Fix inability to destroy snapshot used over NFS

The cache of struct svc_export and struct svc_expkey by nfsd and
rpc.mountd for the snapshot holds references to the mount point.
We need to flush them out before unmounting, otherwise umount
would fail with EBUSY.

Reviewed-by: Don Brady <[email protected]>
Reviewed-by: Brian Behlendorf <[email protected]>
Signed-off-by: Youzhong Yang <[email protected]>
Closes #6000 
Closes #10783 

1 files changed, 19 insertions, 0 deletions

diff --git a/module/os/linux/zfs/zfs_ctldir.c b/module/os/linux/zfs/zfs_ctldir.c
index c2748ce45..26e785a0d 100644
--- a/module/os/linux/zfs/zfs_ctldir.c
+++ b/module/os/linux/zfs/zfs_ctldir.c
@@ -31,6 +31,7 @@
  * Copyright 2015, OmniTI Computer Consulting, Inc. All rights reserved.
  * Copyright (c) 2018 George Melikov. All Rights Reserved.
  * Copyright (c) 2019 Datto, Inc. All rights reserved.
+ * Copyright (c) 2020 The MathWorks, Inc. All rights reserved.
  */
 
 /*
@@ -978,6 +979,22 @@ out:
 }
 
 /*
+ * Flush everything out of the kernel's export table and such.
+ * This is needed as once the snapshot is used over NFS, its
+ * entries in svc_export and svc_expkey caches hold reference
+ * to the snapshot mount point. There is no known way of flushing
+ * only the entries related to the snapshot.
+ */
+static void
+exportfs_flush(void)
+{
+	char *argv[] = { "/usr/sbin/exportfs", "-f", NULL };
+	char *envp[] = { NULL };
+
+	(void) call_usermodehelper(argv[0], argv, envp, UMH_WAIT_PROC);
+}
+
+/*
  * Attempt to unmount a snapshot by making a call to user space.
  * There is no assurance that this can or will succeed, is just a
  * best effort.  In the case where it does fail, perhaps because
@@ -999,6 +1016,8 @@ zfsctl_snapshot_unmount(char *snapname, int flags)
 	}
 	rw_exit(&zfs_snapshot_lock);
 
+	exportfs_flush();
+
 	if (flags & MNT_FORCE)
 		argv[4] = "-fn";
 	argv[5] = se->se_path;