erofs: convert workstn to XArray

XArray has friendly APIs and it will replace the old radix
tree in the near future.

This convert makes use of __xa_cmpxchg when inserting on
a just inserted item by other thread. In detail, instead
of totally looking up again as what we did for the old
radix tree, it will try to legitimize the current in-tree
item in the XArray therefore more effective.

In addition, naming is rather a challenge for non-English
speaker like me. The basic idea of workstn is to provide
a runtime sparse array with items arranged in the physical
block number order. Such items (was called workgroup) can be
used to record compress clusters or for later new features.

However, both workgroup and workstn seem not good names from
whatever point of view, so I'd like to rename them as pslot
and managed_pslots to stand for physical slots. This patch
handles the second as a part of the radix tree convert.

Cc: Matthew Wilcox <willy@infradead.org>
Link: https://lore.kernel.org/r/20200220024642.91529-1-gaoxiang25@huawei.com
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Gao Xiang <gaoxiang25@huawei.com>

1 files changed, 39 insertions, 37 deletions

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index 80e47f07d946..c4b6c9aa87ec 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -67,16 +67,6 @@ static void z_erofs_pcluster_init_once(void *ptr)
 		pcl->compressed_pages[i] = NULL;
 }
 
-static void z_erofs_pcluster_init_always(struct z_erofs_pcluster *pcl)
-{
-	struct z_erofs_collection *cl = z_erofs_primarycollection(pcl);
-
-	atomic_set(&pcl->obj.refcount, 1);
-
-	DBG_BUGON(cl->nr_pages);
-	DBG_BUGON(cl->vcnt);
-}
-
 int __init z_erofs_init_zip_subsystem(void)
 {
 	pcluster_cachep = kmem_cache_create("erofs_compress",
@@ -341,26 +331,19 @@ static int z_erofs_lookup_collection(struct z_erofs_collector *clt,
 				     struct inode *inode,
 				     struct erofs_map_blocks *map)
 {
-	struct erofs_workgroup *grp;
-	struct z_erofs_pcluster *pcl;
+	struct z_erofs_pcluster *pcl = clt->pcl;
 	struct z_erofs_collection *cl;
 	unsigned int length;
 
-	grp = erofs_find_workgroup(inode->i_sb, map->m_pa >> PAGE_SHIFT);
-	if (!grp)
-		return -ENOENT;
-
-	pcl = container_of(grp, struct z_erofs_pcluster, obj);
+	/* to avoid unexpected loop formed by corrupted images */
 	if (clt->owned_head == &pcl->next || pcl == clt->tailpcl) {
 		DBG_BUGON(1);
-		erofs_workgroup_put(grp);
 		return -EFSCORRUPTED;
 	}
 
 	cl = z_erofs_primarycollection(pcl);
 	if (cl->pageofs != (map->m_la & ~PAGE_MASK)) {
 		DBG_BUGON(1);
-		erofs_workgroup_put(grp);
 		return -EFSCORRUPTED;
 	}
 
@@ -368,7 +351,6 @@ static int z_erofs_lookup_collection(struct z_erofs_collector *clt,
 	if (length & Z_EROFS_PCLUSTER_FULL_LENGTH) {
 		if ((map->m_llen << Z_EROFS_PCLUSTER_LENGTH_BIT) > length) {
 			DBG_BUGON(1);
-			erofs_workgroup_put(grp);
 			return -EFSCORRUPTED;
 		}
 	} else {
@@ -391,7 +373,6 @@ static int z_erofs_lookup_collection(struct z_erofs_collector *clt,
 	/* clean tailpcl if the current owned_head is Z_EROFS_PCLUSTER_TAIL */
 	if (clt->owned_head == Z_EROFS_PCLUSTER_TAIL)
 		clt->tailpcl = NULL;
-	clt->pcl = pcl;
 	clt->cl = cl;
 	return 0;
 }
@@ -402,6 +383,7 @@ static int z_erofs_register_collection(struct z_erofs_collector *clt,
 {
 	struct z_erofs_pcluster *pcl;
 	struct z_erofs_collection *cl;
+	struct erofs_workgroup *grp;
 	int err;
 
