Merge tag 'gfs2-for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2

Pull gfs2 updates from Andreas Gruenbacher:

 - Use iomap for non-journaled buffered I/O. This largely eliminates
   buffer heads on filesystems where the block size matches the page
   size. Many thanks to Christoph Hellwig for this patch!

 - Fixes for some more journaled data filesystem bugs, found by running
   xfstests with data journaling on for all files (chattr +j $MNT) (Bob
   Peterson)

 - gfs2_evict_inode refactoring (Bob Peterson)

 - Use the statfs data in the journal during recovery instead of reading
   it in from the local statfs inodes (Abhi Das)

 - Several other minor fixes by various people

* tag 'gfs2-for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2: (30 commits)
  gfs2: Recover statfs info in journal head
  gfs2: lookup local statfs inodes prior to journal recovery
  gfs2: Add fields for statfs info in struct gfs2_log_header_host
  gfs2: Ignore subsequent errors after withdraw in rgrp_go_sync
  gfs2: Eliminate gl_vm
  gfs2: Only access gl_delete for iopen glocks
  gfs2: Fix comments to glock_hash_walk
  gfs2: eliminate GLF_QUEUED flag in favor of list_empty(gl_holders)
  gfs2: Ignore journal log writes for jdata holes
  gfs2: simplify gfs2_block_map
  gfs2: Only set PageChecked if we have a transaction
  gfs2: don't lock sd_ail_lock in gfs2_releasepage
  gfs2: make gfs2_ail1_empty_one return the count of active items
  gfs2: Wipe jdata and ail1 in gfs2_journal_wipe, formerly gfs2_meta_wipe
  gfs2: enhance log_blocks trace point to show log blocks free
  gfs2: add missing log_blocks trace points in gfs2_write_revokes
  gfs2: rename gfs2_write_full_page to gfs2_write_jdata_page, remove parm
  gfs2: add validation checks for size of superblock
  gfs2: use-after-free in sysfs deregistration
  gfs2: Fix NULL pointer dereference in gfs2_rgrp_dump
  ...

fs/gfs2/aops.c

@@ -91,22 +91,13 @@ static int gfs2_writepage(struct page *page, struct writeback_control *wbc)
 	struct inode *inode = page->mapping->host;
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
-	loff_t i_size = i_size_read(inode);
-	pgoff_t end_index = i_size >> PAGE_SHIFT;
-	unsigned offset;
+	struct iomap_writepage_ctx wpc = { };
 
 	if (gfs2_assert_withdraw(sdp, gfs2_glock_is_held_excl(ip->i_gl)))
 		goto out;
 	if (current->journal_info)
 		goto redirty;
-	/* Is the page fully outside i_size? (truncate in progress) */
-	offset = i_size & (PAGE_SIZE-1);
-	if (page->index > end_index || (page->index == end_index && !offset)) {
-		page->mapping->a_ops->invalidatepage(page, 0, PAGE_SIZE);
-		goto out;
-	}
-
-	return nobh_writepage(page, gfs2_get_block_noalloc, wbc);
+	return iomap_writepage(page, wbc, &wpc, &gfs2_writeback_ops);
 
 redirty:
 	redirty_page_for_writepage(wbc, page);
@@ -115,10 +106,15 @@ static int gfs2_writepage(struct page *page, struct writeback_control *wbc)
 	return 0;
 }
 
-/* This is the same as calling block_write_full_page, but it also
+/**
+ * gfs2_write_jdata_page - gfs2 jdata-specific version of block_write_full_page
+ * @page: The page to write
+ * @wbc: The writeback control
+ *
+ * This is the same as calling block_write_full_page, but it also
  * writes pages outside of i_size
  */
-static int gfs2_write_full_page(struct page *page, get_block_t *get_block,
+static int gfs2_write_jdata_page(struct page *page,
 				 struct writeback_control *wbc)
 {
 	struct inode * const inode = page->mapping->host;
@@ -137,7 +133,7 @@ static int gfs2_write_full_page(struct page *page, get_block_t *get_block,
 	if (page->index == end_index && offset)
 		zero_user_segment(page, offset, PAGE_SIZE);
 
-	return __block_write_full_page(inode, page, get_block, wbc,
+	return __block_write_full_page(inode, page, gfs2_get_block_noalloc, wbc,
 				       end_buffer_async_write);
 }
 
@@ -166,7 +162,7 @@ static int __gfs2_jdata_writepage(struct page *page, struct writeback_control *w
 		}
 		gfs2_page_add_databufs(ip, page, 0, sdp->sd_vfs->s_blocksize);
 	}
-	return gfs2_write_full_page(page, gfs2_get_block_noalloc, wbc);
+	return gfs2_write_jdata_page(page, wbc);
 }
 
 /**
@@ -208,7 +204,8 @@ static int gfs2_writepages(struct address_space *mapping,
 			   struct writeback_control *wbc)
 {
 	struct gfs2_sbd *sdp = gfs2_mapping2sbd(mapping);
-	int ret = mpage_writepages(mapping, wbc, gfs2_get_block_noalloc);
+	struct iomap_writepage_ctx wpc = { };
+	int ret;
 
 	/*
 	 * Even if we didn't write any pages here, we might still be holding
@@ -216,9 +213,9 @@ static int gfs2_writepages(struct address_space *mapping,
 	 * want balance_dirty_pages() to loop indefinitely trying to write out
 	 * pages held in the ail that it can't find.
 	 */
+	ret = iomap_writepages(mapping, wbc, &wpc, &gfs2_writeback_ops);
 	if (ret == 0)
 		set_bit(SDF_FORCE_AIL_FLUSH, &sdp->sd_flags);
-
 	return ret;
 }
 
@@ -470,12 +467,13 @@ static int stuffed_readpage(struct gfs2_inode *ip, struct page *page)
 
 static int __gfs2_readpage(void *file, struct page *page)
 {
-	struct gfs2_inode *ip = GFS2_I(page->mapping->host);
-	struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host);
+	struct inode *inode = page->mapping->host;
+	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	int error;
 
-	if (i_blocksize(page->mapping->host) == PAGE_SIZE &&
-	    !page_has_buffers(page)) {
+	if (!gfs2_is_jdata(ip) ||
+	    (i_blocksize(inode) == PAGE_SIZE && !page_has_buffers(page))) {
 		error = iomap_readpage(page, &gfs2_iomap_ops);
 	} else if (gfs2_is_stuffed(ip)) {
 		error = stuffed_readpage(ip, page);
@@ -563,8 +561,12 @@ static void gfs2_readahead(struct readahead_control *rac)
 	struct inode *inode = rac->mapping->host;
 	struct gfs2_inode *ip = GFS2_I(inode);
 
-	if (!gfs2_is_stuffed(ip))
+	if (gfs2_is_stuffed(ip))
+		;
+	else if (gfs2_is_jdata(ip))
 		mpage_readahead(rac, gfs2_block_map);
+	else
+		iomap_readahead(rac, &gfs2_iomap_ops);
 }
 
 /**
@@ -621,6 +623,7 @@ void adjust_fs_space(struct inode *inode)
  
 static int jdata_set_page_dirty(struct page *page)
 {
+	if (current->journal_info)
 		SetPageChecked(page);
 	return __set_page_dirty_buffers(page);
 }
@@ -663,8 +666,11 @@ static void gfs2_discard(struct gfs2_sbd *sdp, struct buffer_head *bh)
 	if (bd) {
 		if (!list_empty(&bd->bd_list) && !buffer_pinned(bh))
 			list_del_init(&bd->bd_list);
-		else
+		else {
+			spin_lock(&sdp->sd_ail_lock);
 			gfs2_remove_from_journal(bh, REMOVE_JDATA);
+			spin_unlock(&sdp->sd_ail_lock);
+		}
 	}
 	bh->b_bdev = NULL;
 	clear_buffer_mapped(bh);
@@ -736,7 +742,6 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 	 */
 
 	gfs2_log_lock(sdp);
-	spin_lock(&sdp->sd_ail_lock);
 	head = bh = page_buffers(page);
 	do {
 		if (atomic_read(&bh->b_count))
@@ -748,7 +753,6 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 			goto cannot_release;
 		bh = bh->b_this_page;
 	} while(bh != head);
-	spin_unlock(&sdp->sd_ail_lock);
 
 	head = bh = page_buffers(page);
 	do {
@@ -774,7 +778,6 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 	return try_to_free_buffers(page);
 
 cannot_release:
-	spin_unlock(&sdp->sd_ail_lock);
 	gfs2_log_unlock(sdp);
 	return 0;
 }
@@ -784,12 +787,13 @@ static const struct address_space_operations gfs2_aops = {
 	.writepages = gfs2_writepages,
 	.readpage = gfs2_readpage,
 	.readahead = gfs2_readahead,
+	.set_page_dirty = iomap_set_page_dirty,
+	.releasepage = iomap_releasepage,
+	.invalidatepage = iomap_invalidatepage,
 	.bmap = gfs2_bmap,
-	.invalidatepage = gfs2_invalidatepage,
-	.releasepage = gfs2_releasepage,
 	.direct_IO = noop_direct_IO,
-	.migratepage = buffer_migrate_page,
-	.is_partially_uptodate = block_is_partially_uptodate,
+	.migratepage = iomap_migrate_page,
+	.is_partially_uptodate = iomap_is_partially_uptodate,
 	.error_remove_page = generic_error_remove_page,
 };
 

fs/gfs2/bmap.c

@@ -56,7 +56,6 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh,
 			       u64 block, struct page *page)
 {
 	struct inode *inode = &ip->i_inode;
-	struct buffer_head *bh;
 	int release = 0;
 
 	if (!page || page->index) {
@@ -80,20 +79,21 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh,
 		SetPageUptodate(page);
 	}
 
