[PATCH 0/7] super block scalabilit patches V3

[PATCH 0/7] super block scalabilit patches V3

[Josef Bacik]

Here are the cleaned up versions of Dave Chinners super block scalability
patches.  I've been testing them locally for a while and they are pretty solid.
They fix a few big issues, such as the global inode list and soft lockups on
boxes on unmount that have lots of inodes in cache.  Al if you would consider
pulling these in that would be great, you can pull from here

git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling

Thanks,

Josef

[PATCH 1/8] writeback: plug writeback at a high level

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

Doing writeback on lots of little files causes terrible IOPS storms
because of the per-mapping writeback plugging we do. This
essentially causes imeediate dispatch of IO for each mapping,
regardless of the context in which writeback is occurring.

IOWs, running a concurrent write-lots-of-small 4k files using fsmark
on XFS results in a huge number of IOPS being issued for data
writes.  Metadata writes are sorted and plugged at a high level by
XFS, so aggregate nicely into large IOs. However, data writeback IOs
are dispatched in individual 4k IOs, even when the blocks of two
consecutively written files are adjacent.

Test VM: 8p, 8GB RAM, 4xSSD in RAID0, 100TB sparse XFS filesystem,
metadata CRCs enabled.

Kernel: 3.10-rc5 + xfsdev + my 3.11 xfs queue (~70 patches)

Test:

$ ./fs_mark  -D  10000  -S0  -n  10000  -s  4096  -L  120  -d
/mnt/scratch/0  -d  /mnt/scratch/1  -d  /mnt/scratch/2  -d
/mnt/scratch/3  -d  /mnt/scratch/4  -d  /mnt/scratch/5  -d
/mnt/scratch/6  -d  /mnt/scratch/7

Result:

		wall	sys	create rate	Physical write IO
		time	CPU	(avg files/s)	 IOPS	Bandwidth
		-----	-----	------------	------	---------
unpatched	6m56s	15m47s	24,000+/-500	26,000	130MB/s
patched		5m06s	13m28s	32,800+/-600	 1,500	180MB/s
improvement	-26.44%	-14.68%	  +36.67%	-94.23%	+38.46%

If I use zero length files, this workload at about 500 IOPS, so
plugging drops the data IOs from roughly 25,500/s to 1000/s.
3 lines of code, 35% better throughput for 15% less CPU.

The benefits of plugging at this layer are likely to be higher for
spinning media as the IO patterns for this workload are going make a
much bigger difference on high IO latency devices.....

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/fs-writeback.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 32a8bbd..43b6555 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -677,7 +677,9 @@ static long writeback_sb_inodes(struct super_block *sb,
 	unsigned long start_time = jiffies;
 	long write_chunk;
 	long wrote = 0;  /* count both pages and inodes */
+	struct blk_plug plug;
 
+	blk_start_plug(&plug);
 	while (!list_empty(&wb->b_io)) {
 		struct inode *inode = wb_inode(wb->b_io.prev);
 
@@ -774,6 +776,7 @@ static long writeback_sb_inodes(struct super_block *sb,
 				break;
 		}
 	}
+	blk_finish_plug(&plug);
 	return wrote;
 }
 
-- 
2.1.0

Re: [PATCH 1/8] writeback: plug writeback at a high level

[Christoph Hellwig]

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 2/8] inode: add hlist_fake to avoid the inode hash lock in evict

[Josef Bacik]

Some filesystems don't use the VFS inode hash and fake the fact they
are hashed so that all the writeback code works correctly. However,
this means the evict() path still tries to remove the inode from the
hash, meaning that the inode_hash_lock() needs to be taken
unnecessarily. Hence under certain workloads the inode_hash_lock can
be contended even if the inode is never actually hashed.

To avoid this add hlist_fake to test if the inode isn't actually
hashed to avoid taking the hash lock on inodes that have never been
hashed.  Based on Dave Chinner's

inode: add IOP_NOTHASHED to avoid inode hash lock in evict

basd on Al's suggestions.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 include/linux/fs.h   | 2 +-
 include/linux/list.h | 5 +++++
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/include/linux/fs.h b/include/linux/fs.h
index 35ec87e..b841718 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2563,7 +2563,7 @@ static inline void insert_inode_hash(struct inode *inode)
 extern void __remove_inode_hash(struct inode *);
 static inline void remove_inode_hash(struct inode *inode)
 {
-	if (!inode_unhashed(inode))
+	if (!inode_unhashed(inode) && !hlist_fake(&inode->i_hash))
 		__remove_inode_hash(inode);
 }
 
diff --git a/include/linux/list.h b/include/linux/list.h
index feb773c..3e3e64a 100644
--- a/include/linux/list.h
+++ b/include/linux/list.h
@@ -672,6 +672,11 @@ static inline void hlist_add_fake(struct hlist_node *n)
 	n->pprev = &n->next;
 }
 
+static inline bool hlist_fake(struct hlist_node *h)
+{
+	return h->pprev == &h->next;
+}
+
 /*
  * Move a list from one list head to another. Fixup the pprev
  * reference of the first entry if it exists.
-- 
2.1.0

Re: [PATCH 2/8] inode: add hlist_fake to avoid the inode hash lock in evict

[Christoph Hellwig]

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 3/8] inode: convert inode_sb_list_lock to per-sb

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

The process of reducing contention on per-superblock inode lists
starts with moving the locking to match the per-superblock inode
list. This takes the global lock out of the picture and reduces the
contention problems to within a single filesystem. This doesn't get
rid of contention as the locks still have global CPU scope, but it
does isolate operations on different superblocks form each other.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/block_dev.c                   | 12 ++++++------
 fs/drop_caches.c                 | 10 ++++++----
 fs/fs-writeback.c                | 12 ++++++------
 fs/inode.c                       | 28 +++++++++++++---------------
 fs/internal.h                    |  1 -
 fs/notify/inode_mark.c           | 20 ++++++++++----------
 fs/quota/dquot.c                 | 16 ++++++++--------
 fs/super.c                       |  3 ++-
 include/linux/fs.h               |  5 ++++-
 include/linux/fsnotify_backend.h |  4 ++--
 10 files changed, 57 insertions(+), 54 deletions(-)

diff --git a/fs/block_dev.c b/fs/block_dev.c
index c7e4163..f2a89be 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -1762,7 +1762,7 @@ void iterate_bdevs(void (*func)(struct block_device *, void *), void *arg)
 {
 	struct inode *inode, *old_inode = NULL;
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&blockdev_superblock->s_inode_list_lock);
 	list_for_each_entry(inode, &blockdev_superblock->s_inodes, i_sb_list) {
 		struct address_space *mapping = inode->i_mapping;
 
@@ -1774,13 +1774,13 @@ void iterate_bdevs(void (*func)(struct block_device *, void *), void *arg)
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
-		spin_unlock(&inode_sb_list_lock);
+		spin_unlock(&blockdev_superblock->s_inode_list_lock);
 		/*
 		 * We hold a reference to 'inode' so it couldn't have been
 		 * removed from s_inodes list while we dropped the
-		 * inode_sb_list_lock.  We cannot iput the inode now as we can
+		 * s_inode_list_lock  We cannot iput the inode now as we can
 		 * be holding the last reference and we cannot iput it under
-		 * inode_sb_list_lock. So we keep the reference and iput it
+		 * s_inode_list_lock. So we keep the reference and iput it
 		 * later.
 		 */
 		iput(old_inode);
@@ -1788,8 +1788,8 @@ void iterate_bdevs(void (*func)(struct block_device *, void *), void *arg)
 
 		func(I_BDEV(inode), arg);
 
-		spin_lock(&inode_sb_list_lock);
+		spin_lock(&blockdev_superblock->s_inode_list_lock);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&blockdev_superblock->s_inode_list_lock);
 	iput(old_inode);
 }
diff --git a/fs/drop_caches.c b/fs/drop_caches.c
index 5718cb9..d72d52b 100644
--- a/fs/drop_caches.c
+++ b/fs/drop_caches.c
@@ -17,7 +17,7 @@ static void drop_pagecache_sb(struct super_block *sb, void *unused)
 {
 	struct inode *inode, *toput_inode = NULL;
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&sb->s_inode_list_lock);
 	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
 		spin_lock(&inode->i_lock);
 		if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) ||
@@ -27,13 +27,15 @@ static void drop_pagecache_sb(struct super_block *sb, void *unused)
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
-		spin_unlock(&inode_sb_list_lock);
+		spin_unlock(&sb->s_inode_list_lock);
+
 		invalidate_mapping_pages(inode->i_mapping, 0, -1);
 		iput(toput_inode);
 		toput_inode = inode;
-		spin_lock(&inode_sb_list_lock);
+
+		spin_lock(&sb->s_inode_list_lock);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 	iput(toput_inode);
 }
 
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 43b6555..333afa3 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1371,7 +1371,7 @@ static void wait_sb_inodes(struct super_block *sb)
 	 */
 	WARN_ON(!rwsem_is_locked(&sb->s_umount));
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&sb->s_inode_list_lock);
 
 	/*
 	 * Data integrity sync. Must wait for all pages under writeback,
@@ -1391,14 +1391,14 @@ static void wait_sb_inodes(struct super_block *sb)
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
-		spin_unlock(&inode_sb_list_lock);
+		spin_unlock(&sb->s_inode_list_lock);
 
 		/*
 		 * We hold a reference to 'inode' so it couldn't have been
 		 * removed from s_inodes list while we dropped the
-		 * inode_sb_list_lock.  We cannot iput the inode now as we can
+		 * s_inode_list_lock.  We cannot iput the inode now as we can
 		 * be holding the last reference and we cannot iput it under
-		 * inode_sb_list_lock. So we keep the reference and iput it
+		 * s_inode_list_lock. So we keep the reference and iput it
 		 * later.
 		 */
 		iput(old_inode);
@@ -1408,9 +1408,9 @@ static void wait_sb_inodes(struct super_block *sb)
 
 		cond_resched();
 
-		spin_lock(&inode_sb_list_lock);
+		spin_lock(&sb->s_inode_list_lock);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 	iput(old_inode);
 }
 
diff --git a/fs/inode.c b/fs/inode.c
index ea37cd1..12d58fa 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -28,8 +28,8 @@
  *   inode->i_state, inode->i_hash, __iget()
  * Inode LRU list locks protect:
  *   inode->i_sb->s_inode_lru, inode->i_lru
- * inode_sb_list_lock protects:
- *   sb->s_inodes, inode->i_sb_list
+ * inode->i_sb->s_inode_list_lock protects:
+ *   inode->i_sb->s_inodes, inode->i_sb_list
  * bdi->wb.list_lock protects:
  *   bdi->wb.b_{dirty,io,more_io,dirty_time}, inode->i_wb_list
  * inode_hash_lock protects:
@@ -37,7 +37,7 @@
  *
  * Lock ordering:
  *
- * inode_sb_list_lock
+ * inode->i_sb->s_inode_list_lock
  *   inode->i_lock
  *     Inode LRU list locks
  *
@@ -45,7 +45,7 @@
  *   inode->i_lock
  *
  * inode_hash_lock
- *   inode_sb_list_lock
+ *   inode->i_sb->s_inode_list_lock
  *   inode->i_lock
  *
  * iunique_lock
@@ -57,8 +57,6 @@ static unsigned int i_hash_shift __read_mostly;
 static struct hlist_head *inode_hashtable __read_mostly;
 static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock);
 
-__cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_sb_list_lock);
-
 /*
  * Empty aops. Can be used for the cases where the user does not
  * define any of the address_space operations.
@@ -424,18 +422,18 @@ static void inode_lru_list_del(struct inode *inode)
  */
 void inode_sb_list_add(struct inode *inode)
 {
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&inode->i_sb->s_inode_list_lock);
 	list_add(&inode->i_sb_list, &inode->i_sb->s_inodes);
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&inode->i_sb->s_inode_list_lock);
 }
 EXPORT_SYMBOL_GPL(inode_sb_list_add);
 
 static inline void inode_sb_list_del(struct inode *inode)
 {
 	if (!list_empty(&inode->i_sb_list)) {
-		spin_lock(&inode_sb_list_lock);
+		spin_lock(&inode->i_sb->s_inode_list_lock);
 		list_del_init(&inode->i_sb_list);
-		spin_unlock(&inode_sb_list_lock);
+		spin_unlock(&inode->i_sb->s_inode_list_lock);
 	}
 }
 
@@ -592,7 +590,7 @@ void evict_inodes(struct super_block *sb)
 	struct inode *inode, *next;
 	LIST_HEAD(dispose);
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&sb->s_inode_list_lock);
 	list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) {
 		if (atomic_read(&inode->i_count))
 			continue;
@@ -608,7 +606,7 @@ void evict_inodes(struct super_block *sb)
 		spin_unlock(&inode->i_lock);
 		list_add(&inode->i_lru, &dispose);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 
 	dispose_list(&dispose);
 }
@@ -629,7 +627,7 @@ int invalidate_inodes(struct super_block *sb, bool kill_dirty)
 	struct inode *inode, *next;
 	LIST_HEAD(dispose);
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&sb->s_inode_list_lock);
 	list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) {
 		spin_lock(&inode->i_lock);
 		if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) {
@@ -652,7 +650,7 @@ int invalidate_inodes(struct super_block *sb, bool kill_dirty)
 		spin_unlock(&inode->i_lock);
 		list_add(&inode->i_lru, &dispose);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 
 	dispose_list(&dispose);
 
@@ -884,7 +882,7 @@ struct inode *new_inode(struct super_block *sb)
 {
 	struct inode *inode;
 
-	spin_lock_prefetch(&inode_sb_list_lock);
+	spin_lock_prefetch(&sb->s_inode_list_lock);
 
 	inode = new_inode_pseudo(sb);
 	if (inode)
diff --git a/fs/internal.h b/fs/internal.h
index 01dce1d..b6c350d 100644
--- a/fs/internal.h
+++ b/fs/internal.h
@@ -111,7 +111,6 @@ extern int open_check_o_direct(struct file *f);
 /*
  * inode.c
  */
-extern spinlock_t inode_sb_list_lock;
 extern long prune_icache_sb(struct super_block *sb, struct shrink_control *sc);
 extern void inode_add_lru(struct inode *inode);
 
diff --git a/fs/notify/inode_mark.c b/fs/notify/inode_mark.c
index 3daf513..a4e1a8f 100644
--- a/fs/notify/inode_mark.c
+++ b/fs/notify/inode_mark.c
@@ -163,17 +163,17 @@ int fsnotify_add_inode_mark(struct fsnotify_mark *mark,
 
 /**
  * fsnotify_unmount_inodes - an sb is unmounting.  handle any watched inodes.
- * @list: list of inodes being unmounted (sb->s_inodes)
+ * @sb: superblock being unmounted.
  *
  * Called during unmount with no locks held, so needs to be safe against
- * concurrent modifiers. We temporarily drop inode_sb_list_lock and CAN block.
+ * concurrent modifiers. We temporarily drop sb->s_inode_list_lock and CAN block.
  */
-void fsnotify_unmount_inodes(struct list_head *list)
+void fsnotify_unmount_inodes(struct super_block *sb)
 {
 	struct inode *inode, *next_i, *need_iput = NULL;
 
-	spin_lock(&inode_sb_list_lock);
-	list_for_each_entry_safe(inode, next_i, list, i_sb_list) {
+	spin_lock(&sb->s_inode_list_lock);
+	list_for_each_entry_safe(inode, next_i, &sb->s_inodes, i_sb_list) {
 		struct inode *need_iput_tmp;
 
 		/*
@@ -209,7 +209,7 @@ void fsnotify_unmount_inodes(struct list_head *list)
 		spin_unlock(&inode->i_lock);
 