 	/* no available workgroup, let's allocate one */
@@ -409,7 +391,7 @@ static int z_erofs_register_collection(struct z_erofs_collector *clt,
 	if (!pcl)
 		return -ENOMEM;
 
-	z_erofs_pcluster_init_always(pcl);
+	atomic_set(&pcl->obj.refcount, 1);
 	pcl->obj.index = map->m_pa >> PAGE_SHIFT;
 
 	pcl->length = (map->m_llen << Z_EROFS_PCLUSTER_LENGTH_BIT) |
@@ -429,19 +411,29 @@ static int z_erofs_register_collection(struct z_erofs_collector *clt,
 	clt->mode = COLLECT_PRIMARY_FOLLOWED;
 
 	cl = z_erofs_primarycollection(pcl);
+
+	/* must be cleaned before freeing to slab */
+	DBG_BUGON(cl->nr_pages);
+	DBG_BUGON(cl->vcnt);
+
 	cl->pageofs = map->m_la & ~PAGE_MASK;
 
 	/*
 	 * lock all primary followed works before visible to others
 	 * and mutex_trylock *never* fails for a new pcluster.
 	 */
-	mutex_trylock(&cl->lock);
+	DBG_BUGON(!mutex_trylock(&cl->lock));
 
-	err = erofs_register_workgroup(inode->i_sb, &pcl->obj);
-	if (err) {
-		mutex_unlock(&cl->lock);
-		kmem_cache_free(pcluster_cachep, pcl);
-		return -EAGAIN;
+	grp = erofs_insert_workgroup(inode->i_sb, &pcl->obj);
+	if (IS_ERR(grp)) {
+		err = PTR_ERR(grp);
+		goto err_out;
+	}
+
+	if (grp != &pcl->obj) {
+		clt->pcl = container_of(grp, struct z_erofs_pcluster, obj);
+		err = -EEXIST;
+		goto err_out;
 	}
 	/* used to check tail merging loop due to corrupted images */
 	if (clt->owned_head == Z_EROFS_PCLUSTER_TAIL)
@@ -450,12 +442,18 @@ static int z_erofs_register_collection(struct z_erofs_collector *clt,
 	clt->pcl = pcl;
 	clt->cl = cl;
 	return 0;
+
+err_out:
+	mutex_unlock(&cl->lock);
+	kmem_cache_free(pcluster_cachep, pcl);
+	return err;
 }
 
 static int z_erofs_collector_begin(struct z_erofs_collector *clt,
 				   struct inode *inode,
 				   struct erofs_map_blocks *map)
 {
+	struct erofs_workgroup *grp;
 	int ret;
 
 	DBG_BUGON(clt->cl);
@@ -469,21 +467,25 @@ static int z_erofs_collector_begin(struct z_erofs_collector *clt,
 		return -EINVAL;
 	}
 
-repeat:
-	ret = z_erofs_lookup_collection(clt, inode, map);
-	if (ret == -ENOENT) {
+	grp = erofs_find_workgroup(inode->i_sb, map->m_pa >> PAGE_SHIFT);
+	if (grp) {
+		clt->pcl = container_of(grp, struct z_erofs_pcluster, obj);
+	} else {
 		ret = z_erofs_register_collection(clt, inode, map);
 
-		/* someone registered at the same time, give another try */
-		if (ret == -EAGAIN) {
-			cond_resched();
-			goto repeat;
-		}
+		if (!ret)
+			goto out;
+		if (ret != -EEXIST)
+			return ret;
 	}
 
-	if (ret)
+	ret = z_erofs_lookup_collection(clt, inode, map);
+	if (ret) {
+		erofs_workgroup_put(&clt->pcl->obj);
 		return ret;
+	}
 
+out:
 	z_erofs_pagevec_ctor_init(&clt->vector, Z_EROFS_NR_INLINE_PAGEVECS,
 				  clt->cl->pagevec, clt->cl->vcnt);