1 files changed, 90 insertions, 318 deletions

diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index 0e55c5d7db5f..7394c7af5de5 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -1136,7 +1136,7 @@ xfs_inactive(
 	 * If the inode is already free, then there can be nothing
 	 * to clean up here.
 	 */
-	if (ip->i_d.di_mode == 0 || VN_BAD(VFS_I(ip))) {
+	if (ip->i_d.di_mode == 0 || is_bad_inode(VFS_I(ip))) {
 		ASSERT(ip->i_df.if_real_bytes == 0);
 		ASSERT(ip->i_df.if_broot_bytes == 0);
 		return VN_INACTIVE_CACHE;
@@ -1387,23 +1387,28 @@ xfs_create(
 	xfs_inode_t		**ipp,
 	cred_t			*credp)
 {
-	xfs_mount_t		*mp = dp->i_mount;
-	xfs_inode_t		*ip;
-	xfs_trans_t		*tp;
+	int			is_dir = S_ISDIR(mode);
+	struct xfs_mount	*mp = dp->i_mount;
+	struct xfs_inode	*ip = NULL;
+	struct xfs_trans	*tp = NULL;
 	int			error;
 	xfs_bmap_free_t		free_list;
 	xfs_fsblock_t		first_block;
 	boolean_t		unlock_dp_on_error = B_FALSE;
-	int			dm_event_sent = 0;
 	uint			cancel_flags;
 	int			committed;
 	xfs_prid_t		prid;
-	struct xfs_dquot	*udqp, *gdqp;
+	struct xfs_dquot	*udqp = NULL;
+	struct xfs_dquot	*gdqp = NULL;
 	uint			resblks;
+	uint			log_res;
+	uint			log_count;
 
-	ASSERT(!*ipp);
 	xfs_itrace_entry(dp);
 
+	if (XFS_FORCED_SHUTDOWN(mp))
+		return XFS_ERROR(EIO);
+
 	if (DM_EVENT_ENABLED(dp, DM_EVENT_CREATE)) {
 		error = XFS_SEND_NAMESP(mp, DM_EVENT_CREATE,
 				dp, DM_RIGHT_NULL, NULL,
@@ -1412,84 +1417,97 @@ xfs_create(
 
 		if (error)
 			return error;
-		dm_event_sent = 1;
 	}
 
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return XFS_ERROR(EIO);
-
-	/* Return through std_return after this point. */
-
-	udqp = gdqp = NULL;
 	if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
 		prid = dp->i_d.di_projid;
 	else
-		prid = (xfs_prid_t)dfltprid;
+		prid = dfltprid;
 
 	/*
 	 * Make sure that we have allocated dquot(s) on disk.
 	 */
 	error = XFS_QM_DQVOPALLOC(mp, dp,
 			current_fsuid(), current_fsgid(), prid,
-			XFS_QMOPT_QUOTALL|XFS_QMOPT_INHERIT, &udqp, &gdqp);
+			XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp);
 	if (error)
 		goto std_return;
 
-	ip = NULL;
+	if (is_dir) {
+		rdev = 0;
+		resblks = XFS_MKDIR_SPACE_RES(mp, name->len);
+		log_res = XFS_MKDIR_LOG_RES(mp);
+		log_count = XFS_MKDIR_LOG_COUNT;
+		tp = xfs_trans_alloc(mp, XFS_TRANS_MKDIR);
+	} else {
+		resblks = XFS_CREATE_SPACE_RES(mp, name->len);
+		log_res = XFS_CREATE_LOG_RES(mp);
+		log_count = XFS_CREATE_LOG_COUNT;
+		tp = xfs_trans_alloc(mp, XFS_TRANS_CREATE);
+	}
 
