PATCH 36/51] [GFS2] Clean up ordered write code

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

The following patch removes the ordered write processing from
databuf_lo_before_commit() and moves it to log.c. This has the effect of
greatly simplyfying databuf_lo_before_commit() and well as potentially
making the ordered write code more efficient.

As a side effect of this, its now possible to remove ordered buffers
from the ordered buffer list at any time, so we now make use of this in
invalidatepage and releasepage to ensure timely release of these
buffers.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 23b611a..388dc1b 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -612,13 +612,13 @@ struct gfs2_sbd {
 	unsigned int sd_log_num_revoke;
 	unsigned int sd_log_num_rg;
 	unsigned int sd_log_num_databuf;
-	unsigned int sd_log_num_jdata;
 
 	struct list_head sd_log_le_gl;
 	struct list_head sd_log_le_buf;
 	struct list_head sd_log_le_revoke;
 	struct list_head sd_log_le_rg;
 	struct list_head sd_log_le_databuf;
+	struct list_head sd_log_le_ordered;
 
 	unsigned int sd_log_blks_free;
 	struct mutex sd_log_reserve_mutex;
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index d8232ec..20fa528 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -620,6 +620,57 @@ static void log_flush_commit(struct gfs2_sbd *sdp)
 	}
 }
 
+static void gfs2_ordered_write(struct gfs2_sbd *sdp)
+{
+	struct gfs2_bufdata *bd;
+	struct buffer_head *bh;
+	LIST_HEAD(written);
+
+	gfs2_log_lock(sdp);
+	while (!list_empty(&sdp->sd_log_le_ordered)) {
+		bd = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_bufdata, bd_le.le_list);
+		list_move(&bd->bd_le.le_list, &written);
+		bh = bd->bd_bh;
+		if (!buffer_dirty(bh))
+			continue;
+		get_bh(bh);
+		gfs2_log_unlock(sdp);
+		lock_buffer(bh);
+		if (test_clear_buffer_dirty(bh)) {
+			bh->b_end_io = end_buffer_write_sync;
+			submit_bh(WRITE, bh);
+		} else {
+			unlock_buffer(bh);
+			brelse(bh);
+		}
+		gfs2_log_lock(sdp);
+	}
+	list_splice(&written, &sdp->sd_log_le_ordered);
+	gfs2_log_unlock(sdp);
+}
+
+static void gfs2_ordered_wait(struct gfs2_sbd *sdp)
+{
+	struct gfs2_bufdata *bd;
+	struct buffer_head *bh;
+
+	gfs2_log_lock(sdp);
+	while (!list_empty(&sdp->sd_log_le_ordered)) {
+		bd = list_entry(sdp->sd_log_le_ordered.prev, struct gfs2_bufdata, bd_le.le_list);
+		bh = bd->bd_bh;
+		if (buffer_locked(bh)) {
+			get_bh(bh);
+			gfs2_log_unlock(sdp);
+			wait_on_buffer(bh);
+			brelse(bh);
+			gfs2_log_lock(sdp);
+			continue;
+		}
+		list_del_init(&bd->bd_le.le_list);
+	}
+	gfs2_log_unlock(sdp);
+}
+
 /**
  * gfs2_log_flush - flush incore transaction(s)
  * @sdp: the filesystem
@@ -648,7 +699,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl)
 	INIT_LIST_HEAD(&ai->ai_ail2_list);
 
 	gfs2_assert_withdraw(sdp,
-			     sdp->sd_log_num_buf + sdp->sd_log_num_jdata ==
+			     sdp->sd_log_num_buf + sdp->sd_log_num_databuf ==
 			     sdp->sd_log_commited_buf +
 			     sdp->sd_log_commited_databuf);
 	gfs2_assert_withdraw(sdp,
@@ -658,7 +709,10 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl)
 	sdp->sd_log_flush_wrapped = 0;
 	ai->ai_first = sdp->sd_log_flush_head;
 
+	gfs2_ordered_write(sdp);
 	lops_before_commit(sdp);
+	gfs2_ordered_wait(sdp);
+
 	if (!list_empty(&sdp->sd_log_flush_list))
 		log_flush_commit(sdp);
 	else if (sdp->sd_log_tail != current_tail(sdp) && !sdp->sd_log_idle){
@@ -751,7 +805,6 @@ void gfs2_log_shutdown(struct gfs2_sbd *sdp)
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_blks_reserved);
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_gl);
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_buf);
-	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_jdata);
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke);
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_rg);
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_databuf);
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 3ec5871..7e2d4e6 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -555,10 +555,11 @@ static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 	if (gfs2_is_jdata(ip)) {
 		gfs2_pin(sdp, bd->bd_bh);
 		tr->tr_num_databuf_new++;
-		sdp->sd_log_num_jdata++;
+		sdp->sd_log_num_databuf++;
+		list_add(&le->le_list, &sdp->sd_log_le_databuf);
+	} else {
+		list_add(&le->le_list, &sdp->sd_log_le_ordered);
 	}
-	sdp->sd_log_num_databuf++;
-	list_add(&le->le_list, &sdp->sd_log_le_databuf);
 out:
 	gfs2_log_unlock(sdp);
 	unlock_buffer(bd->bd_bh);
@@ -583,114 +584,59 @@ static int gfs2_check_magic(struct buffer_head *bh)
 /**
  * databuf_lo_before_commit - Scan the data buffers, writing as we go
  *
- * Here we scan through the lists of buffers and make the assumption
- * that any buffer thats been pinned is being journaled, and that
- * any unpinned buffer is an ordered write data buffer and therefore
- * will be written back rather than journaled.
  */
+
 static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 {
-	LIST_HEAD(started);
-	struct gfs2_bufdata *bd1 = NULL, *bd2, *bdt;
+	struct gfs2_bufdata *bd1 = NULL, *bd2;
 	struct buffer_head *bh = NULL,*bh1 = NULL;
 	struct gfs2_log_descriptor *ld;
 	unsigned int limit;
-	unsigned int total_dbuf;
-	unsigned int total_jdata;
+	unsigned int total;
 	unsigned int num, n;
 	__be64 *ptr = NULL;
+	int magic;
+
 
 	limit = databuf_limit(sdp);
 
-	/*
-	 * Start writing ordered buffers, write journaled buffers
-	 * into the log along with a header
-	 */
 	gfs2_log_lock(sdp);
-	total_dbuf = sdp->sd_log_num_databuf;
-	total_jdata = sdp->sd_log_num_jdata;
+	total = sdp->sd_log_num_databuf;
 	bd2 = bd1 = list_prepare_entry(bd1, &sdp->sd_log_le_databuf,
 				       bd_le.le_list);
-	while(total_dbuf) {
-		num = total_jdata;
+	while(total) {
+		num = total;
 		if (num > limit)
 			num = limit;
+
+		gfs2_log_unlock(sdp);
+		bh = gfs2_log_get_buf(sdp);
+		gfs2_log_lock(sdp);
+
+		ld = (struct gfs2_log_descriptor *)bh->b_data;
+		ptr = (__be64 *)(bh->b_data + DATABUF_OFFSET);
+		ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
+		ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
+		ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
+		ld->ld_type = cpu_to_be32(GFS2_LOG_DESC_JDATA);
+		ld->ld_length = cpu_to_be32(num + 1);
+		ld->ld_data1 = cpu_to_be32(num);
+		ld->ld_data2 = cpu_to_be32(0);
+		memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
+
 		n = 0;
-		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
-			 */
+		list_for_each_entry_continue(bd1, &sdp->sd_log_le_databuf,
+					     bd_le.le_list) {
 			bh1 = bd1->bd_bh;
 
-			/* An ordered write buffer */
-			if (bh1 && !buffer_pinned(bh1)) {
-				list_move(&bd1->bd_le.le_list, &started);
-				if (bd1 == bd2) {
-					bd2 = NULL;
-					bd2 = list_prepare_entry(bd2,
-							&sdp->sd_log_le_databuf,
-							bd_le.le_list);
-				}
-				total_dbuf--;
-				if (bh1) {
-					if (buffer_dirty(bh1)) {
-						get_bh(bh1);
-
-						gfs2_log_unlock(sdp);
-
-						ll_rw_block(SWRITE, 1, &bh1);
-						brelse(bh1);
-
-						gfs2_log_lock(sdp);
-					}
-					continue;
-				}
-				continue;
-			} else if (bh1) { /* A journaled buffer */
-				int magic;
-				gfs2_log_unlock(sdp);
-				if (!bh) {
-					bh = gfs2_log_get_buf(sdp);
-					ld = (struct gfs2_log_descriptor *)
-					     bh->b_data;
-					ptr = (__be64 *)(bh->b_data +
-							 DATABUF_OFFSET);
-					ld->ld_header.mh_magic =
-						cpu_to_be32(GFS2_MAGIC);
-					ld->ld_header.mh_type =
-						cpu_to_be32(GFS2_METATYPE_LD);
-					ld->ld_header.mh_format =
-						cpu_to_be32(GFS2_FORMAT_LD);
-					ld->ld_type =
-						cpu_to_be32(GFS2_LOG_DESC_JDATA);
-					ld->ld_length = cpu_to_be32(num + 1);
-					ld->ld_data1 = cpu_to_be32(num);
-					ld->ld_data2 = cpu_to_be32(0);
-					memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
-				}
-				magic = gfs2_check_magic(bh1);
-				*ptr++ = cpu_to_be64(bh1->b_blocknr);
-				*ptr++ = cpu_to_be64((__u64)magic);
-				clear_buffer_escaped(bh1);
-				if (unlikely(magic != 0))
-					set_buffer_escaped(bh1);
-				gfs2_log_lock(sdp);
-				if (++n >= num)
-					break;
-			} else if (!bh1) {
-				total_dbuf--;
-				sdp->sd_log_num_databuf--;
-				list_del_init(&bd1->bd_le.le_list);
-				if (bd1 == bd2) {
-					bd2 = NULL;
-					bd2 = list_prepare_entry(bd2,
-						&sdp->sd_log_le_databuf,
-						bd_le.le_list);
-                                }
-				kmem_cache_free(gfs2_bufdata_cachep, bd1);
-			}
+			magic = gfs2_check_magic(bh1);
+			*ptr++ = cpu_to_be64(bh1->b_blocknr);
+			*ptr++ = cpu_to_be64((__u64)magic);
+			clear_buffer_escaped(bh1);
+			if (unlikely(magic != 0))
+				set_buffer_escaped(bh1);
+			if (++n >= num)
+				break;
 		}
 		gfs2_log_unlock(sdp);
 		if (bh) {
@@ -727,34 +673,10 @@ static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
 				break;
 		}
 		bh = NULL;
-		BUG_ON(total_dbuf < num);
-		total_dbuf -= num;
-		total_jdata -= num;
+		BUG_ON(total < num);
+		total -= num;
 	}
 	gfs2_log_unlock(sdp);
-
-	/* Wait on all ordered buffers */
-	while (!list_empty(&started)) {
-		gfs2_log_lock(sdp);
-		bd1 = list_entry(started.next, struct gfs2_bufdata,
-				 bd_le.le_list);
-		list_del_init(&bd1->bd_le.le_list);
-		sdp->sd_log_num_databuf--;
-		bh = bd1->bd_bh;
-		if (bh) {
-			bh->b_private = NULL;
-			get_bh(bh);
-			gfs2_log_unlock(sdp);
-			wait_on_buffer(bh);
-			brelse(bh);
-		} else
-			gfs2_log_unlock(sdp);
-
-		kmem_cache_free(gfs2_bufdata_cachep, bd1);
-	}
-
-	/* We've removed all the ordered write bufs here, so only jdata left */
-	gfs2_assert_warn(sdp, sdp->sd_log_num_databuf == sdp->sd_log_num_jdata);
 }
 
 static int databuf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start,
@@ -838,11 +760,9 @@ static void databuf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 		bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list);
 		list_del_init(&bd->bd_le.le_list);
 		sdp->sd_log_num_databuf--;
-		sdp->sd_log_num_jdata--;
 		gfs2_unpin(sdp, bd->bd_bh, ai);
 	}
 	gfs2_assert_warn(sdp, !sdp->sd_log_num_databuf);
-	gfs2_assert_warn(sdp, !sdp->sd_log_num_jdata);
 }
 
 
diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c
index 8407d1d..dd1ea49 100644
--- a/fs/gfs2/ops_address.c
+++ b/fs/gfs2/ops_address.c
@@ -616,13 +616,50 @@ static sector_t gfs2_bmap(struct address_space *mapping, sector_t lblock)
 	return dblock;
 }
 
+static void gfs2_discard(struct gfs2_sbd *sdp, struct buffer_head *bh)
+{
+	struct gfs2_bufdata *bd;
+
+	lock_buffer(bh);
+	gfs2_log_lock(sdp);
+	clear_buffer_dirty(bh);
+	bd = bh->b_private;
+	if (bd) {
+		if (!list_empty(&bd->bd_le.le_list)) {
+			if (!buffer_pinned(bh))
+				list_del_init(&bd->bd_le.le_list);
+		}
+	}
+	bh->b_bdev = NULL;
+	clear_buffer_mapped(bh);
+	clear_buffer_req(bh);
+	clear_buffer_new(bh);
+	gfs2_log_unlock(sdp);
+	unlock_buffer(bh);
+}
+
 static void gfs2_invalidatepage(struct page *page, unsigned long offset)
 {
+	struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host);
+	struct buffer_head *bh, *head;
+	unsigned long pos = 0;
+
 	BUG_ON(!PageLocked(page));
 	if (offset == 0)
 		ClearPageChecked(page);
+	if (!page_has_buffers(page))
+		goto out;
 
-	block_invalidatepage(page, offset);
+	bh = head = page_buffers(page);
+	do {
+		if (offset <= pos)
+			gfs2_discard(sdp, bh);
+		pos += bh->b_size;
+		bh = bh->b_this_page;
+	} while (bh != head);
+out:
+	if (offset == 0)
+		try_to_release_page(page, 0);
 }
 
