diff mbox series

[2/7] syscalls/fsopen: New tests

Message ID e4955c8edfb29c1d0f74a8cd24b1dcc68f3b6814.1581680021.git.viresh.kumar@linaro.org
State Superseded
Headers show
Series Add new LTP tests related to fsmount family of syscalls | expand

Commit Message

Viresh Kumar Feb. 14, 2020, 11:35 a.m. UTC
Add tests to check working of fsopen() syscall.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 configure.ac                                |  1 +
 runtest/syscalls                            |  3 +
 testcases/kernel/syscalls/fsopen/.gitignore |  2 +
 testcases/kernel/syscalls/fsopen/Makefile   |  6 ++
 testcases/kernel/syscalls/fsopen/fsopen01.c | 71 +++++++++++++++++++++
 testcases/kernel/syscalls/fsopen/fsopen02.c | 56 ++++++++++++++++
 6 files changed, 139 insertions(+)
 create mode 100644 testcases/kernel/syscalls/fsopen/.gitignore
 create mode 100644 testcases/kernel/syscalls/fsopen/Makefile
 create mode 100644 testcases/kernel/syscalls/fsopen/fsopen01.c
 create mode 100644 testcases/kernel/syscalls/fsopen/fsopen02.c

Comments

Li Wang Feb. 16, 2020, 10:11 a.m. UTC | #1
On Fri, Feb 14, 2020 at 7:36 PM Viresh Kumar <viresh.kumar@linaro.org>
wrote:

> Add tests to check working of fsopen() syscall.
>
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
>  configure.ac                                |  1 +
>  runtest/syscalls                            |  3 +
>  testcases/kernel/syscalls/fsopen/.gitignore |  2 +
>  testcases/kernel/syscalls/fsopen/Makefile   |  6 ++
>  testcases/kernel/syscalls/fsopen/fsopen01.c | 71 +++++++++++++++++++++
>  testcases/kernel/syscalls/fsopen/fsopen02.c | 56 ++++++++++++++++
>  6 files changed, 139 insertions(+)
>  create mode 100644 testcases/kernel/syscalls/fsopen/.gitignore
>  create mode 100644 testcases/kernel/syscalls/fsopen/Makefile
>  create mode 100644 testcases/kernel/syscalls/fsopen/fsopen01.c
>  create mode 100644 testcases/kernel/syscalls/fsopen/fsopen02.c
>
> diff --git a/configure.ac b/configure.ac
> index df4e8c8322fc..4125160a19bb 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -80,6 +80,7 @@ AC_CHECK_FUNCS([ \
>      execveat \
>      fallocate \
>      fchownat \
> +    fsopen \
>      fstatat \
>      getdents \
>      getdents64 \
> diff --git a/runtest/syscalls b/runtest/syscalls
> index 0743cf4e3f74..0a56599ebad9 100644
> --- a/runtest/syscalls
> +++ b/runtest/syscalls
> @@ -542,6 +542,9 @@ fanotify13 fanotify13
>  fanotify14 fanotify14
>  fanotify15 fanotify15
>
> +fsopen01 fsopen01
> +fsopen02 fsopen02
> +
>  ioperm01 ioperm01
>  ioperm02 ioperm02
>
> diff --git a/testcases/kernel/syscalls/fsopen/.gitignore
> b/testcases/kernel/syscalls/fsopen/.gitignore
> new file mode 100644
> index 000000000000..80089dd137a7
> --- /dev/null
> +++ b/testcases/kernel/syscalls/fsopen/.gitignore
> @@ -0,0 +1,2 @@
> +fsopen01
> +fsopen02
> diff --git a/testcases/kernel/syscalls/fsopen/Makefile
> b/testcases/kernel/syscalls/fsopen/Makefile
> new file mode 100644
> index 000000000000..5ea7d67db123
> --- /dev/null
> +++ b/testcases/kernel/syscalls/fsopen/Makefile
> @@ -0,0 +1,6 @@
> +# SPDX-License-Identifier: GPL-2.0-or-later
> +
> +top_srcdir             ?= ../../../..
> +
> +include $(top_srcdir)/include/mk/testcases.mk
> +include $(top_srcdir)/include/mk/generic_leaf_target.mk
> diff --git a/testcases/kernel/syscalls/fsopen/fsopen01.c
> b/testcases/kernel/syscalls/fsopen/fsopen01.c
> new file mode 100644
> index 000000000000..3587c67636ea
> --- /dev/null
> +++ b/testcases/kernel/syscalls/fsopen/fsopen01.c
> @@ -0,0 +1,71 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (c) 2020 Viresh Kumar <viresh.kumar@linaro.org>
> + *
> + * Description:
> + * Basic fsopen() test which tries to configure and mount the filesystem
> as
> + * well.
> + */
> +#include "tst_test.h"
> +#include "lapi/fsmount.h"
> +
> +#define MNTPOINT       "mntpoint"
> +
> +static void run(void)
> +{
> +       int fd, fsmfd;
> +
> +       TEST(fsopen(tst_device->fs_type, 0));
> +       fd = TST_RET;
> +
> +       if (fd == -1)
> +               tst_brk(TFAIL | TERRNO, "fsopen() failed");
> +
> +       TEST(fsconfig(fd, FSCONFIG_SET_STRING, "source", tst_device->dev,
> 0));
> +       if (TST_RET == -1) {
> +               tst_res(TBROK | TERRNO, "fsconfig() failed");
> +               goto out;
> +       }
> +
> +       TEST(fsconfig(fd, FSCONFIG_CMD_CREATE, NULL, NULL, 0));
> +       if (TST_RET == -1) {
> +               tst_res(TBROK | TERRNO, "fsconfig() failed");
> +               goto out;
> +       }
> +
> +       TEST(fsmount(fd, 0, 0));
> +       if (TST_RET == -1) {
> +               tst_res(TBROK | TERRNO, "fsmount() failed");
> +               goto out;
> +       }
> +
> +       fsmfd = TST_RET;
> +
> +       TEST(move_mount(fsmfd, "", AT_FDCWD, MNTPOINT,
> +                       MOVE_MOUNT_F_EMPTY_PATH));
> +
> +       SAFE_CLOSE(fsmfd);
> +
> +       if (TST_RET == -1) {
> +               tst_res(TBROK | TERRNO, "move_mount() failed");
> +               goto out;
> +       }
> +
> +       SAFE_CLOSE(TST_RET);
> +
> +       TEST(umount(MNTPOINT));
> +
> +       tst_res(TPASS, "fsopen() passed");
> +
> +out:
> +       SAFE_CLOSE(fd);
> +}
> +
> +static struct tst_test test = {
> +       .min_kver = "5.2",
>

I suggest removing .min_kver check in all of the tests to let they can be
running on many distributions(which backport the features).

+       .test_all = run,
> +       .needs_root = 1,
> +       .needs_tmpdir = 1,
>

.needs_tmpdir is not necessary because tst_test->mntpoint helps create that
too.


> +       .format_device = 1,
> +       .mntpoint = MNTPOINT,
>

Maybe to test more filesystems and skip on FUSE? Just like the way in zlang@'s
patch.