 		/* In case the dropping of a reference would nuke next_i. */
-		while (&next_i->i_sb_list != list) {
+		while (&next_i->i_sb_list != &sb->s_inodes) {
 			spin_lock(&next_i->i_lock);
 			if (!(next_i->i_state & (I_FREEING | I_WILL_FREE)) &&
 						atomic_read(&next_i->i_count)) {
@@ -224,12 +224,12 @@ void fsnotify_unmount_inodes(struct list_head *list)
 		}
 
 		/*
-		 * We can safely drop inode_sb_list_lock here because either
+		 * We can safely drop s_inode_list_lock here because either
 		 * we actually hold references on both inode and next_i or
 		 * end of list.  Also no new inodes will be added since the
 		 * umount has begun.
 		 */
-		spin_unlock(&inode_sb_list_lock);
+		spin_unlock(&sb->s_inode_list_lock);
 
 		if (need_iput_tmp)
 			iput(need_iput_tmp);
@@ -241,7 +241,7 @@ void fsnotify_unmount_inodes(struct list_head *list)
 
 		iput(inode);
 
-		spin_lock(&inode_sb_list_lock);
+		spin_lock(&sb->s_inode_list_lock);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 }
diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
index 20d1f74..2863ec6 100644
--- a/fs/quota/dquot.c
+++ b/fs/quota/dquot.c
@@ -923,7 +923,7 @@ static void add_dquot_ref(struct super_block *sb, int type)
 	int reserved = 0;
 #endif
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&sb->s_inode_list_lock);
 	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
 		spin_lock(&inode->i_lock);
 		if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) ||
@@ -934,7 +934,7 @@ static void add_dquot_ref(struct super_block *sb, int type)
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
-		spin_unlock(&inode_sb_list_lock);
+		spin_unlock(&sb->s_inode_list_lock);
 
 #ifdef CONFIG_QUOTA_DEBUG
 		if (unlikely(inode_get_rsv_space(inode) > 0))
@@ -946,15 +946,15 @@ static void add_dquot_ref(struct super_block *sb, int type)
 		/*
 		 * We hold a reference to 'inode' so it couldn't have been
 		 * removed from s_inodes list while we dropped the
-		 * inode_sb_list_lock We cannot iput the inode now as we can be
+		 * s_inode_list_lock. We cannot iput the inode now as we can be
 		 * holding the last reference and we cannot iput it under
-		 * inode_sb_list_lock. So we keep the reference and iput it
+		 * s_inode_list_lock. So we keep the reference and iput it
 		 * later.
 		 */
 		old_inode = inode;
-		spin_lock(&inode_sb_list_lock);
+		spin_lock(&sb->s_inode_list_lock);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 	iput(old_inode);
 
 #ifdef CONFIG_QUOTA_DEBUG
@@ -1023,7 +1023,7 @@ static void remove_dquot_ref(struct super_block *sb, int type,
 	struct inode *inode;
 	int reserved = 0;
 
-	spin_lock(&inode_sb_list_lock);
+	spin_lock(&sb->s_inode_list_lock);
 	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
 		/*
 		 *  We have to scan also I_NEW inodes because they can already
@@ -1039,7 +1039,7 @@ static void remove_dquot_ref(struct super_block *sb, int type,
 		}
 		spin_unlock(&dq_data_lock);
 	}
-	spin_unlock(&inode_sb_list_lock);
+	spin_unlock(&sb->s_inode_list_lock);
 #ifdef CONFIG_QUOTA_DEBUG
 	if (reserved) {
 		printk(KERN_WARNING "VFS (%s): Writes happened after quota"
diff --git a/fs/super.c b/fs/super.c
index 928c20f..15e57f5 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -191,6 +191,7 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags)
 	INIT_HLIST_NODE(&s->s_instances);
 	INIT_HLIST_BL_HEAD(&s->s_anon);
 	INIT_LIST_HEAD(&s->s_inodes);
+	spin_lock_init(&s->s_inode_list_lock);
 
 	if (list_lru_init_memcg(&s->s_dentry_lru))
 		goto fail;
@@ -399,7 +400,7 @@ void generic_shutdown_super(struct super_block *sb)
 		sync_filesystem(sb);
 		sb->s_flags &= ~MS_ACTIVE;
 
-		fsnotify_unmount_inodes(&sb->s_inodes);
+		fsnotify_unmount_inodes(sb);
 
 		evict_inodes(sb);
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index b841718..8607d92 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1280,7 +1280,6 @@ struct super_block {
 #endif
 	const struct xattr_handler **s_xattr;
 
-	struct list_head	s_inodes;	/* all inodes */
 	struct hlist_bl_head	s_anon;		/* anonymous dentries for (nfs) exporting */
 	struct list_head	s_mounts;	/* list of mounts; _not_ for fs use */
 	struct block_device	*s_bdev;
@@ -1351,6 +1350,10 @@ struct super_block {
 	 * Indicates how deep in a filesystem stack this SB is
 	 */
 	int s_stack_depth;
+
+	/* s_inode_list_lock protects s_inodes */
+	spinlock_t		s_inode_list_lock ____cacheline_aligned_in_smp;
+	struct list_head	s_inodes;	/* all inodes */
 };
 
 extern struct timespec current_fs_time(struct super_block *sb);
diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h
index 0f313f9..236cbc4 100644
--- a/include/linux/fsnotify_backend.h
+++ b/include/linux/fsnotify_backend.h
@@ -359,7 +359,7 @@ extern void fsnotify_clear_marks_by_group_flags(struct fsnotify_group *group, un
 extern void fsnotify_clear_marks_by_group(struct fsnotify_group *group);
 extern void fsnotify_get_mark(struct fsnotify_mark *mark);
 extern void fsnotify_put_mark(struct fsnotify_mark *mark);
-extern void fsnotify_unmount_inodes(struct list_head *list);
+extern void fsnotify_unmount_inodes(struct super_block *sb);
 
 /* put here because inotify does some weird stuff when destroying watches */
 extern void fsnotify_init_event(struct fsnotify_event *event,
@@ -395,7 +395,7 @@ static inline u32 fsnotify_get_cookie(void)
 	return 0;
 }
 
-static inline void fsnotify_unmount_inodes(struct list_head *list)
+static inline void fsnotify_unmount_inodes(struct super_block *sb)
 {}
 
 #endif	/* CONFIG_FSNOTIFY */
-- 
2.1.0

Re: [PATCH 3/8] inode: convert inode_sb_list_lock to per-sb

[Christoph Hellwig]

On Thu, Jun 11, 2015 at 03:41:08PM -0400, Josef Bacik wrote:
> From: Dave Chinner <dchinner@redhat.com>
> 
> The process of reducing contention on per-superblock inode lists
> starts with moving the locking to match the per-superblock inode
> list. This takes the global lock out of the picture and reduces the
> contention problems to within a single filesystem. This doesn't get
> rid of contention as the locks still have global CPU scope, but it
> does isolate operations on different superblocks form each other.

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 4/8] sync: serialise per-superblock sync operations

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

When competing sync(2) calls walk the same filesystem, they need to
walk the list of inodes on the superblock to find all the inodes
that we need to wait for IO completion on. However, when multiple
wait_sb_inodes() calls do this at the same time, they contend on the
the inode_sb_list_lock and the contention causes system wide
slowdowns. In effect, concurrent sync(2) calls can take longer and
burn more CPU than if they were serialised.

Stop the worst of the contention by adding a per-sb mutex to wrap
around wait_sb_inodes() so that we only execute one sync(2) IO
completion walk per superblock superblock at a time and hence avoid
contention being triggered by concurrent sync(2) calls.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/fs-writeback.c  | 11 +++++++++++
 fs/super.c         |  1 +
 include/linux/fs.h |  2 ++
 3 files changed, 14 insertions(+)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 333afa3..04fbc8a 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1361,6 +1361,15 @@ out_unlock_inode:
 }
 EXPORT_SYMBOL(__mark_inode_dirty);
 
+/*
+ * The @s_sync_lock is used to serialise concurrent sync operations
+ * to avoid lock contention problems with concurrent wait_sb_inodes() calls.
+ * Concurrent callers will block on the s_sync_lock rather than doing contending
+ * walks. The queueing maintains sync(2) required behaviour as all the IO that
+ * has been issued up to the time this function is enter is guaranteed to be
+ * completed by the time we have gained the lock and waited for all IO that is
+ * in progress regardless of the order callers are granted the lock.
+ */
 static void wait_sb_inodes(struct super_block *sb)
 {
 	struct inode *inode, *old_inode = NULL;
@@ -1371,6 +1380,7 @@ static void wait_sb_inodes(struct super_block *sb)
 	 */
 	WARN_ON(!rwsem_is_locked(&sb->s_umount));
 
+	mutex_lock(&sb->s_sync_lock);
 	spin_lock(&sb->s_inode_list_lock);
 
 	/*
@@ -1412,6 +1422,7 @@ static void wait_sb_inodes(struct super_block *sb)
 	}
 	spin_unlock(&sb->s_inode_list_lock);
 	iput(old_inode);
+	mutex_unlock(&sb->s_sync_lock);
 }
 
 /**
diff --git a/fs/super.c b/fs/super.c
index 15e57f5..ac2b64b 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -190,6 +190,7 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags)
 	s->s_flags = flags;
 	INIT_HLIST_NODE(&s->s_instances);
 	INIT_HLIST_BL_HEAD(&s->s_anon);
+	mutex_init(&s->s_sync_lock);
 	INIT_LIST_HEAD(&s->s_inodes);
 	spin_lock_init(&s->s_inode_list_lock);
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 8607d92..ae23d35 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1346,6 +1346,8 @@ struct super_block {
 	struct list_lru		s_inode_lru ____cacheline_aligned_in_smp;
 	struct rcu_head		rcu;
 
+	struct mutex		s_sync_lock;	/* sync serialisation lock */
+
 	/*
 	 * Indicates how deep in a filesystem stack this SB is
 	 */
-- 
2.1.0

Re: [PATCH 4/8] sync: serialise per-superblock sync operations

[Christoph Hellwig]

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 5/8] inode: rename i_wb_list to i_io_list

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

There's a small consistency problem between the inode and writeback
naming. Writeback calls the "for IO" inode queues b_io and
b_more_io, but the inode calls these the "writeback list" or
i_wb_list. This makes it hard to an new "under writeback" list to
the inode, or call it an "under IO" list on the bdi because either
way we'll have writeback on IO and IO on writeback and it'll just be
confusing. I'm getting confused just writing this!

So, rename the inode "for IO" list variable to i_io_list so we can
add a new "writeback list" in a subsequent patch.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/fs-writeback.c  | 22 +++++++++++-----------
 fs/inode.c         |  6 +++---
 include/linux/fs.h |  2 +-
 mm/backing-dev.c   |  8 ++++----
 4 files changed, 19 insertions(+), 19 deletions(-)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 04fbc8a..aa72536 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -96,7 +96,7 @@ EXPORT_SYMBOL_GPL(inode_to_bdi);
 
 static inline struct inode *wb_inode(struct list_head *head)
 {
-	return list_entry(head, struct inode, i_wb_list);
+	return list_entry(head, struct inode, i_io_list);
 }
 
 /*
@@ -205,7 +205,7 @@ void inode_wb_list_del(struct inode *inode)
 	struct backing_dev_info *bdi = inode_to_bdi(inode);
 
 	spin_lock(&bdi->wb.list_lock);
-	list_del_init(&inode->i_wb_list);
+	list_del_init(&inode->i_io_list);
 	spin_unlock(&bdi->wb.list_lock);
 }
 
@@ -228,7 +228,7 @@ static void redirty_tail(struct inode *inode, struct bdi_writeback *wb)
 		if (time_before(inode->dirtied_when, tail->dirtied_when))
 			inode->dirtied_when = jiffies;
 	}
-	list_move(&inode->i_wb_list, &wb->b_dirty);
+	list_move(&inode->i_io_list, &wb->b_dirty);
 }
 
 /*
@@ -237,7 +237,7 @@ static void redirty_tail(struct inode *inode, struct bdi_writeback *wb)
 static void requeue_io(struct inode *inode, struct bdi_writeback *wb)
 {
 	assert_spin_locked(&wb->list_lock);
-	list_move(&inode->i_wb_list, &wb->b_more_io);
+	list_move(&inode->i_io_list, &wb->b_more_io);
 }
 
 static void inode_sync_complete(struct inode *inode)
@@ -296,7 +296,7 @@ static int move_expired_inodes(struct list_head *delaying_queue,
 		if (older_than_this &&
 		    inode_dirtied_after(inode, *older_than_this))
 			break;
-		list_move(&inode->i_wb_list, &tmp);
+		list_move(&inode->i_io_list, &tmp);
 		moved++;
 		if (flags & EXPIRE_DIRTY_ATIME)
 			set_bit(__I_DIRTY_TIME_EXPIRED, &inode->i_state);
@@ -319,7 +319,7 @@ static int move_expired_inodes(struct list_head *delaying_queue,
 		list_for_each_prev_safe(pos, node, &tmp) {
 			inode = wb_inode(pos);
 			if (inode->i_sb == sb)
-				list_move(&inode->i_wb_list, dispatch_queue);
+				list_move(&inode->i_io_list, dispatch_queue);
 		}
 	}
 out:
@@ -471,10 +471,10 @@ static void requeue_inode(struct inode *inode, struct bdi_writeback *wb,
 		redirty_tail(inode, wb);
 	} else if (inode->i_state & I_DIRTY_TIME) {
 		inode->dirtied_when = jiffies;
-		list_move(&inode->i_wb_list, &wb->b_dirty_time);
+		list_move(&inode->i_io_list, &wb->b_dirty_time);
 	} else {
 		/* The inode is clean. Remove from writeback lists. */
-		list_del_init(&inode->i_wb_list);
+		list_del_init(&inode->i_io_list);
 	}
 }
 
@@ -616,7 +616,7 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb,
 	 * touch it. See comment above for explanation.
 	 */
 	if (!(inode->i_state & I_DIRTY_ALL))
-		list_del_init(&inode->i_wb_list);
+		list_del_init(&inode->i_io_list);
 	spin_unlock(&wb->list_lock);
 	inode_sync_complete(inode);
 out:
@@ -1343,9 +1343,9 @@ void __mark_inode_dirty(struct inode *inode, int flags)
 			if (dirtytime)
 				inode->dirtied_time_when = jiffies;
 			if (inode->i_state & (I_DIRTY_INODE | I_DIRTY_PAGES))
-				list_move(&inode->i_wb_list, &bdi->wb.b_dirty);
+				list_move(&inode->i_io_list, &bdi->wb.b_dirty);
 			else
-				list_move(&inode->i_wb_list,
+				list_move(&inode->i_io_list,
 					  &bdi->wb.b_dirty_time);
 			spin_unlock(&bdi->wb.list_lock);
 			trace_writeback_dirty_inode_enqueue(inode);
diff --git a/fs/inode.c b/fs/inode.c
index 12d58fa..c76a575 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -31,7 +31,7 @@
  * inode->i_sb->s_inode_list_lock protects:
  *   inode->i_sb->s_inodes, inode->i_sb_list
  * bdi->wb.list_lock protects:
- *   bdi->wb.b_{dirty,io,more_io,dirty_time}, inode->i_wb_list
+ *   bdi->wb.b_{dirty,io,more_io,dirty_time}, inode->i_io_list
  * inode_hash_lock protects:
  *   inode_hashtable, inode->i_hash
  *
@@ -355,7 +355,7 @@ void inode_init_once(struct inode *inode)
 	memset(inode, 0, sizeof(*inode));
 	INIT_HLIST_NODE(&inode->i_hash);
 	INIT_LIST_HEAD(&inode->i_devices);
-	INIT_LIST_HEAD(&inode->i_wb_list);
+	INIT_LIST_HEAD(&inode->i_io_list);
 	INIT_LIST_HEAD(&inode->i_lru);
 	address_space_init_once(&inode->i_data);
 	i_size_ordered_init(inode);
@@ -523,7 +523,7 @@ static void evict(struct inode *inode)
 	BUG_ON(!(inode->i_state & I_FREEING));
 	BUG_ON(!list_empty(&inode->i_lru));
 