 /**
@@ -732,9 +769,14 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 		if (bd) {
 			gfs2_assert_warn(sdp, bd->bd_bh == bh);
 			gfs2_assert_warn(sdp, list_empty(&bd->bd_list_tr));
-			bd->bd_bh = NULL;
-			if (!list_empty(&bd->bd_le.le_list))
-				bd = NULL;
+			if (!list_empty(&bd->bd_le.le_list)) {
+				if (!buffer_pinned(bh))
+					list_del_init(&bd->bd_le.le_list);
+				else
+					bd = NULL;
+			}
+			if (bd)
+				bd->bd_bh = NULL;
 			bh->b_private = NULL;
 		}
 		gfs2_log_unlock(sdp);
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 314c113..35f3dfa 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -82,6 +82,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb)
 	INIT_LIST_HEAD(&sdp->sd_log_le_revoke);
 	INIT_LIST_HEAD(&sdp->sd_log_le_rg);
 	INIT_LIST_HEAD(&sdp->sd_log_le_databuf);
+	INIT_LIST_HEAD(&sdp->sd_log_le_ordered);
 
 	mutex_init(&sdp->sd_log_reserve_mutex);
 	INIT_LIST_HEAD(&sdp->sd_ail1_list);
-- 
1.5.1.2

[Cluster-devel] [PATCH 37/51] [GFS2] Fix ordering of dirty/journal for ordered buffer unstuffing

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index 1e56f4d..93fa427 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -93,10 +93,10 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh,
 		map_bh(bh, inode->i_sb, block);
 
 	set_buffer_uptodate(bh);
-	if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED || gfs2_is_jdata(ip))
-		gfs2_trans_add_bh(ip->i_gl, bh, 0);
 	if (!gfs2_is_jdata(ip))
 		mark_buffer_dirty(bh);
+	if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED || gfs2_is_jdata(ip))
+		gfs2_trans_add_bh(ip->i_gl, bh, 0);
 
 	if (release) {
 		unlock_page(page);
-- 
1.5.1.2

[Cluster-devel] [PATCH 38/51] [GFS2] Replace revoke structure with bufdata structure

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

Both the revoke structure and the bufdata structure are quite similar.
They are basically small tags which are put on lists. In addition to
which the revoke structure is always allocated when there is a bufdata
structure which is (or can be) freed. As such it should be possible to
reduce the number of frees and allocations by using the same structure
for both purposes.

This patch is the first step along that path. It replaces existing uses
of the revoke structure with the bufdata structure.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 388dc1b..8aa5780 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -114,7 +114,13 @@ struct gfs2_bufdata {
 	struct buffer_head *bd_bh;
 	struct gfs2_glock *bd_gl;
 
-	struct list_head bd_list_tr;
+	union {
+		struct list_head list_tr;
+		u64 blkno;
+	} u;
+#define bd_list_tr u.list_tr
+#define bd_blkno u.blkno
+
 	struct gfs2_log_element bd_le;
 
 	struct gfs2_ail *bd_ail;
@@ -298,11 +304,6 @@ struct gfs2_file {
 	struct gfs2_holder f_fl_gh;
 };
 
-struct gfs2_revoke {
-	struct gfs2_log_element rv_le;
-	u64 rv_blkno;
-};
-
 struct gfs2_revoke_replay {
 	struct list_head rr_list;
 	u64 rr_blkno;
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 7e2d4e6..cf6fe36 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -357,7 +357,7 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 	struct buffer_head *bh;
 	unsigned int offset;
 	struct list_head *head = &sdp->sd_log_le_revoke;
-	struct gfs2_revoke *rv;
+	struct gfs2_bufdata *bd;
 
 	if (!sdp->sd_log_num_revoke)
 		return;
@@ -376,8 +376,8 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 	offset = sizeof(struct gfs2_log_descriptor);
 
 	while (!list_empty(head)) {
-		rv = list_entry(head->next, struct gfs2_revoke, rv_le.le_list);
-		list_del_init(&rv->rv_le.le_list);
+		bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list);
+		list_del_init(&bd->bd_le.le_list);
 		sdp->sd_log_num_revoke--;
 
 		if (offset + sizeof(u64) > sdp->sd_sb.sb_bsize) {
@@ -392,8 +392,8 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 			offset = sizeof(struct gfs2_meta_header);
 		}
 
-		*(__be64 *)(bh->b_data + offset) = cpu_to_be64(rv->rv_blkno);
-		kfree(rv);
+		*(__be64 *)(bh->b_data + offset) = cpu_to_be64(bd->bd_blkno);
+		kfree(bd);
 
 		offset += sizeof(u64);
 	}
diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index f8dabf8..eadf96e 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -144,23 +144,23 @@ void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta)
 
 void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, u64 blkno)
 {
-	struct gfs2_revoke *rv = kmalloc(sizeof(struct gfs2_revoke),
-					 GFP_NOFS | __GFP_NOFAIL);
-	lops_init_le(&rv->rv_le, &gfs2_revoke_lops);
-	rv->rv_blkno = blkno;
-	lops_add(sdp, &rv->rv_le);
+	struct gfs2_bufdata *bd = kmalloc(sizeof(struct gfs2_bufdata),
+					  GFP_NOFS | __GFP_NOFAIL);
+	lops_init_le(&bd->bd_le, &gfs2_revoke_lops);
+	bd->bd_blkno = blkno;
+	lops_add(sdp, &bd->bd_le);
 }
 
 void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno)
 {
-	struct gfs2_revoke *rv;
+	struct gfs2_bufdata *bd;
 	int found = 0;
 
 	gfs2_log_lock(sdp);
 
-	list_for_each_entry(rv, &sdp->sd_log_le_revoke, rv_le.le_list) {
-		if (rv->rv_blkno == blkno) {
-			list_del(&rv->rv_le.le_list);
+	list_for_each_entry(bd, &sdp->sd_log_le_revoke, bd_le.le_list) {
+		if (bd->bd_blkno == blkno) {
+			list_del(&bd->bd_le.le_list);
 			gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke);
 			sdp->sd_log_num_revoke--;
 			found = 1;
@@ -172,7 +172,7 @@ void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno)
 
 	if (found) {
 		struct gfs2_trans *tr = current->journal_info;
-		kfree(rv);
+		kfree(bd);
 		tr->tr_num_revoke_rm++;
 	}
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 39/51] [GFS2] Use slab operations for all gfs2_bufdata allocations

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

The old revoke structure was allocated using kalloc/kfree but
there is a slab cache for gfs2_bufdata, so we should use that
now that the structures have been converted.

This is part two of the patch series to merge the revoke
and gfs2_bufdata structures.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index cf6fe36..4cbef4c 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -393,7 +393,7 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 		}
 
 		*(__be64 *)(bh->b_data + offset) = cpu_to_be64(bd->bd_blkno);
-		kfree(bd);
+		kmem_cache_free(gfs2_bufdata_cachep, bd);
 
 		offset += sizeof(u64);
 	}
diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index eadf96e..01cc27f 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -144,7 +144,7 @@ void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta)
 
 void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, u64 blkno)
 {
-	struct gfs2_bufdata *bd = kmalloc(sizeof(struct gfs2_bufdata),
+	struct gfs2_bufdata *bd = kmem_cache_alloc(gfs2_bufdata_cachep,
 					  GFP_NOFS | __GFP_NOFAIL);
 	lops_init_le(&bd->bd_le, &gfs2_revoke_lops);
 	bd->bd_blkno = blkno;
@@ -172,7 +172,7 @@ void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno)
 
 	if (found) {
 		struct gfs2_trans *tr = current->journal_info;
-		kfree(bd);
+		kmem_cache_free(gfs2_bufdata_cachep, bd);
 		tr->tr_num_revoke_rm++;
 	}
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 40/51] [GFS2] Clean up gfs2_trans_add_revoke()

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

The following alters gfs2_trans_add_revoke() to take a struct
gfs2_bufdata as an argument. This eliminates the memory allocation which
was previously required by making use of the already existing struct
gfs2_bufdata. It makes some sanity checks to ensure that the
gfs2_bufdata has been removed from all the lists before its recycled as
a revoke structure. This saves one memory allocation and one free per
revoke structure.

Also as a result, and to simplify the locking, since there is no longer
any blocking code in gfs2_trans_add_revoke() we must hold the log lock
whenever this function is called. This reduces the amount of times we
take and unlock the log lock.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index b17346a..4670dcb 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -41,7 +41,6 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 	struct list_head *head = &gl->gl_ail_list;
 	struct gfs2_bufdata *bd;
 	struct buffer_head *bh;
-	u64 blkno;
 	int error;
 
 	blocks = atomic_read(&gl->gl_ail_count);
@@ -57,15 +56,12 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
 		bd = list_entry(head->next, struct gfs2_bufdata,
 				bd_ail_gl_list);
 		bh = bd->bd_bh;
-		blkno = bh->b_blocknr;
-		gfs2_assert_withdraw(sdp, !buffer_busy(bh));
-
 		gfs2_remove_from_ail(NULL, bd);
-		gfs2_log_unlock(sdp);
-
-		gfs2_trans_add_revoke(sdp, blkno);
-
-		gfs2_log_lock(sdp);
+		bd->bd_bh = NULL;
+		bh->b_private = NULL;
+		bd->bd_blkno = bh->b_blocknr;
+		gfs2_assert_withdraw(sdp, !buffer_busy(bh));
+		gfs2_trans_add_revoke(sdp, bd);
 	}
 	gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));
 	gfs2_log_unlock(sdp);
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 20fa528..4d04e6f 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -71,8 +71,8 @@ unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct,
 void gfs2_remove_from_ail(struct address_space *mapping, struct gfs2_bufdata *bd)
 {
 	bd->bd_ail = NULL;
-	list_del(&bd->bd_ail_st_list);
-	list_del(&bd->bd_ail_gl_list);
+	list_del_init(&bd->bd_ail_st_list);
+	list_del_init(&bd->bd_ail_gl_list);
 	atomic_dec(&bd->bd_gl->gl_ail_count);
 	if (mapping)
 		gfs2_meta_cache_flush(GFS2_I(mapping->host));
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 4cbef4c..342c10e 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -343,11 +343,8 @@ static void revoke_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 	tr = current->journal_info;
 	tr->tr_touched = 1;
 	tr->tr_num_revoke++;
-
-	gfs2_log_lock(sdp);
 	sdp->sd_log_num_revoke++;
 	list_add(&le->le_list, &sdp->sd_log_le_revoke);
-	gfs2_log_unlock(sdp);
 }
 
 static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index d762e4f..19097bc 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -313,42 +313,31 @@ void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
 	while (blen) {
 		bh = getbuf(ip->i_gl, bstart, NO_CREATE);
 		if (bh) {
-			struct gfs2_bufdata *bd = bh->b_private;
+			struct gfs2_bufdata *bd;
 
+			lock_buffer(bh);
+			gfs2_log_lock(sdp);
+			bd = bh->b_private;
 			if (test_clear_buffer_pinned(bh)) {
 				struct gfs2_trans *tr = current->journal_info;
-				struct gfs2_inode *bh_ip =
-					GFS2_I(bh->b_page->mapping->host);
-
-				gfs2_log_lock(sdp);
 				list_del_init(&bd->bd_le.le_list);
 				gfs2_assert_warn(sdp, sdp->sd_log_num_buf);
 				sdp->sd_log_num_buf--;
-				gfs2_log_unlock(sdp);
-				if (bh_ip->i_inode.i_private != NULL)
-					tr->tr_num_databuf_rm++;
-				else
-					tr->tr_num_buf_rm++;
+				tr->tr_num_buf_rm++;
 				brelse(bh);
 			}
 			if (bd) {
-				gfs2_log_lock(sdp);
 				if (bd->bd_ail) {
-					u64 blkno = bh->b_blocknr;
-					bd->bd_ail = NULL;
-					list_del(&bd->bd_ail_st_list);
-					list_del(&bd->bd_ail_gl_list);
-					atomic_dec(&bd->bd_gl->gl_ail_count);
-					brelse(bh);
-					gfs2_log_unlock(sdp);
-					gfs2_trans_add_revoke(sdp, blkno);
-				} else
-					gfs2_log_unlock(sdp);
+					gfs2_remove_from_ail(NULL, bd);
+					bh->b_private = NULL;
+					bd->bd_bh = NULL;
+					bd->bd_blkno = bh->b_blocknr;
+					gfs2_trans_add_revoke(sdp, bd);
+				}
 			}
-
-			lock_buffer(bh);
 			clear_buffer_dirty(bh);
 			clear_buffer_uptodate(bh);
+			gfs2_log_unlock(sdp);
 			unlock_buffer(bh);
 
 			brelse(bh);
diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c
index 01cc27f..717983e 100644
--- a/fs/gfs2/trans.c
+++ b/fs/gfs2/trans.c
@@ -142,12 +142,12 @@ void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta)
 	lops_add(sdp, &bd->bd_le);
 }
 
-void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, u64 blkno)
+void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd)
 {
-	struct gfs2_bufdata *bd = kmem_cache_alloc(gfs2_bufdata_cachep,
-					  GFP_NOFS | __GFP_NOFAIL);
+	BUG_ON(!list_empty(&bd->bd_le.le_list));
+	BUG_ON(!list_empty(&bd->bd_ail_st_list));
+	BUG_ON(!list_empty(&bd->bd_ail_gl_list));
 	lops_init_le(&bd->bd_le, &gfs2_revoke_lops);
-	bd->bd_blkno = blkno;
 	lops_add(sdp, &bd->bd_le);
 }
 
@@ -160,7 +160,7 @@ void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno)
 
 	list_for_each_entry(bd, &sdp->sd_log_le_revoke, bd_le.le_list) {
 		if (bd->bd_blkno == blkno) {
-			list_del(&bd->bd_le.le_list);
+			list_del_init(&bd->bd_le.le_list);
 			gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke);
 			sdp->sd_log_num_revoke--;
 			found = 1;
diff --git a/fs/gfs2/trans.h b/fs/gfs2/trans.h
index 23d4cbe..043d5f4 100644
--- a/fs/gfs2/trans.h
+++ b/fs/gfs2/trans.h
@@ -32,7 +32,7 @@ void gfs2_trans_end(struct gfs2_sbd *sdp);
 
 void gfs2_trans_add_gl(struct gfs2_glock *gl);
 void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta);
-void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, u64 blkno);
+void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd);
 void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno);
 void gfs2_trans_add_rg(struct gfs2_rgrpd *rgd);
 
-- 
1.5.1.2

[Cluster-devel] [PATCH 41/51] [GFS2] flocks from same process trip kernel BUG at fs/gfs2/glock.c:1118!