-	tp = xfs_trans_alloc(mp, XFS_TRANS_CREATE);
 	cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
-	resblks = XFS_CREATE_SPACE_RES(mp, name->len);
+
 	/*
 	 * Initially assume that the file does not exist and
 	 * reserve the resources for that case.  If that is not
 	 * the case we'll drop the one we have and get a more
 	 * appropriate transaction later.
 	 */
-	error = xfs_trans_reserve(tp, resblks, XFS_CREATE_LOG_RES(mp), 0,
-			XFS_TRANS_PERM_LOG_RES, XFS_CREATE_LOG_COUNT);
+	error = xfs_trans_reserve(tp, resblks, log_res, 0,
+			XFS_TRANS_PERM_LOG_RES, log_count);
 	if (error == ENOSPC) {
 		resblks = 0;
-		error = xfs_trans_reserve(tp, 0, XFS_CREATE_LOG_RES(mp), 0,
-				XFS_TRANS_PERM_LOG_RES, XFS_CREATE_LOG_COUNT);
+		error = xfs_trans_reserve(tp, 0, log_res, 0,
+				XFS_TRANS_PERM_LOG_RES, log_count);
 	}
 	if (error) {
 		cancel_flags = 0;
-		goto error_return;
+		goto out_trans_cancel;
 	}
 
 	xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
 	unlock_dp_on_error = B_TRUE;
 
-	xfs_bmap_init(&free_list, &first_block);
+	/*
+	 * Check for directory link count overflow.
+	 */
+	if (is_dir && dp->i_d.di_nlink >= XFS_MAXLINK) {
+		error = XFS_ERROR(EMLINK);
+		goto out_trans_cancel;
+	}
 
-	ASSERT(ip == NULL);
+	xfs_bmap_init(&free_list, &first_block);
 
 	/*
 	 * Reserve disk quota and the inode.
 	 */
 	error = XFS_TRANS_RESERVE_QUOTA(mp, tp, udqp, gdqp, resblks, 1, 0);
 	if (error)
-		goto error_return;
+		goto out_trans_cancel;
 
 	error = xfs_dir_canenter(tp, dp, name, resblks);
 	if (error)
-		goto error_return;
-	error = xfs_dir_ialloc(&tp, dp, mode, 1,
-			rdev, credp, prid, resblks > 0,
-			&ip, &committed);
+		goto out_trans_cancel;
+
+	/*
+	 * A newly created regular or special file just has one directory
+	 * entry pointing to them, but a directory also the "." entry
+	 * pointing to itself.
+	 */
+	error = xfs_dir_ialloc(&tp, dp, mode, is_dir ? 2 : 1, rdev, credp,
+			       prid, resblks > 0, &ip, &committed);
 	if (error) {
 		if (error == ENOSPC)
-			goto error_return;
-		goto abort_return;
+			goto out_trans_cancel;
+		goto out_trans_abort;
 	}
-	xfs_itrace_ref(ip);
 
 	/*
 	 * At this point, we've gotten a newly allocated inode.
 	 * It is locked (and joined to the transaction).
 	 */
-
+	xfs_itrace_ref(ip);
 	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
 
 	/*
@@ -1508,19 +1526,28 @@ xfs_create(
 					resblks - XFS_IALLOC_SPACE_RES(mp) : 0);
 	if (error) {
 		ASSERT(error != ENOSPC);
-		goto abort_return;
+		goto out_trans_abort;
 	}
 	xfs_ichgtime(dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
 	xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
 
+	if (is_dir) {
+		error = xfs_dir_init(tp, ip, dp);
+		if (error)
+			goto out_bmap_cancel;
+
+		error = xfs_bumplink(tp, dp);
+		if (error)
+			goto out_bmap_cancel;
+	}
+
 	/*
 	 * If this is a synchronous mount, make sure that the
 	 * create transaction goes to disk before returning to
 	 * the user.
 	 */
-	if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
+	if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC))
 		xfs_trans_set_sync(tp);
-	}
 
 	/*
 	 * Attach the dquot(s) to the inodes and modify them incore.
@@ -1537,16 +1564,13 @@ xfs_create(
 	IHOLD(ip);
 
 	error = xfs_bmap_finish(&tp, &free_list, &committed);
-	if (error) {
-		xfs_bmap_cancel(&free_list);
-		goto abort_rele;
-	}
+	if (error)
+		goto out_abort_rele;
 
 	error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
 	if (error) {
 		IRELE(ip);
-		tp = NULL;
-		goto error_return;
+		goto out_dqrele;
 	}
 