-	if (!list_empty(&inode->i_wb_list))
+	if (!list_empty(&inode->i_io_list))
 		inode_wb_list_del(inode);
 
 	inode_sb_list_del(inode);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index ae23d35..f4d56cc 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -634,7 +634,7 @@ struct inode {
 	unsigned long		dirtied_time_when;
 
 	struct hlist_node	i_hash;
-	struct list_head	i_wb_list;	/* backing dev IO list */
+	struct list_head	i_io_list;	/* backing dev IO list */
 	struct list_head	i_lru;		/* inode LRU list */
 	struct list_head	i_sb_list;
 	union {
diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index 6dc4580..6e7a644 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -54,13 +54,13 @@ static int bdi_debug_stats_show(struct seq_file *m, void *v)
 
 	nr_dirty = nr_io = nr_more_io = nr_dirty_time = 0;
 	spin_lock(&wb->list_lock);
-	list_for_each_entry(inode, &wb->b_dirty, i_wb_list)
+	list_for_each_entry(inode, &wb->b_dirty, i_io_list)
 		nr_dirty++;
-	list_for_each_entry(inode, &wb->b_io, i_wb_list)
+	list_for_each_entry(inode, &wb->b_io, i_io_list)
 		nr_io++;
-	list_for_each_entry(inode, &wb->b_more_io, i_wb_list)
+	list_for_each_entry(inode, &wb->b_more_io, i_io_list)
 		nr_more_io++;
-	list_for_each_entry(inode, &wb->b_dirty_time, i_wb_list)
+	list_for_each_entry(inode, &wb->b_dirty_time, i_io_list)
 		if (inode->i_state & I_DIRTY_TIME)
 			nr_dirty_time++;
 	spin_unlock(&wb->list_lock);
-- 
2.1.0

Re: [PATCH 5/8] inode: rename i_wb_list to i_io_list

[Christoph Hellwig]

On Thu, Jun 11, 2015 at 03:41:10PM -0400, Josef Bacik wrote:
> From: Dave Chinner <dchinner@redhat.com>
> 
> There's a small consistency problem between the inode and writeback
> naming. Writeback calls the "for IO" inode queues b_io and
> b_more_io, but the inode calls these the "writeback list" or
> i_wb_list. This makes it hard to an new "under writeback" list to
> the inode, or call it an "under IO" list on the bdi because either
> way we'll have writeback on IO and IO on writeback and it'll just be
> confusing. I'm getting confused just writing this!
> 
> So, rename the inode "for IO" list variable to i_io_list so we can
> add a new "writeback list" in a subsequent patch.

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 6/8] bdi: add a new writeback list for sync

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

wait_sb_inodes() current does a walk of all inodes in the filesystem
to find dirty one to wait on during sync. This is highly
inefficient and wastes a lot of CPU when there are lots of clean
cached inodes that we don't need to wait on.

To avoid this "all inode" walk, we need to track inodes that are
currently under writeback that we need to wait for. We do this by
adding inodes to a writeback list on the bdi when the mapping is
first tagged as having pages under writeback.  wait_sb_inodes() can
then walk this list of "inodes under IO" and wait specifically just
for the inodes that the current sync(2) needs to wait for.

To avoid needing all the realted locking to be safe against
interrupts, Jan Kara suggested that we be lazy about removal from
the writeback list. That is, we don't remove inodes from the
writeback list on IO completion, but do it directly during a
wait_sb_inodes() walk.

This means that the a rare sync(2) call will have some work to do
skipping clean inodes However, in the current problem case of
concurrent sync workloads, concurrent wait_sb_inodes() calls only
walk the very recently dispatched inodes and hence should have very
little work to do.

This also means that we have to remove the inodes from the writeback
list during eviction. Do this in inode_wait_for_writeback() once
all writeback on the inode is complete.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
---
 fs/fs-writeback.c           | 131 ++++++++++++++++++++++++++++++++++----------
 fs/inode.c                  |   1 +
 include/linux/backing-dev.h |   3 +
 include/linux/fs.h          |   1 +
 mm/backing-dev.c            |   1 +
 mm/page-writeback.c         |  14 +++++
 6 files changed, 122 insertions(+), 29 deletions(-)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index aa72536..3f5b2ff 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -210,6 +210,34 @@ void inode_wb_list_del(struct inode *inode)
 }
 
 /*
+ * mark an inode as under writeback on the given bdi
+ */
+void bdi_mark_inode_writeback(struct backing_dev_info *bdi, struct inode *inode)
+{
+	WARN_ON_ONCE(bdi != inode_to_bdi(inode));
+	if (list_empty(&inode->i_wb_list)) {
+		spin_lock(&bdi->wb.list_lock);
+		if (list_empty(&inode->i_wb_list))
+			list_add_tail(&inode->i_wb_list, &bdi->wb.b_writeback);
+		spin_unlock(&bdi->wb.list_lock);
+	}
+}
+
+/*
+ * clear an inode as under writeback on the given bdi
+ */
+static void bdi_clear_inode_writeback(struct backing_dev_info *bdi,
+				      struct inode *inode)
+{
+	WARN_ON_ONCE(bdi != inode_to_bdi(inode));
+	if (!list_empty(&inode->i_wb_list)) {
+		spin_lock(&bdi->wb.list_lock);
+		list_del_init(&inode->i_wb_list);
+		spin_unlock(&bdi->wb.list_lock);
+	}
+}
+
+/*
  * Redirty an inode: set its when-it-was dirtied timestamp and move it to the
  * furthest end of its superblock's dirty-inode list.
  *
@@ -383,13 +411,23 @@ static void __inode_wait_for_writeback(struct inode *inode)
 }
 
 /*
- * Wait for writeback on an inode to complete. Caller must have inode pinned.
+ * Wait for writeback on an inode to complete during eviction. Caller must have
+ * inode pinned.
  */
 void inode_wait_for_writeback(struct inode *inode)
 {
+	BUG_ON(!(inode->i_state & I_FREEING));
+
 	spin_lock(&inode->i_lock);
 	__inode_wait_for_writeback(inode);
 	spin_unlock(&inode->i_lock);
+
+	/*
+	 * bd_inode's will have already had the bdev free'd so don't bother
+	 * doing the bdi_clear_inode_writeback.
+	 */
+	if (!sb_is_blkdev_sb(inode->i_sb))
+		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
 }
 
 /*
@@ -1372,7 +1410,8 @@ EXPORT_SYMBOL(__mark_inode_dirty);
  */
 static void wait_sb_inodes(struct super_block *sb)
 {
-	struct inode *inode, *old_inode = NULL;
+	struct backing_dev_info *bdi = sb->s_bdi;
+	LIST_HEAD(sync_list);
 
 	/*
 	 * We need to be protected against the filesystem going from
@@ -1380,48 +1419,82 @@ static void wait_sb_inodes(struct super_block *sb)
 	 */
 	WARN_ON(!rwsem_is_locked(&sb->s_umount));
 
-	mutex_lock(&sb->s_sync_lock);
-	spin_lock(&sb->s_inode_list_lock);
-
 	/*
-	 * Data integrity sync. Must wait for all pages under writeback,
-	 * because there may have been pages dirtied before our sync
-	 * call, but which had writeout started before we write it out.
-	 * In which case, the inode may not be on the dirty list, but
-	 * we still have to wait for that writeout.
+	 * Data integrity sync. Must wait for all pages under writeback, because
+	 * there may have been pages dirtied before our sync call, but which had
+	 * writeout started before we write it out.  In which case, the inode
+	 * may not be on the dirty list, but we still have to wait for that
+	 * writeout.
+	 *
+	 * To avoid syncing inodes put under IO after we have started here,
+	 * splice the io list to a temporary list head and walk that. Newly
+	 * dirtied inodes will go onto the primary list so we won't wait for
+	 * them. This is safe to do as we are serialised by the s_sync_lock,
+	 * so we'll complete processing the complete list before the next
+	 * sync operation repeats the splice-and-walk process.
+	 *
+	 * Inodes that have pages under writeback after we've finished the wait
+	 * may or may not be on the primary list. They had pages under put IO
+	 * after we started our wait, so we need to make sure the next sync or
+	 * IO completion treats them correctly, Move them back to the primary
+	 * list and restart the walk.
+	 *
+	 * Inodes that are clean after we have waited for them don't belong on
+	 * any list, so simply remove them from the sync list and move onwards.
 	 */
-	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
+	mutex_lock(&sb->s_sync_lock);
+	spin_lock(&bdi->wb.list_lock);
+	list_splice_init(&bdi->wb.b_writeback, &sync_list);
+
+	while (!list_empty(&sync_list)) {
+		struct inode *inode = list_first_entry(&sync_list, struct inode,
+						       i_wb_list);
 		struct address_space *mapping = inode->i_mapping;
 
+		/*
+		 * We are lazy on IO completion and don't remove inodes from the
+		 * list when they are clean. Detect that immediately and skip
+		 * inodes we don't ahve to wait on.
+		 */
+		if (!mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) {
+			list_del_init(&inode->i_wb_list);
+			cond_resched_lock(&bdi->wb.list_lock);
+			continue;
+		}
+
 		spin_lock(&inode->i_lock);
-		if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) ||
-		    (mapping->nrpages == 0)) {
+		if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) {
+                        list_move(&inode->i_wb_list, &bdi->wb.b_writeback);
 			spin_unlock(&inode->i_lock);
+			cond_resched_lock(&bdi->wb.list_lock);
 			continue;
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
-		spin_unlock(&sb->s_inode_list_lock);
-
-		/*
-		 * We hold a reference to 'inode' so it couldn't have been
-		 * removed from s_inodes list while we dropped the
-		 * s_inode_list_lock.  We cannot iput the inode now as we can
-		 * be holding the last reference and we cannot iput it under
-		 * s_inode_list_lock. So we keep the reference and iput it
-		 * later.
-		 */
-		iput(old_inode);
-		old_inode = inode;
+		spin_unlock(&bdi->wb.list_lock);
 
 		filemap_fdatawait(mapping);
-
 		cond_resched();
 
-		spin_lock(&sb->s_inode_list_lock);
+		/*
+		 * the inode has been written back now, so check whether we
+		 * still have pages under IO and move it back to the primary
+		 * list if necessary.  We really need the mapping->tree_lock
+		 * here because bdi_mark_inode_writeback may have not done
+		 * anything because we were on the spliced list and we need to
+		 * check TAG_WRITEBACK.
+		 */
+		spin_lock_irq(&mapping->tree_lock);
+		spin_lock(&bdi->wb.list_lock);
+		if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) {
+			WARN_ON(list_empty(&inode->i_wb_list));
+			list_move(&inode->i_wb_list, &bdi->wb.b_writeback);
+                } else
+			list_del_init(&inode->i_wb_list);
+		spin_unlock_irq(&mapping->tree_lock);
+		iput(inode);
 	}
-	spin_unlock(&sb->s_inode_list_lock);
-	iput(old_inode);
+	spin_unlock(&bdi->wb.list_lock);
 	mutex_unlock(&sb->s_sync_lock);
 }
 
diff --git a/fs/inode.c b/fs/inode.c
index c76a575..d1e6598 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -356,6 +356,7 @@ void inode_init_once(struct inode *inode)
 	INIT_HLIST_NODE(&inode->i_hash);
 	INIT_LIST_HEAD(&inode->i_devices);
 	INIT_LIST_HEAD(&inode->i_io_list);
+	INIT_LIST_HEAD(&inode->i_wb_list);
 	INIT_LIST_HEAD(&inode->i_lru);
 	address_space_init_once(&inode->i_data);
 	i_size_ordered_init(inode);
diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h
index aff923a..12d8224 100644
--- a/include/linux/backing-dev.h
+++ b/include/linux/backing-dev.h
@@ -56,6 +56,7 @@ struct bdi_writeback {
 	struct list_head b_io;		/* parked for writeback */
 	struct list_head b_more_io;	/* parked for more writeback */
 	struct list_head b_dirty_time;	/* time stamps are dirty */
+	struct list_head b_writeback;	/* inodes under writeback */
 	spinlock_t list_lock;		/* protects the b_* lists */
 };
 
@@ -124,6 +125,8 @@ void bdi_start_background_writeback(struct backing_dev_info *bdi);
 void bdi_writeback_workfn(struct work_struct *work);
 int bdi_has_dirty_io(struct backing_dev_info *bdi);
 void bdi_wakeup_thread_delayed(struct backing_dev_info *bdi);
+void bdi_mark_inode_writeback(struct backing_dev_info *bdi,
+			      struct inode *inode);
 
 extern spinlock_t bdi_lock;
 extern struct list_head bdi_list;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index f4d56cc..f47792f 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -637,6 +637,7 @@ struct inode {
 	struct list_head	i_io_list;	/* backing dev IO list */
 	struct list_head	i_lru;		/* inode LRU list */
 	struct list_head	i_sb_list;
+	struct list_head	i_wb_list;	/* backing dev writeback list */
 	union {
 		struct hlist_head	i_dentry;
 		struct rcu_head		i_rcu;
diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index 6e7a644..df31958 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -386,6 +386,7 @@ static void bdi_wb_init(struct bdi_writeback *wb, struct backing_dev_info *bdi)
 	INIT_LIST_HEAD(&wb->b_io);
 	INIT_LIST_HEAD(&wb->b_more_io);
 	INIT_LIST_HEAD(&wb->b_dirty_time);
+	INIT_LIST_HEAD(&wb->b_writeback);
 	spin_lock_init(&wb->list_lock);
 	INIT_DELAYED_WORK(&wb->dwork, bdi_writeback_workfn);
 }
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index eb59f7e..f3751d1 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -2382,11 +2382,25 @@ int __test_set_page_writeback(struct page *page, bool keep_write)
 		spin_lock_irqsave(&mapping->tree_lock, flags);
 		ret = TestSetPageWriteback(page);
 		if (!ret) {
+			bool on_wblist;
+
+			on_wblist = mapping_tagged(mapping,
+						   PAGECACHE_TAG_WRITEBACK);
+
 			radix_tree_tag_set(&mapping->page_tree,
 						page_index(page),
 						PAGECACHE_TAG_WRITEBACK);
 			if (bdi_cap_account_writeback(bdi))
 				__inc_bdi_stat(bdi, BDI_WRITEBACK);
+
+			/*
+			 * we can come through here when swapping anonymous
+			 * pages, so we don't necessarily have an inode to
+			 * track for sync. Inodes are remove lazily, so there is
+			 * no equivalent in test_clear_page_writeback().
+			 */
+			if (!on_wblist && mapping->host)
+				bdi_mark_inode_writeback(bdi, mapping->host);
 		}
 		if (!PageDirty(page))
 			radix_tree_tag_clear(&mapping->page_tree,
-- 
2.1.0

Re: [PATCH 6/8] bdi: add a new writeback list for sync

[Jan Kara]

On Thu 11-06-15 15:41:11, Josef Bacik wrote:
> From: Dave Chinner <dchinner@redhat.com>
> 
> wait_sb_inodes() current does a walk of all inodes in the filesystem
> to find dirty one to wait on during sync. This is highly
> inefficient and wastes a lot of CPU when there are lots of clean
> cached inodes that we don't need to wait on.
> 
> To avoid this "all inode" walk, we need to track inodes that are
> currently under writeback that we need to wait for. We do this by
> adding inodes to a writeback list on the bdi when the mapping is
> first tagged as having pages under writeback.  wait_sb_inodes() can
> then walk this list of "inodes under IO" and wait specifically just
> for the inodes that the current sync(2) needs to wait for.
> 
> To avoid needing all the realted locking to be safe against
> interrupts, Jan Kara suggested that we be lazy about removal from
> the writeback list. That is, we don't remove inodes from the
> writeback list on IO completion, but do it directly during a
> wait_sb_inodes() walk.
> 
> This means that the a rare sync(2) call will have some work to do
> skipping clean inodes However, in the current problem case of
> concurrent sync workloads, concurrent wait_sb_inodes() calls only
> walk the very recently dispatched inodes and hence should have very
> little work to do.
> 
> This also means that we have to remove the inodes from the writeback
> list during eviction. Do this in inode_wait_for_writeback() once
> all writeback on the inode is complete.
> 
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> Signed-off-by: Josef Bacik <jbacik@fb.com>

I don't think you noticed comments I did when you last posted this
series:

>  /*
> - * Wait for writeback on an inode to complete. Caller must have inode pinned.
> + * Wait for writeback on an inode to complete during eviction. Caller must have
> + * inode pinned.
>   */
>  void inode_wait_for_writeback(struct inode *inode)
>  {
> +	BUG_ON(!(inode->i_state & I_FREEING));
> +
>  	spin_lock(&inode->i_lock);
>  	__inode_wait_for_writeback(inode);
>  	spin_unlock(&inode->i_lock);
> +
> +	/*
> +	 * bd_inode's will have already had the bdev free'd so don't bother
> +	 * doing the bdi_clear_inode_writeback.
> +	 */
> +	if (!sb_is_blkdev_sb(inode->i_sb))
> +		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
>  }

Why do we bother with not putting bdev inode back?