[swhiteho]

From: Abhijith Das <adas@redhat.com>

This patch adds a new flag to the gfs2_holder structure GL_FLOCK.
It is set on holders of glocks representing flocks. This flag is
checked in add_to_queue() and a process is permitted to queue more
than one holder onto a glock if it is set. This solves the issue
of a process not being able to do multiple flocks on the same file.
Through a single descriptor, a process can now promote and demote
flocks. Through multiple descriptors a process can now queue
multiple flocks on the same file. There's still the problem of
a process deadlocking itself (because gfs2 blocking locks are not
interruptible) by queueing incompatible deadlock.

Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 931368a..d631cad 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -1106,24 +1106,31 @@ static void add_to_queue(struct gfs2_holder *gh)
 	if (test_and_set_bit(HIF_WAIT, &gh->gh_iflags))
 		BUG();
 
-	existing = find_holder_by_owner(&gl->gl_holders, gh->gh_owner_pid);
-	if (existing) {
-		print_symbol(KERN_WARNING "original: %s\n", existing->gh_ip);
-		printk(KERN_INFO "pid : %d\n", existing->gh_owner_pid);
-		printk(KERN_INFO "lock type : %d lock state : %d\n",
-				existing->gh_gl->gl_name.ln_type, existing->gh_gl->gl_state);
-		print_symbol(KERN_WARNING "new: %s\n", gh->gh_ip);
-		printk(KERN_INFO "pid : %d\n", gh->gh_owner_pid);
-		printk(KERN_INFO "lock type : %d lock state : %d\n",
-				gl->gl_name.ln_type, gl->gl_state);
-		BUG();
-	}
-
-	existing = find_holder_by_owner(&gl->gl_waiters3, gh->gh_owner_pid);
-	if (existing) {
-		print_symbol(KERN_WARNING "original: %s\n", existing->gh_ip);
-		print_symbol(KERN_WARNING "new: %s\n", gh->gh_ip);
-		BUG();
+	if (!(gh->gh_flags & GL_FLOCK)) {
+		existing = find_holder_by_owner(&gl->gl_holders, 
+						gh->gh_owner_pid);
+		if (existing) {
+			print_symbol(KERN_WARNING "original: %s\n", 
+				     existing->gh_ip);
+			printk(KERN_INFO "pid : %d\n", existing->gh_owner_pid);
+			printk(KERN_INFO "lock type : %d lock state : %d\n",
+			       existing->gh_gl->gl_name.ln_type, 
+			       existing->gh_gl->gl_state);
+			print_symbol(KERN_WARNING "new: %s\n", gh->gh_ip);
+			printk(KERN_INFO "pid : %d\n", gh->gh_owner_pid);
+			printk(KERN_INFO "lock type : %d lock state : %d\n",
+			       gl->gl_name.ln_type, gl->gl_state);
+			BUG();
+		}
+		
+		existing = find_holder_by_owner(&gl->gl_waiters3, 
+						gh->gh_owner_pid);
+		if (existing) {
+			print_symbol(KERN_WARNING "original: %s\n", 
+				     existing->gh_ip);
+			print_symbol(KERN_WARNING "new: %s\n", gh->gh_ip);
+			BUG();
+		}
 	}
 
 	if (gh->gh_flags & LM_FLAG_PRIORITY)
diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h
index f7a8e62..b16f604 100644
--- a/fs/gfs2/glock.h
+++ b/fs/gfs2/glock.h
@@ -26,6 +26,7 @@
 #define GL_SKIP			0x00000100
 #define GL_ATIME		0x00000200
 #define GL_NOCACHE		0x00000400
+#define GL_FLOCK		0x00000800
 #define GL_NOCANCEL		0x00001000
 
 #define GLR_TRYFAILED		13
diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c
index 94d76ac..46a9e10 100644
--- a/fs/gfs2/ops_file.c
+++ b/fs/gfs2/ops_file.c
@@ -571,7 +571,8 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl)
 	int error = 0;
 
 	state = (fl->fl_type == F_WRLCK) ? LM_ST_EXCLUSIVE : LM_ST_SHARED;
-	flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY) | GL_EXACT | GL_NOCACHE;
+	flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY) | GL_EXACT | GL_NOCACHE 
+		| GL_FLOCK;
 
 	mutex_lock(&fp->f_fl_mutex);
 
@@ -579,21 +580,19 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl)
 	if (gl) {
 		if (fl_gh->gh_state == state)
 			goto out;
-		gfs2_glock_hold(gl);
 		flock_lock_file_wait(file,
 				     &(struct file_lock){.fl_type = F_UNLCK});
-		gfs2_glock_dq_uninit(fl_gh);
+		gfs2_glock_dq_wait(fl_gh);
+		gfs2_holder_reinit(state, flags, fl_gh);
 	} else {
 		error = gfs2_glock_get(GFS2_SB(&ip->i_inode),
 				      ip->i_no_addr, &gfs2_flock_glops,
 				      CREATE, &gl);
 		if (error)
 			goto out;
+		gfs2_holder_init(gl, state, flags, fl_gh);
+		gfs2_glock_put(gl);
 	}
-
-	gfs2_holder_init(gl, state, flags, fl_gh);
-	gfs2_glock_put(gl);
-
 	error = gfs2_glock_nq(fl_gh);
 	if (error) {
 		gfs2_holder_uninit(fl_gh);
-- 
1.5.1.2

[Cluster-devel] [PATCH 42/51] [GFS2] Move inode deletion out of blocking_cb

[swhiteho]

From: Wendy Cheng <wcheng@redhat.com>

Move inode deletion code out of blocking_cb handle_callback route to
avoid racy conditions that end up blocking lock_dlm1 thread. Fix
bugzilla 286821.

Signed-off-by: Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index d631cad..a37efe4 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -716,12 +716,8 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state,
 		gl->gl_demote_time = jiffies;
 		if (remote && gl->gl_ops->go_type == LM_TYPE_IOPEN &&
 		    gl->gl_object) {
-			struct inode *inode = igrab(gl->gl_object);
+			gfs2_glock_schedule_for_reclaim(gl);
 			spin_unlock(&gl->gl_spin);
-			if (inode) {
-				d_prune_aliases(inode);
-				iput(inode);
-			}
 			return;
 		}
 	} else if (gl->gl_demote_state != LM_ST_UNLOCKED &&
-- 
1.5.1.2

[Cluster-devel] [PATCH 43/51] [DLM] Make dlm_sendd cond_resched more

[swhiteho]

From: Patrick Caulfield <pcaulfie@redhat.com>

Under high recovery loads dlm_sendd can monopolise the CPU and cause soft lockups.

This one extra and one moved cond_resched() make it yield a little more during
such times keeping work moving.

Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 62a8a6c..58bf3f5 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -1264,14 +1264,15 @@ static void send_to_sock(struct connection *con)
 		if (len) {
 			ret = sendpage(con->sock, e->page, offset, len,
 				       msg_flags);
-			if (ret == -EAGAIN || ret == 0)
+			if (ret == -EAGAIN || ret == 0) {
+				cond_resched();
 				goto out;
+			}
 			if (ret <= 0)
 				goto send_error;
-		} else {
+		}
 			/* Don't starve people filling buffers */
 			cond_resched();
-		}
 
 		spin_lock(&con->writequeue_lock);
 		e->offset += ret;
-- 
1.5.1.2

[Cluster-devel] [PATCH 44/51] [GFS2] GFS2: chmod hung - fix race in thread creation

[swhiteho]

From: Bob Peterson <rpeterso@redhat.com>

The problem boiled down to a race between the gdlm_init_threads()
function initializing thread1 and its setting of blist = 1.
Essentially, "if (current == ls->thread1)" was checked by the thread
before the thread creator set ls->thread1.

Since thread1 is the only thread who is allowed to work on the
blocking queue, and since neither thread thought it was thread1, no one
was working on the queue.  So everything just sat.

This patch reuses the ls->async_lock spin_lock to fix the race,
and it fixes the problem.  I've done more than 2000 iterations of the
loop that was recreating the failure and it seems to work.

Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

--

diff --git a/fs/gfs2/locking/dlm/thread.c b/fs/gfs2/locking/dlm/thread.c
index 1aca51e..bd938f0 100644
--- a/fs/gfs2/locking/dlm/thread.c
+++ b/fs/gfs2/locking/dlm/thread.c
@@ -268,20 +268,16 @@ static inline int check_drop(struct gdlm_ls *ls)
 	return 0;
 }
 
-static int gdlm_thread(void *data)
+static int gdlm_thread(void *data, int blist)
 {
 	struct gdlm_ls *ls = (struct gdlm_ls *) data;
 	struct gdlm_lock *lp = NULL;
-	int blist = 0;
 	uint8_t complete, blocking, submit, drop;
 	DECLARE_WAITQUEUE(wait, current);
 
 	/* Only thread1 is allowed to do blocking callbacks since gfs
 	   may wait for a completion callback within a blocking cb. */
 
-	if (current == ls->thread1)
-		blist = 1;
-
 	while (!kthread_should_stop()) {
 		set_current_state(TASK_INTERRUPTIBLE);
 		add_wait_queue(&ls->thread_wait, &wait);
@@ -333,12 +329,22 @@ static int gdlm_thread(void *data)
 	return 0;
 }
 
+static int gdlm_thread1(void *data)
+{
+	return gdlm_thread(data, 1);
+}
+
+static int gdlm_thread2(void *data)
+{
+	return gdlm_thread(data, 0);
+}
+
 int gdlm_init_threads(struct gdlm_ls *ls)
 {
 	struct task_struct *p;
 	int error;
 
-	p = kthread_run(gdlm_thread, ls, "lock_dlm1");
+	p = kthread_run(gdlm_thread1, ls, "lock_dlm1");
 	error = IS_ERR(p);
 	if (error) {
 		log_error("can't start lock_dlm1 thread %d", error);
@@ -346,7 +352,7 @@ int gdlm_init_threads(struct gdlm_ls *ls)
 	}
 	ls->thread1 = p;
 
-	p = kthread_run(gdlm_thread, ls, "lock_dlm2");
+	p = kthread_run(gdlm_thread2, ls, "lock_dlm2");
 	error = IS_ERR(p);
 	if (error) {
 		log_error("can't start lock_dlm2 thread %d", error);
-- 
1.5.1.2

[Cluster-devel] [PATCH 45/51] [GFS2] Clean up journaled data writing

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

This patch cleans up the code for writing journaled data into the log.
It also removes the need to allocate a small "tag" structure for each
block written into the log. Instead we just keep count of the outstanding
I/O so that we can be sure that its all been written at the correct time.
Another result of this patch is that a number of ll_rw_block() calls
have become submit_bh() calls, closing some races at the same time.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index 8aa5780..eaddfb5 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -341,12 +341,6 @@ struct gfs2_quota_data {
 	unsigned long qd_last_touched;
 };
 
-struct gfs2_log_buf {
-	struct list_head lb_list;
-	struct buffer_head *lb_bh;
-	struct buffer_head *lb_real;
-};
-
 struct gfs2_trans {
 	unsigned long tr_ip;
 
@@ -631,7 +625,8 @@ struct gfs2_sbd {
 
 	unsigned long sd_log_flush_time;
 	struct rw_semaphore sd_log_flush_lock;
-	struct list_head sd_log_flush_list;
+	atomic_t sd_log_in_flight;
+	wait_queue_head_t sd_log_flush_wait;
 
 	unsigned int sd_log_flush_head;
 	u64 sd_log_flush_wrapped;
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 4d04e6f..ee70467 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -104,11 +104,8 @@ static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 			gfs2_assert(sdp, bd->bd_ail == ai);
 
 			if (!buffer_busy(bh)) {
-				if (!buffer_uptodate(bh)) {
-					gfs2_log_unlock(sdp);
+				if (!buffer_uptodate(bh))
 					gfs2_io_error_bh(sdp, bh);
-					gfs2_log_lock(sdp);
-				}
 				list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list);
 				continue;
 			}
@@ -118,9 +115,16 @@ static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai)
 
 			list_move(&bd->bd_ail_st_list, &ai->ai_ail1_list);
 
+			get_bh(bh);
 			gfs2_log_unlock(sdp);
-			wait_on_buffer(bh);
-			ll_rw_block(WRITE, 1, &bh);
+			lock_buffer(bh);
+			if (test_clear_buffer_dirty(bh)) {
+				bh->b_end_io = end_buffer_write_sync;
+				submit_bh(WRITE, bh);
+			} else {
+				unlock_buffer(bh);
+				brelse(bh);
+			}
 			gfs2_log_lock(sdp);
 
 			retry = 1;
@@ -446,10 +450,10 @@ static unsigned int current_tail(struct gfs2_sbd *sdp)
 	return tail;
 }
 
-static inline void log_incr_head(struct gfs2_sbd *sdp)
+void gfs2_log_incr_head(struct gfs2_sbd *sdp)
 {
 	if (sdp->sd_log_flush_head == sdp->sd_log_tail)
-		gfs2_assert_withdraw(sdp, sdp->sd_log_flush_head == sdp->sd_log_head);
+		BUG_ON(sdp->sd_log_flush_head != sdp->sd_log_head);
 
 	if (++sdp->sd_log_flush_head == sdp->sd_jdesc->jd_blocks) {
 		sdp->sd_log_flush_head = 0;
@@ -458,6 +462,23 @@ static inline void log_incr_head(struct gfs2_sbd *sdp)
 }
 