+	if (gfs2_is_jdata(ip)) {
+		struct buffer_head *bh;
+
 		if (!page_has_buffers(page))
 			create_empty_buffers(page, BIT(inode->i_blkbits),
 					     BIT(BH_Uptodate));
 
 		bh = page_buffers(page);
-
 		if (!buffer_mapped(bh))
 			map_bh(bh, inode->i_sb, block);
 
 		set_buffer_uptodate(bh);
-	if (gfs2_is_jdata(ip))
 		gfs2_trans_add_data(ip->i_gl, bh);
-	else {
-		mark_buffer_dirty(bh);
+	} else {
+		set_page_dirty(page);
 		gfs2_ordered_add_inode(ip);
 	}
 
@@ -1158,6 +1158,7 @@ static int gfs2_iomap_begin(struct inode *inode, loff_t pos, loff_t length,
 	struct metapath mp = { .mp_aheight = 1, };
 	int ret;
 
+	if (gfs2_is_jdata(ip))
 		iomap->flags |= IOMAP_F_BUFFER_HEAD;
 
 	trace_gfs2_iomap_start(ip, pos, length, flags);
@@ -1291,6 +1292,7 @@ int gfs2_block_map(struct inode *inode, sector_t lblock,
 	loff_t length = bh_map->b_size;
 	struct metapath mp = { .mp_aheight = 1, };
 	struct iomap iomap = { };
+	int flags = create ? IOMAP_WRITE : 0;
 	int ret;
 
 	clear_buffer_mapped(bh_map);
@@ -1298,15 +1300,14 @@ int gfs2_block_map(struct inode *inode, sector_t lblock,
 	clear_buffer_boundary(bh_map);
 	trace_gfs2_bmap(ip, bh_map, lblock, create, 1);
 
-	if (create) {
-		ret = gfs2_iomap_get(inode, pos, length, IOMAP_WRITE, &iomap, &mp);
-		if (!ret && iomap.type == IOMAP_HOLE)
+	ret = gfs2_iomap_get(inode, pos, length, flags, &iomap, &mp);
+	if (!ret && iomap.type == IOMAP_HOLE) {
+		if (create)
 			ret = gfs2_iomap_alloc(inode, &iomap, &mp);
-		release_metapath(&mp);
-	} else {
-		ret = gfs2_iomap_get(inode, pos, length, 0, &iomap, &mp);
-		release_metapath(&mp);
+		else
+			ret = -ENODATA;
 	}
+	release_metapath(&mp);
 	if (ret)
 		goto out;
 
@@ -2518,3 +2519,26 @@ int __gfs2_punch_hole(struct file *file, loff_t offset, loff_t length)
 		gfs2_trans_end(sdp);
 	return error;
 }
+
+static int gfs2_map_blocks(struct iomap_writepage_ctx *wpc, struct inode *inode,
+		loff_t offset)
+{
+	struct metapath mp = { .mp_aheight = 1, };
+	int ret;
+
+	if (WARN_ON_ONCE(gfs2_is_stuffed(GFS2_I(inode))))
+		return -EIO;
+
+	if (offset >= wpc->iomap.offset &&
+	    offset < wpc->iomap.offset + wpc->iomap.length)
+		return 0;
+
+	memset(&wpc->iomap, 0, sizeof(wpc->iomap));
+	ret = gfs2_iomap_get(inode, offset, INT_MAX, 0, &wpc->iomap, &mp);
+	release_metapath(&mp);
+	return ret;
+}
+
+const struct iomap_writeback_ops gfs2_writeback_ops = {
+	.map_blocks		= gfs2_map_blocks,
+};

fs/gfs2/bmap.h

@@ -44,6 +44,7 @@ static inline void gfs2_write_calc_reserv(const struct gfs2_inode *ip,
 }
 
 extern const struct iomap_ops gfs2_iomap_ops;
+extern const struct iomap_writeback_ops gfs2_writeback_ops;
 
 extern int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page);
 extern int gfs2_block_map(struct inode *inode, sector_t lblock,

fs/gfs2/glock.c

@@ -270,7 +270,12 @@ static void __gfs2_glock_put(struct gfs2_glock *gl)
 	gfs2_glock_remove_from_lru(gl);
 	spin_unlock(&gl->gl_lockref.lock);
 	GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders));
-	GLOCK_BUG_ON(gl, mapping && mapping->nrpages && !gfs2_withdrawn(sdp));
+	if (mapping) {
+		truncate_inode_pages_final(mapping);
+		if (!gfs2_withdrawn(sdp))
+			GLOCK_BUG_ON(gl, mapping->nrpages ||
+				     mapping->nrexceptional);
+	}
 	trace_gfs2_glock_put(gl);
 	sdp->sd_lockstruct.ls_ops->lm_put_lock(gl);
 }
@@ -453,9 +458,6 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state)
 		else
 			gl->gl_lockref.count--;
 	}
-	if (held1 && held2 && list_empty(&gl->gl_holders))
-		clear_bit(GLF_QUEUED, &gl->gl_flags);
-
 	if (new_state != gl->gl_target)
 		/* shorten our minimum hold time */
 		gl->gl_hold_time = max(gl->gl_hold_time - GL_GLOCK_HOLD_DECR,
@@ -1049,6 +1051,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
 	gl->gl_object = NULL;
 	gl->gl_hold_time = GL_GLOCK_DFT_HOLD;
 	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
+	if (gl->gl_name.ln_type == LM_TYPE_IOPEN)
 		INIT_DELAYED_WORK(&gl->gl_delete, delete_work_func);
 
 	mapping = gfs2_glock2aspace(gl);
@@ -1345,7 +1348,6 @@ __acquires(&gl->gl_lockref.lock)
 		if (unlikely((gh->gh_flags & LM_FLAG_PRIORITY) && !insert_pt))
 			insert_pt = &gh2->gh_list;
 	}
-	set_bit(GLF_QUEUED, &gl->gl_flags);
 	trace_gfs2_glock_queue(gh, 1);
 	gfs2_glstats_inc(gl, GFS2_LKS_QCOUNT);
 	gfs2_sbstats_inc(gl, GFS2_LKS_QCOUNT);
@@ -1646,16 +1648,15 @@ void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state)
 	unsigned long now = jiffies;
 
 	gfs2_glock_hold(gl);
+	spin_lock(&gl->gl_lockref.lock);
 	holdtime = gl->gl_tchange + gl->gl_hold_time;
-	if (test_bit(GLF_QUEUED, &gl->gl_flags) &&
+	if (!list_empty(&gl->gl_holders) &&
 	    gl->gl_name.ln_type == LM_TYPE_INODE) {
 		if (time_before(now, holdtime))
 			delay = holdtime - now;
 		if (test_bit(GLF_REPLY_PENDING, &gl->gl_flags))
 			delay = gl->gl_hold_time;
 	}
-
-	spin_lock(&gl->gl_lockref.lock);
 	handle_callback(gl, state, delay, true);
 	__gfs2_glock_queue_work(gl, delay);
 	spin_unlock(&gl->gl_lockref.lock);
@@ -1842,10 +1843,9 @@ static struct shrinker glock_shrinker = {
 };
 
 /**
- * examine_bucket - Call a function for glock in a hash bucket
+ * glock_hash_walk - Call a function for glock in a hash bucket
  * @examiner: the function
  * @sdp: the filesystem
- * @bucket: the bucket
  *
  * Note that the function can be called multiple times on the same
  * object.  So the user must ensure that the function can cope with
@@ -1901,10 +1901,12 @@ bool gfs2_delete_work_queued(const struct gfs2_glock *gl)
 
 static void flush_delete_work(struct gfs2_glock *gl)
 {
+	if (gl->gl_name.ln_type == LM_TYPE_IOPEN) {
 		if (cancel_delayed_work(&gl->gl_delete)) {
 			queue_delayed_work(gfs2_delete_workqueue,
 					   &gl->gl_delete, 0);
 		}
+	}
 	gfs2_glock_queue_work(gl, 0);
 }
 
@@ -2100,7 +2102,7 @@ static const char *gflags2str(char *buf, const struct gfs2_glock *gl)
 		*p++ = 'I';
 	if (test_bit(GLF_FROZEN, gflags))
 		*p++ = 'F';
-	if (test_bit(GLF_QUEUED, gflags))
+	if (!list_empty(&gl->gl_holders))
 		*p++ = 'q';
 	if (test_bit(GLF_LRU, gflags))
 		*p++ = 'L';
@@ -2415,7 +2417,7 @@ static const struct seq_operations gfs2_glstats_seq_ops = {
 	.show  = gfs2_glstats_seq_show,
 };
 
-static const struct seq_operations gfs2_sbstats_seq_ops = {
+static const struct seq_operations gfs2_sbstats_sops = {
 	.start = gfs2_sbstats_seq_start,
 	.next  = gfs2_sbstats_seq_next,
 	.stop  = gfs2_sbstats_seq_stop,
@@ -2468,16 +2470,6 @@ static int gfs2_glstats_open(struct inode *inode, struct file *file)
 	return __gfs2_glocks_open(inode, file, &gfs2_glstats_seq_ops);
 }
 
-static int gfs2_sbstats_open(struct inode *inode, struct file *file)
-{
-	int ret = seq_open(file, &gfs2_sbstats_seq_ops);
-	if (ret == 0) {
-		struct seq_file *seq = file->private_data;
-		seq->private = inode->i_private;  /* sdp */
-	}
-	return ret;
-}
-
 static const struct file_operations gfs2_glocks_fops = {
 	.owner   = THIS_MODULE,
 	.open    = gfs2_glocks_open,
@@ -2494,13 +2486,7 @@ static const struct file_operations gfs2_glstats_fops = {
 	.release = gfs2_glocks_release,
 };
 
