On Tue, Oct 22, 2024 at 05:16:31PM -0400, Rae Moar wrote:
diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c index 34b7b6833df3..340723571b0f 100644 --- a/lib/kunit/executor.c +++ b/lib/kunit/executor.c @@ -29,6 +29,22 @@ const char *kunit_action(void) return action_param; }
+/*
- Run KUnit tests after initialization
- */
+#ifdef CONFIG_KUNIT_AUTORUN_ENABLED +static bool autorun_param = true; +#else +static bool autorun_param; +#endif +module_param_named(autorun, autorun_param, bool, 0); +MODULE_PARM_DESC(autorun, "Run KUnit tests after initialization");
+bool kunit_autorun(void) +{
return autorun_param;+}
static char *filter_glob_param; static char *filter_param; static char *filter_action_param; @@ -266,7 +282,7 @@ void kunit_exec_run_tests(struct kunit_suite_set *suite_set, bool builtin) pr_info("1..%zu\n", num_suites);
When using this feature, I still see some KTAP output that are printed from this function (kunit_exec_run_tests). I think it would be great if we could remove this output as to not clutter the kernel log.
At first, I was confused as to why we needed to call this function and initialize the tests but I realized the debugfs suites need to be created.
So instead, could we check for kunit_autorun() here instead as a condition before printing the output?
Sure, I'll address it in the next revision.
Thanks, Stanislav