 	XFS_QM_DQRELE(mp, udqp);
@@ -1555,26 +1579,22 @@ xfs_create(
 	*ipp = ip;
 
 	/* Fallthrough to std_return with error = 0  */
-
-std_return:
-	if ((*ipp || (error != 0 && dm_event_sent != 0)) &&
-	    DM_EVENT_ENABLED(dp, DM_EVENT_POSTCREATE)) {
-		(void) XFS_SEND_NAMESP(mp, DM_EVENT_POSTCREATE,
-			dp, DM_RIGHT_NULL,
-			*ipp ? ip : NULL,
-			DM_RIGHT_NULL, name->name, NULL,
-			mode, error, 0);
+ std_return:
+	if (DM_EVENT_ENABLED(dp, DM_EVENT_POSTCREATE)) {
+		XFS_SEND_NAMESP(mp, DM_EVENT_POSTCREATE, dp, DM_RIGHT_NULL,
+				ip, DM_RIGHT_NULL, name->name, NULL, mode,
+				error, 0);
 	}
+
 	return error;
 
- abort_return:
+ out_bmap_cancel:
+	xfs_bmap_cancel(&free_list);
+ out_trans_abort:
 	cancel_flags |= XFS_TRANS_ABORT;
-	/* FALLTHROUGH */
-
- error_return:
-	if (tp != NULL)
-		xfs_trans_cancel(tp, cancel_flags);
-
+ out_trans_cancel:
+	xfs_trans_cancel(tp, cancel_flags);
+ out_dqrele:
 	XFS_QM_DQRELE(mp, udqp);
 	XFS_QM_DQRELE(mp, gdqp);
 
@@ -1583,20 +1603,18 @@ std_return:
 
 	goto std_return;
 
- abort_rele:
+ out_abort_rele:
 	/*
 	 * Wait until after the current transaction is aborted to
 	 * release the inode.  This prevents recursive transactions
 	 * and deadlocks from xfs_inactive.
 	 */
+	xfs_bmap_cancel(&free_list);
 	cancel_flags |= XFS_TRANS_ABORT;
 	xfs_trans_cancel(tp, cancel_flags);
 	IRELE(ip);
-
-	XFS_QM_DQRELE(mp, udqp);
-	XFS_QM_DQRELE(mp, gdqp);
-
-	goto std_return;
+	unlock_dp_on_error = B_FALSE;
+	goto out_dqrele;
 }
 
 #ifdef DEBUG
@@ -2004,8 +2022,10 @@ xfs_link(
 	/* Return through std_return after this point. */
 
 	error = XFS_QM_DQATTACH(mp, sip, 0);
-	if (!error && sip != tdp)
-		error = XFS_QM_DQATTACH(mp, tdp, 0);
+	if (error)
+		goto std_return;
+
+	error = XFS_QM_DQATTACH(mp, tdp, 0);
 	if (error)
 		goto std_return;
 