-static const struct file_operations gfs2_sbstats_fops = {
-	.owner   = THIS_MODULE,
-	.open	 = gfs2_sbstats_open,
-	.read    = seq_read,
-	.llseek  = seq_lseek,
-	.release = seq_release,
-};
+DEFINE_SEQ_ATTRIBUTE(gfs2_sbstats);
 
 void gfs2_create_debugfs_file(struct gfs2_sbd *sdp)
 {

fs/gfs2/glops.c

@@ -178,6 +178,9 @@ static int rgrp_go_sync(struct gfs2_glock *gl)
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	struct address_space *mapping = &sdp->sd_aspace;
 	struct gfs2_rgrpd *rgd = gfs2_glock2rgrp(gl);
+	const unsigned bsize = sdp->sd_sb.sb_bsize;
+	loff_t start = (rgd->rd_addr * bsize) & PAGE_MASK;
+	loff_t end = PAGE_ALIGN((rgd->rd_addr + rgd->rd_length) * bsize) - 1;
 	int error;
 
 	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
@@ -186,18 +189,13 @@ static int rgrp_go_sync(struct gfs2_glock *gl)
 
 	gfs2_log_flush(sdp, gl, GFS2_LOG_HEAD_FLUSH_NORMAL |
 		       GFS2_LFC_RGRP_GO_SYNC);
-	filemap_fdatawrite_range(mapping, gl->gl_vm.start, gl->gl_vm.end);
-	error = filemap_fdatawait_range(mapping, gl->gl_vm.start, gl->gl_vm.end);
-	WARN_ON_ONCE(error);
+	filemap_fdatawrite_range(mapping, start, end);
+	error = filemap_fdatawait_range(mapping, start, end);
+	WARN_ON_ONCE(error && !gfs2_withdrawn(sdp));
 	mapping_set_error(mapping, error);
 	if (!error)
 		error = gfs2_ail_empty_gl(gl);
-
-	spin_lock(&gl->gl_lockref.lock);
-	rgd = gl->gl_object;
-	if (rgd)
 	gfs2_free_clones(rgd);
-	spin_unlock(&gl->gl_lockref.lock);
 	return error;
 }
 
@@ -216,15 +214,23 @@ static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;
 	struct address_space *mapping = &sdp->sd_aspace;
 	struct gfs2_rgrpd *rgd = gfs2_glock2rgrp(gl);
+	const unsigned bsize = sdp->sd_sb.sb_bsize;
+	loff_t start = (rgd->rd_addr * bsize) & PAGE_MASK;
+	loff_t end = PAGE_ALIGN((rgd->rd_addr + rgd->rd_length) * bsize) - 1;
 
-	if (rgd)
 	gfs2_rgrp_brelse(rgd);
-
 	WARN_ON_ONCE(!(flags & DIO_METADATA));
-	truncate_inode_pages_range(mapping, gl->gl_vm.start, gl->gl_vm.end);
+	truncate_inode_pages_range(mapping, start, end);
+	rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
+}
+
+static void gfs2_rgrp_go_dump(struct seq_file *seq, struct gfs2_glock *gl,
+			      const char *fs_id_buf)
+{
+	struct gfs2_rgrpd *rgd = gfs2_glock2rgrp(gl);
 
 	if (rgd)
-		rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
+		gfs2_rgrp_dump(seq, rgd, fs_id_buf);
 }
 
 static struct gfs2_inode *gfs2_glock2inode(struct gfs2_glock *gl)
@@ -712,7 +718,7 @@ const struct gfs2_glock_operations gfs2_rgrp_glops = {
 	.go_sync = rgrp_go_sync,
 	.go_inval = rgrp_go_inval,
 	.go_lock = gfs2_rgrp_go_lock,
-	.go_dump = gfs2_rgrp_dump,
+	.go_dump = gfs2_rgrp_go_dump,
 	.go_type = LM_TYPE_RGRP,
 	.go_flags = GLOF_LVB,
 };

fs/gfs2/incore.h

@@ -41,6 +41,10 @@ struct gfs2_log_header_host {
 	u32 lh_flags;		/* GFS2_LOG_HEAD_... */
 	u32 lh_tail;		/* Block number of log tail */
 	u32 lh_blkno;
+
+	s64 lh_local_total;
+	s64 lh_local_free;
+	s64 lh_local_dinodes;
 };
 
 /*
@@ -340,7 +344,6 @@ enum {
 	GLF_REPLY_PENDING		= 9,
 	GLF_INITIAL			= 10,
 	GLF_FROZEN			= 11,
-	GLF_QUEUED			= 12,
 	GLF_LRU				= 13,
 	GLF_OBJECT			= 14, /* Used only for tracing */
 	GLF_BLOCKING			= 15,
@@ -378,18 +381,11 @@ struct gfs2_glock {
 	atomic_t gl_ail_count;
 	atomic_t gl_revokes;
 	struct delayed_work gl_work;
-	union {
 	/* For iopen glocks only */
 	struct {
 		struct delayed_work gl_delete;
 		u64 gl_no_formal_ino;
 	};
-		/* For rgrp glocks only */
-		struct {
-			loff_t start;
-			loff_t end;
-		} gl_vm;
-	};
 	struct rcu_head gl_rcu;
 	struct rhash_head gl_node;
 };
@@ -701,10 +697,18 @@ struct gfs2_pcpu_lkstats {
 	struct gfs2_lkstats lkstats[10];
 };
 
+/* List of local (per node) statfs inodes */
+struct local_statfs_inode {
+	struct list_head si_list;
+	struct inode *si_sc_inode;
+	unsigned int si_jid; /* journal id this statfs inode corresponds to */
+};
+
 struct gfs2_sbd {
 	struct super_block *sd_vfs;
 	struct gfs2_pcpu_lkstats __percpu *sd_lkstats;
 	struct kobject sd_kobj;
+	struct completion sd_kobj_unregister;
 	unsigned long sd_flags;	/* SDF_... */
 	struct gfs2_sb_host sd_sb;
 
@@ -751,6 +755,7 @@ struct gfs2_sbd {
 	struct inode *sd_jindex;
 	struct inode *sd_statfs_inode;
 	struct inode *sd_sc_inode;
+	struct list_head sd_sc_inodes_list;
 	struct inode *sd_qc_inode;
 	struct inode *sd_rindex;
 	struct inode *sd_quota_inode;

fs/gfs2/log.c

@@ -70,7 +70,7 @@ unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct)
  *
  */
 
-static void gfs2_remove_from_ail(struct gfs2_bufdata *bd)
+void gfs2_remove_from_ail(struct gfs2_bufdata *bd)
 {
 	bd->bd_tr = NULL;
 	list_del_init(&bd->bd_ail_st_list);
@@ -244,13 +244,15 @@ static void gfs2_ail1_start(struct gfs2_sbd *sdp)
  * @tr: the transaction
  * @max_revokes: If nonzero, issue revokes for the bd items for written buffers
  *
+ * returns: the transaction's count of remaining active items
  */
 
-static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
+static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
 				int *max_revokes)
 {
 	struct gfs2_bufdata *bd, *s;
 	struct buffer_head *bh;
+	int active_count = 0;
 
 	list_for_each_entry_safe_reverse(bd, s, &tr->tr_ail1_list,
 					 bd_ail_st_list) {
@@ -265,8 +267,10 @@ static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
 		 * If the ail buffer is not busy and caught an error, flag it
 		 * for others.
 		 */
-		if (!sdp->sd_log_error && buffer_busy(bh))
+		if (!sdp->sd_log_error && buffer_busy(bh)) {
+			active_count++;
 			continue;
+		}
 		if (!buffer_uptodate(bh) &&
 		    !cmpxchg(&sdp->sd_log_error, 0, -EIO)) {
 			gfs2_io_error_bh(sdp, bh);
@@ -285,6 +289,7 @@ static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_trans *tr,
 		}
 		list_move(&bd->bd_ail_st_list, &tr->tr_ail2_list);
 	}
+	return active_count;
 }
 
 /**
@@ -303,8 +308,7 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int max_revokes)
 
 	spin_lock(&sdp->sd_ail_lock);
 	list_for_each_entry_safe_reverse(tr, s, &sdp->sd_ail1_list, tr_list) {
-		gfs2_ail1_empty_one(sdp, tr, &max_revokes);
-		if (list_empty(&tr->tr_ail1_list) && oldest_tr)
+		if (!gfs2_ail1_empty_one(sdp, tr, &max_revokes) && oldest_tr)
 			list_move(&tr->tr_list, &sdp->sd_ail2_list);
 		else
 			oldest_tr = 0;
@@ -716,16 +720,24 @@ void gfs2_write_revokes(struct gfs2_sbd *sdp)
 		atomic_dec(&sdp->sd_log_blks_free);
 		/* If no blocks have been reserved, we need to also
 		 * reserve a block for the header */
-		if (!sdp->sd_log_blks_reserved)
+		if (!sdp->sd_log_blks_reserved) {
 			atomic_dec(&sdp->sd_log_blks_free);
+			trace_gfs2_log_blocks(sdp, -2);
+		} else {
+			trace_gfs2_log_blocks(sdp, -1);
+		}
 	}
 	gfs2_ail1_empty(sdp, max_revokes);
 	gfs2_log_unlock(sdp);
 
 	if (!sdp->sd_log_num_revoke) {
 		atomic_inc(&sdp->sd_log_blks_free);
-		if (!sdp->sd_log_blks_reserved)
+		if (!sdp->sd_log_blks_reserved) {
 			atomic_inc(&sdp->sd_log_blks_free);
+			trace_gfs2_log_blocks(sdp, 2);
+		} else {
+			trace_gfs2_log_blocks(sdp, 1);
+		}
 	}
 }
 
@@ -902,7 +914,7 @@ static void empty_ail1_list(struct gfs2_sbd *sdp)
 }
 
 /**
- * drain_bd - drain the buf and databuf queue for a failed transaction
+ * trans_drain - drain the buf and databuf queue for a failed transaction
  * @tr: the transaction to drain
  *
  * When this is called, we're taking an error exit for a log write that failed
@@ -954,10 +966,8 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 		goto out;
 