+       .all_filesystems = 1,
+       .dev_fs_flags = TST_FS_SKIP_FUSE,

+};
> diff --git a/testcases/kernel/syscalls/fsopen/fsopen02.c
> b/testcases/kernel/syscalls/fsopen/fsopen02.c
> new file mode 100644
> index 000000000000..6831a12c696b
> --- /dev/null
> +++ b/testcases/kernel/syscalls/fsopen/fsopen02.c
> @@ -0,0 +1,56 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (c) 2020 Viresh Kumar <viresh.kumar@linaro.org>
> + *
> + * Description:
> + * Basic fsopen() failure tests.
> + */
> +#include "tst_test.h"
> +#include "lapi/fsmount.h"
> +
> +const char *invalid_fs = "invalid";
> +const char *valid_fs;
> +
> +static struct tcase {
> +       char *name;
> +       const char **fs;
> +       unsigned int flags;
> +       int exp_errno;
> +} tcases[] = {
> +       {"invalid-fs", &invalid_fs, 0, ENODEV},
> +       {"invalid-flags", &valid_fs, 0x10, EINVAL},
> +};
> +
> +static void setup(void)
> +{
> +       valid_fs = tst_device->fs_type;
> +}
> +
> +static void run(unsigned int n)
> +{
> +       struct tcase *tc = &tcases[n];
> +
> +       TEST(fsopen(*tc->fs, tc->flags));
> +
> +       if (TST_RET != -1) {
> +               SAFE_CLOSE(TST_RET);
> +               tst_brk(TFAIL, "%s: fsopen() succeeded unexpectedly
> (index: %d)",
> +                       tc->name, n);
> +       }
> +
> +       if (tc->exp_errno != TST_ERR) {
> +               tst_brk(TFAIL | TTERRNO, "%s: fsopen() should fail with
> %s",
> +                       tc->name, tst_strerrno(tc->exp_errno));
> +       }
> +
> +       tst_res(TPASS | TTERRNO, "%s: fsopen() failed as expected",
> tc->name);
> +}
> +
> +static struct tst_test test = {
> +       .min_kver = "5.2",
>

To delete .min_kver.


> +       .tcnt = ARRAY_SIZE(tcases),
> +       .test = run,
> +       .setup = setup,
> +       .needs_root = 1,
> +       .needs_device = 1,
> +};
> --
> 2.21.0.rc0.269.g1a574e7a288b
>
>
> --
> Mailing list info: https://lists.linux.it/listinfo/ltp
>
>
Viresh Kumar Feb. 17, 2020, 8:09 a.m. UTC | #2
On 16-02-20, 18:11, Li Wang wrote:
> > +static struct tst_test test = {
> > +       .min_kver = "5.2",
> >
> 
> I suggest removing .min_kver check in all of the tests to let they can be
> running on many distributions(which backport the features).

I never thought about that earlier, good point. All other comments look fine as
well. Thanks.
Cyril Hrubis Feb. 17, 2020, 1:36 p.m. UTC | #3
Hi!
> > +       SAFE_CLOSE(TST_RET);
> > +
> > +       TEST(umount(MNTPOINT));
> > +
> > +       tst_res(TPASS, "fsopen() passed");
> > +
> > +out:
> > +       SAFE_CLOSE(fd);
> > +}
> > +
> > +static struct tst_test test = {
> > +       .min_kver = "5.2",
> >
> 
> I suggest removing .min_kver check in all of the tests to let they can be
> running on many distributions(which backport the features).

If we do that we have to explicitely check for ENOSYS errno in each
test, quite possibly with a dummy call to the tested syscall in test
setup, because once these calls gets libc wrappers we will no longer
call the tst_syscall() that checks for it.
Li Wang Feb. 18, 2020, 1:15 a.m. UTC | #4
On Mon, Feb 17, 2020 at 9:36 PM Cyril Hrubis <chrubis@suse.cz> wrote:

> ...
> > > +static struct tst_test test = {
> > > +       .min_kver = "5.2",
> > >
> >
> > I suggest removing .min_kver check in all of the tests to let they can be
> > running on many distributions(which backport the features).
>
> If we do that we have to explicitely check for ENOSYS errno in each
> test, quite possibly with a dummy call to the tested syscall in test
> setup, because once these calls gets libc wrappers we will no longer
> call the tst_syscall() that checks for it.
>

+1 add dummy call to the tested syscall in the setup.
Agree, thanks for point out this.
Viresh Kumar Feb. 18, 2020, 8:25 a.m. UTC | #5
On 18-02-20, 09:15, Li Wang wrote:
> On Mon, Feb 17, 2020 at 9:36 PM Cyril Hrubis <chrubis@suse.cz> wrote:
> 
> > ...
> > > > +static struct tst_test test = {
> > > > +       .min_kver = "5.2",
> > > >
> > >
> > > I suggest removing .min_kver check in all of the tests to let they can be
> > > running on many distributions(which backport the features).
> >
> > If we do that we have to explicitely check for ENOSYS errno in each
> > test, quite possibly with a dummy call to the tested syscall in test
> > setup, because once these calls gets libc wrappers we will no longer
> > call the tst_syscall() that checks for it.
> >
> 
> +1 add dummy call to the tested syscall in the setup.
> Agree, thanks for point out this.