 /**
+ * gfs2_log_write_endio - End of I/O for a log buffer
+ * @bh: The buffer head
+ * @uptodate: I/O Status
+ *
+ */
+
+static void gfs2_log_write_endio(struct buffer_head *bh, int uptodate)
+{
+	struct gfs2_sbd *sdp = bh->b_private;
+	bh->b_private = NULL;
+
+	end_buffer_write_sync(bh, uptodate);
+	if (atomic_dec_and_test(&sdp->sd_log_in_flight))
+		wake_up(&sdp->sd_log_flush_wait);
+}
+
+/**
  * gfs2_log_get_buf - Get and initialize a buffer to use for log control data
  * @sdp: The GFS2 superblock
  *
@@ -467,25 +488,42 @@ static inline void log_incr_head(struct gfs2_sbd *sdp)
 struct buffer_head *gfs2_log_get_buf(struct gfs2_sbd *sdp)
 {
 	u64 blkno = log_bmap(sdp, sdp->sd_log_flush_head);
-	struct gfs2_log_buf *lb;
 	struct buffer_head *bh;
 
-	lb = kzalloc(sizeof(struct gfs2_log_buf), GFP_NOFS | __GFP_NOFAIL);
-	list_add(&lb->lb_list, &sdp->sd_log_flush_list);
-
-	bh = lb->lb_bh = sb_getblk(sdp->sd_vfs, blkno);
+	bh = sb_getblk(sdp->sd_vfs, blkno);
 	lock_buffer(bh);
 	memset(bh->b_data, 0, bh->b_size);
 	set_buffer_uptodate(bh);
 	clear_buffer_dirty(bh);
-	unlock_buffer(bh);
-
-	log_incr_head(sdp);
+	gfs2_log_incr_head(sdp);
+	atomic_inc(&sdp->sd_log_in_flight);
+	bh->b_private = sdp;
+	bh->b_end_io = gfs2_log_write_endio;
 
 	return bh;
 }
 
 /**
+ * gfs2_fake_write_endio - 
+ * @bh: The buffer head
+ * @uptodate: The I/O Status
+ *
+ */
+
+static void gfs2_fake_write_endio(struct buffer_head *bh, int uptodate)
+{
+	struct buffer_head *real_bh = bh->b_private;
+	struct gfs2_sbd *sdp = GFS2_SB(real_bh->b_page->mapping->host);
+
+	end_buffer_write_sync(bh, uptodate);
+	free_buffer_head(bh);
+	unlock_buffer(real_bh);
+	brelse(real_bh);
+	if (atomic_dec_and_test(&sdp->sd_log_in_flight))
+		wake_up(&sdp->sd_log_flush_wait);
+}
+
+/**
  * gfs2_log_fake_buf - Build a fake buffer head to write metadata buffer to log
  * @sdp: the filesystem
  * @data: the data the buffer_head should point to
@@ -497,22 +535,20 @@ struct buffer_head *gfs2_log_fake_buf(struct gfs2_sbd *sdp,
 				      struct buffer_head *real)
 {
 	u64 blkno = log_bmap(sdp, sdp->sd_log_flush_head);
-	struct gfs2_log_buf *lb;
 	struct buffer_head *bh;
 
-	lb = kzalloc(sizeof(struct gfs2_log_buf), GFP_NOFS | __GFP_NOFAIL);
-	list_add(&lb->lb_list, &sdp->sd_log_flush_list);
-	lb->lb_real = real;
-
-	bh = lb->lb_bh = alloc_buffer_head(GFP_NOFS | __GFP_NOFAIL);
+	bh = alloc_buffer_head(GFP_NOFS | __GFP_NOFAIL);
 	atomic_set(&bh->b_count, 1);
-	bh->b_state = (1 << BH_Mapped) | (1 << BH_Uptodate);
+	bh->b_state = (1 << BH_Mapped) | (1 << BH_Uptodate) | (1 << BH_Lock);
 	set_bh_page(bh, real->b_page, bh_offset(real));
 	bh->b_blocknr = blkno;
 	bh->b_size = sdp->sd_sb.sb_bsize;
 	bh->b_bdev = sdp->sd_vfs->s_bdev;
+	bh->b_private = real;
+	bh->b_end_io = gfs2_fake_write_endio;
 
-	log_incr_head(sdp);
+	gfs2_log_incr_head(sdp);
+	atomic_inc(&sdp->sd_log_in_flight);
 
 	return bh;
 }
@@ -579,45 +615,24 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull)
 		gfs2_assert_withdraw(sdp, !pull);
 
 	sdp->sd_log_idle = (tail == sdp->sd_log_flush_head);
-	log_incr_head(sdp);
+	gfs2_log_incr_head(sdp);
 }
 
 static void log_flush_commit(struct gfs2_sbd *sdp)
 {
-	struct list_head *head = &sdp->sd_log_flush_list;
-	struct gfs2_log_buf *lb;
-	struct buffer_head *bh;
-	int flushcount = 0;
-
-	while (!list_empty(head)) {
-		lb = list_entry(head->next, struct gfs2_log_buf, lb_list);
-		list_del(&lb->lb_list);
-		bh = lb->lb_bh;
-
-		wait_on_buffer(bh);
-		if (!buffer_uptodate(bh))
-			gfs2_io_error_bh(sdp, bh);
-		if (lb->lb_real) {
-			while (atomic_read(&bh->b_count) != 1)  /* Grrrr... */
-				schedule();
-			free_buffer_head(bh);
-		} else
-			brelse(bh);
-		kfree(lb);
-		flushcount++;
+	DEFINE_WAIT(wait);
+
+	if (atomic_read(&sdp->sd_log_in_flight)) {
+		do {
+			prepare_to_wait(&sdp->sd_log_flush_wait, &wait,
+					TASK_UNINTERRUPTIBLE);
+			if (atomic_read(&sdp->sd_log_in_flight))
+				io_schedule();
+		} while(atomic_read(&sdp->sd_log_in_flight));
+		finish_wait(&sdp->sd_log_flush_wait, &wait);
 	}
 
-	/* If nothing was journaled, the header is unplanned and unwanted. */
-	if (flushcount) {
-		log_write_header(sdp, 0, 0);
-	} else {
-		unsigned int tail;
-		tail = current_tail(sdp);
-
-		gfs2_ail1_empty(sdp, 0);
-		if (sdp->sd_log_tail != tail)
-			log_pull_tail(sdp, tail);
-	}
+	log_write_header(sdp, 0, 0);
 }
 
 static void gfs2_ordered_write(struct gfs2_sbd *sdp)
@@ -698,10 +713,16 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl)
 	INIT_LIST_HEAD(&ai->ai_ail1_list);
 	INIT_LIST_HEAD(&ai->ai_ail2_list);
 
-	gfs2_assert_withdraw(sdp,
-			     sdp->sd_log_num_buf + sdp->sd_log_num_databuf ==
-			     sdp->sd_log_commited_buf +
-			     sdp->sd_log_commited_databuf);
+	if (sdp->sd_log_num_buf != sdp->sd_log_commited_buf) {
+		printk(KERN_INFO "GFS2: log buf %u %u\n", sdp->sd_log_num_buf,
+		       sdp->sd_log_commited_buf);
+		gfs2_assert_withdraw(sdp, 0);
+	}
+	if (sdp->sd_log_num_databuf != sdp->sd_log_commited_databuf) {
+		printk(KERN_INFO "GFS2: log databuf %u %u\n",
+		       sdp->sd_log_num_databuf, sdp->sd_log_commited_databuf);
+		gfs2_assert_withdraw(sdp, 0);
+	}
 	gfs2_assert_withdraw(sdp,
 			sdp->sd_log_num_revoke == sdp->sd_log_commited_revoke);
 
@@ -713,7 +734,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl)
 	lops_before_commit(sdp);
 	gfs2_ordered_wait(sdp);
 
-	if (!list_empty(&sdp->sd_log_flush_list))
+	if (sdp->sd_log_head != sdp->sd_log_flush_head)
 		log_flush_commit(sdp);
 	else if (sdp->sd_log_tail != current_tail(sdp) && !sdp->sd_log_idle){
 		gfs2_log_lock(sdp);
diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h
index 6394235..dae2824 100644
--- a/fs/gfs2/log.h
+++ b/fs/gfs2/log.h
@@ -52,6 +52,7 @@ int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags);
 
 int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks);
 void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks);
+void gfs2_log_incr_head(struct gfs2_sbd *sdp);
 
 struct buffer_head *gfs2_log_get_buf(struct gfs2_sbd *sdp);
 struct buffer_head *gfs2_log_fake_buf(struct gfs2_sbd *sdp,
diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c
index 342c10e..6c27cea 100644
--- a/fs/gfs2/lops.c
+++ b/fs/gfs2/lops.c
@@ -91,6 +91,39 @@ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh,
 	unlock_buffer(bh);
 }
 
+
+static inline struct gfs2_log_descriptor *bh_log_desc(struct buffer_head *bh)
+{
+	return (struct gfs2_log_descriptor *)bh->b_data;
+}
+
+static inline __be64 *bh_log_ptr(struct buffer_head *bh)
+{
+	struct gfs2_log_descriptor *ld = bh_log_desc(bh);
+	return (__force __be64 *)(ld + 1);
+}
+
+static inline __be64 *bh_ptr_end(struct buffer_head *bh)
+{
+	return (__force __be64 *)(bh->b_data + bh->b_size);
+}
+
+
+static struct buffer_head *gfs2_get_log_desc(struct gfs2_sbd *sdp, u32 ld_type)
+{
+	struct buffer_head *bh = gfs2_log_get_buf(sdp);
+	struct gfs2_log_descriptor *ld = bh_log_desc(bh);
+	ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
+	ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
+	ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
+	ld->ld_type = cpu_to_be32(ld_type);
+	ld->ld_length = 0;
+	ld->ld_data1 = 0;
+	ld->ld_data2 = 0;
+	memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
+	return bh;
+}
+
 static void __glock_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le)
 {
 	struct gfs2_glock *gl;
@@ -181,7 +214,6 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 	struct gfs2_log_descriptor *ld;
 	struct gfs2_bufdata *bd1 = NULL, *bd2;
 	unsigned int total;
-	unsigned int offset = BUF_OFFSET;
 	unsigned int limit;
 	unsigned int num;
 	unsigned n;
@@ -198,18 +230,12 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 		if (total > limit)
 			num = limit;
 		gfs2_log_unlock(sdp);
-		bh = gfs2_log_get_buf(sdp);
+		bh = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_METADATA);
 		gfs2_log_lock(sdp);
-		ld = (struct gfs2_log_descriptor *)bh->b_data;
-		ptr = (__be64 *)(bh->b_data + offset);
-		ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
-		ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
-		ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
-		ld->ld_type = cpu_to_be32(GFS2_LOG_DESC_METADATA);
+		ld = bh_log_desc(bh);
+		ptr = bh_log_ptr(bh);
 		ld->ld_length = cpu_to_be32(num + 1);
 		ld->ld_data1 = cpu_to_be32(num);
-		ld->ld_data2 = cpu_to_be32(0);
-		memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
 
 		n = 0;
 		list_for_each_entry_continue(bd1, &sdp->sd_log_le_buf,
@@ -220,17 +246,17 @@ static void buf_lo_before_commit(struct gfs2_sbd *sdp)
 		}
 
 		gfs2_log_unlock(sdp);
-		set_buffer_dirty(bh);
-		ll_rw_block(WRITE, 1, &bh);
+		submit_bh(WRITE, bh);
 		gfs2_log_lock(sdp);
 
 		n = 0;
 		list_for_each_entry_continue(bd2, &sdp->sd_log_le_buf,
 					     bd_le.le_list) {
+			get_bh(bd2->bd_bh);
 			gfs2_log_unlock(sdp);
+			lock_buffer(bd2->bd_bh);
 			bh = gfs2_log_fake_buf(sdp, bd2->bd_bh);
-			set_buffer_dirty(bh);
-			ll_rw_block(WRITE, 1, &bh);
+			submit_bh(WRITE, bh);
 			gfs2_log_lock(sdp);
 			if (++n >= num)
 				break;
@@ -359,17 +385,11 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 	if (!sdp->sd_log_num_revoke)
 		return;
 
-	bh = gfs2_log_get_buf(sdp);
-	ld = (struct gfs2_log_descriptor *)bh->b_data;
-	ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
-	ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
-	ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
-	ld->ld_type = cpu_to_be32(GFS2_LOG_DESC_REVOKE);
+	bh = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_REVOKE);
+	ld = bh_log_desc(bh);
 	ld->ld_length = cpu_to_be32(gfs2_struct2blk(sdp, sdp->sd_log_num_revoke,
 						    sizeof(u64)));
 	ld->ld_data1 = cpu_to_be32(sdp->sd_log_num_revoke);
-	ld->ld_data2 = cpu_to_be32(0);
-	memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
 	offset = sizeof(struct gfs2_log_descriptor);
 
 	while (!list_empty(head)) {
@@ -378,8 +398,7 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 		sdp->sd_log_num_revoke--;
 
 		if (offset + sizeof(u64) > sdp->sd_sb.sb_bsize) {
-			set_buffer_dirty(bh);
-			ll_rw_block(WRITE, 1, &bh);
+			submit_bh(WRITE, bh);
 
 			bh = gfs2_log_get_buf(sdp);
 			mh = (struct gfs2_meta_header *)bh->b_data;
@@ -396,8 +415,7 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp)
 	}
 	gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke);
 
-	set_buffer_dirty(bh);
-	ll_rw_block(WRITE, 1, &bh);
+	submit_bh(WRITE, bh);
 }
 
 static void revoke_lo_before_scan(struct gfs2_jdesc *jd,
@@ -562,118 +580,110 @@ out:
 	unlock_buffer(bd->bd_bh);
 }
 
-static int gfs2_check_magic(struct buffer_head *bh)
+static void gfs2_check_magic(struct buffer_head *bh)
 {
-	struct page *page = bh->b_page;
 	void *kaddr;
 	__be32 *ptr;
-	int rv = 0;
 
-	kaddr = kmap_atomic(page, KM_USER0);
+	clear_buffer_escaped(bh);
+	kaddr = kmap_atomic(bh->b_page, KM_USER0);
 	ptr = kaddr + bh_offset(bh);
 	if (*ptr == cpu_to_be32(GFS2_MAGIC))
-		rv = 1;
+		set_buffer_escaped(bh);
 	kunmap_atomic(kaddr, KM_USER0);
-
-	return rv;
 }
 
