Re: [RFC PATCH 2/2] ext4: improve extents status tree shrinker to avoid scanning delayed entries

[Zheng Liu <gnehzuil.liu@gmail.com>]

On Mon, Dec 23, 2013 at 09:54:19AM +0100, Jan Kara wrote:
> On Fri 20-12-13 18:42:45, Zheng Liu wrote:
> > From: Zheng Liu <wenqing.lz@taobao.com>
> > 
> > The extents status tree shrinker will scan all inodes on sbi->s_es_lru
> > under heavy memory pressure, and try to reclaim the entry from extents
> > status tree.  During this process it couldn't reclaim the delayed entry
> > because ext4 needs to use these entries to do delayed allocation space
> > reservation, seek_data/hole, etc....  So if a system has done a huge
> > number of writes and these dirty pages don't be written out.  There will
> > be a lot of delayed entries on extents status tree.  If shrinker tries
> > to reclaim memory from the tree, it will burn some CPU time to iterate
> > on these non-reclaimable entries.  At some circumstances it could cause
> > excessive stall time.
> > 
> > In this commit a new list is used to track reclaimable entries of extent
> > status tree (e.g. written/unwritten/hole entries).  The shrinker will
> > scan reclaimable entry on this list.  So it won't encouter any delayed
> > entry and don't need to take too much time to spin.  But the defect is
> > that we need to cost extra 1/3 memory space for one entry.  Before this
> > commit, 'struct extent_status' occupies 48 bytes on a 64bits platform.
> > After that it will occupy 64 bytes. :(
>   This looks sensible. I was just wondering about one thing: One incorrect
> thing the old extent shrinker does is that it tries to reclaim 'nr_to_scan'
> objects. That is wrong - it should *scan* 'nr_to_scan' objects and reclaim
> objects it can find. Now we shouldn't always start scanning at the end of
> the LRU because if delayed extents accumulate there we would never reclaim
> anything. Rather we should cycle through the list of entries we have. But
> that doesn't play well with the fact we have LRU list and thus want to
> reclaim from the end of the list. In the end what you do might be the best
> we can do but I wanted to mention the above just in case someone has some
> idea.

Ah, thanks for pointing it out.  So maybe we can fix this issue before
we are sure that the new improvement is acceptable because it makes us
avoid scanning too many objects.  What do you think?