@@ -2110,209 +2130,6 @@ std_return:
 	goto std_return;
 }
 
-
-int
-xfs_mkdir(
-	xfs_inode_t             *dp,
-	struct xfs_name		*dir_name,
-	mode_t			mode,
-	xfs_inode_t		**ipp,
-	cred_t			*credp)
-{
-	xfs_mount_t		*mp = dp->i_mount;
-	xfs_inode_t		*cdp;	/* inode of created dir */
-	xfs_trans_t		*tp;
-	int			cancel_flags;
-	int			error;
-	int			committed;
-	xfs_bmap_free_t         free_list;
-	xfs_fsblock_t           first_block;
-	boolean_t		unlock_dp_on_error = B_FALSE;
-	boolean_t		created = B_FALSE;
-	int			dm_event_sent = 0;
-	xfs_prid_t		prid;
-	struct xfs_dquot	*udqp, *gdqp;
-	uint			resblks;
-
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return XFS_ERROR(EIO);
-
-	tp = NULL;
-
-	if (DM_EVENT_ENABLED(dp, DM_EVENT_CREATE)) {
-		error = XFS_SEND_NAMESP(mp, DM_EVENT_CREATE,
-					dp, DM_RIGHT_NULL, NULL,
-					DM_RIGHT_NULL, dir_name->name, NULL,
-					mode, 0, 0);
-		if (error)
-			return error;
-		dm_event_sent = 1;
-	}
-
-	/* Return through std_return after this point. */
-
-	xfs_itrace_entry(dp);
-
-	mp = dp->i_mount;
-	udqp = gdqp = NULL;
-	if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
-		prid = dp->i_d.di_projid;
-	else
-		prid = (xfs_prid_t)dfltprid;
-
-	/*
-	 * Make sure that we have allocated dquot(s) on disk.
-	 */
-	error = XFS_QM_DQVOPALLOC(mp, dp,
-			current_fsuid(), current_fsgid(), prid,
-			XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp);
-	if (error)
-		goto std_return;
-
-	tp = xfs_trans_alloc(mp, XFS_TRANS_MKDIR);
-	cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
-	resblks = XFS_MKDIR_SPACE_RES(mp, dir_name->len);
-	error = xfs_trans_reserve(tp, resblks, XFS_MKDIR_LOG_RES(mp), 0,
-				  XFS_TRANS_PERM_LOG_RES, XFS_MKDIR_LOG_COUNT);
-	if (error == ENOSPC) {
-		resblks = 0;
-		error = xfs_trans_reserve(tp, 0, XFS_MKDIR_LOG_RES(mp), 0,
-					  XFS_TRANS_PERM_LOG_RES,
-					  XFS_MKDIR_LOG_COUNT);
-	}
-	if (error) {
-		cancel_flags = 0;
-		goto error_return;
-	}
-
-	xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
-	unlock_dp_on_error = B_TRUE;
-
-	/*
-	 * Check for directory link count overflow.
-	 */
-	if (dp->i_d.di_nlink >= XFS_MAXLINK) {
-		error = XFS_ERROR(EMLINK);
-		goto error_return;
-	}
-
-	/*
-	 * Reserve disk quota and the inode.
-	 */
-	error = XFS_TRANS_RESERVE_QUOTA(mp, tp, udqp, gdqp, resblks, 1, 0);
-	if (error)
-		goto error_return;
-
-	error = xfs_dir_canenter(tp, dp, dir_name, resblks);
-	if (error)
-		goto error_return;
-	/*
-	 * create the directory inode.
-	 */
-	error = xfs_dir_ialloc(&tp, dp, mode, 2,
-			0, credp, prid, resblks > 0,
-		&cdp, NULL);
-	if (error) {
-		if (error == ENOSPC)
-			goto error_return;
-		goto abort_return;
-	}
-	xfs_itrace_ref(cdp);
-
-	/*
-	 * Now we add the directory inode to the transaction.
-	 * We waited until now since xfs_dir_ialloc might start
-	 * a new transaction.  Had we joined the transaction
-	 * earlier, the locks might have gotten released. An error
-	 * from here on will result in the transaction cancel
-	 * unlocking dp so don't do it explicitly in the error path.
-	 */
-	IHOLD(dp);
-	xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
-	unlock_dp_on_error = B_FALSE;
-
-	xfs_bmap_init(&free_list, &first_block);
-
-	error = xfs_dir_createname(tp, dp, dir_name, cdp->i_ino,
-					&first_block, &free_list, resblks ?
-					resblks - XFS_IALLOC_SPACE_RES(mp) : 0);
-	if (error) {
-		ASSERT(error != ENOSPC);
-		goto error1;
-	}
-	xfs_ichgtime(dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
-
-	error = xfs_dir_init(tp, cdp, dp);
-	if (error)
-		goto error2;
-
-	error = xfs_bumplink(tp, dp);
-	if (error)
-		goto error2;
-
-	created = B_TRUE;
-
-	*ipp = cdp;
-	IHOLD(cdp);
-
-	/*
-	 * Attach the dquots to the new inode and modify the icount incore.
-	 */
-	XFS_QM_DQVOPCREATE(mp, tp, cdp, udqp, gdqp);
-
-	/*
-	 * If this is a synchronous mount, make sure that the
-	 * mkdir transaction goes to disk before returning to
-	 * the user.
-	 */
-	if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
-		xfs_trans_set_sync(tp);
-	}
-
-	error = xfs_bmap_finish(&tp, &free_list, &committed);
-	if (error) {
-		IRELE(cdp);
-		goto error2;
-	}
-
-	error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
-	XFS_QM_DQRELE(mp, udqp);
-	XFS_QM_DQRELE(mp, gdqp);
-	if (error) {
-		IRELE(cdp);
-	}
-
-	/* Fall through to std_return with error = 0 or errno from
-	 * xfs_trans_commit. */
-
-std_return:
-	if ((created || (error != 0 && dm_event_sent != 0)) &&
-	    DM_EVENT_ENABLED(dp, DM_EVENT_POSTCREATE)) {
-		(void) XFS_SEND_NAMESP(mp, DM_EVENT_POSTCREATE,
-					dp, DM_RIGHT_NULL,
-					created ? cdp : NULL,
-					DM_RIGHT_NULL,
-					dir_name->name, NULL,
-					mode, error, 0);
-	}
-	return error;
-
- error2:
- error1:
-	xfs_bmap_cancel(&free_list);
- abort_return:
-	cancel_flags |= XFS_TRANS_ABORT;
- error_return:
-	xfs_trans_cancel(tp, cancel_flags);
-	XFS_QM_DQRELE(mp, udqp);
-	XFS_QM_DQRELE(mp, gdqp);
-
-	if (unlock_dp_on_error)
-		xfs_iunlock(dp, XFS_ILOCK_EXCL);
-
-	goto std_return;
-}
-
 int
 xfs_symlink(
 	xfs_inode_t		*dp,
@@ -2587,51 +2404,6 @@ std_return:
 }
 