-/**
- * databuf_lo_before_commit - Scan the data buffers, writing as we go
- *
- */
-
-static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
+static void gfs2_write_blocks(struct gfs2_sbd *sdp, struct buffer_head *bh,
+			      struct list_head *list, struct list_head *done,
+			      unsigned int n)
 {
-	struct gfs2_bufdata *bd1 = NULL, *bd2;
-	struct buffer_head *bh = NULL,*bh1 = NULL;
+	struct buffer_head *bh1;
 	struct gfs2_log_descriptor *ld;
-	unsigned int limit;
-	unsigned int total;
-	unsigned int num, n;
-	__be64 *ptr = NULL;
-	int magic;
+	struct gfs2_bufdata *bd;
+	__be64 *ptr;
 
+	if (!bh)
+		return;
 
-	limit = databuf_limit(sdp);
+	ld = bh_log_desc(bh);
+	ld->ld_length = cpu_to_be32(n + 1);
+	ld->ld_data1 = cpu_to_be32(n);
 
+	ptr = bh_log_ptr(bh);
+	
+	get_bh(bh);
+	submit_bh(WRITE, bh);
 	gfs2_log_lock(sdp);
-	total = sdp->sd_log_num_databuf;
-	bd2 = bd1 = list_prepare_entry(bd1, &sdp->sd_log_le_databuf,
-				       bd_le.le_list);
-	while(total) {
-		num = total;
-		if (num > limit)
-			num = limit;
-
-		gfs2_log_unlock(sdp);
-		bh = gfs2_log_get_buf(sdp);
-		gfs2_log_lock(sdp);
-
-		ld = (struct gfs2_log_descriptor *)bh->b_data;
-		ptr = (__be64 *)(bh->b_data + DATABUF_OFFSET);
-		ld->ld_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
-		ld->ld_header.mh_type = cpu_to_be32(GFS2_METATYPE_LD);
-		ld->ld_header.mh_format = cpu_to_be32(GFS2_FORMAT_LD);
-		ld->ld_type = cpu_to_be32(GFS2_LOG_DESC_JDATA);
-		ld->ld_length = cpu_to_be32(num + 1);
-		ld->ld_data1 = cpu_to_be32(num);
-		ld->ld_data2 = cpu_to_be32(0);
-		memset(ld->ld_reserved, 0, sizeof(ld->ld_reserved));
-
-		n = 0;
-		list_for_each_entry_continue(bd1, &sdp->sd_log_le_databuf,
-					     bd_le.le_list) {
-			bh1 = bd1->bd_bh;
-
-			magic = gfs2_check_magic(bh1);
-			*ptr++ = cpu_to_be64(bh1->b_blocknr);
-			*ptr++ = cpu_to_be64((__u64)magic);
-			clear_buffer_escaped(bh1);
-			if (unlikely(magic != 0))
-				set_buffer_escaped(bh1);
-			if (++n >= num)
-				break;
+	while(!list_empty(list)) {
+		bd = list_entry(list->next, struct gfs2_bufdata, bd_le.le_list);
+		list_move_tail(&bd->bd_le.le_list, done);
+		get_bh(bd->bd_bh);
+		while (be64_to_cpu(*ptr) != bd->bd_bh->b_blocknr) {
+			gfs2_log_incr_head(sdp);
+			ptr += 2;
 		}
 		gfs2_log_unlock(sdp);
-		if (bh) {
-			set_buffer_dirty(bh);
-			ll_rw_block(WRITE, 1, &bh);
-			bh = NULL;
-			ptr = NULL;
+		lock_buffer(bd->bd_bh);
+		if (buffer_escaped(bd->bd_bh)) {
+			void *kaddr;
+			bh1 = gfs2_log_get_buf(sdp);
+			kaddr = kmap_atomic(bd->bd_bh->b_page, KM_USER0);
+			memcpy(bh1->b_data, kaddr + bh_offset(bd->bd_bh),
+			       bh1->b_size);
+			kunmap_atomic(kaddr, KM_USER0);
+			*(__be32 *)bh1->b_data = 0;
+			clear_buffer_escaped(bd->bd_bh);
+			unlock_buffer(bd->bd_bh);
+			brelse(bd->bd_bh);
+		} else {
+			bh1 = gfs2_log_fake_buf(sdp, bd->bd_bh);
 		}
-		n = 0;
+		submit_bh(WRITE, bh1);
 		gfs2_log_lock(sdp);
-		list_for_each_entry_continue(bd2, &sdp->sd_log_le_databuf,
-					     bd_le.le_list) {
-			if (!bd2->bd_bh)
-				continue;
-			/* copy buffer if it needs escaping */
+		ptr += 2;
+	}
+	gfs2_log_unlock(sdp);
+	brelse(bh);
+}
+
+/**
+ * databuf_lo_before_commit - Scan the data buffers, writing as we go
+ *
+ */
+
+static void databuf_lo_before_commit(struct gfs2_sbd *sdp)
+{
+	struct gfs2_bufdata *bd = NULL;
+	struct buffer_head *bh = NULL;
+	unsigned int n = 0;
+	__be64 *ptr = NULL, *end = NULL;
+	LIST_HEAD(processed);
+	LIST_HEAD(in_progress);
+
+	gfs2_log_lock(sdp);
+	while (!list_empty(&sdp->sd_log_le_databuf)) {
+		if (ptr == end) {
 			gfs2_log_unlock(sdp);
-			if (unlikely(buffer_escaped(bd2->bd_bh))) {
-				void *kaddr;
-				struct page *page = bd2->bd_bh->b_page;
-				bh = gfs2_log_get_buf(sdp);
-				kaddr = kmap_atomic(page, KM_USER0);
-				memcpy(bh->b_data,
-				       kaddr + bh_offset(bd2->bd_bh),
-				       sdp->sd_sb.sb_bsize);
-				kunmap_atomic(kaddr, KM_USER0);
-				*(__be32 *)bh->b_data = 0;
-			} else {
-				bh = gfs2_log_fake_buf(sdp, bd2->bd_bh);
-			}
-			set_buffer_dirty(bh);
-			ll_rw_block(WRITE, 1, &bh);
+			gfs2_write_blocks(sdp, bh, &in_progress, &processed, n);
+			n = 0;
+			bh = gfs2_get_log_desc(sdp, GFS2_LOG_DESC_JDATA);
+			ptr = bh_log_ptr(bh);
+			end = bh_ptr_end(bh) - 1;
 			gfs2_log_lock(sdp);
-			if (++n >= num)
-				break;
+			continue;
 		}
-		bh = NULL;
-		BUG_ON(total < num);
-		total -= num;
+		bd = list_entry(sdp->sd_log_le_databuf.next, struct gfs2_bufdata, bd_le.le_list);
+		list_move_tail(&bd->bd_le.le_list, &in_progress);
+		gfs2_check_magic(bd->bd_bh);
+		*ptr++ = cpu_to_be64(bd->bd_bh->b_blocknr);
+		*ptr++ = cpu_to_be64(buffer_escaped(bh) ? 1 : 0);
+		n++;
 	}
 	gfs2_log_unlock(sdp);
+	gfs2_write_blocks(sdp, bh, &in_progress, &processed, n);
+	gfs2_log_lock(sdp);
+	list_splice(&processed, &sdp->sd_log_le_databuf);
+	gfs2_log_unlock(sdp);
 }
 
 static int databuf_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start,
@@ -807,10 +817,10 @@ const struct gfs2_log_operations gfs2_databuf_lops = {
 
 const struct gfs2_log_operations *gfs2_log_ops[] = {
 	&gfs2_glock_lops,
+	&gfs2_databuf_lops,
 	&gfs2_buf_lops,
-	&gfs2_revoke_lops,
 	&gfs2_rg_lops,
-	&gfs2_databuf_lops,
+	&gfs2_revoke_lops,
 	NULL,
 };
 
diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index 19097bc..1d80f2d 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -297,6 +297,37 @@ void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh,
 		unlock_page(bh->b_page);
 }
 
+void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int meta)
+{
+	struct gfs2_sbd *sdp = GFS2_SB(bh->b_page->mapping->host);
+	struct gfs2_bufdata *bd = bh->b_private;
+	if (test_clear_buffer_pinned(bh)) {
+		list_del_init(&bd->bd_le.le_list);
+		if (meta) {
+			gfs2_assert_warn(sdp, sdp->sd_log_num_buf);
+			sdp->sd_log_num_buf--;
+			tr->tr_num_buf_rm++;
+		} else {
+			gfs2_assert_warn(sdp, sdp->sd_log_num_databuf);
+			sdp->sd_log_num_databuf--;
+			tr->tr_num_databuf_rm++;
+		}
+		tr->tr_touched = 1;
+		brelse(bh);
+	}
+	if (bd) {
+		if (bd->bd_ail) {
+			gfs2_remove_from_ail(NULL, bd);
+			bh->b_private = NULL;
+			bd->bd_bh = NULL;
+			bd->bd_blkno = bh->b_blocknr;
+			gfs2_trans_add_revoke(sdp, bd);
+		}
+	}
+	clear_buffer_dirty(bh);
+	clear_buffer_uptodate(bh);
+}
+
 /**
  * gfs2_meta_wipe - make inode's buffers so they aren't dirty/pinned anymore
  * @ip: the inode who owns the buffers
@@ -313,33 +344,11 @@ void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
 	while (blen) {
 		bh = getbuf(ip->i_gl, bstart, NO_CREATE);
 		if (bh) {
-			struct gfs2_bufdata *bd;
-
 			lock_buffer(bh);
 			gfs2_log_lock(sdp);
-			bd = bh->b_private;
-			if (test_clear_buffer_pinned(bh)) {
-				struct gfs2_trans *tr = current->journal_info;
-				list_del_init(&bd->bd_le.le_list);
-				gfs2_assert_warn(sdp, sdp->sd_log_num_buf);
-				sdp->sd_log_num_buf--;
-				tr->tr_num_buf_rm++;
-				brelse(bh);
-			}
-			if (bd) {
-				if (bd->bd_ail) {
-					gfs2_remove_from_ail(NULL, bd);
-					bh->b_private = NULL;
-					bd->bd_bh = NULL;
-					bd->bd_blkno = bh->b_blocknr;
-					gfs2_trans_add_revoke(sdp, bd);
-				}
-			}
-			clear_buffer_dirty(bh);
-			clear_buffer_uptodate(bh);
+			gfs2_remove_from_journal(bh, current->journal_info, 1);
 			gfs2_log_unlock(sdp);
 			unlock_buffer(bh);
-
 			brelse(bh);
 		}
 
diff --git a/fs/gfs2/meta_io.h b/fs/gfs2/meta_io.h
index fd67f07..b704822 100644
--- a/fs/gfs2/meta_io.h
+++ b/fs/gfs2/meta_io.h
@@ -51,6 +51,9 @@ int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh);
 void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh,
 			 int meta);
 
+void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr,
+			      int meta);
+
 void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen);
 
 void gfs2_meta_cache_flush(struct gfs2_inode *ip);
diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c
index dd1ea49..b7baf18 100644
--- a/fs/gfs2/ops_address.c
+++ b/fs/gfs2/ops_address.c
@@ -414,7 +414,8 @@ static int gfs2_prepare_write(struct file *file, struct page *page,
 	if (ind_blocks || data_blocks)
 		rblocks += RES_STATFS + RES_QUOTA;
 
-	error = gfs2_trans_begin(sdp, rblocks, 0);
+	error = gfs2_trans_begin(sdp, rblocks,
+				 PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize);
 	if (error)
 		goto out_trans_fail;
 
@@ -625,10 +626,10 @@ static void gfs2_discard(struct gfs2_sbd *sdp, struct buffer_head *bh)
 	clear_buffer_dirty(bh);
 	bd = bh->b_private;
 	if (bd) {
-		if (!list_empty(&bd->bd_le.le_list)) {
-			if (!buffer_pinned(bh))
-				list_del_init(&bd->bd_le.le_list);
-		}
+		if (!list_empty(&bd->bd_le.le_list) && !buffer_pinned(bh))
+			list_del_init(&bd->bd_le.le_list);
+		else
+			gfs2_remove_from_journal(bh, current->journal_info, 0);
 	}
 	bh->b_bdev = NULL;
 	clear_buffer_mapped(bh);
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 35f3dfa..1ac9afa 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -89,7 +89,8 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb)
 	INIT_LIST_HEAD(&sdp->sd_ail2_list);
 
 	init_rwsem(&sdp->sd_log_flush_lock);
-	INIT_LIST_HEAD(&sdp->sd_log_flush_list);
+	atomic_set(&sdp->sd_log_in_flight, 0);
+	init_waitqueue_head(&sdp->sd_log_flush_wait);
 
 	INIT_LIST_HEAD(&sdp->sd_revoke_list);
 
diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c
index 2cbe5a3..291f0c7 100644
--- a/fs/gfs2/ops_inode.c
+++ b/fs/gfs2/ops_inode.c
@@ -905,12 +905,17 @@ static int gfs2_permission(struct inode *inode, int mask, struct nameidata *nd)
 static int setattr_size(struct inode *inode, struct iattr *attr)
 {
 	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	int error;
 
 	if (attr->ia_size != ip->i_di.di_size) {
-		error = vmtruncate(inode, attr->ia_size);
+		error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);
 		if (error)
 			return error;
+		error = vmtruncate(inode, attr->ia_size);
+		gfs2_trans_end(sdp);
+		if (error) 
+			return error;
 	}
 
 	error = gfs2_truncatei(ip, attr->ia_size);
diff --git a/fs/gfs2/ops_super.c b/fs/gfs2/ops_super.c
index 4316690..950f314 100644
--- a/fs/gfs2/ops_super.c
+++ b/fs/gfs2/ops_super.c
@@ -455,12 +455,15 @@ static void gfs2_delete_inode(struct inode *inode)
 	}
 
 	error = gfs2_dinode_dealloc(ip);
-	/*
-	 * Must do this before unlock to avoid trying to write back
-	 * potentially dirty data now that inode no longer exists
-	 * on disk.
-	 */
+	if (error)
+		goto out_unlock;
+
+	error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);
+	if (error)
+		goto out_unlock;
+	/* Needs to be done before glock release & also in a transaction */
 	truncate_inode_pages(&inode->i_data, 0);
+	gfs2_trans_end(sdp);
 
 out_unlock:
 	gfs2_glock_dq(&ip->i_iopen_gh);
-- 
1.5.1.2

[Cluster-devel] [PATCH 46/51] [GFS2] Data corruption fix

[swhiteho]

From: Wendy Cheng <wcheng@redhat.com>

* GFS2 has been using i_cache array to store its indirect meta blocks.
Its flush routine doesn't correctly clean up all the entries. The
problem would show while multiple nodes do simultaneous writes to the
same file. Upon glock exclusive lock transfer, if the file is a sparse
file with large file size where the indirect meta blocks span multiple
array entries with "zero" entries in between. The flush routine
prematurely stops the flushing that leaves old (stale) entries around.
This leads to several nasty issues, including data corruption.
* Fix gfs2_get_block_noalloc checking to correctly return EIO upon
unmapped buffer.

