Re: [PATCH 3/5] xfs: check for cow blocks before trying to clear them during inode reclaim
From: Brian Foster <hidden>
Date: 2018-02-27 13:55:50
On Thu, Feb 22, 2018 at 06:00:00PM -0800, Darrick J. Wong wrote:
From: Darrick J. Wong <redacted> There's no point in allocating a transaction and locking the inode in preparation to clear cow blocks if there actually are any cow fork extents. Therefore, move the xfs_reflink_cancel_cow_range hunk to xfs_inactive and check the cow ifp first. This makes inode reclamation run faster. Signed-off-by: Darrick J. Wong <redacted> ---
Reviewed-by: Brian Foster <redacted>
quoted hunk ↗ jump to hunk
fs/xfs/xfs_inode.c | 5 +++++ fs/xfs/xfs_super.c | 9 --------- 2 files changed, 5 insertions(+), 9 deletions(-)diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 604ee38..50fbbf5 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c@@ -1872,6 +1872,7 @@ xfs_inactive( xfs_inode_t *ip) { struct xfs_mount *mp; + struct xfs_ifork *cow_ifp = XFS_IFORK_PTR(ip, XFS_COW_FORK); int error; int truncate = 0;@@ -1892,6 +1893,10 @@ xfs_inactive( if (mp->m_flags & XFS_MOUNT_RDONLY) return; + /* Try to clean out the cow blocks if there are any. */ + if (xfs_is_reflink_inode(ip) && cow_ifp->if_bytes > 0) + xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, true); + if (VFS_I(ip)->i_nlink != 0) { /* * force is true because we are evicting an inode from thediff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 7aba628..624a802 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c@@ -970,7 +970,6 @@ xfs_fs_destroy_inode( struct inode *inode) { struct xfs_inode *ip = XFS_I(inode); - int error; trace_xfs_destroy_inode(ip);@@ -978,14 +977,6 @@ xfs_fs_destroy_inode( XFS_STATS_INC(ip->i_mount, vn_rele); XFS_STATS_INC(ip->i_mount, vn_remove); - if (xfs_is_reflink_inode(ip)) { - error = xfs_reflink_cancel_cow_range(ip, 0, NULLFILEOFF, true); - if (error && !XFS_FORCED_SHUTDOWN(ip->i_mount)) - xfs_warn(ip->i_mount, -"Error %d while evicting CoW blocks for inode %llu.", - error, ip->i_ino); - } - xfs_inactive(ip); ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0); --To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html