1 files changed, 214 insertions, 29 deletions

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 65b2e3f85f52..14796b744e0a 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -33,6 +33,7 @@
 #include "xfs_bmap_util.h"
 #include "xfs_dquot_item.h"
 #include "xfs_dquot.h"
+#include "xfs_reflink.h"
 
 #include <linux/kthread.h>
 #include <linux/freezer.h>
@@ -76,6 +77,9 @@ xfs_inode_alloc(
 	ip->i_mount = mp;
 	memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
 	ip->i_afp = NULL;
+	ip->i_cowfp = NULL;
+	ip->i_cnextents = 0;
+	ip->i_cformat = XFS_DINODE_FMT_EXTENTS;
 	memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
 	ip->i_flags = 0;
 	ip->i_delayed_blks = 0;
@@ -101,6 +105,8 @@ xfs_inode_free_callback(
 
 	if (ip->i_afp)
 		xfs_idestroy_fork(ip, XFS_ATTR_FORK);
+	if (ip->i_cowfp)
+		xfs_idestroy_fork(ip, XFS_COW_FORK);
 
 	if (ip->i_itemp) {
 		ASSERT(!(ip->i_itemp->ili_item.li_flags & XFS_LI_IN_AIL));
@@ -787,6 +793,33 @@ xfs_eofblocks_worker(
 	xfs_queue_eofblocks(mp);
 }
 
+/*
+ * Background scanning to trim preallocated CoW space. This is queued
+ * based on the 'speculative_cow_prealloc_lifetime' tunable (5m by default).
+ * (We'll just piggyback on the post-EOF prealloc space workqueue.)
+ */
+STATIC void
+xfs_queue_cowblocks(
+	struct xfs_mount *mp)
+{
+	rcu_read_lock();
+	if (radix_tree_tagged(&mp->m_perag_tree, XFS_ICI_COWBLOCKS_TAG))
+		queue_delayed_work(mp->m_eofblocks_workqueue,
+				   &mp->m_cowblocks_work,
+				   msecs_to_jiffies(xfs_cowb_secs * 1000));
+	rcu_read_unlock();
+}
+
+void
+xfs_cowblocks_worker(
+	struct work_struct *work)
+{
+	struct xfs_mount *mp = container_of(to_delayed_work(work),
+				struct xfs_mount, m_cowblocks_work);
+	xfs_icache_free_cowblocks(mp, NULL);
+	xfs_queue_cowblocks(mp);
+}
+
 int
 xfs_inode_ag_iterator(
 	struct xfs_mount	*mp,
@@ -1343,18 +1376,30 @@ xfs_inode_free_eofblocks(
 	return ret;
 }
 
-int
-xfs_icache_free_eofblocks(
+static int
+__xfs_icache_free_eofblocks(
 	struct xfs_mount	*mp,
-	struct xfs_eofblocks	*eofb)
+	struct xfs_eofblocks	*eofb,
+	int			(*execute)(struct xfs_inode *ip, int flags,
+					   void *args),
+	int			tag)
 {
 	int flags = SYNC_TRYLOCK;
 
 	if (eofb && (eofb->eof_flags & XFS_EOF_FLAGS_SYNC))
 		flags = SYNC_WAIT;
 
-	return xfs_inode_ag_iterator_tag(mp, xfs_inode_free_eofblocks, flags,
-					 eofb, XFS_ICI_EOFBLOCKS_TAG);
+	return xfs_inode_ag_iterator_tag(mp, execute, flags,
+					 eofb, tag);
+}
+
+int
+xfs_icache_free_eofblocks(
+	struct xfs_mount	*mp,
+	struct xfs_eofblocks	*eofb)
+{
+	return __xfs_icache_free_eofblocks(mp, eofb, xfs_inode_free_eofblocks,
+			XFS_ICI_EOFBLOCKS_TAG);
 }
 
 /*
@@ -1363,9 +1408,11 @@ xfs_icache_free_eofblocks(
  * failure. We make a best effort by including each quota under low free space
  * conditions (less than 1% free space) in the scan.
  */
-int
-xfs_inode_free_quota_eofblocks(
-	struct xfs_inode *ip)
+static int
+__xfs_inode_free_quota_eofblocks(
+	struct xfs_inode	*ip,
+	int			(*execute)(struct xfs_mount *mp,
+					   struct xfs_eofblocks	*eofb))
 {
 	int scan = 0;
 	struct xfs_eofblocks eofb = {0};
@@ -1401,14 +1448,25 @@ xfs_inode_free_quota_eofblocks(
 	}
 
 	if (scan)
-		xfs_icache_free_eofblocks(ip->i_mount, &eofb);
+		execute(ip->i_mount, &eofb);
 