Signed-off-by: Wendy Cheng <wcheng@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index 1d80f2d..4da4239 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -374,10 +374,10 @@ void gfs2_meta_cache_flush(struct gfs2_inode *ip)
 
 	for (x = 0; x < GFS2_MAX_META_HEIGHT; x++) {
 		bh_slot = &ip->i_cache[x];
-		if (!*bh_slot)
-			break;
-		brelse(*bh_slot);
-		*bh_slot = NULL;
+		if (*bh_slot) {
+			brelse(*bh_slot);
+			*bh_slot = NULL;
+		}
 	}
 
 	spin_unlock(&ip->i_spin);
diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c
index b7baf18..4002f41 100644
--- a/fs/gfs2/ops_address.c
+++ b/fs/gfs2/ops_address.c
@@ -90,7 +90,7 @@ static int gfs2_get_block_noalloc(struct inode *inode, sector_t lblock,
 	error = gfs2_block_map(inode, lblock, 0, bh_result);
 	if (error)
 		return error;
-	if (bh_result->b_blocknr == 0)
+	if (!buffer_mapped(bh_result))
 		return -EIO;
 	return 0;
 }
-- 
1.5.1.2

[Cluster-devel] [PATCH 47/51] [GFS2] Alternate gfs2_iget to avoid looking up inodes being freed

[swhiteho]

From: Benjamin Marzinski <bmarzins@redhat.com>

There is a possible deadlock between two processes on the same node, where one
process is deleting an inode, and another process is looking for allocated but
unused inodes to delete in order to create more space.

process A does an iput() on inode X, and it's i_count drops to 0. This causes
iput_final() to be called, which puts an inode into state I_FREEING at
generic_delete_inode(). There no point between when iput_final() is called, and
when I_FREEING is set where GFS2 could acquire any glocks. Once I_FREEING is
set, no other process on that node can successfully look up that inode until
the delete finishes.

process B locks the the resource group for the same inode in get_local_rgrp(),
which is called by gfs2_inplace_reserve_i()

process A tries to lock the resource group for the inode in
gfs2_dinode_dealloc(), but it's already locked by process B

process B waits in find_inode for the inode to have the I_FREEING state cleared.

Deadlock.

This patch solves the problem by adding an alternative to gfs2_iget(),
gfs2_iget_skip(), that simply skips any inodes that are in the I_FREEING
state.o The alternate test function is just like the original one, except that
it fails if the inode is being freed, and sets a skipped flag. The alternate
set function is just like the original, except that it fails if the skipped
flag is set. Only try_rgrp_unlink() calls gfs2_iget_skip() instead of
gfs2_iget().

Signed-off-by: Benjamin E. Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c
index 08c6dd0..9949bb7 100644
--- a/fs/gfs2/dir.c
+++ b/fs/gfs2/dir.c
@@ -1502,7 +1502,7 @@ struct inode *gfs2_dir_search(struct inode *dir, const struct qstr *name)
 		inode = gfs2_inode_lookup(dir->i_sb, 
 				be16_to_cpu(dent->de_type),
 				be64_to_cpu(dent->de_inum.no_addr),
-				be64_to_cpu(dent->de_inum.no_formal_ino));
+				be64_to_cpu(dent->de_inum.no_formal_ino), 0);
 		brelse(bh);
 		return inode;
 	}
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 013f00b..5f6dc32 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -77,6 +77,49 @@ static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr)
 	return iget5_locked(sb, hash, iget_test, iget_set, &no_addr);
 }
 
+struct gfs2_skip_data {
+	u64	no_addr;
+	int	skipped;
+};
+
+static int iget_skip_test(struct inode *inode, void *opaque)
+{
+	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_skip_data *data = opaque;
+
+	if (ip->i_no_addr == data->no_addr && inode->i_private != NULL){
+		if (inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE)){
+			data->skipped = 1;
+			return 0;
+		}
+		return 1;
+	}
+	return 0;
+}
+
+static int iget_skip_set(struct inode *inode, void *opaque)
+{
+	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_skip_data *data = opaque;
+
+	if (data->skipped)
+		return 1;
+	inode->i_ino = (unsigned long)(data->no_addr);
+	ip->i_no_addr = data->no_addr;
+	return 0;
+}
+
+static struct inode *gfs2_iget_skip(struct super_block *sb,
+				    u64 no_addr)
+{
+	struct gfs2_skip_data data;
+	unsigned long hash = (unsigned long)no_addr;
+
+	data.no_addr = no_addr;
+	data.skipped = 0;
+	return iget5_locked(sb, hash, iget_skip_test, iget_skip_set, &data);
+}
+
 /**
  * GFS2 lookup code fills in vfs inode contents based on info obtained
  * from directory entry inside gfs2_inode_lookup(). This has caused issues
@@ -112,6 +155,7 @@ void gfs2_set_iop(struct inode *inode)
  * @sb: The super block
  * @no_addr: The inode number
  * @type: The type of the inode
+ * @skip_freeing: set this not return an inode if it is currently being freed.
  *
  * Returns: A VFS inode, or an error
  */
@@ -119,13 +163,19 @@ void gfs2_set_iop(struct inode *inode)
 struct inode *gfs2_inode_lookup(struct super_block *sb, 
 				unsigned int type,
 				u64 no_addr,
-				u64 no_formal_ino)
+				u64 no_formal_ino, int skip_freeing)
 {
-	struct inode *inode = gfs2_iget(sb, no_addr);
-	struct gfs2_inode *ip = GFS2_I(inode);
+	struct inode *inode;
+	struct gfs2_inode *ip;
 	struct gfs2_glock *io_gl;
 	int error;
 
+	if (skip_freeing)
+		inode = gfs2_iget_skip(sb, no_addr);
+	else
+		inode = gfs2_iget(sb, no_addr);
+	ip = GFS2_I(inode);
+
 	if (!inode)
 		return ERR_PTR(-ENOBUFS);
 
@@ -949,7 +999,7 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name,
 
 	inode = gfs2_inode_lookup(dir->i_sb, IF2DT(mode),
 					inum.no_addr,
-					inum.no_formal_ino);
+					inum.no_formal_ino, 0);
 	if (IS_ERR(inode))
 		goto fail_gunlock2;
 
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index 4517ac8..351ac87 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -49,7 +49,8 @@ static inline void gfs2_inum_out(const struct gfs2_inode *ip,
 void gfs2_inode_attr_in(struct gfs2_inode *ip);
 void gfs2_set_iop(struct inode *inode);
 struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, 
-				u64 no_addr, u64 no_formal_ino);
+				u64 no_addr, u64 no_formal_ino,
+				int skip_freeing);
 struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr);
 
 int gfs2_inode_refresh(struct gfs2_inode *ip);
diff --git a/fs/gfs2/ops_export.c b/fs/gfs2/ops_export.c
index b8312ed..e2d1347 100644
--- a/fs/gfs2/ops_export.c
+++ b/fs/gfs2/ops_export.c
@@ -237,7 +237,7 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb, void *inum_obj)
 
 	inode = gfs2_inode_lookup(sb, DT_UNKNOWN,
 					inum->no_addr,
-					0);
+					0, 0);
 	if (!inode)
 		goto fail;
 	if (IS_ERR(inode)) {
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 1ac9afa..17de58e 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -230,7 +230,7 @@ fail:
 static inline struct inode *gfs2_lookup_root(struct super_block *sb,
 					     u64 no_addr)
 {
-	return gfs2_inode_lookup(sb, DT_DIR, no_addr, 0);
+	return gfs2_inode_lookup(sb, DT_DIR, no_addr, 0, 0);
 }
 
 static int init_sb(struct gfs2_sbd *sdp, int silent, int undo)
diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index 2d7f7ea..708c287 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -884,7 +884,7 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked)
 			continue;
 		*last_unlinked = no_addr;
 		inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN,
-					  no_addr, -1);
+					  no_addr, -1, 1);
 		if (!IS_ERR(inode))
 			return inode;
 	}
-- 
1.5.1.2

[Cluster-devel] [PATCH 48/51] [GFS2] Don't try to remove buffers that don't exist

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c
index 4002f41..873a511 100644
--- a/fs/gfs2/ops_address.c
+++ b/fs/gfs2/ops_address.c
@@ -747,7 +747,7 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 	struct gfs2_bufdata *bd;
 
 	if (!page_has_buffers(page))
-		goto out;
+		return 0;
 
 	gfs2_log_lock(sdp);
 	head = bh = page_buffers(page);
@@ -787,7 +787,6 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 		bh = bh->b_this_page;
 	} while (bh != head);
 
-out:
 	return try_to_free_buffers(page);
 cannot_release:
 	gfs2_log_unlock(sdp);
-- 
1.5.1.2

[Cluster-devel] [PATCH 49/51] [GFS2] Get superblock a different way

[swhiteho]

From: Steven Whitehouse <swhiteho@redhat.com>

The mapping may be NULL by the time the I/O has completed, so
we now get the superblock by a different route (via the bd and glock)
to avoid this problem.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Wendy Cheng <wcheng@redhat.com>

diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index ee70467..7df7024 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -513,7 +513,8 @@ struct buffer_head *gfs2_log_get_buf(struct gfs2_sbd *sdp)
 static void gfs2_fake_write_endio(struct buffer_head *bh, int uptodate)
 {
 	struct buffer_head *real_bh = bh->b_private;
-	struct gfs2_sbd *sdp = GFS2_SB(real_bh->b_page->mapping->host);
+	struct gfs2_bufdata *bd = real_bh->b_private;
+	struct gfs2_sbd *sdp = bd->bd_gl->gl_sbd;
 
 	end_buffer_write_sync(bh, uptodate);
 	free_buffer_head(bh);
-- 
1.5.1.2

[Cluster-devel] [PATCH 50/51] [DLM] don't overwrite castparam if it's NULL

[swhiteho]

From: Patrick Caulfield <pcaulfie@redhat.com>

If the castaddr passed to the userland API is NULL then don't overwrite the
existing castparam. This allows a different thread to cancel a lock request and
the CANCEL AST gets delivered to the original thread.

bz#306391 (for RHEL4) refers.

Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c
index 2082daf..031229f 100644
--- a/fs/dlm/lock.c
+++ b/fs/dlm/lock.c
@@ -4429,7 +4429,8 @@ int dlm_user_unlock(struct dlm_ls *ls, struct dlm_user_args *ua_tmp,
 
 	if (lvb_in && ua->lksb.sb_lvbptr)
 		memcpy(ua->lksb.sb_lvbptr, lvb_in, DLM_USER_LVB_LEN);
-	ua->castparam = ua_tmp->castparam;
+	if (ua_tmp->castparam)
+		ua->castparam = ua_tmp->castparam;
 	ua->user_lksb = ua_tmp->user_lksb;
 
 	error = set_unlock_args(flags, ua, &args);
@@ -4474,7 +4475,8 @@ int dlm_user_cancel(struct dlm_ls *ls, struct dlm_user_args *ua_tmp,
 		goto out;
 
 	ua = (struct dlm_user_args *)lkb->lkb_astparam;
-	ua->castparam = ua_tmp->castparam;
+	if (ua_tmp->castparam)
+		ua->castparam = ua_tmp->castparam;
 	ua->user_lksb = ua_tmp->user_lksb;
 
 	error = set_unlock_args(flags, ua, &args);
-- 
1.5.1.2

[Cluster-devel] [PATCH 51/51] [DLM] block dlm_recv in recovery transition

[swhiteho]

From: David Teigland <teigland@redhat.com>

Introduce a per-lockspace rwsem that's held in read mode by dlm_recv
threads while working in the dlm.  This allows dlm_recv activity to be
suspended when the lockspace transitions to, from and between recovery
cycles.

The specific bug prompting this change is one where an in-progress
recovery cycle is aborted by a new recovery cycle.  While dlm_recv was
processing a recovery message, the recovery cycle was aborted and
dlm_recoverd began cleaning up.  dlm_recv decremented recover_locks_count
on an rsb after dlm_recoverd had reset it to zero.  This is fixed by
suspending dlm_recv (taking write lock on the rwsem) before aborting the
current recovery.