Not sure if I understood it very clearly and don't want to spam everyone with an
incorrect patchset still missing this point and so asking here :)

How exactly would the setup() routine look like? Something like this ?

        syscall(#NR, ##__VA_ARGS__);
Li Wang Feb. 18, 2020, 9:02 a.m. UTC | #6
On Tue, Feb 18, 2020 at 4:25 PM Viresh Kumar <viresh.kumar@linaro.org>
wrote:

> On 18-02-20, 09:15, Li Wang wrote:
> > On Mon, Feb 17, 2020 at 9:36 PM Cyril Hrubis <chrubis@suse.cz> wrote:
> >
> > > ...
> > > > > +static struct tst_test test = {
> > > > > +       .min_kver = "5.2",
> > > > >
> > > >
> > > > I suggest removing .min_kver check in all of the tests to let they
> can be
> > > > running on many distributions(which backport the features).
> > >
> > > If we do that we have to explicitely check for ENOSYS errno in each
> > > test, quite possibly with a dummy call to the tested syscall in test
> > > setup, because once these calls gets libc wrappers we will no longer
> > > call the tst_syscall() that checks for it.
> > >
> >
> > +1 add dummy call to the tested syscall in the setup.
> > Agree, thanks for point out this.
>
> Not sure if I understood it very clearly and don't want to spam everyone
> with an
> incorrect patchset still missing this point and so asking here :)
>
> How exactly would the setup() routine look like? Something like this ?
>
>         syscall(#NR, ##__VA_ARGS__);
>

I suggest to use tst_syscall(...);.

A dummy call in setup() just helps check the ENOSYS errno, something like:

$ cat testcases/cve/cve-2016-7117.c |grep setup -A 10
static void setup(void)
{
        ...
        tst_syscall(__NR_recvmmsg, 0, 0, 0, 0, 0);
}
Viresh Kumar Feb. 18, 2020, 9:08 a.m. UTC | #7
On 18-02-20, 17:02, Li Wang wrote:
> On Tue, Feb 18, 2020 at 4:25 PM Viresh Kumar <viresh.kumar@linaro.org>
> wrote:
> 
> > On 18-02-20, 09:15, Li Wang wrote:
> > > On Mon, Feb 17, 2020 at 9:36 PM Cyril Hrubis <chrubis@suse.cz> wrote:
> > >
> > > > ...
> > > > > > +static struct tst_test test = {
> > > > > > +       .min_kver = "5.2",
> > > > > >
> > > > >
> > > > > I suggest removing .min_kver check in all of the tests to let they
> > can be
> > > > > running on many distributions(which backport the features).
> > > >
> > > > If we do that we have to explicitely check for ENOSYS errno in each
> > > > test, quite possibly with a dummy call to the tested syscall in test
> > > > setup, because once these calls gets libc wrappers we will no longer
> > > > call the tst_syscall() that checks for it.
> > > >
> > >
> > > +1 add dummy call to the tested syscall in the setup.
> > > Agree, thanks for point out this.
> >
> > Not sure if I understood it very clearly and don't want to spam everyone
> > with an
> > incorrect patchset still missing this point and so asking here :)
> >
> > How exactly would the setup() routine look like? Something like this ?
> >
> >         syscall(#NR, ##__VA_ARGS__);
> >
> 
> I suggest to use tst_syscall(...);.
> 
> A dummy call in setup() just helps check the ENOSYS errno, something like:
> 
> $ cat testcases/cve/cve-2016-7117.c |grep setup -A 10
> static void setup(void)
> {
>         ...
>         tst_syscall(__NR_recvmmsg, 0, 0, 0, 0, 0);
> }

