selftests/sched_ext: Fail interrupted test runs

SIGINT and SIGTERM set exit_req, causing the runner to stop before
executing the remaining tests. However, the exit status only reflects
the number of failed tests, so an interrupted run returns success when
none of the completed tests failed.

Include exit_req in the failure condition so callers can distinguish
an interrupted run from a successful run. Keep the result counts
limited to the tests that actually ran.

Tested the actual runner under GDB, delivering each signal before the
first test and after the example test passed. The original runner
returns 0 in all four cases; the fixed runner returns 1. Normal -h,
-l and -t example invocations still return 0.

Fixes: 9d851afa4826 ("selftests/sched_ext: Abort test loop on signal")
Assisted-by: LLM
Signed-off-by: Tianyi Chen <hi@tychen.cc>
Signed-off-by: Tejun Heo <tj@kernel.org>

1 files changed, 1 insertions, 1 deletions

diff --git a/tools/testing/selftests/sched_ext/runner.c b/tools/testing/selftests/sched_ext/runner.c
index c264807caa91..57dda75384f1 100644
--- a/tools/testing/selftests/sched_ext/runner.c
+++ b/tools/testing/selftests/sched_ext/runner.c
@@ -243,7 +243,7 @@ int main(int argc, char **argv)
 			printf("  - %s\n", failed_tests[i]);
 	}
 
-	return failed > 0 ? 1 : 0;
+	return failed > 0 || exit_req ? 1 : 0;
 }
 
 void scx_test_register(struct scx_test *test)