Module Name: src
Committed By: riz
Date: Wed Nov 4 17:55:23 UTC 2015
Modified Files:
src/sys/kern [netbsd-7]: vfs_mount.c
Log Message:
Pull up following revision(s) (requested by manu in ticket #962):
sys/kern/vfs_mount.c: revision 1.36
Do not VFS_SYNC before VFS_UNMOUNT on force unmount
VFS_SYNC does not consider whether we are performing a force unmount or not,
and therefore it can wait for a while if the filesytstem is misbehaving.
Removing VFS_SYNC before VFS_UNMOUNT on forced unmount fixes the problem.
This should not cause harm as the VFS_SYNC seems just useless.
As noted by Chuck Silvers in
http://mail-index.netbsd.org/tech-kern/2015/07/13/msg019156.html
- Nothing seems to prevent vnodes from getting dirty again after VFS_SYNC call.
- Filesystems do flush data through vflush() in VFS_UNMOUNT anyway.
As a consequence, the VFS_SYNC call in do_unmount() could probably be
completely removed. But since such a change is quite dangerous, we just
remove it in the case of forced unmounts, which are situations where
the risk of data loss is known to the operator.
To generate a diff of this commit:
cvs rdiff -u -r1.30.2.2 -r1.30.2.3 src/sys/kern/vfs_mount.c
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/kern/vfs_mount.c
diff -u src/sys/kern/vfs_mount.c:1.30.2.2 src/sys/kern/vfs_mount.c:1.30.2.3
--- src/sys/kern/vfs_mount.c:1.30.2.2 Fri Jan 9 14:50:35 2015
+++ src/sys/kern/vfs_mount.c Wed Nov 4 17:55:23 2015
@@ -1,4 +1,4 @@
-/* $NetBSD: vfs_mount.c,v 1.30.2.2 2015/01/09 14:50:35 martin Exp $ */
+/* $NetBSD: vfs_mount.c,v 1.30.2.3 2015/11/04 17:55:23 riz Exp $ */
/*-
* Copyright (c) 1997-2011 The NetBSD Foundation, Inc.
@@ -67,7 +67,7 @@
*/
#include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: vfs_mount.c,v 1.30.2.2 2015/01/09 14:50:35 martin Exp $");
+__KERNEL_RCSID(0, "$NetBSD: vfs_mount.c,v 1.30.2.3 2015/11/04 17:55:23 riz Exp $");
#define _VFS_VNODE_PRIVATE
@@ -835,7 +835,7 @@ dounmount(struct mount *mp, int flags, s
if (mp->mnt_syncer != NULL)
vfs_deallocate_syncvnode(mp);
error = 0;
- if ((mp->mnt_flag & MNT_RDONLY) == 0) {
+ if (((mp->mnt_flag & MNT_RDONLY) == 0) && ((flags & MNT_FORCE) == 0)) {
error = VFS_SYNC(mp, MNT_WAIT, l->l_cred);
}
if (error == 0 || (flags & MNT_FORCE)) {