 	return scan;
 }
 
-void
-xfs_inode_set_eofblocks_tag(
-	xfs_inode_t	*ip)
+int
+xfs_inode_free_quota_eofblocks(
+	struct xfs_inode *ip)
+{
+	return __xfs_inode_free_quota_eofblocks(ip, xfs_icache_free_eofblocks);
+}
+
+static void
+__xfs_inode_set_eofblocks_tag(
+	xfs_inode_t	*ip,
+	void		(*execute)(struct xfs_mount *mp),
+	void		(*set_tp)(struct xfs_mount *mp, xfs_agnumber_t agno,
+				  int error, unsigned long caller_ip),
+	int		tag)
 {
 	struct xfs_mount *mp = ip->i_mount;
 	struct xfs_perag *pag;
@@ -1426,26 +1484,22 @@ xfs_inode_set_eofblocks_tag(
 
 	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
 	spin_lock(&pag->pag_ici_lock);
-	trace_xfs_inode_set_eofblocks_tag(ip);
 
-	tagged = radix_tree_tagged(&pag->pag_ici_root,
-				   XFS_ICI_EOFBLOCKS_TAG);
+	tagged = radix_tree_tagged(&pag->pag_ici_root, tag);
 	radix_tree_tag_set(&pag->pag_ici_root,
-			   XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino),
-			   XFS_ICI_EOFBLOCKS_TAG);
+			   XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino), tag);
 	if (!tagged) {
 		/* propagate the eofblocks tag up into the perag radix tree */
 		spin_lock(&ip->i_mount->m_perag_lock);
 		radix_tree_tag_set(&ip->i_mount->m_perag_tree,
 				   XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino),
-				   XFS_ICI_EOFBLOCKS_TAG);
+				   tag);
 		spin_unlock(&ip->i_mount->m_perag_lock);
 
 		/* kick off background trimming */
-		xfs_queue_eofblocks(ip->i_mount);
+		execute(ip->i_mount);
 
-		trace_xfs_perag_set_eofblocks(ip->i_mount, pag->pag_agno,
-					      -1, _RET_IP_);
+		set_tp(ip->i_mount, pag->pag_agno, -1, _RET_IP_);
 	}
 
 	spin_unlock(&pag->pag_ici_lock);
@@ -1453,9 +1507,22 @@ xfs_inode_set_eofblocks_tag(
 }
 
 void
-xfs_inode_clear_eofblocks_tag(
+xfs_inode_set_eofblocks_tag(
 	xfs_inode_t	*ip)
 {
+	trace_xfs_inode_set_eofblocks_tag(ip);
+	return __xfs_inode_set_eofblocks_tag(ip, xfs_queue_eofblocks,
+			trace_xfs_perag_set_eofblocks,
+			XFS_ICI_EOFBLOCKS_TAG);
+}
+
+static void
+__xfs_inode_clear_eofblocks_tag(
+	xfs_inode_t	*ip,
+	void		(*clear_tp)(struct xfs_mount *mp, xfs_agnumber_t agno,
+				    int error, unsigned long caller_ip),
+	int		tag)
+{
 	struct xfs_mount *mp = ip->i_mount;
 	struct xfs_perag *pag;
 
@@ -1465,23 +1532,141 @@ xfs_inode_clear_eofblocks_tag(
 
 	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino));
 	spin_lock(&pag->pag_ici_lock);
-	trace_xfs_inode_clear_eofblocks_tag(ip);
 
 	radix_tree_tag_clear(&pag->pag_ici_root,
-			     XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino),
-			     XFS_ICI_EOFBLOCKS_TAG);
-	if (!radix_tree_tagged(&pag->pag_ici_root, XFS_ICI_EOFBLOCKS_TAG)) {
+			     XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino), tag);
+	if (!radix_tree_tagged(&pag->pag_ici_root, tag)) {
 		/* clear the eofblocks tag from the perag radix tree */
 		spin_lock(&ip->i_mount->m_perag_lock);
 		radix_tree_tag_clear(&ip->i_mount->m_perag_tree,
 				     XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino),
-				     XFS_ICI_EOFBLOCKS_TAG);
+				     tag);
 		spin_unlock(&ip->i_mount->m_perag_lock);
-		trace_xfs_perag_clear_eofblocks(ip->i_mount, pag->pag_agno,
-					       -1, _RET_IP_);
+		clear_tp(ip->i_mount, pag->pag_agno, -1, _RET_IP_);
 	}
 
