[src/trunk]: src/tests/fs/vfs Apparently this way of triggering the msdosfs r...

[pooka <pooka%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/5b76d8362256
branches:  trunk
changeset: 763239:5b76d8362256
user:      pooka <pooka%NetBSD.org@localhost>
date:      Mon Mar 14 19:05:19 2011 +0000

description:
Apparently this way of triggering the msdosfs rename vnode leak
does not bite every time (most commonly observed on the amd64/qemu
runs), so add a race condition catcher.

diffstat:

 tests/fs/vfs/t_renamerace.c |  14 ++++++++++++--
 1 files changed, 12 insertions(+), 2 deletions(-)

diffs (29 lines):

diff -r 53b58977eefa -r 5b76d8362256 tests/fs/vfs/t_renamerace.c
--- a/tests/fs/vfs/t_renamerace.c       Mon Mar 14 17:18:12 2011 +0000
+++ b/tests/fs/vfs/t_renamerace.c       Mon Mar 14 19:05:19 2011 +0000
@@ -1,4 +1,4 @@
-/*     $NetBSD: t_renamerace.c,v 1.21 2011/03/06 16:00:16 pooka Exp $  */
+/*     $NetBSD: t_renamerace.c,v 1.22 2011/03/14 19:05:19 pooka Exp $  */
 
 /*
  * Modified for rump and atf from a program supplied
@@ -113,8 +113,18 @@
        if (FSTYPE_LFS(tc))
                abort();
 
-       if (FSTYPE_MSDOS(tc))
+       if (FSTYPE_MSDOS(tc)) {
                atf_tc_expect_fail("PR kern/44661");
+               /*
+                * XXX: race does not trigger every time at least
+                * on amd64/qemu.
+                */
+               if (msdosfs_fstest_unmount(tc, mp, 0) != 0) {
+                       rump_pub_vfs_mount_print(mp, 1);
+                       atf_tc_fail_errno("unmount failed");
+               }
+               atf_tc_fail("race did not trigger this time");
+       }
 }
 
 static void