[PATCH] selftests: drivers/s390x: Use SKIP() during FIXTURE_SETUP

[PATCH] selftests: drivers/s390x: Use SKIP() during FIXTURE_SETUP

[Kees Cook]

Instead of mixing selftest harness and ksft helpers, perform SKIP
testing from the FIXTURE_SETUPs. This also means TEST_HARNESS_MAIN does
not need to be open-coded.

Signed-off-by: Kees Cook <keescook@chromium.org>
---
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: Janosch Frank <frankja@linux.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: kvm@vger.kernel.org
Cc: linux-kselftest@vger.kernel.org
---
I wasn't able to build or run test this, since it's very s390 specific, it seems?
---
 .../drivers/s390x/uvdevice/test_uvdevice.c    | 32 ++++++++-----------
 1 file changed, 13 insertions(+), 19 deletions(-)

diff --git a/tools/testing/selftests/drivers/s390x/uvdevice/test_uvdevice.c b/tools/testing/selftests/drivers/s390x/uvdevice/test_uvdevice.c
index ea0cdc37b44f..9622fac42597 100644
--- a/tools/testing/selftests/drivers/s390x/uvdevice/test_uvdevice.c
+++ b/tools/testing/selftests/drivers/s390x/uvdevice/test_uvdevice.c
@@ -18,6 +18,16 @@
 
 #define UV_PATH  "/dev/uv"
 #define BUFFER_SIZE 0x200
+
+#define open_uv() \
+do { \
+	self->uv_fd = open(UV_PATH, O_ACCMODE);	\
+	if (self->uv_fd < 0) \
+		SKIP(return, "No uv-device or cannot access " UV_PATH  "\n" \
+			     "Enable CONFIG_S390_UV_UAPI and check the access rights on " \
+			     UV_PATH ".\n"); \
+} while (0)
+
 FIXTURE(uvio_fixture) {
 	int uv_fd;
 	struct uvio_ioctl_cb uvio_ioctl;
@@ -37,7 +47,7 @@ FIXTURE_VARIANT_ADD(uvio_fixture, att) {
 
 FIXTURE_SETUP(uvio_fixture)
 {
-	self->uv_fd = open(UV_PATH, O_ACCMODE);
+	open_uv();
 
 	self->uvio_ioctl.argument_addr = (__u64)self->buffer;
 	self->uvio_ioctl.argument_len = variant->arg_size;
@@ -159,7 +169,7 @@ FIXTURE(attest_fixture) {
 
 FIXTURE_SETUP(attest_fixture)
 {
-	self->uv_fd = open(UV_PATH, O_ACCMODE);
+	open_uv();
 
 	self->uvio_ioctl.argument_addr = (__u64)&self->uvio_attest;
 	self->uvio_ioctl.argument_len = sizeof(self->uvio_attest);
@@ -257,20 +267,4 @@ TEST_F(attest_fixture, att_inval_addr)
 	att_inval_addr_test(&self->uvio_attest.meas_addr, _metadata, self);
 }
 
-static void __attribute__((constructor)) __constructor_order_last(void)
-{
-	if (!__constructor_order)
-		__constructor_order = _CONSTRUCTOR_ORDER_BACKWARD;
-}
-
-int main(int argc, char **argv)
-{
-	int fd = open(UV_PATH, O_ACCMODE);
-
-	if (fd < 0)
-		ksft_exit_skip("No uv-device or cannot access " UV_PATH  "\n"
-			       "Enable CONFIG_S390_UV_UAPI and check the access rights on "
-			       UV_PATH ".\n");
-	close(fd);
-	return test_harness_run(argc, argv);
-}
+TEST_HARNESS_MAIN
-- 
2.34.1

Re: [PATCH] selftests: drivers/s390x: Use SKIP() during FIXTURE_SETUP

[Janosch Frank]

On 5/18/24 02:18, Kees Cook wrote:
> Instead of mixing selftest harness and ksft helpers, perform SKIP
> testing from the FIXTURE_SETUPs. This also means TEST_HARNESS_MAIN does
> not need to be open-coded.
> 
> Signed-off-by: Kees Cook <keescook@chromium.org>
> ---

Adding the author in CC.


This changes the skip behavior from one single skip to a skip per test.
Not an issue, but a change.

But also we're generating invalid TAP AFAIK which we also have been 
before but we can fix that in this patch.

 From what I understand this line should have a "#" prefix to show that 
it's a comment:
Enable CONFIG_S390_UV_UAPI and check the access rights on /dev/uv.