...
>  		spin_lock(&inode->i_lock);
> -		if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) ||
> -		    (mapping->nrpages == 0)) {
> +		if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) {
> +                        list_move(&inode->i_wb_list, &bdi->wb.b_writeback);

Indenting is broken here...

>  			spin_unlock(&inode->i_lock);
> +			cond_resched_lock(&bdi->wb.list_lock);
>  			continue;
>  		}
>  		__iget(inode);
>  		spin_unlock(&inode->i_lock);


								Honza
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

Re: [PATCH 6/8] bdi: add a new writeback list for sync

[Josef Bacik]

On 06/15/2015 07:12 AM, Jan Kara wrote:
> On Thu 11-06-15 15:41:11, Josef Bacik wrote:
>> From: Dave Chinner <dchinner@redhat.com>
>>
>> wait_sb_inodes() current does a walk of all inodes in the filesystem
>> to find dirty one to wait on during sync. This is highly
>> inefficient and wastes a lot of CPU when there are lots of clean
>> cached inodes that we don't need to wait on.
>>
>> To avoid this "all inode" walk, we need to track inodes that are
>> currently under writeback that we need to wait for. We do this by
>> adding inodes to a writeback list on the bdi when the mapping is
>> first tagged as having pages under writeback.  wait_sb_inodes() can
>> then walk this list of "inodes under IO" and wait specifically just
>> for the inodes that the current sync(2) needs to wait for.
>>
>> To avoid needing all the realted locking to be safe against
>> interrupts, Jan Kara suggested that we be lazy about removal from
>> the writeback list. That is, we don't remove inodes from the
>> writeback list on IO completion, but do it directly during a
>> wait_sb_inodes() walk.
>>
>> This means that the a rare sync(2) call will have some work to do
>> skipping clean inodes However, in the current problem case of
>> concurrent sync workloads, concurrent wait_sb_inodes() calls only
>> walk the very recently dispatched inodes and hence should have very
>> little work to do.
>>
>> This also means that we have to remove the inodes from the writeback
>> list during eviction. Do this in inode_wait_for_writeback() once
>> all writeback on the inode is complete.
>>
>> Signed-off-by: Dave Chinner <dchinner@redhat.com>
>> Signed-off-by: Josef Bacik <jbacik@fb.com>
>
> I don't think you noticed comments I did when you last posted this
> series:

You're right I missed them completely, sorry.

>
>>   /*
>> - * Wait for writeback on an inode to complete. Caller must have inode pinned.
>> + * Wait for writeback on an inode to complete during eviction. Caller must have
>> + * inode pinned.
>>    */
>>   void inode_wait_for_writeback(struct inode *inode)
>>   {
>> +	BUG_ON(!(inode->i_state & I_FREEING));
>> +
>>   	spin_lock(&inode->i_lock);
>>   	__inode_wait_for_writeback(inode);
>>   	spin_unlock(&inode->i_lock);
>> +
>> +	/*
>> +	 * bd_inode's will have already had the bdev free'd so don't bother
>> +	 * doing the bdi_clear_inode_writeback.
>> +	 */
>> +	if (!sb_is_blkdev_sb(inode->i_sb))
>> +		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
>>   }
>
> Why do we bother with not putting bdev inode back?
>

My memory is rusty on this, but if the inode is the inode for a bdev we 
will have already free'd the bdev at this point and we get a null 
pointer deref, so this just skips that bit.  I'll fix up the indenting. 
  Thanks,

Josef

Re: [PATCH 6/8] bdi: add a new writeback list for sync

[Jan Kara]

On Tue 16-06-15 08:42:27, Josef Bacik wrote:
> >>  /*
> >>- * Wait for writeback on an inode to complete. Caller must have inode pinned.
> >>+ * Wait for writeback on an inode to complete during eviction. Caller must have
> >>+ * inode pinned.
> >>   */
> >>  void inode_wait_for_writeback(struct inode *inode)
> >>  {
> >>+	BUG_ON(!(inode->i_state & I_FREEING));
> >>+
> >>  	spin_lock(&inode->i_lock);
> >>  	__inode_wait_for_writeback(inode);
> >>  	spin_unlock(&inode->i_lock);
> >>+
> >>+	/*
> >>+	 * bd_inode's will have already had the bdev free'd so don't bother
> >>+	 * doing the bdi_clear_inode_writeback.
> >>+	 */
> >>+	if (!sb_is_blkdev_sb(inode->i_sb))
> >>+		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
> >>  }
> >
> >Why do we bother with not putting bdev inode back?
> >
> 
> My memory is rusty on this, but if the inode is the inode for a bdev
> we will have already free'd the bdev at this point and we get a null
> pointer deref, so this just skips that bit.

Ah, the reason likely is that bdev->bd_disk is NULL (already cleaned up in
__blkdev_put()) at this moment and thus bdev_get_queue() called from
inode_to_bdi() will oops. Can you please add these details to the comment?
It's a bit subtle...

Also we shouldn't have any pages in the block device mapping anymore
because of the work done in __blkdev_put() (and thus inode shouldn't be in
the writeback list) but I'd be calmer if we asserted
list_empty(&inode->i_wb_list). Can you please add that? Thanks!

								Honza
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

Re: [PATCH 6/8] bdi: add a new writeback list for sync

[Josef Bacik]

On 06/17/2015 03:34 AM, Jan Kara wrote:
> On Tue 16-06-15 08:42:27, Josef Bacik wrote:
>>>>   /*
>>>> - * Wait for writeback on an inode to complete. Caller must have inode pinned.
>>>> + * Wait for writeback on an inode to complete during eviction. Caller must have
>>>> + * inode pinned.
>>>>    */
>>>>   void inode_wait_for_writeback(struct inode *inode)
>>>>   {
>>>> +	BUG_ON(!(inode->i_state & I_FREEING));
>>>> +
>>>>   	spin_lock(&inode->i_lock);
>>>>   	__inode_wait_for_writeback(inode);
>>>>   	spin_unlock(&inode->i_lock);
>>>> +
>>>> +	/*
>>>> +	 * bd_inode's will have already had the bdev free'd so don't bother
>>>> +	 * doing the bdi_clear_inode_writeback.
>>>> +	 */
>>>> +	if (!sb_is_blkdev_sb(inode->i_sb))
>>>> +		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
>>>>   }
>>>
>>> Why do we bother with not putting bdev inode back?
>>>
>>
>> My memory is rusty on this, but if the inode is the inode for a bdev
>> we will have already free'd the bdev at this point and we get a null
>> pointer deref, so this just skips that bit.
>
> Ah, the reason likely is that bdev->bd_disk is NULL (already cleaned up in
> __blkdev_put()) at this moment and thus bdev_get_queue() called from
> inode_to_bdi() will oops. Can you please add these details to the comment?
> It's a bit subtle...
>
> Also we shouldn't have any pages in the block device mapping anymore
> because of the work done in __blkdev_put() (and thus inode shouldn't be in
> the writeback list) but I'd be calmer if we asserted
> list_empty(&inode->i_wb_list). Can you please add that? Thanks!

Won't it trip if we never sync before we drop the device tho?  So we 
write some stuff to the block device, it gets written out, we then drop 
the device for whatever reason and boom, hit BUG_ON(&inode->i_wb_list) 
because we're still on the writeback list even though it doesn't matter 
because this disk is going away.  Just an untested theory, what do you 
think?  If it is possible I suppose I could just add the clear'ing bit 
for the bd inode to before we drop bd_disk if that would make you happy. 
  Thanks,

Josef

Re: [PATCH 6/8] bdi: add a new writeback list for sync

[Jan Kara]

On Wed 17-06-15 10:55:58, Josef Bacik wrote:
> On 06/17/2015 03:34 AM, Jan Kara wrote:
> >On Tue 16-06-15 08:42:27, Josef Bacik wrote:
> >>>>  /*
> >>>>- * Wait for writeback on an inode to complete. Caller must have inode pinned.
> >>>>+ * Wait for writeback on an inode to complete during eviction. Caller must have
> >>>>+ * inode pinned.
> >>>>   */
> >>>>  void inode_wait_for_writeback(struct inode *inode)
> >>>>  {
> >>>>+	BUG_ON(!(inode->i_state & I_FREEING));
> >>>>+
> >>>>  	spin_lock(&inode->i_lock);
> >>>>  	__inode_wait_for_writeback(inode);
> >>>>  	spin_unlock(&inode->i_lock);
> >>>>+
> >>>>+	/*
> >>>>+	 * bd_inode's will have already had the bdev free'd so don't bother
> >>>>+	 * doing the bdi_clear_inode_writeback.
> >>>>+	 */
> >>>>+	if (!sb_is_blkdev_sb(inode->i_sb))
> >>>>+		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
> >>>>  }
> >>>
> >>>Why do we bother with not putting bdev inode back?
> >>>
> >>
> >>My memory is rusty on this, but if the inode is the inode for a bdev
> >>we will have already free'd the bdev at this point and we get a null
> >>pointer deref, so this just skips that bit.
> >
> >Ah, the reason likely is that bdev->bd_disk is NULL (already cleaned up in
> >__blkdev_put()) at this moment and thus bdev_get_queue() called from
> >inode_to_bdi() will oops. Can you please add these details to the comment?
> >It's a bit subtle...
> >
> >Also we shouldn't have any pages in the block device mapping anymore
> >because of the work done in __blkdev_put() (and thus inode shouldn't be in
> >the writeback list) but I'd be calmer if we asserted
> >list_empty(&inode->i_wb_list). Can you please add that? Thanks!
> 
> Won't it trip if we never sync before we drop the device tho?  So we
> write some stuff to the block device, it gets written out, we then
> drop the device for whatever reason and boom, hit
> BUG_ON(&inode->i_wb_list) because we're still on the writeback list
> even though it doesn't matter because this disk is going away.  Just
> an untested theory, what do you think?

Well, we are going to free the inode. If it is still linked into
the writeback list, we are in trouble as the list will now contain freed
element. So better BUG_ON earlier than chase memory corruption later.

Calling bdi_clear_inode_writeback() in kill_bdev() is probably a good idea
and we can then just have the assertion in evict() to make sure nothing
went wrong. OK?

								Honza
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

[PATCH 6/8 V4] bdi: add a new writeback list for sync

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

wait_sb_inodes() current does a walk of all inodes in the filesystem
to find dirty one to wait on during sync. This is highly
inefficient and wastes a lot of CPU when there are lots of clean
cached inodes that we don't need to wait on.

To avoid this "all inode" walk, we need to track inodes that are
currently under writeback that we need to wait for. We do this by
adding inodes to a writeback list on the bdi when the mapping is
first tagged as having pages under writeback.  wait_sb_inodes() can
then walk this list of "inodes under IO" and wait specifically just
for the inodes that the current sync(2) needs to wait for.

To avoid needing all the realted locking to be safe against
interrupts, Jan Kara suggested that we be lazy about removal from
the writeback list. That is, we don't remove inodes from the
writeback list on IO completion, but do it directly during a
wait_sb_inodes() walk.

This means that the a rare sync(2) call will have some work to do
skipping clean inodes However, in the current problem case of
concurrent sync workloads, concurrent wait_sb_inodes() calls only
walk the very recently dispatched inodes and hence should have very
little work to do.

This also means that we have to remove the inodes from the writeback
list during eviction. Do this in inode_wait_for_writeback() once
all writeback on the inode is complete.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Tested-by: Dave Chinner <dchinner@redhat.com>
---
V3->V4:
-added bdi_clear_inode_writeback to kill_bdev.
-added BUG_ON(!list_emtpy(&inode->i_wb_list)) to inode_wait_writeback
-clarified the comment in inode_wait_writeback
-fixed some indenting

 fs/block_dev.c              |   2 +
 fs/fs-writeback.c           | 136 ++++++++++++++++++++++++++++++++++----------
 fs/inode.c                  |   1 +
 include/linux/backing-dev.h |   5 ++
 include/linux/fs.h          |   1 +
 mm/backing-dev.c            |   1 +
 mm/page-writeback.c         |  14 +++++
 7 files changed, 131 insertions(+), 29 deletions(-)

diff --git a/fs/block_dev.c b/fs/block_dev.c
index f2a89be..2726ff1 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -64,6 +64,8 @@ void kill_bdev(struct block_device *bdev)
 {
 	struct address_space *mapping = bdev->bd_inode->i_mapping;
 
+	bdi_clear_inode_writeback(inode_to_bdi(bdev->bd_inode),
+				  bdev->bd_inode);
 	if (mapping->nrpages == 0 && mapping->nrshadows == 0)
 		return;
 
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index aa72536..3958772 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -210,6 +210,34 @@ void inode_wb_list_del(struct inode *inode)
 }
 
 /*
+ * mark an inode as under writeback on the given bdi
+ */
+void bdi_mark_inode_writeback(struct backing_dev_info *bdi, struct inode *inode)
+{
+	WARN_ON_ONCE(bdi != inode_to_bdi(inode));
+	if (list_empty(&inode->i_wb_list)) {
+		spin_lock(&bdi->wb.list_lock);
+		if (list_empty(&inode->i_wb_list))
+			list_add_tail(&inode->i_wb_list, &bdi->wb.b_writeback);
+		spin_unlock(&bdi->wb.list_lock);
+	}
+}
+
+/*
+ * clear an inode as under writeback on the given bdi
+ */
+void bdi_clear_inode_writeback(struct backing_dev_info *bdi,
+			       struct inode *inode)
+{
+	WARN_ON_ONCE(bdi != inode_to_bdi(inode));
+	if (!list_empty(&inode->i_wb_list)) {
+		spin_lock(&bdi->wb.list_lock);
+		list_del_init(&inode->i_wb_list);
+		spin_unlock(&bdi->wb.list_lock);
+	}
+}
+
+/*
  * Redirty an inode: set its when-it-was dirtied timestamp and move it to the
  * furthest end of its superblock's dirty-inode list.
  *
@@ -383,13 +411,28 @@ static void __inode_wait_for_writeback(struct inode *inode)
 }
 