 	/* Log might have been flushed while we waited for the flush lock */
-	if (gl && !test_bit(GLF_LFLUSH, &gl->gl_flags)) {
-		up_write(&sdp->sd_log_flush_lock);
-		return;
-	}
+	if (gl && !test_bit(GLF_LFLUSH, &gl->gl_flags))
+		goto out;
 	trace_gfs2_log_flush(sdp, 1, flags);
 
 	if (flags & GFS2_LOG_HEAD_FLUSH_SHUTDOWN)
@@ -971,25 +981,25 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 		if (unlikely (state == SFS_FROZEN))
 			if (gfs2_assert_withdraw_delayed(sdp,
 			       !tr->tr_num_buf_new && !tr->tr_num_databuf_new))
-				goto out;
+				goto out_withdraw;
 	}
 
 	if (unlikely(state == SFS_FROZEN))
 		if (gfs2_assert_withdraw_delayed(sdp, !sdp->sd_log_num_revoke))
-			goto out;
+			goto out_withdraw;
 	if (gfs2_assert_withdraw_delayed(sdp,
 			sdp->sd_log_num_revoke == sdp->sd_log_committed_revoke))
-		goto out;
+		goto out_withdraw;
 
 	gfs2_ordered_write(sdp);
 	if (gfs2_withdrawn(sdp))
-		goto out;
+		goto out_withdraw;
 	lops_before_commit(sdp, tr);
 	if (gfs2_withdrawn(sdp))
-		goto out;
+		goto out_withdraw;
 	gfs2_log_submit_bio(&sdp->sd_log_bio, REQ_OP_WRITE);
 	if (gfs2_withdrawn(sdp))
-		goto out;
+		goto out_withdraw;
 
 	if (sdp->sd_log_head != sdp->sd_log_flush_head) {
 		log_flush_wait(sdp);
@@ -1000,7 +1010,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 		log_write_header(sdp, flags);
 	}
 	if (gfs2_withdrawn(sdp))
-		goto out;
+		goto out_withdraw;
 	lops_after_commit(sdp, tr);
 
 	gfs2_log_lock(sdp);
@@ -1020,7 +1030,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 		if (!sdp->sd_log_idle) {
 			empty_ail1_list(sdp);
 			if (gfs2_withdrawn(sdp))
-				goto out;
+				goto out_withdraw;
 			atomic_dec(&sdp->sd_log_blks_free); /* Adjust for unreserved buffer */
 			trace_gfs2_log_blocks(sdp, -1);
 			log_write_header(sdp, flags);
@@ -1033,8 +1043,16 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 			atomic_set(&sdp->sd_freeze_state, SFS_FROZEN);
 	}
 
+out_end:
+	trace_gfs2_log_flush(sdp, 0, flags);
 out:
-	if (gfs2_withdrawn(sdp)) {
+	up_write(&sdp->sd_log_flush_lock);
+	gfs2_trans_free(sdp, tr);
+	if (gfs2_withdrawing(sdp))
+		gfs2_withdraw(sdp);
+	return;
+
+out_withdraw:
 	trans_drain(tr);
 	/**
 	 * If the tr_list is empty, we're withdrawing during a log
@@ -1048,12 +1066,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
 	spin_unlock(&sdp->sd_ail_lock);
 	ail_drain(sdp); /* frees all transactions */
 	tr = NULL;
-	}
-
-	trace_gfs2_log_flush(sdp, 0, flags);
-	up_write(&sdp->sd_log_flush_lock);
-
-	gfs2_trans_free(sdp, tr);
+	goto out_end;
 }
 
 /**

fs/gfs2/log.h

@@ -63,7 +63,7 @@ static inline void gfs2_ordered_add_inode(struct gfs2_inode *ip)
 
 extern void gfs2_ordered_del_inode(struct gfs2_inode *ip);
 extern unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct);
-
+extern void gfs2_remove_from_ail(struct gfs2_bufdata *bd);
 extern void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks);
 extern int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks);
 extern void gfs2_write_log_header(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd,

fs/gfs2/lops.c

@@ -823,7 +823,7 @@ static int buf_lo_scan_elements(struct gfs2_jdesc *jd, u32 start,
  *
  */
 
-static void gfs2_meta_sync(struct gfs2_glock *gl)
+void gfs2_meta_sync(struct gfs2_glock *gl)
 {
 	struct address_space *mapping = gfs2_glock2aspace(gl);
 	struct gfs2_sbd *sdp = gl->gl_name.ln_sbd;

fs/gfs2/lops.h

@@ -27,6 +27,7 @@ extern void gfs2_log_submit_bio(struct bio **biop, int opf);
 extern void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh);
 extern int gfs2_find_jhead(struct gfs2_jdesc *jd,
 			   struct gfs2_log_header_host *head, bool keep_cache);
+extern void gfs2_meta_sync(struct gfs2_glock *gl);
 
 static inline unsigned int buf_limit(struct gfs2_sbd *sdp)
 {

fs/gfs2/meta_io.c

@@ -348,38 +348,109 @@ void gfs2_remove_from_journal(struct buffer_head *bh, int meta)
 		brelse(bh);
 	}
 	if (bd) {
-		spin_lock(&sdp->sd_ail_lock);
 		if (bd->bd_tr) {
 			gfs2_trans_add_revoke(sdp, bd);
 		} else if (was_pinned) {
 			bh->b_private = NULL;
 			kmem_cache_free(gfs2_bufdata_cachep, bd);
+		} else if (!list_empty(&bd->bd_ail_st_list) &&
+					!list_empty(&bd->bd_ail_gl_list)) {
+			gfs2_remove_from_ail(bd);
 		}
-		spin_unlock(&sdp->sd_ail_lock);
 	}
 	clear_buffer_dirty(bh);
 	clear_buffer_uptodate(bh);
 }
 
 /**
- * gfs2_meta_wipe - make inode's buffers so they aren't dirty/pinned anymore
+ * gfs2_ail1_wipe - remove deleted/freed buffers from the ail1 list
+ * @sdp: superblock
+ * @bstart: starting block address of buffers to remove
+ * @blen: length of buffers to be removed
+ *
+ * This function is called from gfs2_journal wipe, whose job is to remove
+ * buffers, corresponding to deleted blocks, from the journal. If we find any
+ * bufdata elements on the system ail1 list, they haven't been written to
+ * the journal yet. So we remove them.
+ */
+static void gfs2_ail1_wipe(struct gfs2_sbd *sdp, u64 bstart, u32 blen)
+{
+	struct gfs2_trans *tr, *s;
+	struct gfs2_bufdata *bd, *bs;
+	struct buffer_head *bh;
+	u64 end = bstart + blen;
+
+	gfs2_log_lock(sdp);
+	spin_lock(&sdp->sd_ail_lock);
+	list_for_each_entry_safe(tr, s, &sdp->sd_ail1_list, tr_list) {
+		list_for_each_entry_safe(bd, bs, &tr->tr_ail1_list,
+					 bd_ail_st_list) {
+			bh = bd->bd_bh;
+			if (bh->b_blocknr < bstart || bh->b_blocknr >= end)
+				continue;
+
+			gfs2_remove_from_journal(bh, REMOVE_JDATA);
+		}
+	}
+	spin_unlock(&sdp->sd_ail_lock);
+	gfs2_log_unlock(sdp);
+}
+
+static struct buffer_head *gfs2_getjdatabuf(struct gfs2_inode *ip, u64 blkno)
+{
+	struct address_space *mapping = ip->i_inode.i_mapping;
+	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
+	struct page *page;
+	struct buffer_head *bh;
+	unsigned int shift = PAGE_SHIFT - sdp->sd_sb.sb_bsize_shift;
+	unsigned long index = blkno >> shift; /* convert block to page */
+	unsigned int bufnum = blkno - (index << shift);
+
+	page = find_get_page_flags(mapping, index, FGP_LOCK|FGP_ACCESSED);
+	if (!page)
+		return NULL;
+	if (!page_has_buffers(page)) {
+		unlock_page(page);
+		put_page(page);
+		return NULL;
+	}
+	/* Locate header for our buffer within our page */
+	for (bh = page_buffers(page); bufnum--; bh = bh->b_this_page)
+		/* Do nothing */;
+	get_bh(bh);
+	unlock_page(page);
+	put_page(page);
+	return bh;
+}
+
+/**
+ * gfs2_journal_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
  * @blen: the number of buffers in the run
  *
  */
 
-void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
+void gfs2_journal_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 	struct buffer_head *bh;
+	int ty;
 
+	gfs2_ail1_wipe(sdp, bstart, blen);
 	while (blen) {
+		ty = REMOVE_META;
 		bh = gfs2_getbuf(ip->i_gl, bstart, NO_CREATE);
+		if (!bh && gfs2_is_jdata(ip)) {
+			bh = gfs2_getjdatabuf(ip, bstart);
+			ty = REMOVE_JDATA;
+		}
 		if (bh) {
 			lock_buffer(bh);
 			gfs2_log_lock(sdp);
-			gfs2_remove_from_journal(bh, REMOVE_META);
+			spin_lock(&sdp->sd_ail_lock);
+			gfs2_remove_from_journal(bh, ty);
+			spin_unlock(&sdp->sd_ail_lock);
 			gfs2_log_unlock(sdp);
 			unlock_buffer(bh);
 			brelse(bh);

fs/gfs2/meta_io.h

@@ -60,7 +60,7 @@ enum {
 };
 
 extern void gfs2_remove_from_journal(struct buffer_head *bh, int meta);
-extern void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen);
+extern void gfs2_journal_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen);
 extern int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num,
 				     struct buffer_head **bhp);
 

fs/gfs2/ops_fstype.c

@@ -110,6 +110,8 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb)
 	spin_lock_init(&sdp->sd_trunc_lock);
 	spin_lock_init(&sdp->sd_bitmap_lock);
 