Okay, thanks.
Petr Vorel Feb. 19, 2020, 8:23 a.m. UTC | #8
Hi Viresh,

as somebody (probably Li) pointed out fsopen/fsopen01.c is similar to fsmount/fsmount01.c,
so please during rebase drop it. BTW for this dropped test I'd use different
approach (close fd in cleanup function and use tst_brk(TFAIL instead of
tst_res(TBROK and goto):

static void cleanup(void)
{
	if (fd > 0)
		SAFE_CLOSE(fd);
}

...
	if (fd == -1)
		tst_brk(TFAIL | TERRNO, "fsopen() failed");

Also you added autotools check for functions (e.g. fsopen) during commits, but I
move them to the first (together with lapi header).

Kind regards,
Petr
Viresh Kumar Feb. 19, 2020, 8:34 a.m. UTC | #9
On 19-02-20, 09:23, Petr Vorel wrote:
> Hi Viresh,
> 
> as somebody (probably Li) pointed out fsopen/fsopen01.c is similar to fsmount/fsmount01.c,
> so please during rebase drop it. BTW for this dropped test I'd use different
> approach (close fd in cleanup function and use tst_brk(TFAIL instead of
> tst_res(TBROK and goto):
> 
> static void cleanup(void)
> {
> 	if (fd > 0)
> 		SAFE_CLOSE(fd);
> }
> 
> ...
> 	if (fd == -1)
> 		tst_brk(TFAIL | TERRNO, "fsopen() failed");

Lets review this comment after my V2 series is posted today. It may be better to
keep separate tests.

> Also you added autotools check for functions (e.g. fsopen) during commits, but I
> move them to the first (together with lapi header).

Yeah, I already saw that. Thanks for that, you did the right thing :)
Petr Vorel Feb. 19, 2020, 8:50 a.m. UTC | #10
Hi,

> On Mon, Feb 17, 2020 at 9:36 PM Cyril Hrubis <chrubis@suse.cz> wrote:

> > ...
> > > > +static struct tst_test test = {
> > > > +       .min_kver = "5.2",


> > > I suggest removing .min_kver check in all of the tests to let they can be
> > > running on many distributions(which backport the features).

> > If we do that we have to explicitely check for ENOSYS errno in each
> > test, quite possibly with a dummy call to the tested syscall in test
> > setup, because once these calls gets libc wrappers we will no longer
> > call the tst_syscall() that checks for it.


> +1 add dummy call to the tested syscall in the setup.
> Agree, thanks for point out this.
Could anybody add it to fsmount/fsmount01.c instead?


If anybody don't mind, I'll rename fsopen02.c to fsopen01.c,
remove .min_kver = "5.2" and replace tst_brk with tst_res + return, and merge it:

// SPDX-License-Identifier: GPL-2.0-or-later
/*
 * Copyright (c) 2020 Viresh Kumar <viresh.kumar@linaro.org>
 *
 * Description:
 * Basic fsopen() failure tests.
 */

#include "tst_test.h"
#include "lapi/fsmount.h"

const char *invalid_fs = "invalid";
const char *valid_fs;

static struct tcase {
	char *name;
	const char **fs;
	unsigned int flags;
	int exp_errno;
} tcases[] = {
	{"invalid-fs", &invalid_fs, 0, ENODEV},
	{"invalid-flags", &valid_fs, 0x10, EINVAL},
};

static void setup(void)
{
	valid_fs = tst_device->fs_type;
}

static void run(unsigned int n)
{
	struct tcase *tc = &tcases[n];