 /*
- * Wait for writeback on an inode to complete. Caller must have inode pinned.
+ * Wait for writeback on an inode to complete during eviction. Caller must have
+ * inode pinned.
  */
 void inode_wait_for_writeback(struct inode *inode)
 {
+	BUG_ON(!(inode->i_state & I_FREEING));
+
 	spin_lock(&inode->i_lock);
 	__inode_wait_for_writeback(inode);
 	spin_unlock(&inode->i_lock);
+
+	/*
+	 * For a bd_inode when we do inode_to_bdi we'll want to get the bdev for
+	 * the inode and then deref bdev->bd_disk, which at this point has been
+	 * set to NULL, so we would panic.  At the point we are dropping our
+	 * bd_inode we won't have any pages under writeback on the device so
+	 * this is safe.  But just in case we'll assert to make sure we don't
+	 * screw this up.
+	 */
+	if (!sb_is_blkdev_sb(inode->i_sb))
+		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
+	BUG_ON(!list_empty(&inode->i_wb_list));
 }
 
 /*
@@ -1372,7 +1415,8 @@ EXPORT_SYMBOL(__mark_inode_dirty);
  */
 static void wait_sb_inodes(struct super_block *sb)
 {
-	struct inode *inode, *old_inode = NULL;
+	struct backing_dev_info *bdi = sb->s_bdi;
+	LIST_HEAD(sync_list);
 
 	/*
 	 * We need to be protected against the filesystem going from
@@ -1380,48 +1424,82 @@ static void wait_sb_inodes(struct super_block *sb)
 	 */
 	WARN_ON(!rwsem_is_locked(&sb->s_umount));
 
-	mutex_lock(&sb->s_sync_lock);
-	spin_lock(&sb->s_inode_list_lock);
-
 	/*
-	 * Data integrity sync. Must wait for all pages under writeback,
-	 * because there may have been pages dirtied before our sync
-	 * call, but which had writeout started before we write it out.
-	 * In which case, the inode may not be on the dirty list, but
-	 * we still have to wait for that writeout.
+	 * Data integrity sync. Must wait for all pages under writeback, because
+	 * there may have been pages dirtied before our sync call, but which had
+	 * writeout started before we write it out.  In which case, the inode
+	 * may not be on the dirty list, but we still have to wait for that
+	 * writeout.
+	 *
+	 * To avoid syncing inodes put under IO after we have started here,
+	 * splice the io list to a temporary list head and walk that. Newly
+	 * dirtied inodes will go onto the primary list so we won't wait for
+	 * them. This is safe to do as we are serialised by the s_sync_lock,
+	 * so we'll complete processing the complete list before the next
+	 * sync operation repeats the splice-and-walk process.
+	 *
+	 * Inodes that have pages under writeback after we've finished the wait
+	 * may or may not be on the primary list. They had pages under put IO
+	 * after we started our wait, so we need to make sure the next sync or
+	 * IO completion treats them correctly, Move them back to the primary
+	 * list and restart the walk.
+	 *
+	 * Inodes that are clean after we have waited for them don't belong on
+	 * any list, so simply remove them from the sync list and move onwards.
 	 */
-	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
+	mutex_lock(&sb->s_sync_lock);
+	spin_lock(&bdi->wb.list_lock);
+	list_splice_init(&bdi->wb.b_writeback, &sync_list);
+
+	while (!list_empty(&sync_list)) {
+		struct inode *inode = list_first_entry(&sync_list, struct inode,
+						       i_wb_list);
 		struct address_space *mapping = inode->i_mapping;
 
+		/*
+		 * We are lazy on IO completion and don't remove inodes from the
+		 * list when they are clean. Detect that immediately and skip
+		 * inodes we don't ahve to wait on.
+		 */
+		if (!mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) {
+			list_del_init(&inode->i_wb_list);
+			cond_resched_lock(&bdi->wb.list_lock);
+			continue;
+		}
+
 		spin_lock(&inode->i_lock);
-		if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) ||
-		    (mapping->nrpages == 0)) {
+		if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) {
+			list_move(&inode->i_wb_list, &bdi->wb.b_writeback);
 			spin_unlock(&inode->i_lock);
+			cond_resched_lock(&bdi->wb.list_lock);
 			continue;
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
-		spin_unlock(&sb->s_inode_list_lock);
-
-		/*
-		 * We hold a reference to 'inode' so it couldn't have been
-		 * removed from s_inodes list while we dropped the
-		 * s_inode_list_lock.  We cannot iput the inode now as we can
-		 * be holding the last reference and we cannot iput it under
-		 * s_inode_list_lock. So we keep the reference and iput it
-		 * later.
-		 */
-		iput(old_inode);
-		old_inode = inode;
+		spin_unlock(&bdi->wb.list_lock);
 
 		filemap_fdatawait(mapping);
-
 		cond_resched();
 
-		spin_lock(&sb->s_inode_list_lock);
+		/*
+		 * the inode has been written back now, so check whether we
+		 * still have pages under IO and move it back to the primary
+		 * list if necessary.  We really need the mapping->tree_lock
+		 * here because bdi_mark_inode_writeback may have not done
+		 * anything because we were on the spliced list and we need to
+		 * check TAG_WRITEBACK.
+		 */
+		spin_lock_irq(&mapping->tree_lock);
+		spin_lock(&bdi->wb.list_lock);
+		if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) {
+			WARN_ON(list_empty(&inode->i_wb_list));
+			list_move(&inode->i_wb_list, &bdi->wb.b_writeback);
+                } else
+			list_del_init(&inode->i_wb_list);
+		spin_unlock_irq(&mapping->tree_lock);
+		iput(inode);
 	}
-	spin_unlock(&sb->s_inode_list_lock);
-	iput(old_inode);
+	spin_unlock(&bdi->wb.list_lock);
 	mutex_unlock(&sb->s_sync_lock);
 }
 
diff --git a/fs/inode.c b/fs/inode.c
index c76a575..d1e6598 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -356,6 +356,7 @@ void inode_init_once(struct inode *inode)
 	INIT_HLIST_NODE(&inode->i_hash);
 	INIT_LIST_HEAD(&inode->i_devices);
 	INIT_LIST_HEAD(&inode->i_io_list);
+	INIT_LIST_HEAD(&inode->i_wb_list);
 	INIT_LIST_HEAD(&inode->i_lru);
 	address_space_init_once(&inode->i_data);
 	i_size_ordered_init(inode);
diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h
index aff923a..7764d1e 100644
--- a/include/linux/backing-dev.h
+++ b/include/linux/backing-dev.h
@@ -56,6 +56,7 @@ struct bdi_writeback {
 	struct list_head b_io;		/* parked for writeback */
 	struct list_head b_more_io;	/* parked for more writeback */
 	struct list_head b_dirty_time;	/* time stamps are dirty */
+	struct list_head b_writeback;	/* inodes under writeback */
 	spinlock_t list_lock;		/* protects the b_* lists */
 };
 
@@ -124,6 +125,10 @@ void bdi_start_background_writeback(struct backing_dev_info *bdi);
 void bdi_writeback_workfn(struct work_struct *work);
 int bdi_has_dirty_io(struct backing_dev_info *bdi);
 void bdi_wakeup_thread_delayed(struct backing_dev_info *bdi);
+void bdi_mark_inode_writeback(struct backing_dev_info *bdi,
+			      struct inode *inode);
+void bdi_clear_inode_writeback(struct backing_dev_info *bdi,
+			       struct inode *inode);
 
 extern spinlock_t bdi_lock;
 extern struct list_head bdi_list;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index f4d56cc..f47792f 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -637,6 +637,7 @@ struct inode {
 	struct list_head	i_io_list;	/* backing dev IO list */
 	struct list_head	i_lru;		/* inode LRU list */
 	struct list_head	i_sb_list;
+	struct list_head	i_wb_list;	/* backing dev writeback list */
 	union {
 		struct hlist_head	i_dentry;
 		struct rcu_head		i_rcu;
diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index 6e7a644..df31958 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -386,6 +386,7 @@ static void bdi_wb_init(struct bdi_writeback *wb, struct backing_dev_info *bdi)
 	INIT_LIST_HEAD(&wb->b_io);
 	INIT_LIST_HEAD(&wb->b_more_io);
 	INIT_LIST_HEAD(&wb->b_dirty_time);
+	INIT_LIST_HEAD(&wb->b_writeback);
 	spin_lock_init(&wb->list_lock);
 	INIT_DELAYED_WORK(&wb->dwork, bdi_writeback_workfn);
 }
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index eb59f7e..f3751d1 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -2382,11 +2382,25 @@ int __test_set_page_writeback(struct page *page, bool keep_write)
 		spin_lock_irqsave(&mapping->tree_lock, flags);
 		ret = TestSetPageWriteback(page);
 		if (!ret) {
+			bool on_wblist;
+
+			on_wblist = mapping_tagged(mapping,
+						   PAGECACHE_TAG_WRITEBACK);
+
 			radix_tree_tag_set(&mapping->page_tree,
 						page_index(page),
 						PAGECACHE_TAG_WRITEBACK);
 			if (bdi_cap_account_writeback(bdi))
 				__inc_bdi_stat(bdi, BDI_WRITEBACK);
+
+			/*
+			 * we can come through here when swapping anonymous
+			 * pages, so we don't necessarily have an inode to
+			 * track for sync. Inodes are remove lazily, so there is
+			 * no equivalent in test_clear_page_writeback().
+			 */
+			if (!on_wblist && mapping->host)
+				bdi_mark_inode_writeback(bdi, mapping->host);
 		}
 		if (!PageDirty(page))
 			radix_tree_tag_clear(&mapping->page_tree,
-- 
2.1.0

Re: [PATCH 6/8 V4] bdi: add a new writeback list for sync

[Jan Kara]

On Thu 18-06-15 15:18:43, Josef Bacik wrote:
> From: Dave Chinner <dchinner@redhat.com>
> 
> wait_sb_inodes() current does a walk of all inodes in the filesystem
> to find dirty one to wait on during sync. This is highly
> inefficient and wastes a lot of CPU when there are lots of clean
> cached inodes that we don't need to wait on.
> 
> To avoid this "all inode" walk, we need to track inodes that are
> currently under writeback that we need to wait for. We do this by
> adding inodes to a writeback list on the bdi when the mapping is
> first tagged as having pages under writeback.  wait_sb_inodes() can
> then walk this list of "inodes under IO" and wait specifically just
> for the inodes that the current sync(2) needs to wait for.
> 
> To avoid needing all the realted locking to be safe against
> interrupts, Jan Kara suggested that we be lazy about removal from
> the writeback list. That is, we don't remove inodes from the
> writeback list on IO completion, but do it directly during a
> wait_sb_inodes() walk.
> 
> This means that the a rare sync(2) call will have some work to do
> skipping clean inodes However, in the current problem case of
> concurrent sync workloads, concurrent wait_sb_inodes() calls only
> walk the very recently dispatched inodes and hence should have very
> little work to do.
> 
> This also means that we have to remove the inodes from the writeback
> list during eviction. Do this in inode_wait_for_writeback() once
> all writeback on the inode is complete.
> 
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> Signed-off-by: Josef Bacik <jbacik@fb.com>
> Tested-by: Dave Chinner <dchinner@redhat.com>
  Looks good. You can add:
Reviewed-by: Jan Kara <jack@suse.cz>