+	INIT_LIST_HEAD(&sdp->sd_sc_inodes_list);
+
 	mapping = &sdp->sd_aspace;
 
 	address_space_init_once(mapping);
@@ -169,15 +171,19 @@ static int gfs2_check_sb(struct gfs2_sbd *sdp, int silent)
 		return -EINVAL;
 	}
 
-	/*  If format numbers match exactly, we're done.  */
-
-	if (sb->sb_fs_format == GFS2_FORMAT_FS &&
-	    sb->sb_multihost_format == GFS2_FORMAT_MULTI)
-		return 0;
-
+	if (sb->sb_fs_format != GFS2_FORMAT_FS ||
+	    sb->sb_multihost_format != GFS2_FORMAT_MULTI) {
 		fs_warn(sdp, "Unknown on-disk format, unable to mount\n");
+		return -EINVAL;
+	}
 
+	if (sb->sb_bsize < 512 || sb->sb_bsize > PAGE_SIZE ||
+	    (sb->sb_bsize & (sb->sb_bsize - 1))) {
+		pr_warn("Invalid superblock size\n");
 		return -EINVAL;
+	}
+
+	return 0;
 }
 
 static void end_bio_io_page(struct bio *bio)
@@ -604,6 +610,90 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
 	return error;
 }
 
+/**
+ * init_statfs - look up and initialize master and local (per node) statfs inodes
+ * @sdp: The GFS2 superblock
+ *
+ * This should be called after the jindex is initialized in init_journal() and
+ * before gfs2_journal_recovery() is called because we need to be able to write
+ * to these inodes during recovery.
+ *
+ * Returns: errno
+ */
+static int init_statfs(struct gfs2_sbd *sdp)
+{
+	int error = 0;
+	struct inode *master = d_inode(sdp->sd_master_dir);
+	struct inode *pn = NULL;
+	char buf[30];
+	struct gfs2_jdesc *jd;
+	struct gfs2_inode *ip;
+
+	sdp->sd_statfs_inode = gfs2_lookup_simple(master, "statfs");
+	if (IS_ERR(sdp->sd_statfs_inode)) {
+		error = PTR_ERR(sdp->sd_statfs_inode);
+		fs_err(sdp, "can't read in statfs inode: %d\n", error);
+		goto fail;
+	}
+
+	pn = gfs2_lookup_simple(master, "per_node");
+	if (IS_ERR(pn)) {
+		error = PTR_ERR(pn);
+		fs_err(sdp, "can't find per_node directory: %d\n", error);
+		goto put_statfs;
+	}
+
+	/* For each jid, lookup the corresponding local statfs inode in the
+	 * per_node metafs directory and save it in the sdp->sd_sc_inodes_list. */
+	list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
+		struct local_statfs_inode *lsi =
+			kmalloc(sizeof(struct local_statfs_inode), GFP_NOFS);
+		if (!lsi) {
+			error = -ENOMEM;
+			goto free_local;
+		}
+		sprintf(buf, "statfs_change%u", jd->jd_jid);
+		lsi->si_sc_inode = gfs2_lookup_simple(pn, buf);
+		if (IS_ERR(lsi->si_sc_inode)) {
+			error = PTR_ERR(lsi->si_sc_inode);
+			fs_err(sdp, "can't find local \"sc\" file#%u: %d\n",
+			       jd->jd_jid, error);
+			goto free_local;
+		}
+		lsi->si_jid = jd->jd_jid;
+		if (jd->jd_jid == sdp->sd_jdesc->jd_jid)
+			sdp->sd_sc_inode = lsi->si_sc_inode;
+
+		list_add_tail(&lsi->si_list, &sdp->sd_sc_inodes_list);
+	}
+
+	iput(pn);
+	ip = GFS2_I(sdp->sd_sc_inode);
+	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0,
+				   &sdp->sd_sc_gh);
+	if (error) {
+		fs_err(sdp, "can't lock local \"sc\" file: %d\n", error);
+		goto free_local;
+	}
+	return 0;
+
+free_local:
+	free_local_statfs_inodes(sdp);
+	iput(pn);
+put_statfs:
+	iput(sdp->sd_statfs_inode);
+fail:
+	return error;
+}
+
+/* Uninitialize and free up memory used by the list of statfs inodes */
+static void uninit_statfs(struct gfs2_sbd *sdp)
+{
+	gfs2_glock_dq_uninit(&sdp->sd_sc_gh);
+	free_local_statfs_inodes(sdp);
+	iput(sdp->sd_statfs_inode);
+}
+
 static int init_journal(struct gfs2_sbd *sdp, int undo)
 {
 	struct inode *master = d_inode(sdp->sd_master_dir);
@@ -690,6 +780,11 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 	}
 	trace_gfs2_log_blocks(sdp, atomic_read(&sdp->sd_log_blks_free));
 
+	/* Lookup statfs inodes here so journal recovery can use them. */
+	error = init_statfs(sdp);
+	if (error)
+		goto fail_jinode_gh;
+
 	if (sdp->sd_lockstruct.ls_first) {
 		unsigned int x;
 		for (x = 0; x < sdp->sd_journals; x++) {
@@ -698,14 +793,14 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 			if (sdp->sd_args.ar_spectator) {
 				error = check_journal_clean(sdp, jd, true);
 				if (error)
-					goto fail_jinode_gh;
+					goto fail_statfs;
 				continue;
 			}
 			error = gfs2_recover_journal(jd, true);
 			if (error) {
 				fs_err(sdp, "error recovering journal %u: %d\n",
 				       x, error);
-				goto fail_jinode_gh;
+				goto fail_statfs;
 			}
 		}
 
@@ -714,7 +809,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 		error = gfs2_recover_journal(sdp->sd_jdesc, true);
 		if (error) {
 			fs_err(sdp, "error recovering my journal: %d\n", error);
-			goto fail_jinode_gh;
+			goto fail_statfs;
 		}
 	}
 
@@ -725,6 +820,8 @@ static int init_journal(struct gfs2_sbd *sdp, int undo)
 	INIT_WORK(&sdp->sd_freeze_work, gfs2_freeze_func);
 	return 0;
 
+fail_statfs:
+	uninit_statfs(sdp);
 fail_jinode_gh:
 	/* A withdraw may have done dq/uninit so now we need to check it */
 	if (!sdp->sd_args.ar_spectator &&
@@ -758,20 +855,12 @@ static int init_inodes(struct gfs2_sbd *sdp, int undo)
 	if (error)
 		goto fail;
 
-	/* Read in the master statfs inode */
-	sdp->sd_statfs_inode = gfs2_lookup_simple(master, "statfs");
-	if (IS_ERR(sdp->sd_statfs_inode)) {
-		error = PTR_ERR(sdp->sd_statfs_inode);
-		fs_err(sdp, "can't read in statfs inode: %d\n", error);
-		goto fail_journal;
-	}
-
 	/* Read in the resource index inode */
 	sdp->sd_rindex = gfs2_lookup_simple(master, "rindex");
 	if (IS_ERR(sdp->sd_rindex)) {
 		error = PTR_ERR(sdp->sd_rindex);
 		fs_err(sdp, "can't get resource index inode: %d\n", error);
-		goto fail_statfs;
+		goto fail_journal;
 	}
 	sdp->sd_rindex_uptodate = 0;
 
@@ -800,8 +889,6 @@ static int init_inodes(struct gfs2_sbd *sdp, int undo)
 fail_rindex:
 	gfs2_clear_rgrpd(sdp);
 	iput(sdp->sd_rindex);
-fail_statfs:
-	iput(sdp->sd_statfs_inode);
 fail_journal:
 	init_journal(sdp, UNDO);
 fail:
@@ -829,14 +916,6 @@ static int init_per_node(struct gfs2_sbd *sdp, int undo)
 		return error;
 	}
 
-	sprintf(buf, "statfs_change%u", sdp->sd_jdesc->jd_jid);
-	sdp->sd_sc_inode = gfs2_lookup_simple(pn, buf);
-	if (IS_ERR(sdp->sd_sc_inode)) {
-		error = PTR_ERR(sdp->sd_sc_inode);
-		fs_err(sdp, "can't find local \"sc\" file: %d\n", error);
-		goto fail;
-	}
-
 	sprintf(buf, "quota_change%u", sdp->sd_jdesc->jd_jid);
 	sdp->sd_qc_inode = gfs2_lookup_simple(pn, buf);
 	if (IS_ERR(sdp->sd_qc_inode)) {
@@ -848,33 +927,21 @@ static int init_per_node(struct gfs2_sbd *sdp, int undo)
 	iput(pn);
 	pn = NULL;
 
-	ip = GFS2_I(sdp->sd_sc_inode);
-	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0,
-				   &sdp->sd_sc_gh);
-	if (error) {
-		fs_err(sdp, "can't lock local \"sc\" file: %d\n", error);
-		goto fail_qc_i;
-	}
-
 	ip = GFS2_I(sdp->sd_qc_inode);
 	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0,
 				   &sdp->sd_qc_gh);
 	if (error) {
 		fs_err(sdp, "can't lock local \"qc\" file: %d\n", error);
-		goto fail_ut_gh;
+		goto fail_qc_i;
 	}
 
 	return 0;
 
 fail_qc_gh:
 	gfs2_glock_dq_uninit(&sdp->sd_qc_gh);
-fail_ut_gh:
-	gfs2_glock_dq_uninit(&sdp->sd_sc_gh);
 fail_qc_i:
 	iput(sdp->sd_qc_inode);
 fail_ut_i:
-	iput(sdp->sd_sc_inode);
-fail:
 	iput(pn);
 	return error;
 }
@@ -1062,26 +1129,14 @@ static int gfs2_fill_super(struct super_block *sb, struct fs_context *fc)
 	}
 
 	error = init_names(sdp, silent);
-	if (error) {
-		/* In this case, we haven't initialized sysfs, so we have to
-		   manually free the sdp. */
-		free_sbd(sdp);
-		sb->s_fs_info = NULL;
-		return error;
-	}
+	if (error)
+		goto fail_free;
 
 	snprintf(sdp->sd_fsname, sizeof(sdp->sd_fsname), "%s", sdp->sd_table_name);
 
 	error = gfs2_sys_fs_add(sdp);
-	/*
-	 * If we hit an error here, gfs2_sys_fs_add will have called function
-	 * kobject_put which causes the sysfs usage count to go to zero, which
-	 * causes sysfs to call function gfs2_sbd_release, which frees sdp.
-	 * Subsequent error paths here will call gfs2_sys_fs_del, which also
-	 * kobject_put to free sdp.
-	 */
 	if (error)
-		return error;
+		goto fail_free;
 
 	gfs2_create_debugfs_file(sdp);
 