	TEST(fsopen(*tc->fs, tc->flags));

	if (TST_RET != -1) {
		SAFE_CLOSE(TST_RET);
		tst_res(TFAIL, "%s: fsopen() succeeded unexpectedly (index: %d)",
			tc->name, n);
		return;
	}

	if (tc->exp_errno != TST_ERR) {
		tst_res(TFAIL | TTERRNO, "%s: fsopen() should fail with %s",
			tc->name, tst_strerrno(tc->exp_errno));
		return;
	}

	tst_res(TPASS | TTERRNO, "%s: fsopen() failed as expected", tc->name);
}

static struct tst_test test = {
	.tcnt = ARRAY_SIZE(tcases),
	.test = run,
	.setup = setup,
	.needs_root = 1,
	.needs_device = 1,
};
Viresh Kumar Feb. 19, 2020, 8:51 a.m. UTC | #11
On 19-02-20, 09:50, Petr Vorel wrote:
> Hi,
> 
> > On Mon, Feb 17, 2020 at 9:36 PM Cyril Hrubis <chrubis@suse.cz> wrote:
> 
> > > ...
> > > > > +static struct tst_test test = {
> > > > > +       .min_kver = "5.2",
> 
> 
> > > > I suggest removing .min_kver check in all of the tests to let they can be
> > > > running on many distributions(which backport the features).
> 
> > > If we do that we have to explicitely check for ENOSYS errno in each
> > > test, quite possibly with a dummy call to the tested syscall in test
> > > setup, because once these calls gets libc wrappers we will no longer
> > > call the tst_syscall() that checks for it.
> 
> 
> > +1 add dummy call to the tested syscall in the setup.
> > Agree, thanks for point out this.
> Could anybody add it to fsmount/fsmount01.c instead?

I am doing it.
diff mbox series

Patch

diff --git a/configure.ac b/configure.ac
index df4e8c8322fc..4125160a19bb 100644
--- a/configure.ac
+++ b/configure.ac
@@ -80,6 +80,7 @@  AC_CHECK_FUNCS([ \
     execveat \
     fallocate \
     fchownat \
+    fsopen \
     fstatat \
     getdents \
     getdents64 \
diff --git a/runtest/syscalls b/runtest/syscalls
index 0743cf4e3f74..0a56599ebad9 100644
--- a/runtest/syscalls
+++ b/runtest/syscalls
@@ -542,6 +542,9 @@  fanotify13 fanotify13
 fanotify14 fanotify14
 fanotify15 fanotify15
 