 	spin_unlock(&pag->pag_ici_lock);
 	xfs_perag_put(pag);
 }
 
+void
+xfs_inode_clear_eofblocks_tag(
+	xfs_inode_t	*ip)
+{
+	trace_xfs_inode_clear_eofblocks_tag(ip);
+	return __xfs_inode_clear_eofblocks_tag(ip,
+			trace_xfs_perag_clear_eofblocks, XFS_ICI_EOFBLOCKS_TAG);
+}
+
+/*
+ * Automatic CoW Reservation Freeing
+ *
+ * These functions automatically garbage collect leftover CoW reservations
+ * that were made on behalf of a cowextsize hint when we start to run out
+ * of quota or when the reservations sit around for too long.  If the file
+ * has dirty pages or is undergoing writeback, its CoW reservations will
+ * be retained.
+ *
+ * The actual garbage collection piggybacks off the same code that runs
+ * the speculative EOF preallocation garbage collector.
+ */
+STATIC int
+xfs_inode_free_cowblocks(
+	struct xfs_inode	*ip,
+	int			flags,
+	void			*args)
+{
+	int ret;
+	struct xfs_eofblocks *eofb = args;
+	bool need_iolock = true;
+	int match;
+
+	ASSERT(!eofb || (eofb && eofb->eof_scan_owner != 0));
+
+	if (!xfs_reflink_has_real_cow_blocks(ip)) {
+		trace_xfs_inode_free_cowblocks_invalid(ip);
+		xfs_inode_clear_cowblocks_tag(ip);
+		return 0;
+	}
+
+	/*
+	 * If the mapping is dirty or under writeback we cannot touch the
+	 * CoW fork.  Leave it alone if we're in the midst of a directio.
+	 */
+	if (mapping_tagged(VFS_I(ip)->i_mapping, PAGECACHE_TAG_DIRTY) ||
+	    mapping_tagged(VFS_I(ip)->i_mapping, PAGECACHE_TAG_WRITEBACK) ||
+	    atomic_read(&VFS_I(ip)->i_dio_count))
+		return 0;
+
+	if (eofb) {
+		if (eofb->eof_flags & XFS_EOF_FLAGS_UNION)
+			match = xfs_inode_match_id_union(ip, eofb);
+		else
+			match = xfs_inode_match_id(ip, eofb);
+		if (!match)
+			return 0;
+
+		/* skip the inode if the file size is too small */
+		if (eofb->eof_flags & XFS_EOF_FLAGS_MINFILESIZE &&
+		    XFS_ISIZE(ip) < eofb->eof_min_file_size)
+			return 0;
+
+		/*
+		 * A scan owner implies we already hold the iolock. Skip it in
+		 * xfs_free_eofblocks() to avoid deadlock. This also eliminates
+		 * the possibility of EAGAIN being returned.
+		 */
+		if (eofb->eof_scan_owner == ip->i_ino)
+			need_iolock = false;
+	}
+
+	/* Free the CoW blocks */
+	if (need_iolock) {
+		xfs_ilock(ip, XFS_IOLOCK_EXCL);
+		xfs_ilock(ip, XFS_MMAPLOCK_EXCL);
+	}
+
+	ret = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF);
+
+	if (need_iolock) {
+		xfs_iunlock(ip, XFS_MMAPLOCK_EXCL);
+		xfs_iunlock(ip, XFS_IOLOCK_EXCL);
+	}
+
+	return ret;
+}
+
+int
+xfs_icache_free_cowblocks(
+	struct xfs_mount	*mp,
+	struct xfs_eofblocks	*eofb)
+{
+	return __xfs_icache_free_eofblocks(mp, eofb, xfs_inode_free_cowblocks,
+			XFS_ICI_COWBLOCKS_TAG);
+}
+
+int
+xfs_inode_free_quota_cowblocks(
+	struct xfs_inode *ip)
+{
+	return __xfs_inode_free_quota_eofblocks(ip, xfs_icache_free_cowblocks);
+}
+
+void
+xfs_inode_set_cowblocks_tag(
+	xfs_inode_t	*ip)
+{
+	trace_xfs_inode_set_eofblocks_tag(ip);
+	return __xfs_inode_set_eofblocks_tag(ip, xfs_queue_cowblocks,
+			trace_xfs_perag_set_eofblocks,
+			XFS_ICI_COWBLOCKS_TAG);
+}
+
+void
+xfs_inode_clear_cowblocks_tag(
+	xfs_inode_t	*ip)
+{
+	trace_xfs_inode_clear_eofblocks_tag(ip);
+	return __xfs_inode_clear_eofblocks_tag(ip,
+			trace_xfs_perag_clear_eofblocks, XFS_ICI_COWBLOCKS_TAG);
+}