								Honza

> ---
> V3->V4:
> -added bdi_clear_inode_writeback to kill_bdev.
> -added BUG_ON(!list_emtpy(&inode->i_wb_list)) to inode_wait_writeback
> -clarified the comment in inode_wait_writeback
> -fixed some indenting
> 
>  fs/block_dev.c              |   2 +
>  fs/fs-writeback.c           | 136 ++++++++++++++++++++++++++++++++++----------
>  fs/inode.c                  |   1 +
>  include/linux/backing-dev.h |   5 ++
>  include/linux/fs.h          |   1 +
>  mm/backing-dev.c            |   1 +
>  mm/page-writeback.c         |  14 +++++
>  7 files changed, 131 insertions(+), 29 deletions(-)
> 
> diff --git a/fs/block_dev.c b/fs/block_dev.c
> index f2a89be..2726ff1 100644
> --- a/fs/block_dev.c
> +++ b/fs/block_dev.c
> @@ -64,6 +64,8 @@ void kill_bdev(struct block_device *bdev)
>  {
>  	struct address_space *mapping = bdev->bd_inode->i_mapping;
>  
> +	bdi_clear_inode_writeback(inode_to_bdi(bdev->bd_inode),
> +				  bdev->bd_inode);
>  	if (mapping->nrpages == 0 && mapping->nrshadows == 0)
>  		return;
>  
> diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
> index aa72536..3958772 100644
> --- a/fs/fs-writeback.c
> +++ b/fs/fs-writeback.c
> @@ -210,6 +210,34 @@ void inode_wb_list_del(struct inode *inode)
>  }
>  
>  /*
> + * mark an inode as under writeback on the given bdi
> + */
> +void bdi_mark_inode_writeback(struct backing_dev_info *bdi, struct inode *inode)
> +{
> +	WARN_ON_ONCE(bdi != inode_to_bdi(inode));
> +	if (list_empty(&inode->i_wb_list)) {
> +		spin_lock(&bdi->wb.list_lock);
> +		if (list_empty(&inode->i_wb_list))
> +			list_add_tail(&inode->i_wb_list, &bdi->wb.b_writeback);
> +		spin_unlock(&bdi->wb.list_lock);
> +	}
> +}
> +
> +/*
> + * clear an inode as under writeback on the given bdi
> + */
> +void bdi_clear_inode_writeback(struct backing_dev_info *bdi,
> +			       struct inode *inode)
> +{
> +	WARN_ON_ONCE(bdi != inode_to_bdi(inode));
> +	if (!list_empty(&inode->i_wb_list)) {
> +		spin_lock(&bdi->wb.list_lock);
> +		list_del_init(&inode->i_wb_list);
> +		spin_unlock(&bdi->wb.list_lock);
> +	}
> +}
> +
> +/*
>   * Redirty an inode: set its when-it-was dirtied timestamp and move it to the
>   * furthest end of its superblock's dirty-inode list.
>   *
> @@ -383,13 +411,28 @@ static void __inode_wait_for_writeback(struct inode *inode)
>  }
>  
>  /*
> - * Wait for writeback on an inode to complete. Caller must have inode pinned.
> + * Wait for writeback on an inode to complete during eviction. Caller must have
> + * inode pinned.
>   */
>  void inode_wait_for_writeback(struct inode *inode)
>  {
> +	BUG_ON(!(inode->i_state & I_FREEING));
> +
>  	spin_lock(&inode->i_lock);
>  	__inode_wait_for_writeback(inode);
>  	spin_unlock(&inode->i_lock);
> +
> +	/*
> +	 * For a bd_inode when we do inode_to_bdi we'll want to get the bdev for
> +	 * the inode and then deref bdev->bd_disk, which at this point has been
> +	 * set to NULL, so we would panic.  At the point we are dropping our
> +	 * bd_inode we won't have any pages under writeback on the device so
> +	 * this is safe.  But just in case we'll assert to make sure we don't
> +	 * screw this up.
> +	 */
> +	if (!sb_is_blkdev_sb(inode->i_sb))
> +		bdi_clear_inode_writeback(inode_to_bdi(inode), inode);
> +	BUG_ON(!list_empty(&inode->i_wb_list));
>  }
>  
>  /*
> @@ -1372,7 +1415,8 @@ EXPORT_SYMBOL(__mark_inode_dirty);
>   */
>  static void wait_sb_inodes(struct super_block *sb)
>  {
> -	struct inode *inode, *old_inode = NULL;
> +	struct backing_dev_info *bdi = sb->s_bdi;
> +	LIST_HEAD(sync_list);
>  
>  	/*
>  	 * We need to be protected against the filesystem going from
> @@ -1380,48 +1424,82 @@ static void wait_sb_inodes(struct super_block *sb)
>  	 */
>  	WARN_ON(!rwsem_is_locked(&sb->s_umount));
>  
> -	mutex_lock(&sb->s_sync_lock);
> -	spin_lock(&sb->s_inode_list_lock);
> -
>  	/*
> -	 * Data integrity sync. Must wait for all pages under writeback,
> -	 * because there may have been pages dirtied before our sync
> -	 * call, but which had writeout started before we write it out.
> -	 * In which case, the inode may not be on the dirty list, but
> -	 * we still have to wait for that writeout.
> +	 * Data integrity sync. Must wait for all pages under writeback, because
> +	 * there may have been pages dirtied before our sync call, but which had
> +	 * writeout started before we write it out.  In which case, the inode
> +	 * may not be on the dirty list, but we still have to wait for that
> +	 * writeout.
> +	 *
> +	 * To avoid syncing inodes put under IO after we have started here,
> +	 * splice the io list to a temporary list head and walk that. Newly
> +	 * dirtied inodes will go onto the primary list so we won't wait for
> +	 * them. This is safe to do as we are serialised by the s_sync_lock,
> +	 * so we'll complete processing the complete list before the next
> +	 * sync operation repeats the splice-and-walk process.
> +	 *
> +	 * Inodes that have pages under writeback after we've finished the wait
> +	 * may or may not be on the primary list. They had pages under put IO
> +	 * after we started our wait, so we need to make sure the next sync or
> +	 * IO completion treats them correctly, Move them back to the primary
> +	 * list and restart the walk.
> +	 *
> +	 * Inodes that are clean after we have waited for them don't belong on
> +	 * any list, so simply remove them from the sync list and move onwards.
>  	 */
> -	list_for_each_entry(inode, &sb->s_inodes, i_sb_list) {
> +	mutex_lock(&sb->s_sync_lock);
> +	spin_lock(&bdi->wb.list_lock);
> +	list_splice_init(&bdi->wb.b_writeback, &sync_list);
> +
> +	while (!list_empty(&sync_list)) {
> +		struct inode *inode = list_first_entry(&sync_list, struct inode,
> +						       i_wb_list);
>  		struct address_space *mapping = inode->i_mapping;
>  
> +		/*
> +		 * We are lazy on IO completion and don't remove inodes from the
> +		 * list when they are clean. Detect that immediately and skip
> +		 * inodes we don't ahve to wait on.
> +		 */
> +		if (!mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) {
> +			list_del_init(&inode->i_wb_list);
> +			cond_resched_lock(&bdi->wb.list_lock);
> +			continue;
> +		}
> +
>  		spin_lock(&inode->i_lock);
> -		if ((inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) ||
> -		    (mapping->nrpages == 0)) {
> +		if (inode->i_state & (I_FREEING|I_WILL_FREE|I_NEW)) {
> +			list_move(&inode->i_wb_list, &bdi->wb.b_writeback);
>  			spin_unlock(&inode->i_lock);
> +			cond_resched_lock(&bdi->wb.list_lock);
>  			continue;
>  		}
>  		__iget(inode);
>  		spin_unlock(&inode->i_lock);
> -		spin_unlock(&sb->s_inode_list_lock);
> -
> -		/*
> -		 * We hold a reference to 'inode' so it couldn't have been
> -		 * removed from s_inodes list while we dropped the
> -		 * s_inode_list_lock.  We cannot iput the inode now as we can
> -		 * be holding the last reference and we cannot iput it under
> -		 * s_inode_list_lock. So we keep the reference and iput it
> -		 * later.
> -		 */
> -		iput(old_inode);
> -		old_inode = inode;
> +		spin_unlock(&bdi->wb.list_lock);
>  
>  		filemap_fdatawait(mapping);
> -
>  		cond_resched();
>  
> -		spin_lock(&sb->s_inode_list_lock);
> +		/*
> +		 * the inode has been written back now, so check whether we
> +		 * still have pages under IO and move it back to the primary
> +		 * list if necessary.  We really need the mapping->tree_lock
> +		 * here because bdi_mark_inode_writeback may have not done
> +		 * anything because we were on the spliced list and we need to
> +		 * check TAG_WRITEBACK.
> +		 */
> +		spin_lock_irq(&mapping->tree_lock);
> +		spin_lock(&bdi->wb.list_lock);
> +		if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) {
> +			WARN_ON(list_empty(&inode->i_wb_list));
> +			list_move(&inode->i_wb_list, &bdi->wb.b_writeback);
> +                } else
> +			list_del_init(&inode->i_wb_list);
> +		spin_unlock_irq(&mapping->tree_lock);
> +		iput(inode);
>  	}
> -	spin_unlock(&sb->s_inode_list_lock);
> -	iput(old_inode);
> +	spin_unlock(&bdi->wb.list_lock);
>  	mutex_unlock(&sb->s_sync_lock);
>  }
>  
> diff --git a/fs/inode.c b/fs/inode.c
> index c76a575..d1e6598 100644
> --- a/fs/inode.c
> +++ b/fs/inode.c
> @@ -356,6 +356,7 @@ void inode_init_once(struct inode *inode)
>  	INIT_HLIST_NODE(&inode->i_hash);
>  	INIT_LIST_HEAD(&inode->i_devices);
>  	INIT_LIST_HEAD(&inode->i_io_list);
> +	INIT_LIST_HEAD(&inode->i_wb_list);
>  	INIT_LIST_HEAD(&inode->i_lru);
>  	address_space_init_once(&inode->i_data);
>  	i_size_ordered_init(inode);
> diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h
> index aff923a..7764d1e 100644
> --- a/include/linux/backing-dev.h
> +++ b/include/linux/backing-dev.h
> @@ -56,6 +56,7 @@ struct bdi_writeback {
>  	struct list_head b_io;		/* parked for writeback */
>  	struct list_head b_more_io;	/* parked for more writeback */
>  	struct list_head b_dirty_time;	/* time stamps are dirty */
> +	struct list_head b_writeback;	/* inodes under writeback */
>  	spinlock_t list_lock;		/* protects the b_* lists */
>  };
>  
> @@ -124,6 +125,10 @@ void bdi_start_background_writeback(struct backing_dev_info *bdi);
>  void bdi_writeback_workfn(struct work_struct *work);
>  int bdi_has_dirty_io(struct backing_dev_info *bdi);
>  void bdi_wakeup_thread_delayed(struct backing_dev_info *bdi);
> +void bdi_mark_inode_writeback(struct backing_dev_info *bdi,
> +			      struct inode *inode);
> +void bdi_clear_inode_writeback(struct backing_dev_info *bdi,
> +			       struct inode *inode);
>  
>  extern spinlock_t bdi_lock;
>  extern struct list_head bdi_list;
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index f4d56cc..f47792f 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -637,6 +637,7 @@ struct inode {
>  	struct list_head	i_io_list;	/* backing dev IO list */
>  	struct list_head	i_lru;		/* inode LRU list */
>  	struct list_head	i_sb_list;
> +	struct list_head	i_wb_list;	/* backing dev writeback list */
>  	union {
>  		struct hlist_head	i_dentry;
>  		struct rcu_head		i_rcu;
> diff --git a/mm/backing-dev.c b/mm/backing-dev.c
> index 6e7a644..df31958 100644
> --- a/mm/backing-dev.c
> +++ b/mm/backing-dev.c
> @@ -386,6 +386,7 @@ static void bdi_wb_init(struct bdi_writeback *wb, struct backing_dev_info *bdi)
>  	INIT_LIST_HEAD(&wb->b_io);
>  	INIT_LIST_HEAD(&wb->b_more_io);
>  	INIT_LIST_HEAD(&wb->b_dirty_time);
> +	INIT_LIST_HEAD(&wb->b_writeback);
>  	spin_lock_init(&wb->list_lock);
>  	INIT_DELAYED_WORK(&wb->dwork, bdi_writeback_workfn);
>  }
> diff --git a/mm/page-writeback.c b/mm/page-writeback.c
> index eb59f7e..f3751d1 100644
> --- a/mm/page-writeback.c
> +++ b/mm/page-writeback.c
> @@ -2382,11 +2382,25 @@ int __test_set_page_writeback(struct page *page, bool keep_write)
>  		spin_lock_irqsave(&mapping->tree_lock, flags);
>  		ret = TestSetPageWriteback(page);
>  		if (!ret) {
> +			bool on_wblist;
> +
> +			on_wblist = mapping_tagged(mapping,
> +						   PAGECACHE_TAG_WRITEBACK);
> +
>  			radix_tree_tag_set(&mapping->page_tree,
>  						page_index(page),
>  						PAGECACHE_TAG_WRITEBACK);
>  			if (bdi_cap_account_writeback(bdi))
>  				__inc_bdi_stat(bdi, BDI_WRITEBACK);
> +
> +			/*
> +			 * we can come through here when swapping anonymous
> +			 * pages, so we don't necessarily have an inode to
> +			 * track for sync. Inodes are remove lazily, so there is
> +			 * no equivalent in test_clear_page_writeback().
> +			 */
> +			if (!on_wblist && mapping->host)
> +				bdi_mark_inode_writeback(bdi, mapping->host);
>  		}
>  		if (!PageDirty(page))
>  			radix_tree_tag_clear(&mapping->page_tree,
> -- 
> 2.1.0
> 
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

[PATCH 7/8] writeback: periodically trim the writeback list

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

Inodes are removed lazily from the bdi writeback list, so in the
absence of sync(2) work inodes will build up on the bdi writback
list even though they are no longer under IO. Use the periodic
kupdate work check to remove inodes no longer under IO from the
writeback list.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/fs-writeback.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 3f5b2ff..383b469 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1055,6 +1055,23 @@ static long wb_check_background_flush(struct bdi_writeback *wb)
 	return 0;
 }
 
+/*
+ * clean out writeback list for all inodes that don't have IO in progress
+ */
+static void wb_trim_writeback_list(struct bdi_writeback *wb)
+{
+	struct inode *inode;
+	struct inode *tmp;
+
+	spin_lock(&wb->list_lock);
+	list_for_each_entry_safe(inode, tmp, &wb->b_writeback, i_wb_list) {
+		if (!mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK))
+			list_del_init(&inode->i_wb_list);
+	}
+	spin_unlock(&wb->list_lock);
+
+}
+
 static long wb_check_old_data_flush(struct bdi_writeback *wb)
 {
 	unsigned long expired;
@@ -1071,6 +1088,8 @@ static long wb_check_old_data_flush(struct bdi_writeback *wb)
 	if (time_before(jiffies, expired))
 		return 0;
 
+	wb_trim_writeback_list(wb);
+
 	wb->last_old_flush = jiffies;
 	nr_pages = get_nr_dirty_pages();
 
-- 
2.1.0

[PATCH 8/8] inode: don't softlockup when evicting inodes

[Josef Bacik]

On a box with a lot of ram (148gb) I can make the box softlockup after running
an fs_mark job that creates hundreds of millions of empty files.  This is
because we never generate enough memory pressure to keep the number of inodes on
our unused list low, so when we go to unmount we have to evict ~100 million
inodes.  This makes one processor a very unhappy person, so add a cond_resched()
in dispose_list() and if we need a resched when processing the s_inodes list do
that and run dispose_list() on what we've currently culled.  Thanks,

Signed-off-by: Josef Bacik <jbacik@fb.com>
---
 fs/inode.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/fs/inode.c b/fs/inode.c
index d1e6598..53a1224 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -574,6 +574,7 @@ static void dispose_list(struct list_head *head)
 		list_del_init(&inode->i_lru);
 
 		evict(inode);
+		cond_resched();
 	}
 }
 
@@ -591,6 +592,7 @@ void evict_inodes(struct super_block *sb)
 	struct inode *inode, *next;
 	LIST_HEAD(dispose);
 
+again:
 	spin_lock(&sb->s_inode_list_lock);
 	list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) {
 		if (atomic_read(&inode->i_count))
@@ -606,6 +608,18 @@ void evict_inodes(struct super_block *sb)
 		inode_lru_list_del(inode);
 		spin_unlock(&inode->i_lock);
 		list_add(&inode->i_lru, &dispose);
+
+		/*
+		 * We can have a ton of inodes to evict at unmount time given
+		 * enough memory, check to see if we need to go to sleep for a
+		 * bit so we don't livelock.
+		 */
+		if (need_resched()) {
+			spin_unlock(&sb->s_inode_list_lock);
+			cond_resched();
+			dispose_list(&dispose);
+			goto again;
+		}
 	}
 	spin_unlock(&sb->s_inode_list_lock);
 
-- 
2.1.0

Re: [PATCH 8/8] inode: don't softlockup when evicting inodes

[Jan Kara]

On Thu 11-06-15 15:41:13, Josef Bacik wrote:
> On a box with a lot of ram (148gb) I can make the box softlockup after running
> an fs_mark job that creates hundreds of millions of empty files.  This is
> because we never generate enough memory pressure to keep the number of inodes on
> our unused list low, so when we go to unmount we have to evict ~100 million
> inodes.  This makes one processor a very unhappy person, so add a cond_resched()
> in dispose_list() and if we need a resched when processing the s_inodes list do
> that and run dispose_list() on what we've currently culled.  Thanks,
> 
> Signed-off-by: Josef Bacik <jbacik@fb.com>
  Looks good. You can add:
Reviewed-by: Jan Kara <jack@suse.cz>

								Honza
> ---
>  fs/inode.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
> 
> diff --git a/fs/inode.c b/fs/inode.c
> index d1e6598..53a1224 100644
> --- a/fs/inode.c
> +++ b/fs/inode.c
> @@ -574,6 +574,7 @@ static void dispose_list(struct list_head *head)
>  		list_del_init(&inode->i_lru);
>  
>  		evict(inode);
> +		cond_resched();
>  	}
>  }
>  
> @@ -591,6 +592,7 @@ void evict_inodes(struct super_block *sb)
>  	struct inode *inode, *next;
>  	LIST_HEAD(dispose);
>  
> +again:
>  	spin_lock(&sb->s_inode_list_lock);
>  	list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) {
>  		if (atomic_read(&inode->i_count))
> @@ -606,6 +608,18 @@ void evict_inodes(struct super_block *sb)
>  		inode_lru_list_del(inode);
>  		spin_unlock(&inode->i_lock);
>  		list_add(&inode->i_lru, &dispose);
> +
> +		/*
> +		 * We can have a ton of inodes to evict at unmount time given
> +		 * enough memory, check to see if we need to go to sleep for a
> +		 * bit so we don't livelock.
> +		 */
> +		if (need_resched()) {
> +			spin_unlock(&sb->s_inode_list_lock);
> +			cond_resched();
> +			dispose_list(&dispose);
> +			goto again;
> +		}
>  	}
>  	spin_unlock(&sb->s_inode_list_lock);
>  
> -- 
> 2.1.0
> 
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

Re: [PATCH 0/7] super block scalabilit patches V3

[Tejun Heo]

Hello,

(cc'ing Jens)

On Thu, Jun 11, 2015 at 03:41:05PM -0400, Josef Bacik wrote:
> Here are the cleaned up versions of Dave Chinners super block scalability
> patches.  I've been testing them locally for a while and they are pretty solid.
> They fix a few big issues, such as the global inode list and soft lockups on
> boxes on unmount that have lots of inodes in cache.  Al if you would consider
> pulling these in that would be great, you can pull from here

FYI, the bdi / writeback changes will conflict with the cgroup
writeback support queued in block/for-4.2/writeback.  I think updating
the patchset should be fairly straight forward tho.

Thanks.

-- 
tejun

Re: [PATCH 0/7] super block scalabilit patches V3

[Dave Chinner]

On Thu, Jun 11, 2015 at 03:41:05PM -0400, Josef Bacik wrote:
> Here are the cleaned up versions of Dave Chinners super block scalability
> patches.  I've been testing them locally for a while and they are pretty solid.
> They fix a few big issues, such as the global inode list and soft lockups on
> boxes on unmount that have lots of inodes in cache.  Al if you would consider
> pulling these in that would be great, you can pull from here
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling

Passes all my smoke tests.