+fsopen01 fsopen01
+fsopen02 fsopen02
+
 ioperm01 ioperm01
 ioperm02 ioperm02
 
diff --git a/testcases/kernel/syscalls/fsopen/.gitignore b/testcases/kernel/syscalls/fsopen/.gitignore
new file mode 100644
index 000000000000..80089dd137a7
--- /dev/null
+++ b/testcases/kernel/syscalls/fsopen/.gitignore
@@ -0,0 +1,2 @@ 
+fsopen01
+fsopen02
diff --git a/testcases/kernel/syscalls/fsopen/Makefile b/testcases/kernel/syscalls/fsopen/Makefile
new file mode 100644
index 000000000000..5ea7d67db123
--- /dev/null
+++ b/testcases/kernel/syscalls/fsopen/Makefile
@@ -0,0 +1,6 @@ 
+# SPDX-License-Identifier: GPL-2.0-or-later
+
+top_srcdir		?= ../../../..
+
+include $(top_srcdir)/include/mk/testcases.mk
+include $(top_srcdir)/include/mk/generic_leaf_target.mk
diff --git a/testcases/kernel/syscalls/fsopen/fsopen01.c b/testcases/kernel/syscalls/fsopen/fsopen01.c
new file mode 100644
index 000000000000..3587c67636ea
--- /dev/null
+++ b/testcases/kernel/syscalls/fsopen/fsopen01.c
@@ -0,0 +1,71 @@ 
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2020 Viresh Kumar <viresh.kumar@linaro.org>
+ *
+ * Description:
+ * Basic fsopen() test which tries to configure and mount the filesystem as
+ * well.
+ */
+#include "tst_test.h"
+#include "lapi/fsmount.h"
+
+#define MNTPOINT	"mntpoint"
+
+static void run(void)
+{
+	int fd, fsmfd;
+
+	TEST(fsopen(tst_device->fs_type, 0));
+	fd = TST_RET;
+
+	if (fd == -1)
+		tst_brk(TFAIL | TERRNO, "fsopen() failed");
+
+	TEST(fsconfig(fd, FSCONFIG_SET_STRING, "source", tst_device->dev, 0));
+	if (TST_RET == -1) {
+		tst_res(TBROK | TERRNO, "fsconfig() failed");
+		goto out;
+	}
+
+	TEST(fsconfig(fd, FSCONFIG_CMD_CREATE, NULL, NULL, 0));
+	if (TST_RET == -1) {
+		tst_res(TBROK | TERRNO, "fsconfig() failed");
+		goto out;
+	}
+
+	TEST(fsmount(fd, 0, 0));
+	if (TST_RET == -1) {
+		tst_res(TBROK | TERRNO, "fsmount() failed");
+		goto out;
+	}
+
+	fsmfd = TST_RET;
+
+	TEST(move_mount(fsmfd, "", AT_FDCWD, MNTPOINT,
+			MOVE_MOUNT_F_EMPTY_PATH));
+
+	SAFE_CLOSE(fsmfd);
+
+	if (TST_RET == -1) {
+		tst_res(TBROK | TERRNO, "move_mount() failed");
+		goto out;
+	}
+
+	SAFE_CLOSE(TST_RET);
+
+	TEST(umount(MNTPOINT));
+
+	tst_res(TPASS, "fsopen() passed");
+
+out:
+	SAFE_CLOSE(fd);
+}
+
+static struct tst_test test = {
+	.min_kver = "5.2",
+	.test_all = run,
+	.needs_root = 1,
+	.needs_tmpdir = 1,
+	.format_device = 1,
+	.mntpoint = MNTPOINT,
+};
diff --git a/testcases/kernel/syscalls/fsopen/fsopen02.c b/testcases/kernel/syscalls/fsopen/fsopen02.c
new file mode 100644
index 000000000000..6831a12c696b
--- /dev/null
+++ b/testcases/kernel/syscalls/fsopen/fsopen02.c
@@ -0,0 +1,56 @@ 
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2020 Viresh Kumar <viresh.kumar@linaro.org>
+ *
+ * Description:
+ * Basic fsopen() failure tests.
+ */
+#include "tst_test.h"
+#include "lapi/fsmount.h"
+
+const char *invalid_fs = "invalid";
+const char *valid_fs;
+
+static struct tcase {
+	char *name;
+	const char **fs;
+	unsigned int flags;
+	int exp_errno;
+} tcases[] = {
+	{"invalid-fs", &invalid_fs, 0, ENODEV},
+	{"invalid-flags", &valid_fs, 0x10, EINVAL},
+};
+
+static void setup(void)
+{
+	valid_fs = tst_device->fs_type;
+}
+
+static void run(unsigned int n)
+{
+	struct tcase *tc = &tcases[n];
+
+	TEST(fsopen(*tc->fs, tc->flags));
+
+	if (TST_RET != -1) {
+		SAFE_CLOSE(TST_RET);
+		tst_brk(TFAIL, "%s: fsopen() succeeded unexpectedly (index: %d)",
+			tc->name, n);
+	}
+
+	if (tc->exp_errno != TST_ERR) {
+		tst_brk(TFAIL | TTERRNO, "%s: fsopen() should fail with %s",
+			tc->name, tst_strerrno(tc->exp_errno));
+	}
+
+	tst_res(TPASS | TTERRNO, "%s: fsopen() failed as expected", tc->name);
+}
+
+static struct tst_test test = {
+	.min_kver = "5.2",
+	.tcnt = ARRAY_SIZE(tcases),
+	.test = run,
+	.setup = setup,
+	.needs_root = 1,
+	.needs_device = 1,
+};