Re: [PATCH 1/2] kunit: Add ability to skip entire test suites

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] The newly added `suite->status` field is not reset in `kunit_init_suite()`, leading to a sticky skipped state across multiple debugfs runs.
- [Medium] Bypassing `suite_exit()` when `suite_init()` skips a test suite violates standard kernel initialization contracts and forces resource leaks.
--

commit b745c49805409babd4338da1946b7eb586892485
Author: Vaibhav Jain <vaibhav@linux.ibm.com>

kunit: Add ability to skip entire test suites

This commit adds a status field to struct kunit_suite to allow suite_init
callbacks to mark a suite as skipped. When a suite is marked as skipped,
all test cases within the suite are bypassed without execution.

> diff --git a/lib/kunit/test.c b/lib/kunit/test.c
> index 99773e000e1b7..989acc770265d 100644
> --- a/lib/kunit/test.c
> +++ b/lib/kunit/test.c

[ ... ]

> @@ -795,12 +798,20 @@ int kunit_run_tests(struct kunit_suite *suite)
>  	/* Taint the kernel so we know we've run tests. */
>  	add_taint(TAINT_TEST, LOCKDEP_STILL_OK);
>  
> +	if (suite->status == KUNIT_SKIPPED)
> +		goto suite_end;
> +

[Severity: Medium]
Does this cause the suite to remain permanently skipped on subsequent runs?

If a test suite is executed via debugfs and gets skipped, suite->status
becomes KUNIT_SKIPPED. Because kunit_init_suite() does not reset
suite->status back to KUNIT_SUCCESS, this check will immediately bypass
the suite on all subsequent debugfs runs, preventing re-evaluation
of runtime prerequisites.

>  	if (suite->suite_init) {
>  		suite->suite_init_err = suite->suite_init(suite);
>  		if (suite->suite_init_err) {
> +			suite->status = KUNIT_FAILURE;
>  			kunit_err(suite, KUNIT_SUBTEST_INDENT
>  				  "# failed to initialize (%d)", suite->suite_init_err);
>  			goto suite_end;
> +
> +		} else if (suite->status == KUNIT_SKIPPED) {
> +			/* Skip this kunit suite */
> +			goto suite_end;

[Severity: Medium]
Will this leak resources allocated during suite_init() prior to the skip?

Standard kernel initialization patterns guarantee that if an init function
returns 0 (success), the corresponding exit function will be called.

If suite_init() allocates resources, then sets suite->status to
KUNIT_SKIPPED and returns 0, the goto suite_end bypasses the
suite->suite_exit(suite) call.

Does this force test authors to manually free resources before returning 0
when skipping, or should the exit function still be called here?

>  		}
>  	}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260604162805.556135-1-vaibhav@linux.ibm.com?part=1