Tested-by: Dave Chinner <dchinner@redhat.com>

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

[PATCH] sync: wait_sb_inodes() calls iput() with spinlock held (was Re: [PATCH 0/7] super block scalabilit patches V3)

[Dave Chinner]

On Tue, Jun 16, 2015 at 07:34:29AM +1000, Dave Chinner wrote:
> On Thu, Jun 11, 2015 at 03:41:05PM -0400, Josef Bacik wrote:
> > Here are the cleaned up versions of Dave Chinners super block scalability
> > patches.  I've been testing them locally for a while and they are pretty solid.
> > They fix a few big issues, such as the global inode list and soft lockups on
> > boxes on unmount that have lots of inodes in cache.  Al if you would consider
> > pulling these in that would be great, you can pull from here
> > 
> > git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling
> 
> Passes all my smoke tests.
> 
> Tested-by: Dave Chinner <dchinner@redhat.com>

FWIW, I just updated my trees to whatever is in the above branch and
v4.1-rc8, and now I'm seeing problems with wb.list_lock recursion
and "sleeping in atomic" scehduling issues. generic/269 produced
this:

 BUG: spinlock cpu recursion on CPU#1, fsstress/3852
  lock: 0xffff88042a650c28, .magic: dead4ead, .owner: fsstress/3804, .owner_cpu: 1
 CPU: 1 PID: 3852 Comm: fsstress Tainted: G        W       4.1.0-rc8-dgc+ #263
 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
  ffff88042a650c28 ffff88039898b8e8 ffffffff81e18ffd ffff88042f250fb0
  ffff880428f6b8e0 ffff88039898b908 ffffffff81e12f09 ffff88042a650c28
  ffffffff8221337b ffff88039898b928 ffffffff81e12f34 ffff88042a650c28
 Call Trace:
  [<ffffffff81e18ffd>] dump_stack+0x4c/0x6e
  [<ffffffff81e12f09>] spin_dump+0x90/0x95
  [<ffffffff81e12f34>] spin_bug+0x26/0x2b
  [<ffffffff810e762d>] do_raw_spin_lock+0x10d/0x150
  [<ffffffff81e24975>] _raw_spin_lock+0x15/0x20
  [<ffffffff811f8ba0>] __mark_inode_dirty+0x2b0/0x450
  [<ffffffff812003b8>] __set_page_dirty+0x78/0xd0
  [<ffffffff81200531>] mark_buffer_dirty+0x61/0xf0
  [<ffffffff81200d91>] __block_commit_write.isra.24+0x81/0xb0
  [<ffffffff81202406>] block_write_end+0x36/0x70
  [<ffffffff814fa110>] ? __xfs_get_blocks+0x8a0/0x8a0
  [<ffffffff81202474>] generic_write_end+0x34/0xb0
  [<ffffffff8118af3d>] ? wait_for_stable_page+0x1d/0x50
  [<ffffffff814fa317>] xfs_vm_write_end+0x67/0xc0
  [<ffffffff811813af>] pagecache_write_end+0x1f/0x30
  [<ffffffff815060dd>] xfs_iozero+0x10d/0x190
  [<ffffffff8150666b>] xfs_zero_last_block+0xdb/0x110
  [<ffffffff815067ba>] xfs_zero_eof+0x11a/0x290
  [<ffffffff811d69e0>] ? complete_walk+0x60/0x100
  [<ffffffff811da25f>] ? path_lookupat+0x5f/0x660
  [<ffffffff81506a6e>] xfs_file_aio_write_checks+0x13e/0x160
  [<ffffffff81506f15>] xfs_file_buffered_aio_write+0x75/0x250
  [<ffffffff811ddb0f>] ? user_path_at_empty+0x5f/0xa0
  [<ffffffff810c601d>] ? __might_sleep+0x4d/0x90
  [<ffffffff815071f5>] xfs_file_write_iter+0x105/0x120
  [<ffffffff811cc5ce>] __vfs_write+0xae/0xf0
  [<ffffffff811ccc01>] vfs_write+0xa1/0x190
  [<ffffffff811cd999>] SyS_write+0x49/0xb0
  [<ffffffff811cc781>] ? SyS_lseek+0x91/0xb0
  [<ffffffff81e24fee>] system_call_fastpath+0x12/0x71

And there are a few tests (including generic/269) producing
in_atomic/"scheduling while atomic" bugs in the evict() path such as:

 in_atomic(): 1, irqs_disabled(): 0, pid: 3852, name: fsstress
 CPU: 12 PID: 3852 Comm: fsstress Not tainted 4.1.0-rc8-dgc+ #263
 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
  000000000000015d ffff88039898b6d8 ffffffff81e18ffd 0000000000000000
  ffff880398865550 ffff88039898b6f8 ffffffff810c5f89 ffff8803f15c45c0
  ffffffff8227a3bf ffff88039898b728 ffffffff810c601d ffff88039898b758
 Call Trace:
  [<ffffffff81e18ffd>] dump_stack+0x4c/0x6e
  [<ffffffff810c5f89>] ___might_sleep+0xf9/0x140
  [<ffffffff810c601d>] __might_sleep+0x4d/0x90
  [<ffffffff81201e8b>] block_invalidatepage+0xab/0x140
  [<ffffffff814f7579>] xfs_vm_invalidatepage+0x39/0xb0
  [<ffffffff8118fa77>] truncate_inode_page+0x67/0xa0
  [<ffffffff8118fc92>] truncate_inode_pages_range+0x1a2/0x6f0
  [<ffffffff811828d1>] ? find_get_pages_tag+0xf1/0x1b0
  [<ffffffff8104a663>] ? __switch_to+0x1e3/0x5a0
  [<ffffffff8118dd05>] ? pagevec_lookup_tag+0x25/0x40
  [<ffffffff811f620d>] ? __inode_wait_for_writeback+0x6d/0xc0
  [<ffffffff8119024c>] truncate_inode_pages_final+0x4c/0x60
  [<ffffffff8151c47f>] xfs_fs_evict_inode+0x4f/0x100
  [<ffffffff811e8330>] evict+0xc0/0x1a0
  [<ffffffff811e8d7b>] iput+0x1bb/0x220
  [<ffffffff811f68b3>] sync_inodes_sb+0x353/0x3d0
  [<ffffffff8151def8>] xfs_flush_inodes+0x28/0x40
  [<ffffffff81514648>] xfs_create+0x638/0x770
  [<ffffffff814e9049>] ? xfs_dir2_sf_lookup+0x199/0x330
  [<ffffffff81511091>] xfs_generic_create+0xd1/0x300
  [<ffffffff817a059c>] ? security_inode_permission+0x1c/0x30
  [<ffffffff815112f6>] xfs_vn_create+0x16/0x20
  [<ffffffff811d8665>] vfs_create+0xd5/0x140
  [<ffffffff811dbea3>] do_last+0xff3/0x1200
  [<ffffffff811d9f36>] ? path_init+0x186/0x450
  [<ffffffff811dc130>] path_openat+0x80/0x610
  [<ffffffff81512a24>] ? xfs_iunlock+0xc4/0x210
  [<ffffffff811ddbfa>] do_filp_open+0x3a/0x90
  [<ffffffff811dc8bf>] ? getname_flags+0x4f/0x200
  [<ffffffff81e249ce>] ? _raw_spin_unlock+0xe/0x30
  [<ffffffff811eab17>] ? __alloc_fd+0xa7/0x130
  [<ffffffff811cbcf8>] do_sys_open+0x128/0x220
  [<ffffffff811cbe4e>] SyS_creat+0x1e/0x20
  [<ffffffff81e24fee>] system_call_fastpath+0x12/0x71

It looks to me like iput() is being called with the wb.list_lock
held in wait_sb_inodes(), and everything is going downhill from
there.  Patch below fixes the problem for me.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com


sync: wait_sb_inodes() calls iput() with spinlock held.

From: Dave Chinner <dchinner@redhat.com>

wait_sb_inodes() is triggering "sleeping in atomic" problems with
blocking operations in iput() processing when wait_sb_inodes()
releases the last reference to an inode.  Fix it by delaying the
iput() until the next loop pass when we aren't holding any
spinlocks.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
 fs/fs-writeback.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 1718702..a2cd363 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1436,6 +1436,7 @@ static void wait_sb_inodes(struct super_block *sb)
 {
 	struct backing_dev_info *bdi = sb->s_bdi;
 	LIST_HEAD(sync_list);
+	struct inode *iput_inode = NULL;
 
 	/*
 	 * We need to be protected against the filesystem going from
@@ -1497,6 +1498,9 @@ static void wait_sb_inodes(struct super_block *sb)
 		spin_unlock(&inode->i_lock);
 		spin_unlock(&bdi->wb.list_lock);
 
+		if (iput_inode)
+			iput(iput_inode);
+
 		filemap_fdatawait(mapping);
 		cond_resched();
 
@@ -1516,9 +1520,19 @@ static void wait_sb_inodes(struct super_block *sb)
                 } else
 			list_del_init(&inode->i_wb_list);
 		spin_unlock_irq(&mapping->tree_lock);
-		iput(inode);
+
+		/*
+		 * can't iput inode while holding the wb.list_lock. Save it for
+		 * the next time through the loop when we drop all our spin
+		 * locks.
+		 */
+		iput_inode = inode;
 	}
 	spin_unlock(&bdi->wb.list_lock);
+
+	if (iput_inode)
+		iput(iput_inode);
+
 	mutex_unlock(&sb->s_sync_lock);
 }
 
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in

Re: [PATCH] sync: wait_sb_inodes() calls iput() with spinlock held (was Re: [PATCH 0/7] super block scalabilit patches V3)

[Josef Bacik]

On 06/21/2015 07:26 PM, Dave Chinner wrote:
> On Tue, Jun 16, 2015 at 07:34:29AM +1000, Dave Chinner wrote:
>> On Thu, Jun 11, 2015 at 03:41:05PM -0400, Josef Bacik wrote:
>>> Here are the cleaned up versions of Dave Chinners super block scalability
>>> patches.  I've been testing them locally for a while and they are pretty solid.
>>> They fix a few big issues, such as the global inode list and soft lockups on
>>> boxes on unmount that have lots of inodes in cache.  Al if you would consider
>>> pulling these in that would be great, you can pull from here
>>>
>>> git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling
>>
>> Passes all my smoke tests.
>>
>> Tested-by: Dave Chinner <dchinner@redhat.com>
>
> FWIW, I just updated my trees to whatever is in the above branch and
> v4.1-rc8, and now I'm seeing problems with wb.list_lock recursion
> and "sleeping in atomic" scehduling issues. generic/269 produced
> this:
>
>   BUG: spinlock cpu recursion on CPU#1, fsstress/3852
>    lock: 0xffff88042a650c28, .magic: dead4ead, .owner: fsstress/3804, .owner_cpu: 1
>   CPU: 1 PID: 3852 Comm: fsstress Tainted: G        W       4.1.0-rc8-dgc+ #263
>   Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>    ffff88042a650c28 ffff88039898b8e8 ffffffff81e18ffd ffff88042f250fb0
>    ffff880428f6b8e0 ffff88039898b908 ffffffff81e12f09 ffff88042a650c28
>    ffffffff8221337b ffff88039898b928 ffffffff81e12f34 ffff88042a650c28
>   Call Trace:
>    [<ffffffff81e18ffd>] dump_stack+0x4c/0x6e
>    [<ffffffff81e12f09>] spin_dump+0x90/0x95
>    [<ffffffff81e12f34>] spin_bug+0x26/0x2b
>    [<ffffffff810e762d>] do_raw_spin_lock+0x10d/0x150
>    [<ffffffff81e24975>] _raw_spin_lock+0x15/0x20
>    [<ffffffff811f8ba0>] __mark_inode_dirty+0x2b0/0x450
>    [<ffffffff812003b8>] __set_page_dirty+0x78/0xd0
>    [<ffffffff81200531>] mark_buffer_dirty+0x61/0xf0
>    [<ffffffff81200d91>] __block_commit_write.isra.24+0x81/0xb0
>    [<ffffffff81202406>] block_write_end+0x36/0x70
>    [<ffffffff814fa110>] ? __xfs_get_blocks+0x8a0/0x8a0
>    [<ffffffff81202474>] generic_write_end+0x34/0xb0
>    [<ffffffff8118af3d>] ? wait_for_stable_page+0x1d/0x50
>    [<ffffffff814fa317>] xfs_vm_write_end+0x67/0xc0
>    [<ffffffff811813af>] pagecache_write_end+0x1f/0x30
>    [<ffffffff815060dd>] xfs_iozero+0x10d/0x190
>    [<ffffffff8150666b>] xfs_zero_last_block+0xdb/0x110
>    [<ffffffff815067ba>] xfs_zero_eof+0x11a/0x290
>    [<ffffffff811d69e0>] ? complete_walk+0x60/0x100
>    [<ffffffff811da25f>] ? path_lookupat+0x5f/0x660
>    [<ffffffff81506a6e>] xfs_file_aio_write_checks+0x13e/0x160
>    [<ffffffff81506f15>] xfs_file_buffered_aio_write+0x75/0x250
>    [<ffffffff811ddb0f>] ? user_path_at_empty+0x5f/0xa0
>    [<ffffffff810c601d>] ? __might_sleep+0x4d/0x90
>    [<ffffffff815071f5>] xfs_file_write_iter+0x105/0x120
>    [<ffffffff811cc5ce>] __vfs_write+0xae/0xf0
>    [<ffffffff811ccc01>] vfs_write+0xa1/0x190
>    [<ffffffff811cd999>] SyS_write+0x49/0xb0
>    [<ffffffff811cc781>] ? SyS_lseek+0x91/0xb0
>    [<ffffffff81e24fee>] system_call_fastpath+0x12/0x71
>
> And there are a few tests (including generic/269) producing
> in_atomic/"scheduling while atomic" bugs in the evict() path such as:
>
>   in_atomic(): 1, irqs_disabled(): 0, pid: 3852, name: fsstress
>   CPU: 12 PID: 3852 Comm: fsstress Not tainted 4.1.0-rc8-dgc+ #263
>   Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>    000000000000015d ffff88039898b6d8 ffffffff81e18ffd 0000000000000000
>    ffff880398865550 ffff88039898b6f8 ffffffff810c5f89 ffff8803f15c45c0
>    ffffffff8227a3bf ffff88039898b728 ffffffff810c601d ffff88039898b758
>   Call Trace:
>    [<ffffffff81e18ffd>] dump_stack+0x4c/0x6e
>    [<ffffffff810c5f89>] ___might_sleep+0xf9/0x140
>    [<ffffffff810c601d>] __might_sleep+0x4d/0x90
>    [<ffffffff81201e8b>] block_invalidatepage+0xab/0x140
>    [<ffffffff814f7579>] xfs_vm_invalidatepage+0x39/0xb0
>    [<ffffffff8118fa77>] truncate_inode_page+0x67/0xa0
>    [<ffffffff8118fc92>] truncate_inode_pages_range+0x1a2/0x6f0
>    [<ffffffff811828d1>] ? find_get_pages_tag+0xf1/0x1b0
>    [<ffffffff8104a663>] ? __switch_to+0x1e3/0x5a0
>    [<ffffffff8118dd05>] ? pagevec_lookup_tag+0x25/0x40
>    [<ffffffff811f620d>] ? __inode_wait_for_writeback+0x6d/0xc0
>    [<ffffffff8119024c>] truncate_inode_pages_final+0x4c/0x60
>    [<ffffffff8151c47f>] xfs_fs_evict_inode+0x4f/0x100
>    [<ffffffff811e8330>] evict+0xc0/0x1a0
>    [<ffffffff811e8d7b>] iput+0x1bb/0x220
>    [<ffffffff811f68b3>] sync_inodes_sb+0x353/0x3d0
>    [<ffffffff8151def8>] xfs_flush_inodes+0x28/0x40
>    [<ffffffff81514648>] xfs_create+0x638/0x770
>    [<ffffffff814e9049>] ? xfs_dir2_sf_lookup+0x199/0x330
>    [<ffffffff81511091>] xfs_generic_create+0xd1/0x300
>    [<ffffffff817a059c>] ? security_inode_permission+0x1c/0x30
>    [<ffffffff815112f6>] xfs_vn_create+0x16/0x20
>    [<ffffffff811d8665>] vfs_create+0xd5/0x140
>    [<ffffffff811dbea3>] do_last+0xff3/0x1200
>    [<ffffffff811d9f36>] ? path_init+0x186/0x450
>    [<ffffffff811dc130>] path_openat+0x80/0x610
>    [<ffffffff81512a24>] ? xfs_iunlock+0xc4/0x210
>    [<ffffffff811ddbfa>] do_filp_open+0x3a/0x90
>    [<ffffffff811dc8bf>] ? getname_flags+0x4f/0x200
>    [<ffffffff81e249ce>] ? _raw_spin_unlock+0xe/0x30
>    [<ffffffff811eab17>] ? __alloc_fd+0xa7/0x130
>    [<ffffffff811cbcf8>] do_sys_open+0x128/0x220
>    [<ffffffff811cbe4e>] SyS_creat+0x1e/0x20
>    [<ffffffff81e24fee>] system_call_fastpath+0x12/0x71
>
> It looks to me like iput() is being called with the wb.list_lock
> held in wait_sb_inodes(), and everything is going downhill from
> there.  Patch below fixes the problem for me.
>
> Cheers,
>
> Dave.
>