@@ -1179,9 +1234,9 @@ static int gfs2_fill_super(struct super_block *sb, struct fs_context *fc)
 	gfs2_lm_unmount(sdp);
 fail_debug:
 	gfs2_delete_debugfs_file(sdp);
-	/* gfs2_sys_fs_del must be the last thing we do, since it causes
-	 * sysfs to call function gfs2_sbd_release, which frees sdp. */
 	gfs2_sys_fs_del(sdp);
+fail_free:
+	free_sbd(sdp);
 	sb->s_fs_info = NULL;
 	return error;
 }

fs/gfs2/recovery.c

@@ -144,6 +144,10 @@ int __get_log_header(struct gfs2_sbd *sdp, const struct gfs2_log_header *lh,
 	head->lh_tail = be32_to_cpu(lh->lh_tail);
 	head->lh_blkno = be32_to_cpu(lh->lh_blkno);
 
+	head->lh_local_total = be64_to_cpu(lh->lh_local_total);
+	head->lh_local_free = be64_to_cpu(lh->lh_local_free);
+	head->lh_local_dinodes = be64_to_cpu(lh->lh_local_dinodes);
+
 	return 0;
 }
 /**
@@ -292,6 +296,109 @@ static void gfs2_recovery_done(struct gfs2_sbd *sdp, unsigned int jid,
 		sdp->sd_lockstruct.ls_ops->lm_recovery_result(sdp, jid, message);
 }
 
+/**
+ * update_statfs_inode - Update the master statfs inode or zero out the local
+ *			 statfs inode for a given journal.
+ * @jd: The journal
+ * @head: If NULL, @inode is the local statfs inode and we need to zero it out.
+ *	  Otherwise, it @head contains the statfs change info that needs to be
+ *	  synced to the master statfs inode (pointed to by @inode).
+ * @inode: statfs inode to update.
+ */
+static int update_statfs_inode(struct gfs2_jdesc *jd,
+			       struct gfs2_log_header_host *head,
+			       struct inode *inode)
+{
+	struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
+	struct gfs2_inode *ip;
+	struct buffer_head *bh;
+	struct gfs2_statfs_change_host sc;
+	int error = 0;
+
+	BUG_ON(!inode);
+	ip = GFS2_I(inode);
+
+	error = gfs2_meta_inode_buffer(ip, &bh);
+	if (error)
+		goto out;
+
+	spin_lock(&sdp->sd_statfs_spin);
+
+	if (head) { /* Update the master statfs inode */
+		gfs2_statfs_change_in(&sc, bh->b_data + sizeof(struct gfs2_dinode));
+		sc.sc_total += head->lh_local_total;
+		sc.sc_free += head->lh_local_free;
+		sc.sc_dinodes += head->lh_local_dinodes;
+		gfs2_statfs_change_out(&sc, bh->b_data + sizeof(struct gfs2_dinode));
+
+		fs_info(sdp, "jid=%u: Updated master statfs Total:%lld, "
+			"Free:%lld, Dinodes:%lld after change "
+			"[%+lld,%+lld,%+lld]\n", jd->jd_jid, sc.sc_total,
+			sc.sc_free, sc.sc_dinodes, head->lh_local_total,
+			head->lh_local_free, head->lh_local_dinodes);
+	} else { /* Zero out the local statfs inode */
+		memset(bh->b_data + sizeof(struct gfs2_dinode), 0,
+		       sizeof(struct gfs2_statfs_change));
+		/* If it's our own journal, reset any in-memory changes too */
+		if (jd->jd_jid == sdp->sd_lockstruct.ls_jid) {
+			memset(&sdp->sd_statfs_local, 0,
+			       sizeof(struct gfs2_statfs_change_host));
+		}
+	}
+	spin_unlock(&sdp->sd_statfs_spin);
+
+	mark_buffer_dirty(bh);
+	brelse(bh);
+	gfs2_meta_sync(ip->i_gl);
+
+out:
+	return error;
+}
+
+/**
+ * recover_local_statfs - Update the master and local statfs changes for this
+ *			  journal.
+ *
+ * Previously, statfs updates would be read in from the local statfs inode and
+ * synced to the master statfs inode during recovery.
+ *
+ * We now use the statfs updates in the journal head to update the master statfs
+ * inode instead of reading in from the local statfs inode. To preserve backward
+ * compatibility with kernels that can't do this, we still need to keep the
+ * local statfs inode up to date by writing changes to it. At some point in the
+ * future, we can do away with the local statfs inodes altogether and keep the
+ * statfs changes solely in the journal.
+ *
+ * @jd: the journal
+ * @head: the journal head
+ *
+ * Returns: errno
+ */
+static void recover_local_statfs(struct gfs2_jdesc *jd,
+				 struct gfs2_log_header_host *head)
+{
+	int error;
+	struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
+
+	if (!head->lh_local_total && !head->lh_local_free
+	    && !head->lh_local_dinodes) /* No change */
+		goto zero_local;
+
+	 /* First update the master statfs inode with the changes we
+	  * found in the journal. */
+	error = update_statfs_inode(jd, head, sdp->sd_statfs_inode);
+	if (error)
+		goto out;
+
+zero_local:
+	/* Zero out the local statfs inode so any changes in there
+	 * are not re-recovered. */
+	error = update_statfs_inode(jd, NULL,
+				    find_local_statfs_inode(sdp, jd->jd_jid));
+out:
+	return;
+}
+
 void gfs2_recover_func(struct work_struct *work)
 {
 	struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work);
@@ -411,6 +518,7 @@ void gfs2_recover_func(struct work_struct *work)
 				goto fail_gunlock_thaw;
 		}
 
+		recover_local_statfs(jd, &head);
 		clean_journal(jd, &head);
 		up_read(&sdp->sd_log_flush_lock);
 

fs/gfs2/rgrp.c

@@ -878,7 +878,6 @@ static int rgd_insert(struct gfs2_rgrpd *rgd)
 static int read_rindex_entry(struct gfs2_inode *ip)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-	const unsigned bsize = sdp->sd_sb.sb_bsize;
 	loff_t pos = sdp->sd_rgrps * sizeof(struct gfs2_rindex);
 	struct gfs2_rindex buf;
 	int error;
@@ -924,9 +923,6 @@ static int read_rindex_entry(struct gfs2_inode *ip)
 	spin_unlock(&sdp->sd_rindex_spin);
 	if (!error) {
 		glock_set_object(rgd->rd_gl, rgd);
-		rgd->rd_gl->gl_vm.start = (rgd->rd_addr * bsize) & PAGE_MASK;
-		rgd->rd_gl->gl_vm.end = PAGE_ALIGN((rgd->rd_addr +
-						    rgd->rd_length) * bsize) - 1;
 		return 0;
 	}
 
@@ -2209,20 +2205,17 @@ static void rgblk_free(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd,
 /**
  * gfs2_rgrp_dump - print out an rgrp
  * @seq: The iterator
- * @gl: The glock in question
+ * @rgd: The rgrp in question
  * @fs_id_buf: pointer to file system id (if requested)
  *
  */
 
-void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_glock *gl,
+void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_rgrpd *rgd,
 		    const char *fs_id_buf)
 {
-	struct gfs2_rgrpd *rgd = gl->gl_object;
 	struct gfs2_blkreserv *trs;
 	const struct rb_node *n;
 
-	if (rgd == NULL)
-		return;
 	gfs2_print_dbg(seq, "%s R: n:%llu f:%02x b:%u/%u i:%u r:%u e:%u\n",
 		       fs_id_buf,
 		       (unsigned long long)rgd->rd_addr, rgd->rd_flags,
@@ -2253,7 +2246,7 @@ static void gfs2_rgrp_error(struct gfs2_rgrpd *rgd)
 		(unsigned long long)rgd->rd_addr);
 	fs_warn(sdp, "umount on all nodes and run fsck.gfs2 to fix the error\n");
 	sprintf(fs_id_buf, "fsid=%s: ", sdp->sd_fsname);
-	gfs2_rgrp_dump(NULL, rgd->rd_gl, fs_id_buf);
+	gfs2_rgrp_dump(NULL, rgd, fs_id_buf);
 	rgd->rd_flags |= GFS2_RDF_ERROR;
 }
 
@@ -2445,8 +2438,8 @@ void __gfs2_free_blocks(struct gfs2_inode *ip, struct gfs2_rgrpd *rgd,
 	gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data);
 
 	/* Directories keep their data in the metadata address space */
-	if (meta || ip->i_depth)
-		gfs2_meta_wipe(ip, bstart, blen);
+	if (meta || ip->i_depth || gfs2_is_jdata(ip))
+		gfs2_journal_wipe(ip, bstart, blen);
 }
 
 /**
@@ -2502,7 +2495,7 @@ void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip)
 	gfs2_statfs_change(sdp, 0, +1, -1);
 	trace_gfs2_block_alloc(ip, rgd, ip->i_no_addr, 1, GFS2_BLKST_FREE);
 	gfs2_quota_change(ip, -1, ip->i_inode.i_uid, ip->i_inode.i_gid);
-	gfs2_meta_wipe(ip, ip->i_no_addr, 1);
+	gfs2_journal_wipe(ip, ip->i_no_addr, 1);
 }
 