The transitions to/from normal and recovery modes are simplified by using
this new ability to block dlm_recv.  The switch from normal to recovery
mode means dlm_recv goes from processing locking messages, to saving them
for later, and vice versa.  Races are avoided by blocking dlm_recv when
setting the flag that switches between modes.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>

diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h
index 74901e9..d2fc238 100644
--- a/fs/dlm/dlm_internal.h
+++ b/fs/dlm/dlm_internal.h
@@ -491,6 +491,7 @@ struct dlm_ls {
 	uint64_t		ls_recover_seq;
 	struct dlm_recover	*ls_recover_args;
 	struct rw_semaphore	ls_in_recovery;	/* block local requests */
+	struct rw_semaphore	ls_recv_active;	/* block dlm_recv */
 	struct list_head	ls_requestqueue;/* queue remote requests */
 	struct mutex		ls_requestqueue_mutex;
 	char			*ls_recover_buf;
diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c
index 031229f..3915b8e 100644
--- a/fs/dlm/lock.c
+++ b/fs/dlm/lock.c
@@ -3638,55 +3638,8 @@ static void receive_lookup_reply(struct dlm_ls *ls, struct dlm_message *ms)
 	dlm_put_lkb(lkb);
 }
 
-int dlm_receive_message(struct dlm_header *hd, int nodeid, int recovery)
+static void _receive_message(struct dlm_ls *ls, struct dlm_message *ms)
 {
-	struct dlm_message *ms = (struct dlm_message *) hd;
-	struct dlm_ls *ls;
-	int error = 0;
-
-	if (!recovery)
-		dlm_message_in(ms);
-
-	ls = dlm_find_lockspace_global(hd->h_lockspace);
-	if (!ls) {
-		log_print("drop message %d from %d for unknown lockspace %d",
-			  ms->m_type, nodeid, hd->h_lockspace);
-		return -EINVAL;
-	}
-
-	/* recovery may have just ended leaving a bunch of backed-up requests
-	   in the requestqueue; wait while dlm_recoverd clears them */
-
-	if (!recovery)
-		dlm_wait_requestqueue(ls);
-
-	/* recovery may have just started while there were a bunch of
-	   in-flight requests -- save them in requestqueue to be processed
-	   after recovery.  we can't let dlm_recvd block on the recovery
-	   lock.  if dlm_recoverd is calling this function to clear the
-	   requestqueue, it needs to be interrupted (-EINTR) if another
-	   recovery operation is starting. */
-
-	while (1) {
-		if (dlm_locking_stopped(ls)) {
-			if (recovery) {
-				error = -EINTR;
-				goto out;
-			}
-			error = dlm_add_requestqueue(ls, nodeid, hd);
-			if (error == -EAGAIN)
-				continue;
-			else {
-				error = -EINTR;
-				goto out;
-			}
-		}
-
-		if (dlm_lock_recovery_try(ls))
-			break;
-		schedule();
-	}
-
 	switch (ms->m_type) {
 
 	/* messages sent to a master node */
@@ -3761,17 +3714,90 @@ int dlm_receive_message(struct dlm_header *hd, int nodeid, int recovery)
 		log_error(ls, "unknown message type %d", ms->m_type);
 	}
 
-	dlm_unlock_recovery(ls);
- out:
-	dlm_put_lockspace(ls);
 	dlm_astd_wake();
-	return error;
 }
 
+/* If the lockspace is in recovery mode (locking stopped), then normal
+   messages are saved on the requestqueue for processing after recovery is
+   done.  When not in recovery mode, we wait for dlm_recoverd to drain saved
+   messages off the requestqueue before we process new ones. This occurs right
+   after recovery completes when we transition from saving all messages on
+   requestqueue, to processing all the saved messages, to processing new
+   messages as they arrive. */
 
-/*
- * Recovery related
- */
+static void dlm_receive_message(struct dlm_ls *ls, struct dlm_message *ms,
+				int nodeid)
+{
+	if (dlm_locking_stopped(ls)) {
+		dlm_add_requestqueue(ls, nodeid, (struct dlm_header *) ms);
+	} else {
+		dlm_wait_requestqueue(ls);
+		_receive_message(ls, ms);
+	}
+}
+
+/* This is called by dlm_recoverd to process messages that were saved on
+   the requestqueue. */
+
+void dlm_receive_message_saved(struct dlm_ls *ls, struct dlm_message *ms)
+{
+	_receive_message(ls, ms);
+}
+
+/* This is called by the midcomms layer when something is received for
+   the lockspace.  It could be either a MSG (normal message sent as part of
+   standard locking activity) or an RCOM (recovery message sent as part of
+   lockspace recovery). */
+
+void dlm_receive_buffer(struct dlm_header *hd, int nodeid)
+{
+	struct dlm_message *ms = (struct dlm_message *) hd;
+	struct dlm_rcom *rc = (struct dlm_rcom *) hd;
+	struct dlm_ls *ls;
+	int type = 0;
+
+	switch (hd->h_cmd) {
+	case DLM_MSG:
+		dlm_message_in(ms);
+		type = ms->m_type;
+		break;
+	case DLM_RCOM:
+		dlm_rcom_in(rc);
+		type = rc->rc_type;
+		break;
+	default:
+		log_print("invalid h_cmd %d from %u", hd->h_cmd, nodeid);
+		return;
+	}
+
+	if (hd->h_nodeid != nodeid) {
+		log_print("invalid h_nodeid %d from %d lockspace %x",
+			  hd->h_nodeid, nodeid, hd->h_lockspace);
+		return;
+	}
+
+	ls = dlm_find_lockspace_global(hd->h_lockspace);
+	if (!ls) {
+		log_print("invalid h_lockspace %x from %d cmd %d type %d",
+			  hd->h_lockspace, nodeid, hd->h_cmd, type);
+
+		if (hd->h_cmd == DLM_RCOM && type == DLM_RCOM_STATUS)
+			dlm_send_ls_not_ready(nodeid, rc);
+		return;
+	}
+
+	/* this rwsem allows dlm_ls_stop() to wait for all dlm_recv threads to
+	   be inactive (in this ls) before transitioning to recovery mode */
+
+	down_read(&ls->ls_recv_active);
+	if (hd->h_cmd == DLM_MSG)
+		dlm_receive_message(ls, ms, nodeid);
+	else
+		dlm_receive_rcom(ls, rc, nodeid);
+	up_read(&ls->ls_recv_active);
+
+	dlm_put_lockspace(ls);
+}
 
 static void recover_convert_waiter(struct dlm_ls *ls, struct dlm_lkb *lkb)
 {
diff --git a/fs/dlm/lock.h b/fs/dlm/lock.h
index 1720313..ada0468 100644
--- a/fs/dlm/lock.h
+++ b/fs/dlm/lock.h
@@ -16,7 +16,8 @@
 void dlm_print_rsb(struct dlm_rsb *r);
 void dlm_dump_rsb(struct dlm_rsb *r);
 void dlm_print_lkb(struct dlm_lkb *lkb);
-int dlm_receive_message(struct dlm_header *hd, int nodeid, int recovery);
+void dlm_receive_message_saved(struct dlm_ls *ls, struct dlm_message *ms);
+void dlm_receive_buffer(struct dlm_header *hd, int nodeid);
 int dlm_modes_compat(int mode1, int mode2);
 int dlm_find_rsb(struct dlm_ls *ls, char *name, int namelen,
 	unsigned int flags, struct dlm_rsb **r_ret);
diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c
index 1dc7210..628eaa6 100644
--- a/fs/dlm/lockspace.c
+++ b/fs/dlm/lockspace.c
@@ -519,6 +519,7 @@ static int new_lockspace(char *name, int namelen, void **lockspace,
 	ls->ls_recover_seq = 0;
 	ls->ls_recover_args = NULL;
 	init_rwsem(&ls->ls_in_recovery);
+	init_rwsem(&ls->ls_recv_active);
 	INIT_LIST_HEAD(&ls->ls_requestqueue);
 	mutex_init(&ls->ls_requestqueue_mutex);
 	mutex_init(&ls->ls_clear_proc_locks);
diff --git a/fs/dlm/member.c b/fs/dlm/member.c
index d099775..e9cdcab 100644
--- a/fs/dlm/member.c
+++ b/fs/dlm/member.c
@@ -18,10 +18,6 @@
 #include "rcom.h"
 #include "config.h"
 
-/*
- * Following called by dlm_recoverd thread
- */
-
 static void add_ordered_member(struct dlm_ls *ls, struct dlm_member *new)
 {
 	struct dlm_member *memb = NULL;
@@ -250,18 +246,30 @@ int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv, int *neg_out)
 	return error;
 }
 
-/*
- * Following called from lockspace.c
- */
+/* Userspace guarantees that dlm_ls_stop() has completed on all nodes before
+   dlm_ls_start() is called on any of them to start the new recovery. */
 
 int dlm_ls_stop(struct dlm_ls *ls)
 {
 	int new;
 
 	/*
-	 * A stop cancels any recovery that's in progress (see RECOVERY_STOP,
-	 * dlm_recovery_stopped()) and prevents any new locks from being
-	 * processed (see RUNNING, dlm_locking_stopped()).
+	 * Prevent dlm_recv from being in the middle of something when we do
+	 * the stop.  This includes ensuring dlm_recv isn't processing a
+	 * recovery message (rcom), while dlm_recoverd is aborting and
+	 * resetting things from an in-progress recovery.  i.e. we want
+	 * dlm_recoverd to abort its recovery without worrying about dlm_recv
+	 * processing an rcom at the same time.  Stopping dlm_recv also makes
+	 * it easy for dlm_receive_message() to check locking stopped and add a
+	 * message to the requestqueue without races.
+	 */
+
+	down_write(&ls->ls_recv_active);
+
+	/*
+	 * Abort any recovery that's in progress (see RECOVERY_STOP,
+	 * dlm_recovery_stopped()) and tell any other threads running in the
+	 * dlm to quit any processing (see RUNNING, dlm_locking_stopped()).
 	 */
 
 	spin_lock(&ls->ls_recover_lock);
@@ -271,8 +279,14 @@ int dlm_ls_stop(struct dlm_ls *ls)
 	spin_unlock(&ls->ls_recover_lock);
 
 	/*
+	 * Let dlm_recv run again, now any normal messages will be saved on the
+	 * requestqueue for later.
+	 */
+
+	up_write(&ls->ls_recv_active);
+
+	/*
 	 * This in_recovery lock does two things:
-	 *
 	 * 1) Keeps this function from returning until all threads are out
 	 *    of locking routines and locking is truely stopped.
 	 * 2) Keeps any new requests from being processed until it's unlocked
@@ -284,9 +298,8 @@ int dlm_ls_stop(struct dlm_ls *ls)
 
 	/*
 	 * The recoverd suspend/resume makes sure that dlm_recoverd (if
-	 * running) has noticed the clearing of RUNNING above and quit
-	 * processing the previous recovery.  This will be true for all nodes
-	 * before any nodes start the new recovery.
+	 * running) has noticed RECOVERY_STOP above and quit processing the
+	 * previous recovery.
 	 */
 
 	dlm_recoverd_suspend(ls);
diff --git a/fs/dlm/midcomms.c b/fs/dlm/midcomms.c
index a5126e0..f8c69dd 100644
--- a/fs/dlm/midcomms.c
+++ b/fs/dlm/midcomms.c
@@ -2,7 +2,7 @@
 *******************************************************************************
 **
 **  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
-**  Copyright (C) 2004-2005 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2004-2007 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -27,7 +27,6 @@
 #include "dlm_internal.h"
 #include "lowcomms.h"
 #include "config.h"
-#include "rcom.h"
 #include "lock.h"
 #include "midcomms.h"
 
@@ -117,19 +116,7 @@ int dlm_process_incoming_buffer(int nodeid, const void *base,
 		offset &= (limit - 1);
 		len -= msglen;
 
-		switch (msg->h_cmd) {
-		case DLM_MSG:
-			dlm_receive_message(msg, nodeid, 0);
-			break;
-
-		case DLM_RCOM:
-			dlm_receive_rcom(msg, nodeid);
-			break;
-
-		default:
-			log_print("unknown msg type %x from %u: %u %u %u %u",
-				  msg->h_cmd, nodeid, msglen, len, offset, ret);
-		}
+		dlm_receive_buffer(msg, nodeid);
 	}
 
 	if (msg != (struct dlm_header *) __tmp)
diff --git a/fs/dlm/rcom.c b/fs/dlm/rcom.c
index 188b91c..ae2fd97 100644
--- a/fs/dlm/rcom.c
+++ b/fs/dlm/rcom.c
@@ -2,7 +2,7 @@
 *******************************************************************************
 **
 **  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
-**  Copyright (C) 2005 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2005-2007 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -386,7 +386,10 @@ static void receive_rcom_lock_reply(struct dlm_ls *ls, struct dlm_rcom *rc_in)
 	dlm_recover_process_copy(ls, rc_in);
 }
 
-static int send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in)
+/* If the lockspace doesn't exist then still send a status message
+   back; it's possible that it just doesn't have its global_id yet. */
+
+int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in)
 {
 	struct dlm_rcom *rc;
 	struct rcom_config *rf;
@@ -446,28 +449,11 @@ static int is_old_reply(struct dlm_ls *ls, struct dlm_rcom *rc)
 	return rv;
 }
 
-/* Called by dlm_recvd; corresponds to dlm_receive_message() but special
+/* Called by dlm_recv; corresponds to dlm_receive_message() but special
    recovery-only comms are sent through here. */
 
-void dlm_receive_rcom(struct dlm_header *hd, int nodeid)
+void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid)
 {
-	struct dlm_rcom *rc = (struct dlm_rcom *) hd;
-	struct dlm_ls *ls;
-
-	dlm_rcom_in(rc);
-
-	/* If the lockspace doesn't exist then still send a status message
-	   back; it's possible that it just doesn't have its global_id yet. */
-
-	ls = dlm_find_lockspace_global(hd->h_lockspace);
-	if (!ls) {
-		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;
-	}
-
 	if (dlm_recovery_stopped(ls) && (rc->rc_type != DLM_RCOM_STATUS)) {
 		log_debug(ls, "ignoring recovery message %x from %d",
 			  rc->rc_type, nodeid);
@@ -477,12 +463,6 @@ void dlm_receive_rcom(struct dlm_header *hd, int nodeid)
 	if (is_old_reply(ls, rc))
 		goto out;
 
-	if (nodeid != rc->rc_header.h_nodeid) {
-		log_error(ls, "bad rcom nodeid %d from %d",
-			  rc->rc_header.h_nodeid, nodeid);
-		goto out;
-	}
-
 	switch (rc->rc_type) {
 	case DLM_RCOM_STATUS:
 		receive_rcom_status(ls, rc);
@@ -520,6 +500,6 @@ void dlm_receive_rcom(struct dlm_header *hd, int nodeid)
 		DLM_ASSERT(0, printk("rc_type=%x\n", rc->rc_type););
 	}
  out:
-	dlm_put_lockspace(ls);
+	return;
 }
 
diff --git a/fs/dlm/rcom.h b/fs/dlm/rcom.h
index d798432..b09abd2 100644
--- a/fs/dlm/rcom.h
+++ b/fs/dlm/rcom.h
@@ -2,7 +2,7 @@
 *******************************************************************************
 **
 **  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
-**  Copyright (C) 2005 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2005-2007 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -18,7 +18,8 @@ int dlm_rcom_status(struct dlm_ls *ls, int nodeid);
 int dlm_rcom_names(struct dlm_ls *ls, int nodeid, char *last_name,int last_len);
 int dlm_send_rcom_lookup(struct dlm_rsb *r, int dir_nodeid);
 int dlm_send_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb);
-void dlm_receive_rcom(struct dlm_header *hd, int nodeid);
+void dlm_receive_rcom(struct dlm_ls *ls, struct dlm_rcom *rc, int nodeid);
+int dlm_send_ls_not_ready(int nodeid, struct dlm_rcom *rc_in);
 
