When printing nothing for the diagnostic on PASS/FAIL, use a string format
with an empty string, not an empty format with a NULL parameter.  Clang
complains about the empty format string, which in turn breaks building
with -Werror.

../kselftest_harness.h:1205:30: error: format string is empty [-Werror,-Wformat-zero-length]
                              diagnostic ? "%s" : "", diagnostic);
                                                  ^~
Fixes: 378193eff339 ("selftests: kselftest_harness: let PASS / FAIL provide diagnostic")
Cc: Kees Cook <keescook@chromium.org>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sean Christopherson <seanjc@google.com>
---
 tools/testing/selftests/kselftest_harness.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/tools/testing/selftests/kselftest_harness.h b/tools/testing/selftests/kselftest_harness.h
index 4fd735e48ee7..79ac9e9ada33 100644
--- a/tools/testing/selftests/kselftest_harness.h
+++ b/tools/testing/selftests/kselftest_harness.h
@@ -1197,12 +1197,11 @@ void __run_test(struct __fixture_metadata *f,
 	if (t->results->reason[0])
 		diagnostic = t->results->reason;
 	else if (t->exit_code == KSFT_PASS || t->exit_code == KSFT_FAIL)
-		diagnostic = NULL;
+		diagnostic = "";
 	else
 		diagnostic = "unknown";
 
-	ksft_test_result_code(t->exit_code, test_name,
-			      diagnostic ? "%s" : "", diagnostic);
+	ksft_test_result_code(t->exit_code, test_name, "%s", diagnostic);
 }
 
 static int test_harness_run(int argc, char **argv)

base-commit: 2c71fdf02a95b3dd425b42f28fd47fb2b1d22702
-- 
2.44.0.478.gd926399ef9-goog

On Tue,  9 Apr 2024 15:42:56 -0700 Sean Christopherson wrote:
> -	ksft_test_result_code(t->exit_code, test_name,
> -			      diagnostic ? "%s" : "", diagnostic);
> +	ksft_test_result_code(t->exit_code, test_name, "%s", diagnostic);

Have you tested that to make sure it doesn't change the output?

.. warning: ^^ leading question ;)

On Tue, Apr 09, 2024, Jakub Kicinski wrote:
> On Tue,  9 Apr 2024 15:42:56 -0700 Sean Christopherson wrote:
> > -	ksft_test_result_code(t->exit_code, test_name,
> > -			      diagnostic ? "%s" : "", diagnostic);
> > +	ksft_test_result_code(t->exit_code, test_name, "%s", diagnostic);
> 
> Have you tested that to make sure it doesn't change the output?
> 
> .. warning: ^^ leading question ;)

Heh, I was *this* close to adding a blurb saying this was probably only compile
tested.

On Tue, Apr 09, 2024 at 04:09:20PM -0700, Jakub Kicinski wrote:
> On Tue,  9 Apr 2024 15:42:56 -0700 Sean Christopherson wrote:
> > -	ksft_test_result_code(t->exit_code, test_name,
> > -			      diagnostic ? "%s" : "", diagnostic);
> > +	ksft_test_result_code(t->exit_code, test_name, "%s", diagnostic);
> 
> Have you tested that to make sure it doesn't change the output?
> 
> .. warning: ^^ leading question ;)

Probably should be just this, without changing the NULL init?

	ksft_test_result_code(t->exit_code, test_name, "%s",
			      diagnostic ?: "");

-- 
Kees Cook

On Tue, 9 Apr 2024 16:46:40 -0700 Kees Cook wrote:
> 	ksft_test_result_code(t->exit_code, test_name, "%s",
> 			      diagnostic ?: "");

Could work, but we need to change the callee to extract the first arg
after format from va_arg, which I wasn't sufficiently familiar with :(