 /**

fs/gfs2/rgrp.h

@@ -67,7 +67,7 @@ extern void gfs2_rlist_add(struct gfs2_inode *ip, struct gfs2_rgrp_list *rlist,
 extern void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist);
 extern void gfs2_rlist_free(struct gfs2_rgrp_list *rlist);
 extern u64 gfs2_ri_total(struct gfs2_sbd *sdp);
-extern void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_glock *gl,
+extern void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_rgrpd *rgd,
 			   const char *fs_id_buf);
 extern int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset,
 				   struct buffer_head *bh,

fs/gfs2/super.c

@@ -44,6 +44,12 @@
 #include "xattr.h"
 #include "lops.h"
 
+enum dinode_demise {
+	SHOULD_DELETE_DINODE,
+	SHOULD_NOT_DELETE_DINODE,
+	SHOULD_DEFER_EVICTION,
+};
+
 /**
  * gfs2_jindex_free - Clear all the journal index information
  * @sdp: The GFS2 superblock
@@ -224,7 +230,7 @@ void gfs2_statfs_change_in(struct gfs2_statfs_change_host *sc, const void *buf)
 	sc->sc_dinodes = be64_to_cpu(str->sc_dinodes);
 }
 
-static void gfs2_statfs_change_out(const struct gfs2_statfs_change_host *sc, void *buf)
+void gfs2_statfs_change_out(const struct gfs2_statfs_change_host *sc, void *buf)
 {
 	struct gfs2_statfs_change *str = buf;
 
@@ -702,6 +708,8 @@ static void gfs2_put_super(struct super_block *sb)
 		if (error)
 			gfs2_io_error(sdp);
 	}
+	WARN_ON(gfs2_withdrawing(sdp));
+
 	/*  At this point, we're through modifying the disk  */
 
 	/*  Release stuff  */
@@ -721,7 +729,7 @@ static void gfs2_put_super(struct super_block *sb)
 			gfs2_glock_dq_uninit(&sdp->sd_jinode_gh);
 		gfs2_glock_dq_uninit(&sdp->sd_sc_gh);
 		gfs2_glock_dq_uninit(&sdp->sd_qc_gh);
-		iput(sdp->sd_sc_inode);
+		free_local_statfs_inodes(sdp);
 		iput(sdp->sd_qc_inode);
 	}
 
@@ -736,6 +744,7 @@ static void gfs2_put_super(struct super_block *sb)
 
 	/*  At this point, we're through participating in the lockspace  */
 	gfs2_sys_fs_del(sdp);
+	free_sbd(sdp);
 }
 
 /**
@@ -1309,109 +1318,98 @@ static bool gfs2_upgrade_iopen_glock(struct inode *inode)
 }
 
 /**
- * gfs2_evict_inode - Remove an inode from cache
+ * evict_should_delete - determine whether the inode is eligible for deletion
  * @inode: The inode to evict
  *
- * There are three cases to consider:
- * 1. i_nlink == 0, we are final opener (and must deallocate)
- * 2. i_nlink == 0, we are not the final opener (and cannot deallocate)
- * 3. i_nlink > 0
+ * This function determines whether the evicted inode is eligible to be deleted
+ * and locks the inode glock.
  *
- * If the fs is read only, then we have to treat all cases as per #3
- * since we are unable to do any deallocation. The inode will be
- * deallocated by the next read/write node to attempt an allocation
- * in the same resource group
- *
- * We have to (at the moment) hold the inodes main lock to cover
- * the gap between unlocking the shared lock on the iopen lock and
- * taking the exclusive lock. I'd rather do a shared -> exclusive
- * conversion on the iopen lock, but we can change that later. This
- * is safe, just less efficient.
+ * Returns: the fate of the dinode
  */
-
-static void gfs2_evict_inode(struct inode *inode)
+static enum dinode_demise evict_should_delete(struct inode *inode,
+					      struct gfs2_holder *gh)
 {
+	struct gfs2_inode *ip = GFS2_I(inode);
 	struct super_block *sb = inode->i_sb;
 	struct gfs2_sbd *sdp = sb->s_fs_info;
-	struct gfs2_inode *ip = GFS2_I(inode);
-	struct gfs2_holder gh;
-	struct address_space *metamapping;
-	int error;
-
-	if (test_bit(GIF_FREE_VFS_INODE, &ip->i_flags)) {
-		clear_inode(inode);
-		return;
-	}
-
-	if (inode->i_nlink || sb_rdonly(sb))
-		goto out;
+	int ret;
 
 	if (test_bit(GIF_ALLOC_FAILED, &ip->i_flags)) {
 		BUG_ON(!gfs2_glock_is_locked_by_me(ip->i_gl));
-		gfs2_holder_mark_uninitialized(&gh);
-		goto out_delete;
+		goto should_delete;
 	}
 
 	if (test_bit(GIF_DEFERRED_DELETE, &ip->i_flags))
-		goto out;
+		return SHOULD_DEFER_EVICTION;
 
 	/* Deletes should never happen under memory pressure anymore.  */
 	if (WARN_ON_ONCE(current->flags & PF_MEMALLOC))
-		goto out;
+		return SHOULD_DEFER_EVICTION;
 
 	/* Must not read inode block until block type has been verified */
-	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, &gh);
-	if (unlikely(error)) {
+	ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, gh);
+	if (unlikely(ret)) {
 		glock_clear_object(ip->i_iopen_gh.gh_gl, ip);
 		ip->i_iopen_gh.gh_flags |= GL_NOCACHE;
 		gfs2_glock_dq_uninit(&ip->i_iopen_gh);
-		goto out;
+		return SHOULD_DEFER_EVICTION;
 	}
 
 	if (gfs2_inode_already_deleted(ip->i_gl, ip->i_no_formal_ino))
-		goto out_truncate;
-	error = gfs2_check_blk_type(sdp, ip->i_no_addr, GFS2_BLKST_UNLINKED);
-	if (error)
-		goto out_truncate;
+		return SHOULD_NOT_DELETE_DINODE;
+	ret = gfs2_check_blk_type(sdp, ip->i_no_addr, GFS2_BLKST_UNLINKED);
+	if (ret)
+		return SHOULD_NOT_DELETE_DINODE;
 
 	if (test_bit(GIF_INVALID, &ip->i_flags)) {
-		error = gfs2_inode_refresh(ip);
-		if (error)
-			goto out_truncate;
+		ret = gfs2_inode_refresh(ip);
+		if (ret)
+			return SHOULD_NOT_DELETE_DINODE;
 	}
 
 	/*
 	 * The inode may have been recreated in the meantime.
 	 */
 	if (inode->i_nlink)
-		goto out_truncate;
+		return SHOULD_NOT_DELETE_DINODE;
 
-out_delete:
+should_delete:
 	if (gfs2_holder_initialized(&ip->i_iopen_gh) &&
 	    test_bit(HIF_HOLDER, &ip->i_iopen_gh.gh_iflags)) {
 		if (!gfs2_upgrade_iopen_glock(inode)) {
 			gfs2_holder_uninit(&ip->i_iopen_gh);
-			goto out_truncate;
+			return SHOULD_NOT_DELETE_DINODE;
 		}
 	}
+	return SHOULD_DELETE_DINODE;
+}
+
+/**
+ * evict_unlinked_inode - delete the pieces of an unlinked evicted inode
+ * @inode: The inode to evict
+ */
+static int evict_unlinked_inode(struct inode *inode)
+{
+	struct gfs2_inode *ip = GFS2_I(inode);
+	int ret;
 
 	if (S_ISDIR(inode->i_mode) &&
 	    (ip->i_diskflags & GFS2_DIF_EXHASH)) {
-		error = gfs2_dir_exhash_dealloc(ip);
-		if (error)
-			goto out_unlock;
+		ret = gfs2_dir_exhash_dealloc(ip);
+		if (ret)
+			goto out;
 	}
 
 	if (ip->i_eattr) {
-		error = gfs2_ea_dealloc(ip);
-		if (error)
-			goto out_unlock;
+		ret = gfs2_ea_dealloc(ip);
+		if (ret)
+			goto out;
 	}
 
 	if (!gfs2_is_stuffed(ip)) {
-		error = gfs2_file_dealloc(ip);
-		if (error)
-			goto out_unlock;
+		ret = gfs2_file_dealloc(ip);
+		if (ret)
+			goto out;
 	}
 
 	/* We're about to clear the bitmap for the dinode, but as soon as we
@@ -1419,11 +1417,24 @@ static void gfs2_evict_inode(struct inode *inode)
 	   location and try to set gl_object again. We clear gl_object here so
 	   that subsequent inode creates don't see an old gl_object. */
 	glock_clear_object(ip->i_gl, ip);
-	error = gfs2_dinode_dealloc(ip);
+	ret = gfs2_dinode_dealloc(ip);
 	gfs2_inode_remember_delete(ip->i_gl, ip->i_no_formal_ino);
-	goto out_unlock;
+out:
+	return ret;
+}
+
+/*
+ * evict_linked_inode - evict an inode whose dinode has not been unlinked
+ * @inode: The inode to evict
+ */
+static int evict_linked_inode(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct gfs2_sbd *sdp = sb->s_fs_info;
+	struct gfs2_inode *ip = GFS2_I(inode);
+	struct address_space *metamapping;
+	int ret;
 
-out_truncate:
 	gfs2_log_flush(sdp, ip->i_gl, GFS2_LOG_HEAD_FLUSH_NORMAL |
 		       GFS2_LFC_EVICT_INODE);
 	metamapping = gfs2_glock2aspace(ip->i_gl);
@@ -1434,15 +1445,63 @@ static void gfs2_evict_inode(struct inode *inode)
 	write_inode_now(inode, 1);
 	gfs2_ail_flush(ip->i_gl, 0);
 