 #endif
 
diff --git a/fs/dlm/recoverd.c b/fs/dlm/recoverd.c
index 6657599..4b89e20 100644
--- a/fs/dlm/recoverd.c
+++ b/fs/dlm/recoverd.c
@@ -24,19 +24,28 @@
 
 
 /* If the start for which we're re-enabling locking (seq) has been superseded
-   by a newer stop (ls_recover_seq), we need to leave locking disabled. */
+   by a newer stop (ls_recover_seq), we need to leave locking disabled.
+
+   We suspend dlm_recv threads here to avoid the race where dlm_recv a) sees
+   locking stopped and b) adds a message to the requestqueue, but dlm_recoverd
+   enables locking and clears the requestqueue between a and b. */
 
 static int enable_locking(struct dlm_ls *ls, uint64_t seq)
 {
 	int error = -EINTR;
 
+	down_write(&ls->ls_recv_active);
+
 	spin_lock(&ls->ls_recover_lock);
 	if (ls->ls_recover_seq == seq) {
 		set_bit(LSFL_RUNNING, &ls->ls_flags);
+		/* unblocks processes waiting to enter the dlm */
 		up_write(&ls->ls_in_recovery);
 		error = 0;
 	}
 	spin_unlock(&ls->ls_recover_lock);
+
+	up_write(&ls->ls_recv_active);
 	return error;
 }
 
diff --git a/fs/dlm/requestqueue.c b/fs/dlm/requestqueue.c
index 65008d7..0de04f1 100644
--- a/fs/dlm/requestqueue.c
+++ b/fs/dlm/requestqueue.c
@@ -1,7 +1,7 @@
 /******************************************************************************
 *******************************************************************************
 **
-**  Copyright (C) 2005 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2005-2007 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -20,7 +20,7 @@
 struct rq_entry {
 	struct list_head list;
 	int nodeid;
-	char request[1];
+	char request[0];
 };
 
 /*
@@ -30,42 +30,39 @@ struct rq_entry {
  * lockspace is enabled on some while still suspended on others.
  */
 
-int dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd)
+void dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd)
 {
 	struct rq_entry *e;
 	int length = hd->h_length;
-	int rv = 0;
 
 	e = kmalloc(sizeof(struct rq_entry) + length, GFP_KERNEL);
 	if (!e) {
-		log_print("dlm_add_requestqueue: out of memory\n");
-		return 0;
+		log_print("dlm_add_requestqueue: out of memory len %d", length);
+		return;
 	}
 
 	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);
-	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;
-	}
+	list_add_tail(&e->list, &ls->ls_requestqueue);
 	mutex_unlock(&ls->ls_requestqueue_mutex);
-	return rv;
 }
 
+/*
+ * Called by dlm_recoverd to process normal messages saved while recovery was
+ * happening.  Normal locking has been enabled before this is called.  dlm_recv
+ * upon receiving a message, will wait for all saved messages to be drained
+ * here before processing the message it got.  If a new dlm_ls_stop() arrives
+ * while we're processing these saved messages, it may block trying to suspend
+ * dlm_recv if dlm_recv is waiting for us in dlm_wait_requestqueue.  In that
+ * case, we don't abort since locking_stopped is still 0.  If dlm_recv is not
+ * waiting for us, then this processing may be aborted due to locking_stopped.
+ */
+
 int dlm_process_requestqueue(struct dlm_ls *ls)
 {
 	struct rq_entry *e;
-	struct dlm_header *hd;
 	int error = 0;
 
 	mutex_lock(&ls->ls_requestqueue_mutex);
@@ -79,14 +76,7 @@ int dlm_process_requestqueue(struct dlm_ls *ls)
 		e = list_entry(ls->ls_requestqueue.next, struct rq_entry, list);
 		mutex_unlock(&ls->ls_requestqueue_mutex);
 
-		hd = (struct dlm_header *) e->request;
-		error = dlm_receive_message(hd, e->nodeid, 1);
-
-		if (error == -EINTR) {
-			/* entry is left on requestqueue */
-			log_debug(ls, "process_requestqueue abort eintr");
-			break;
-		}
+		dlm_receive_message_saved(ls, (struct dlm_message *)e->request);
 
 		mutex_lock(&ls->ls_requestqueue_mutex);
 		list_del(&e->list);
@@ -106,10 +96,12 @@ int dlm_process_requestqueue(struct dlm_ls *ls)
 
 /*
  * After recovery is done, locking is resumed and dlm_recoverd takes all the
- * saved requests and processes them as they would have been by dlm_recvd.  At
- * the same time, dlm_recvd will start receiving new requests from remote
- * nodes.  We want to delay dlm_recvd processing new requests until
- * dlm_recoverd has finished processing the old saved requests.
+ * saved requests and processes them as they would have been by dlm_recv.  At
+ * the same time, dlm_recv will start receiving new requests from remote nodes.
+ * We want to delay dlm_recv processing new requests until dlm_recoverd has
+ * finished processing the old saved requests.  We don't check for locking
+ * stopped here because dlm_ls_stop won't stop locking until it's suspended us
+ * (dlm_recv).
  */
 
 void dlm_wait_requestqueue(struct dlm_ls *ls)
@@ -118,8 +110,6 @@ void dlm_wait_requestqueue(struct dlm_ls *ls)
 		mutex_lock(&ls->ls_requestqueue_mutex);
 		if (list_empty(&ls->ls_requestqueue))
 			break;
-		if (dlm_locking_stopped(ls))
-			break;
 		mutex_unlock(&ls->ls_requestqueue_mutex);
 		schedule();
 	}
diff --git a/fs/dlm/requestqueue.h b/fs/dlm/requestqueue.h
index 6a53ea0..aba34fc 100644
--- a/fs/dlm/requestqueue.h
+++ b/fs/dlm/requestqueue.h
@@ -1,7 +1,7 @@
 /******************************************************************************
 *******************************************************************************
 **
-**  Copyright (C) 2005 Red Hat, Inc.  All rights reserved.
+**  Copyright (C) 2005-2007 Red Hat, Inc.  All rights reserved.
 **
 **  This copyrighted material is made available to anyone wishing to use,
 **  modify, copy, or redistribute it subject to the terms and conditions
@@ -13,7 +13,7 @@
 #ifndef __REQUESTQUEUE_DOT_H__
 #define __REQUESTQUEUE_DOT_H__
 
-int dlm_add_requestqueue(struct dlm_ls *ls, int nodeid, struct dlm_header *hd);
+void 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);
-- 
1.5.1.2

[Cluster-devel] [GFS2/DLM] Pull request

[Steven Whitehouse]

Hi,

There have been no changes since I posted the patches for review a few
days ago. Please consider pulling the following changes from the
GFS2/DLM -nmw git tree:

The following changes since commit bbf25010f1a6b761914430f5fca081ec8c7accd1:
  Linus Torvalds (1):
        Linux 2.6.23

are found in the git repository at:

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

Abhijith Das (5):
      [GFS2] Force unstuff of hidden quota inode
      [GFS2] Fix quota do_list operation hang
      [GFS2] Wendy's dump lockname in hex & fix glock dump
      [GFS2] panic after can't parse mount arguments
      [GFS2] flocks from same process trip kernel BUG at fs/gfs2/glock.c:1118!

Benjamin Marzinski (3):
      [GFS2] Add NULL entry to token table
      [GFS2] delay glock demote for a minimum hold time
      [GFS2] Alternate gfs2_iget to avoid looking up inodes being freed

Bob Peterson (9):
      [GFS2] Move some code inside the log lock
      [GFS2] Revert part of earlier log.c changes
      [GFS2] Prevent infinite loop in try_rgrp_unlink()
      [GFS2] Detach buf data during in-place writeback
      [GFS2] invalid metadata block - REVISED
      [GFS2] Ensure journal file cache is flushed after recovery
      [GFS2] Patch to protect sd_log_num_jdata
      [GFS2] Fix ordering of dirty/journal for ordered buffer unstuffing
      [GFS2] GFS2: chmod hung - fix race in thread creation

David Teigland (1):
      [DLM] block dlm_recv in recovery transition

Denis Cheng (7):
      [GFS2] use an temp variable to reduce a spin_unlock
      [GFS2] mark struct *_operations const
      [GFS2] use the declaration of gfs2_dops in the header file instead
      [GFS2] use list_for_each_entry instead
      [GFS2] unneeded typecast
      [GFS2] better code for translating characters
      [GFS2] fixed a NULL pointer assignment BUG

Jesper Juhl (1):
      [GFS2] Clean up duplicate includes in fs/gfs2/

Josef Whiter (1):
      [GFS2] Fix calculation of demote state

Patrick Caulfield (3):
      [DLM] Fix lowcomms socket closing
      [DLM] Make dlm_sendd cond_resched more
      [DLM] don't overwrite castparam if it's NULL

Steve French (1):
      [GFS2] GFS2 not checking pointer on create when running under nfsd

Steven Whitehouse (16):
      [GFS2] Fix two races relating to glock callbacks
      [GFS2] Fix an oops in glock dumping
      [GFS2] Reduce number of gfs2_scand processes to one
      [GFS2] Clean up invalidatepage/releasepage
      [GFS2] Add a missing gfs2_trans_add_bh()
      [GFS2] Correct lock ordering in unlink
      [GFS2] Introduce gfs2_remove_from_ail
      [GFS2] Don't mark jdata dirty in gfs2_unstuffer_page()
      [GFS2] Move pin/unpin into lops.c, clean up locking
      [GFS2] Clean up ordered write code
      [GFS2] Replace revoke structure with bufdata structure
      [GFS2] Use slab operations for all gfs2_bufdata allocations
      [GFS2] Clean up gfs2_trans_add_revoke()
      [GFS2] Clean up journaled data writing
      [GFS2] Don't try to remove buffers that don't exist
      [GFS2] Get superblock a different way

Wendy Cheng (4):
      [GFS2] Reduce truncate IO traffic
      [GFS2] fix inode meta data corruption
      [GFS2] Move inode deletion out of blocking_cb
      [GFS2] Data corruption fix

 fs/dlm/dlm_internal.h          |    1 +
 fs/dlm/lock.c                  |  142 +++++++-----
 fs/dlm/lock.h                  |    3 +-
 fs/dlm/lockspace.c             |    1 +
 fs/dlm/lowcomms.c              |   23 +--
 fs/dlm/member.c                |   41 +++--
 fs/dlm/midcomms.c              |   17 +--
 fs/dlm/rcom.c                  |   36 +---
 fs/dlm/rcom.h                  |    5 +-
 fs/dlm/recoverd.c              |   11 +-
 fs/dlm/requestqueue.c          |   58 ++---
 fs/dlm/requestqueue.h          |    4 +-
 fs/gfs2/bmap.c                 |   35 +++-
 fs/gfs2/daemon.c               |   24 --
 fs/gfs2/daemon.h               |    1 -
 fs/gfs2/dir.c                  |    3 +-
 fs/gfs2/eaops.c                |    8 +-
 fs/gfs2/eaops.h                |    4 +-
 fs/gfs2/glock.c                |  293 ++++++++++++++++---------
 fs/gfs2/glock.h                |    5 +-
 fs/gfs2/glops.c                |   24 +--
 fs/gfs2/incore.h               |   31 ++--
 fs/gfs2/inode.c                |   78 ++++++-
 fs/gfs2/inode.h                |    3 +-
 fs/gfs2/locking/dlm/lock_dlm.h |    1 -
 fs/gfs2/locking/dlm/plock.c    |   11 +-
 fs/gfs2/locking/dlm/thread.c   |   20 +-
 fs/gfs2/locking/nolock/main.c  |    1 -
 fs/gfs2/log.c                  |  230 +++++++++++++-------
 fs/gfs2/log.h                  |    2 +
 fs/gfs2/lops.c                 |  470 ++++++++++++++++++++--------------------
 fs/gfs2/main.c                 |    3 +
 fs/gfs2/meta_io.c              |  136 +++---------
 fs/gfs2/meta_io.h              |    6 +-
 fs/gfs2/mount.c                |    5 +-
 fs/gfs2/ops_address.c          |  146 ++++---------
 fs/gfs2/ops_export.c           |    2 +-
 fs/gfs2/ops_file.c             |   13 +-
 fs/gfs2/ops_fstype.c           |   40 ++--
 fs/gfs2/ops_inode.c            |   38 +++-
 fs/gfs2/ops_super.c            |   14 +-
 fs/gfs2/quota.c                |   13 +
 fs/gfs2/recovery.c             |    2 +-
 fs/gfs2/rgrp.c                 |   39 ++--
 fs/gfs2/super.c                |    1 -
 fs/gfs2/sys.c                  |    2 -
 fs/gfs2/trans.c                |   22 +-
 fs/gfs2/trans.h                |    2 +-
 include/linux/gfs2_ondisk.h    |   30 +++-
 49 files changed, 1139 insertions(+), 961 deletions(-)

[Cluster-devel] [GFS2/DLM] Pre-pull Patch Posting

[swhiteho]

Hi,

This is the current set of patches from the GFS2/DLM -nmw git tree which
are pending inclusion in the current merge window. There are quite a
few mainly as I was a bit lazy in pushing some of the smaller bug fixes
before.

There are a couple of things in -mm which depend upon changes in the
current GFS2 tree, so my plan is to request a merge very shortly to
leave time for those other items to be merged later.

All the changes here only relate to GFS2 and/or DLM, there are no
changes which affect any of the core code. Most of the patches are
in fatc bug fixes and/or cleanups. The only "new" feature is GFS2 is
the nanosecond timestamps feature.

Steve.