Regards,
                                                - Zheng

> 
> 								Honza
> 
> > Cc: "Theodore Ts'o" <tytso@mit.edu>
> > Cc: Andreas Dilger <adilger.kernel@dilger.ca>
> > Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
> > ---
> >  fs/ext4/extents_status.c |   38 +++++++++++++++++++-------------------
> >  fs/ext4/extents_status.h |    2 ++
> >  2 files changed, 21 insertions(+), 19 deletions(-)
> > 
> > diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c
> > index e842d74..11bdb2f 100644
> > --- a/fs/ext4/extents_status.c
> > +++ b/fs/ext4/extents_status.c
> > @@ -169,6 +169,7 @@ void ext4_exit_es(void)
> >  void ext4_es_init_tree(struct ext4_es_tree *tree)
> >  {
> >  	tree->root = RB_ROOT;
> > +	INIT_HLIST_HEAD(&tree->evictable_list);
> >  	tree->cache_es = NULL;
> >  }
> >  
> > @@ -300,10 +301,14 @@ static struct extent_status *
> >  ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len,
> >  		     ext4_fsblk_t pblk)
> >  {
> > +	struct ext4_inode_info *ei = EXT4_I(inode);
> >  	struct extent_status *es;
> > +
> >  	es = kmem_cache_alloc(ext4_es_cachep, GFP_ATOMIC);
> >  	if (es == NULL)
> >  		return NULL;
> > +
> > +	INIT_HLIST_NODE(&es->es_list);
> >  	es->es_lblk = lblk;
> >  	es->es_len = len;
> >  	es->es_pblk = pblk;
> > @@ -312,8 +317,9 @@ ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len,
> >  	 * We don't count delayed extent because we never try to reclaim them
> >  	 */
> >  	if (!ext4_es_is_delayed(es)) {
> > -		EXT4_I(inode)->i_es_lru_nr++;
> > +		ei->i_es_lru_nr++;
> >  		percpu_counter_inc(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt);
> > +		hlist_add_head(&es->es_list, &ei->i_es_tree.evictable_list);
> >  	}
> >  
> >  	return es;
> > @@ -321,10 +327,12 @@ ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len,
> >  
> >  static void ext4_es_free_extent(struct inode *inode, struct extent_status *es)
> >  {
> > +	struct ext4_inode_info *ei = EXT4_I(inode);
> > +
> >  	/* Decrease the lru counter when this es is not delayed */
> >  	if (!ext4_es_is_delayed(es)) {
> > -		BUG_ON(EXT4_I(inode)->i_es_lru_nr == 0);
> > -		EXT4_I(inode)->i_es_lru_nr--;
> > +		BUG_ON(ei->i_es_lru_nr-- == 0);
> > +		hlist_del_init(&es->es_list);
> >  		percpu_counter_dec(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt);
> >  	}
> >  
> > @@ -1092,8 +1100,8 @@ static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei,
> >  {
> >  	struct inode *inode = &ei->vfs_inode;
> >  	struct ext4_es_tree *tree = &ei->i_es_tree;
> > -	struct rb_node *node;
> >  	struct extent_status *es;
> > +	struct hlist_node *tmp;
> >  	unsigned long nr_shrunk = 0;
> >  	static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL,
> >  				      DEFAULT_RATELIMIT_BURST);
> > @@ -1105,21 +1113,13 @@ static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei,
> >  	    __ratelimit(&_rs))
> >  		ext4_warning(inode->i_sb, "forced shrink of precached extents");
> >  
> > -	node = rb_first(&tree->root);
> > -	while (node != NULL) {
> > -		es = rb_entry(node, struct extent_status, rb_node);
> > -		node = rb_next(&es->rb_node);
> > -		/*
> > -		 * We can't reclaim delayed extent from status tree because
> > -		 * fiemap, bigallic, and seek_data/hole need to use it.
> > -		 */
> > -		if (!ext4_es_is_delayed(es)) {
> > -			rb_erase(&es->rb_node, &tree->root);
> > -			ext4_es_free_extent(inode, es);
> > -			nr_shrunk++;
> > -			if (--nr_to_scan == 0)
> > -				break;
> > -		}
> > +	hlist_for_each_entry_safe(es, tmp, &tree->evictable_list, es_list) {
> > +		BUG_ON(ext4_es_is_delayed(es));
> > +		rb_erase(&es->rb_node, &tree->root);
> > +		ext4_es_free_extent(inode, es);
> > +		nr_shrunk++;
> > +		if (--nr_to_scan == 0)
> > +			break;
> >  	}
> >  	tree->cache_es = NULL;
> >  	return nr_shrunk;
> > diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h
> > index 167f4ab8..38ca83e 100644
> > --- a/fs/ext4/extents_status.h
> > +++ b/fs/ext4/extents_status.h
> > @@ -54,6 +54,7 @@ struct ext4_extent;
> >  
> >  struct extent_status {
> >  	struct rb_node rb_node;
> > +	struct hlist_node es_list;
> >  	ext4_lblk_t es_lblk;	/* first logical block extent covers */
> >  	ext4_lblk_t es_len;	/* length of extent in block */
> >  	ext4_fsblk_t es_pblk;	/* first physical block */
> > @@ -61,6 +62,7 @@ struct extent_status {
> >  
> >  struct ext4_es_tree {
> >  	struct rb_root root;
> > +	struct hlist_head evictable_list;
> >  	struct extent_status *cache_es;	/* recently accessed extent */
> >  };
> >  
> > -- 
> > 1.7.9.7
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> -- 
> Jan Kara <jack@suse.cz>
> SUSE Labs, CR