-	error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);
-	if (error)
-		goto out_unlock;
+	ret = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);
+	if (ret)
+		return ret;
+
 	/* Needs to be done before glock release & also in a transaction */
 	truncate_inode_pages(&inode->i_data, 0);
 	truncate_inode_pages(metamapping, 0);
 	gfs2_trans_end(sdp);
+	return 0;
+}
+
+/**
+ * gfs2_evict_inode - Remove an inode from cache
+ * @inode: The inode to evict
+ *
+ * There are three cases to consider:
+ * 1. i_nlink == 0, we are final opener (and must deallocate)
+ * 2. i_nlink == 0, we are not the final opener (and cannot deallocate)
+ * 3. i_nlink > 0
+ *
+ * If the fs is read only, then we have to treat all cases as per #3
+ * since we are unable to do any deallocation. The inode will be
+ * deallocated by the next read/write node to attempt an allocation
+ * in the same resource group
+ *
+ * We have to (at the moment) hold the inodes main lock to cover
+ * the gap between unlocking the shared lock on the iopen lock and
+ * taking the exclusive lock. I'd rather do a shared -> exclusive
+ * conversion on the iopen lock, but we can change that later. This
+ * is safe, just less efficient.
+ */
+
+static void gfs2_evict_inode(struct inode *inode)
+{
+	struct super_block *sb = inode->i_sb;
+	struct gfs2_sbd *sdp = sb->s_fs_info;
+	struct gfs2_inode *ip = GFS2_I(inode);
+	struct gfs2_holder gh;
+	int ret;
+
+	if (test_bit(GIF_FREE_VFS_INODE, &ip->i_flags)) {
+		clear_inode(inode);
+		return;
+	}
+
+	if (inode->i_nlink || sb_rdonly(sb))
+		goto out;
+
+	gfs2_holder_mark_uninitialized(&gh);
+	ret = evict_should_delete(inode, &gh);
+	if (ret == SHOULD_DEFER_EVICTION)
+		goto out;
+	if (ret == SHOULD_DELETE_DINODE)
+		ret = evict_unlinked_inode(inode);
+	else
+		ret = evict_linked_inode(inode);
 
-out_unlock:
 	if (gfs2_rs_active(&ip->i_res))
 		gfs2_rs_deltree(&ip->i_res);
 
@@ -1450,8 +1509,8 @@ static void gfs2_evict_inode(struct inode *inode)
 		glock_clear_object(ip->i_gl, ip);
 		gfs2_glock_dq_uninit(&gh);
 	}
-	if (error && error != GLR_TRYFAILED && error != -EROFS)
-		fs_warn(sdp, "gfs2_evict_inode: %d\n", error);
+	if (ret && ret != GLR_TRYFAILED && ret != -EROFS)
+		fs_warn(sdp, "gfs2_evict_inode: %d\n", ret);
 out:
 	truncate_inode_pages_final(&inode->i_data);
 	if (ip->i_qadata)
@@ -1502,6 +1561,35 @@ static void gfs2_free_inode(struct inode *inode)
 	kmem_cache_free(gfs2_inode_cachep, GFS2_I(inode));
 }
 
+extern void free_local_statfs_inodes(struct gfs2_sbd *sdp)
+{
+	struct local_statfs_inode *lsi, *safe;
+
+	/* Run through the statfs inodes list to iput and free memory */
+	list_for_each_entry_safe(lsi, safe, &sdp->sd_sc_inodes_list, si_list) {
+		if (lsi->si_jid == sdp->sd_jdesc->jd_jid)
+			sdp->sd_sc_inode = NULL; /* belongs to this node */
+		if (lsi->si_sc_inode)
+			iput(lsi->si_sc_inode);
+		list_del(&lsi->si_list);
+		kfree(lsi);
+	}
+}
+
+extern struct inode *find_local_statfs_inode(struct gfs2_sbd *sdp,
+					     unsigned int index)
+{
+	struct local_statfs_inode *lsi;
+
+	/* Return the local (per node) statfs inode in the
+	 * sdp->sd_sc_inodes_list corresponding to the 'index'. */
+	list_for_each_entry(lsi, &sdp->sd_sc_inodes_list, si_list) {
+		if (lsi->si_jid == index)
+			return lsi->si_sc_inode;
+	}
+	return NULL;
+}
+
 const struct super_operations gfs2_super_ops = {
 	.alloc_inode		= gfs2_alloc_inode,
 	.free_inode		= gfs2_free_inode,

fs/gfs2/super.h

@@ -37,11 +37,16 @@ extern void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free,
 			       s64 dinodes);
 extern void gfs2_statfs_change_in(struct gfs2_statfs_change_host *sc,
 				  const void *buf);
+extern void gfs2_statfs_change_out(const struct gfs2_statfs_change_host *sc,
+				   void *buf);
 extern void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh,
 			  struct buffer_head *l_bh);
 extern int gfs2_statfs_sync(struct super_block *sb, int type);
 extern void gfs2_freeze_func(struct work_struct *work);
 
+extern void free_local_statfs_inodes(struct gfs2_sbd *sdp);
+extern struct inode *find_local_statfs_inode(struct gfs2_sbd *sdp,
+					     unsigned int index);
 extern void free_sbd(struct gfs2_sbd *sdp);
 
 extern struct file_system_type gfs2_fs_type;

fs/gfs2/sys.c

@@ -303,7 +303,7 @@ static void gfs2_sbd_release(struct kobject *kobj)
 {
 	struct gfs2_sbd *sdp = container_of(kobj, struct gfs2_sbd, sd_kobj);
 
-	free_sbd(sdp);
+	complete(&sdp->sd_kobj_unregister);
 }
 
 static struct kobj_type gfs2_ktype = {
@@ -655,6 +655,7 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 	sprintf(ro, "RDONLY=%d", sb_rdonly(sb));
 	sprintf(spectator, "SPECTATOR=%d", sdp->sd_args.ar_spectator ? 1 : 0);
 
+	init_completion(&sdp->sd_kobj_unregister);
 	sdp->sd_kobj.kset = gfs2_kset;
 	error = kobject_init_and_add(&sdp->sd_kobj, &gfs2_ktype, NULL,
 				     "%s", sdp->sd_table_name);
@@ -685,6 +686,7 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 fail_reg:
 	fs_err(sdp, "error %d adding sysfs files\n", error);
 	kobject_put(&sdp->sd_kobj);
+	wait_for_completion(&sdp->sd_kobj_unregister);
 	sb->s_fs_info = NULL;
 	return error;
 }
@@ -695,6 +697,7 @@ void gfs2_sys_fs_del(struct gfs2_sbd *sdp)
 	sysfs_remove_group(&sdp->sd_kobj, &tune_group);
 	sysfs_remove_group(&sdp->sd_kobj, &lock_module_group);
 	kobject_put(&sdp->sd_kobj);
+	wait_for_completion(&sdp->sd_kobj_unregister);
 }
 
 static int gfs2_uevent(struct kset *kset, struct kobject *kobj,

fs/gfs2/trace_gfs2.h

@@ -56,7 +56,6 @@
 	{(1UL << GLF_REPLY_PENDING),		"r" },		\
 	{(1UL << GLF_INITIAL),			"I" },		\
 	{(1UL << GLF_FROZEN),			"F" },		\
-	{(1UL << GLF_QUEUED),			"q" },		\
 	{(1UL << GLF_LRU),			"L" },		\
 	{(1UL << GLF_OBJECT),			"o" },		\
 	{(1UL << GLF_BLOCKING),			"b" })
@@ -388,15 +387,17 @@ TRACE_EVENT(gfs2_log_blocks,
 	TP_STRUCT__entry(
 		__field(        dev_t,  dev                     )
 		__field(	int,	blocks			)
+		__field(	int,	blks_free		)
 	),
 
 	TP_fast_assign(
 		__entry->dev		= sdp->sd_vfs->s_dev;
 		__entry->blocks		= blocks;
+		__entry->blks_free	= atomic_read(&sdp->sd_log_blks_free);
 	),
 
-	TP_printk("%u,%u log reserve %d", MAJOR(__entry->dev),
-		  MINOR(__entry->dev), __entry->blocks)
+	TP_printk("%u,%u log reserve %d %d", MAJOR(__entry->dev),
+		  MINOR(__entry->dev), __entry->blocks, __entry->blks_free)
 );
 
 /* Writing back the AIL */

fs/gfs2/util.c

@@ -419,7 +419,7 @@ void gfs2_consist_rgrpd_i(struct gfs2_rgrpd *rgd,
 	char fs_id_buf[sizeof(sdp->sd_fsname) + 7];
 
 	sprintf(fs_id_buf, "fsid=%s: ", sdp->sd_fsname);
-	gfs2_rgrp_dump(NULL, rgd->rd_gl, fs_id_buf);
+	gfs2_rgrp_dump(NULL, rgd, fs_id_buf);
 	gfs2_lm(sdp,
 		"fatal: filesystem consistency error\n"
 		"  RG = %llu\n"

fs/gfs2/util.h

@@ -205,6 +205,16 @@ static inline bool gfs2_withdrawn(struct gfs2_sbd *sdp)
 		test_bit(SDF_WITHDRAWING, &sdp->sd_flags);
 }
 
+/**
+ * gfs2_withdrawing - check if a withdraw is pending
+ * @sdp: the superblock
+ */
+static inline bool gfs2_withdrawing(struct gfs2_sbd *sdp)
+{
+	return test_bit(SDF_WITHDRAWING, &sdp->sd_flags) &&
+	       !test_bit(SDF_WITHDRAWN, &sdp->sd_flags);
+}
+
 #define gfs2_tune_get(sdp, field) \
 gfs2_tune_get_i(&(sdp)->sd_tune, &(sdp)->sd_tune.field)