 int
-xfs_inode_flush(
-	xfs_inode_t	*ip,
-	int		flags)
-{
-	xfs_mount_t	*mp = ip->i_mount;
-	int		error = 0;
-
-	if (XFS_FORCED_SHUTDOWN(mp))
-		return XFS_ERROR(EIO);
-
-	/*
-	 * Bypass inodes which have already been cleaned by
-	 * the inode flush clustering code inside xfs_iflush
-	 */
-	if (xfs_inode_clean(ip))
-		return 0;
-
-	/*
-	 * We make this non-blocking if the inode is contended,
-	 * return EAGAIN to indicate to the caller that they
-	 * did not succeed. This prevents the flush path from
-	 * blocking on inodes inside another operation right
-	 * now, they get caught later by xfs_sync.
-	 */
-	if (flags & FLUSH_SYNC) {
-		xfs_ilock(ip, XFS_ILOCK_SHARED);
-		xfs_iflock(ip);
-	} else if (xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) {
-		if (xfs_ipincount(ip) || !xfs_iflock_nowait(ip)) {
-			xfs_iunlock(ip, XFS_ILOCK_SHARED);
-			return EAGAIN;
-		}
-	} else {
-		return EAGAIN;
-	}
-
-	error = xfs_iflush(ip, (flags & FLUSH_SYNC) ? XFS_IFLUSH_SYNC
-						    : XFS_IFLUSH_ASYNC_NOBLOCK);
-	xfs_iunlock(ip, XFS_ILOCK_SHARED);
-
-	return error;
-}
-
-
-int
 xfs_set_dmattrs(
 	xfs_inode_t     *ip,
 	u_int		evmask,
@@ -2676,7 +2448,7 @@ xfs_reclaim(
 	ASSERT(!VN_MAPPED(VFS_I(ip)));
 
 	/* bad inode, get out here ASAP */
-	if (VN_BAD(VFS_I(ip))) {
+	if (is_bad_inode(VFS_I(ip))) {
 		xfs_ireclaim(ip);
 		return 0;
 	}
@@ -3090,7 +2862,7 @@ xfs_free_file_space(
 
 	/*
 	 * Need to zero the stuff we're not freeing, on disk.
-	 * If its a realtime file & can't use unwritten extents then we
+	 * If it's a realtime file & can't use unwritten extents then we
 	 * actually need to zero the extent edges.  Otherwise xfs_bunmapi
 	 * will take care of it for us.
 	 */