Skip to contentSkip to conversation

selftests/mm: fix incorrect skip output in pkey_sighandler_tests

PATCH Applied Zenghui Yu 2 participants 2 messages Aug 25

Subsystems (3)
Memory management Kselftest MM misc
lore
Share & tools
lore

Zenghui Yu

author
#0 lore raw

PATCH 1/1 selftests/mm: fix incorrect skip output in pkey_sighandler_tests

From: "Zenghui Yu (Huawei)" <[email protected]>

When pkeys is not supported, ksft_exit_skip() runs with ksft_plan already set, which takes the "ok N # SKIP" branch intended for skipping a single test case. The result is a TAP plan of 5 but only one result line.

 $ ./pkey_sighandler_tests
 TAP version 13
 1..5
 ok 1 # SKIP pkeys not supported
 # 1 skipped test(s) detected. Consider enabling relevant config options to improve coverage.
 # Planned tests != run tests (5 != 1)
 # Totals: pass:0 fail:0 xfail:0 xpass:0 skip:1 error:0

Move ksft_set_plan() after the skip check so ksft_exit_skip() takes the "1..0 # SKIP" branch, the correct TAP representation for skipping an entire test file.

 $ ./pkey_sighandler_tests
 TAP version 13
 1..0 # SKIP pkeys not supported
Signed-off-by Zenghui Yu (Huawei) <[email protected]>
---
 tools/testing/selftests/mm/pkey_sighandler_tests.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
tools/testing/selftests/mm/pkey_sighandler_tests.c +2 -1
index c218d0510a2a..74bf79a5399d 100644
@@ -543,11 +543,12 @@ static void (*pkey_tests[])(void) = {
 int main(int argc, char *argv[])
 {
 	ksft_print_header();
-	ksft_set_plan(ARRAY_SIZE(pkey_tests));
 
 	if (!is_pkeys_supported())
 		ksft_exit_skip("pkeys not supported\n");
 
+	ksft_set_plan(ARRAY_SIZE(pkey_tests));
+
 	for (test_nr = 0; test_nr < ARRAY_SIZE(pkey_tests); test_nr++) {
 		tracing_on();
 		(*pkey_tests[test_nr])();
Signature · 1 line
2.53.0
maintainer Acked-by on v1 1/1
#1 lore raw
Quoted 35 lines · Zenghui Yu

From: "Zenghui Yu (Huawei)" <[email protected]>

When pkeys is not supported, ksft_exit_skip() runs with ksft_plan already set, which takes the "ok N # SKIP" branch intended for skipping a single test case. The result is a TAP plan of 5 but only one result line.

 $ ./pkey_sighandler_tests
 TAP version 13
 1..5
 ok 1 # SKIP pkeys not supported
 # 1 skipped test(s) detected. Consider enabling relevant config options to improve coverage.
 # Planned tests != run tests (5 != 1)
 # Totals: pass:0 fail:0 xfail:0 xpass:0 skip:1 error:0

Move ksft_set_plan() after the skip check so ksft_exit_skip() takes the "1..0 # SKIP" branch, the correct TAP representation for skipping an entire test file.

 $ ./pkey_sighandler_tests
 TAP version 13
 1..0 # SKIP pkeys not supported
Signed-off-by Zenghui Yu (Huawei) <[email protected]>
---
 tools/testing/selftests/mm/pkey_sighandler_tests.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
tools/testing/selftests/mm/pkey_sighandler_tests.c +0 -0
index c218d0510a2a..74bf79a5399d 100644
@@ -543,11 +543,12 @@ static void (*pkey_tests[])(void) = {
 int main(int argc, char *argv[])
 {
 	ksft_print_header();

tools/testing/selftests/mm/pkey_sighandler_tests.c static void (*pkey_tests[])(void) = { · line 554

Quoted 10 lines
tools/testing/selftests/mm/pkey_sighandler_tests.c +2 -1
-	ksft_set_plan(ARRAY_SIZE(pkey_tests));
 
 	if (!is_pkeys_supported())
 		ksft_exit_skip("pkeys not supported\n");
 
+	ksft_set_plan(ARRAY_SIZE(pkey_tests));
+
 	for (test_nr = 0; test_nr < ARRAY_SIZE(pkey_tests); test_nr++) {
 		tracing_on();
 		(*pkey_tests[test_nr])();

We also handle it like that in soft-dirty.c, so why not

Acked-by David Hildenbrand (Arm) <[email protected]>
Signature · 3 lines
Cheers,

David