Thanks Dave I'll add it.  I think this is what we were doing at first 
but then I changed it, didn't notice the wb.list_lock.  Thanks,

Josef
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in

Re: [PATCH] sync: wait_sb_inodes() calls iput() with spinlock held (was Re: [PATCH 0/7] super block scalabilit patches V3)

[Josef Bacik]

On 06/21/2015 07:26 PM, Dave Chinner wrote:
> On Tue, Jun 16, 2015 at 07:34:29AM +1000, Dave Chinner wrote:
>> On Thu, Jun 11, 2015 at 03:41:05PM -0400, Josef Bacik wrote:
>>> Here are the cleaned up versions of Dave Chinners super block scalability
>>> patches.  I've been testing them locally for a while and they are pretty solid.
>>> They fix a few big issues, such as the global inode list and soft lockups on
>>> boxes on unmount that have lots of inodes in cache.  Al if you would consider
>>> pulling these in that would be great, you can pull from here
>>>
>>> git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling
>>
>> Passes all my smoke tests.
>>
>> Tested-by: Dave Chinner <dchinner@redhat.com>
>
> FWIW, I just updated my trees to whatever is in the above branch and
> v4.1-rc8, and now I'm seeing problems with wb.list_lock recursion
> and "sleeping in atomic" scehduling issues. generic/269 produced
> this:
>
>   BUG: spinlock cpu recursion on CPU#1, fsstress/3852
>    lock: 0xffff88042a650c28, .magic: dead4ead, .owner: fsstress/3804, .owner_cpu: 1
>   CPU: 1 PID: 3852 Comm: fsstress Tainted: G        W       4.1.0-rc8-dgc+ #263
>   Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>    ffff88042a650c28 ffff88039898b8e8 ffffffff81e18ffd ffff88042f250fb0
>    ffff880428f6b8e0 ffff88039898b908 ffffffff81e12f09 ffff88042a650c28
>    ffffffff8221337b ffff88039898b928 ffffffff81e12f34 ffff88042a650c28
>   Call Trace:
>    [<ffffffff81e18ffd>] dump_stack+0x4c/0x6e
>    [<ffffffff81e12f09>] spin_dump+0x90/0x95
>    [<ffffffff81e12f34>] spin_bug+0x26/0x2b
>    [<ffffffff810e762d>] do_raw_spin_lock+0x10d/0x150
>    [<ffffffff81e24975>] _raw_spin_lock+0x15/0x20
>    [<ffffffff811f8ba0>] __mark_inode_dirty+0x2b0/0x450
>    [<ffffffff812003b8>] __set_page_dirty+0x78/0xd0
>    [<ffffffff81200531>] mark_buffer_dirty+0x61/0xf0
>    [<ffffffff81200d91>] __block_commit_write.isra.24+0x81/0xb0
>    [<ffffffff81202406>] block_write_end+0x36/0x70
>    [<ffffffff814fa110>] ? __xfs_get_blocks+0x8a0/0x8a0
>    [<ffffffff81202474>] generic_write_end+0x34/0xb0
>    [<ffffffff8118af3d>] ? wait_for_stable_page+0x1d/0x50
>    [<ffffffff814fa317>] xfs_vm_write_end+0x67/0xc0
>    [<ffffffff811813af>] pagecache_write_end+0x1f/0x30
>    [<ffffffff815060dd>] xfs_iozero+0x10d/0x190
>    [<ffffffff8150666b>] xfs_zero_last_block+0xdb/0x110
>    [<ffffffff815067ba>] xfs_zero_eof+0x11a/0x290
>    [<ffffffff811d69e0>] ? complete_walk+0x60/0x100
>    [<ffffffff811da25f>] ? path_lookupat+0x5f/0x660
>    [<ffffffff81506a6e>] xfs_file_aio_write_checks+0x13e/0x160
>    [<ffffffff81506f15>] xfs_file_buffered_aio_write+0x75/0x250
>    [<ffffffff811ddb0f>] ? user_path_at_empty+0x5f/0xa0
>    [<ffffffff810c601d>] ? __might_sleep+0x4d/0x90
>    [<ffffffff815071f5>] xfs_file_write_iter+0x105/0x120
>    [<ffffffff811cc5ce>] __vfs_write+0xae/0xf0
>    [<ffffffff811ccc01>] vfs_write+0xa1/0x190
>    [<ffffffff811cd999>] SyS_write+0x49/0xb0
>    [<ffffffff811cc781>] ? SyS_lseek+0x91/0xb0
>    [<ffffffff81e24fee>] system_call_fastpath+0x12/0x71
>
> And there are a few tests (including generic/269) producing
> in_atomic/"scheduling while atomic" bugs in the evict() path such as:
>
>   in_atomic(): 1, irqs_disabled(): 0, pid: 3852, name: fsstress
>   CPU: 12 PID: 3852 Comm: fsstress Not tainted 4.1.0-rc8-dgc+ #263
>   Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>    000000000000015d ffff88039898b6d8 ffffffff81e18ffd 0000000000000000
>    ffff880398865550 ffff88039898b6f8 ffffffff810c5f89 ffff8803f15c45c0
>    ffffffff8227a3bf ffff88039898b728 ffffffff810c601d ffff88039898b758
>   Call Trace:
>    [<ffffffff81e18ffd>] dump_stack+0x4c/0x6e
>    [<ffffffff810c5f89>] ___might_sleep+0xf9/0x140
>    [<ffffffff810c601d>] __might_sleep+0x4d/0x90
>    [<ffffffff81201e8b>] block_invalidatepage+0xab/0x140
>    [<ffffffff814f7579>] xfs_vm_invalidatepage+0x39/0xb0
>    [<ffffffff8118fa77>] truncate_inode_page+0x67/0xa0
>    [<ffffffff8118fc92>] truncate_inode_pages_range+0x1a2/0x6f0
>    [<ffffffff811828d1>] ? find_get_pages_tag+0xf1/0x1b0
>    [<ffffffff8104a663>] ? __switch_to+0x1e3/0x5a0
>    [<ffffffff8118dd05>] ? pagevec_lookup_tag+0x25/0x40
>    [<ffffffff811f620d>] ? __inode_wait_for_writeback+0x6d/0xc0
>    [<ffffffff8119024c>] truncate_inode_pages_final+0x4c/0x60
>    [<ffffffff8151c47f>] xfs_fs_evict_inode+0x4f/0x100
>    [<ffffffff811e8330>] evict+0xc0/0x1a0
>    [<ffffffff811e8d7b>] iput+0x1bb/0x220
>    [<ffffffff811f68b3>] sync_inodes_sb+0x353/0x3d0
>    [<ffffffff8151def8>] xfs_flush_inodes+0x28/0x40
>    [<ffffffff81514648>] xfs_create+0x638/0x770
>    [<ffffffff814e9049>] ? xfs_dir2_sf_lookup+0x199/0x330
>    [<ffffffff81511091>] xfs_generic_create+0xd1/0x300
>    [<ffffffff817a059c>] ? security_inode_permission+0x1c/0x30
>    [<ffffffff815112f6>] xfs_vn_create+0x16/0x20
>    [<ffffffff811d8665>] vfs_create+0xd5/0x140
>    [<ffffffff811dbea3>] do_last+0xff3/0x1200
>    [<ffffffff811d9f36>] ? path_init+0x186/0x450
>    [<ffffffff811dc130>] path_openat+0x80/0x610
>    [<ffffffff81512a24>] ? xfs_iunlock+0xc4/0x210
>    [<ffffffff811ddbfa>] do_filp_open+0x3a/0x90
>    [<ffffffff811dc8bf>] ? getname_flags+0x4f/0x200
>    [<ffffffff81e249ce>] ? _raw_spin_unlock+0xe/0x30
>    [<ffffffff811eab17>] ? __alloc_fd+0xa7/0x130
>    [<ffffffff811cbcf8>] do_sys_open+0x128/0x220
>    [<ffffffff811cbe4e>] SyS_creat+0x1e/0x20
>    [<ffffffff81e24fee>] system_call_fastpath+0x12/0x71
>
> It looks to me like iput() is being called with the wb.list_lock
> held in wait_sb_inodes(), and everything is going downhill from
> there.  Patch below fixes the problem for me.
>

I folded this into "bdi: add a new writeback list for sync" since it was 
there before and to be more bisect friendly.  Let me know if this isn't 
ok with you and I'll undo it.  Thanks,

Josef

Re: [PATCH] sync: wait_sb_inodes() calls iput() with spinlock held (was Re: [PATCH 0/7] super block scalabilit patches V3)

[Dave Chinner]

On Tue, Jun 23, 2015 at 04:14:42PM -0700, Josef Bacik wrote:
> On 06/21/2015 07:26 PM, Dave Chinner wrote:
> >On Tue, Jun 16, 2015 at 07:34:29AM +1000, Dave Chinner wrote:
> >>On Thu, Jun 11, 2015 at 03:41:05PM -0400, Josef Bacik wrote:
> >>>Here are the cleaned up versions of Dave Chinners super block scalability
> >>>patches.  I've been testing them locally for a while and they are pretty solid.
> >>>They fix a few big issues, such as the global inode list and soft lockups on
> >>>boxes on unmount that have lots of inodes in cache.  Al if you would consider
> >>>pulling these in that would be great, you can pull from here
> >>>
> >>>git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling
> >>
> >>Passes all my smoke tests.
> >>
> >>Tested-by: Dave Chinner <dchinner@redhat.com>
> >
> >FWIW, I just updated my trees to whatever is in the above branch and
> >v4.1-rc8, and now I'm seeing problems with wb.list_lock recursion
> >and "sleeping in atomic" scehduling issues. generic/269 produced
> >this:
....
> >It looks to me like iput() is being called with the wb.list_lock
> >held in wait_sb_inodes(), and everything is going downhill from
> >there.  Patch below fixes the problem for me.
> >
> 
> I folded this into "bdi: add a new writeback list for sync" since it
> was there before and to be more bisect friendly.  Let me know if
> this isn't ok with you and I'll undo it.  Thanks,

That's fine - I was expecting you would fold it back in... ;)

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

[PATCH 0/8] super block scalability patches V4

[Josef Bacik]

This is the latest round of patches, it includes all the new reviewed-by's, the
fixed up 'bdi: add a new writeback list for sync' patch that has Jan's
suggestions included and Dave's patch folded in to address the problem he saw in
testing.  I've rebased these on Linus's master from this evening.  Al would you
please consider pulling


git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next.git superblock-scaling

Thanks,

Josef

[PATCH 7/8] writeback: periodically trim the writeback list

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

Inodes are removed lazily from the bdi writeback list, so in the
absence of sync(2) work inodes will build up on the bdi writback
list even though they are no longer under IO. Use the periodic
kupdate work check to remove inodes no longer under IO from the
writeback list.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: Dave Chinner <dchinner@redhat.com>
---
 fs/fs-writeback.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 5c005ad..a2cd363 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1060,6 +1060,23 @@ static long wb_check_background_flush(struct bdi_writeback *wb)
 	return 0;
 }
 
+/*
+ * clean out writeback list for all inodes that don't have IO in progress
+ */
+static void wb_trim_writeback_list(struct bdi_writeback *wb)
+{
+	struct inode *inode;
+	struct inode *tmp;
+
+	spin_lock(&wb->list_lock);
+	list_for_each_entry_safe(inode, tmp, &wb->b_writeback, i_wb_list) {
+		if (!mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK))
+			list_del_init(&inode->i_wb_list);
+	}
+	spin_unlock(&wb->list_lock);
+
+}
+
 static long wb_check_old_data_flush(struct bdi_writeback *wb)
 {
 	unsigned long expired;
@@ -1076,6 +1093,8 @@ static long wb_check_old_data_flush(struct bdi_writeback *wb)
 	if (time_before(jiffies, expired))
 		return 0;
 
+	wb_trim_writeback_list(wb);
+
 	wb->last_old_flush = jiffies;
 	nr_pages = get_nr_dirty_pages();
 
-- 
2.1.0

[PATCH 0/8] Sync and VFS scalability improvements V2

[Josef Bacik]

Here is the update to the patches based on comments from Al and Jan.  The
changes since V1 are

- Dropped the IOP_NOTHASHED patch and replaced it wit the hlist_fake patch as
  per Al's suggestion.
- Dropped "inode: convert per-sb inode list to a list_lru" since Jan has found
  some issues with it and it doesn't seem as awesome as we want.
- Fixed "inode: don't softlockup when evicting inodes" since it was based on the
  list_lru patch.
- Added my SOB to all the patches as per Jan's suggestion and added his relevant
  Reviewed-by's.

I'm getting new performance numbers for these patches now, hopefully that stuff
will finish soon and I can get them posted.  I wanted to get the updated
patchset up as quickly as possible since we're getting close to the merge window
and I'm getting ready to disappear into vacation for a little bit.  Thanks,

Josef

[PATCH 7/8] writeback: periodically trim the writeback list

[Josef Bacik]

From: Dave Chinner <dchinner@redhat.com>

Inodes are removed lazily from the bdi writeback list, so in the
absence of sync(2) work inodes will build up on the bdi writback
list even though they are no longer under IO. Use the periodic
kupdate work check to remove inodes no longer under IO from the
writeback list.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Jan Kara <jack@suse.cz>
---
 fs/fs-writeback.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 82b0f43..aa0de0f 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1037,6 +1037,23 @@ static long wb_check_background_flush(struct bdi_writeback *wb)
 	return 0;
 }
 
+/*
+ * clean out writeback list for all inodes that don't have IO in progress
+ */
+static void wb_trim_writeback_list(struct bdi_writeback *wb)
+{
+	struct inode *inode;
+	struct inode *tmp;
+
+	spin_lock(&wb->list_lock);
+	list_for_each_entry_safe(inode, tmp, &wb->b_writeback, i_wb_list) {
+		if (!mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK))
+			list_del_init(&inode->i_wb_list);
+	}
+	spin_unlock(&wb->list_lock);
+
+}
+
 static long wb_check_old_data_flush(struct bdi_writeback *wb)
 {
 	unsigned long expired;
@@ -1053,6 +1070,8 @@ static long wb_check_old_data_flush(struct bdi_writeback *wb)
 	if (time_before(jiffies, expired))
 		return 0;
 
+	wb_trim_writeback_list(wb);
+
 	wb->last_old_flush = jiffies;
 	nr_pages = get_nr_dirty_pages();
 
-- 
1.8.3.1