All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 0/4] kunit: Fix some bugs in kunit
@ 2023-09-14 11:46 Jinjie Ruan
  2023-09-14 11:46 ` [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set() Jinjie Ruan
                   ` (3 more replies)
  0 siblings, 4 replies; 13+ messages in thread
From: Jinjie Ruan @ 2023-09-14 11:46 UTC (permalink / raw)
  To: brendan.higgins, davidgow, skhan, dlatypov, rmoar,
	janusz.krzysztofik, linux-kselftest, kunit-dev
  Cc: ruanjinjie

The test_cases is not freed in kunit_free_suite_set().

And the copy pointer may be moved in kunit_filter_suites().

The filtered_suite and filtered_suite->test_cases allocated in the last
kunit_filter_attr_tests() in last inner for loop may be leaked if
kunit_filter_suites() fails.

If kunit_filter_suites() succeeds, not only copy but also filtered_suite
and filtered_suite->test_cases should be freed.

Jinjie Ruan (4):
  kunit: Fix missed memory release in kunit_free_suite_set()
  kunit: Fix the wrong kfree of copy for kunit_filter_suites()
  kunit: Fix possible memory leak in kunit_filter_suites()
  kunit: test: Fix the possible memory leak in executor_test

 lib/kunit/executor.c      | 23 +++++++++++++++++------
 lib/kunit/executor_test.c | 24 ++++++++++++++++++------
 2 files changed, 35 insertions(+), 12 deletions(-)

-- 
2.34.1


^ permalink raw reply	[flat|nested] 13+ messages in thread

* [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set()
  2023-09-14 11:46 [PATCH 0/4] kunit: Fix some bugs in kunit Jinjie Ruan
@ 2023-09-14 11:46 ` Jinjie Ruan
  2023-09-19 21:18   ` Rae Moar
  2023-09-14 11:46 ` [PATCH 2/4] kunit: Fix the wrong kfree of copy for kunit_filter_suites() Jinjie Ruan
                   ` (2 subsequent siblings)
  3 siblings, 1 reply; 13+ messages in thread
From: Jinjie Ruan @ 2023-09-14 11:46 UTC (permalink / raw)
  To: brendan.higgins, davidgow, skhan, dlatypov, rmoar,
	janusz.krzysztofik, linux-kselftest, kunit-dev
  Cc: ruanjinjie

modprobe cpumask_kunit and rmmod cpumask_kunit, kmemleak detect
a suspected memory leak as below.

If kunit_filter_suites() in kunit_module_init() succeeds, the
suite_set.start will not be NULL and the kunit_free_suite_set() in
kunit_module_exit() should free all the memory which has not
been freed. However the test_cases in suites is left out.

unreferenced object 0xffff54ac47e83200 (size 512):
  comm "modprobe", pid 592, jiffies 4294913238 (age 1367.612s)
  hex dump (first 32 bytes):
    84 13 1a f0 d3 b6 ff ff 30 68 1a f0 d3 b6 ff ff  ........0h......
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<000000008dec63a2>] slab_post_alloc_hook+0xb8/0x368
    [<00000000ec280d8e>] __kmem_cache_alloc_node+0x174/0x290
    [<00000000896c7740>] __kmalloc+0x60/0x2c0
    [<000000007a50fa06>] kunit_filter_suites+0x254/0x5b8
    [<0000000078cc98e2>] kunit_module_notify+0xf4/0x240
    [<0000000033cea952>] notifier_call_chain+0x98/0x17c
    [<00000000973d05cc>] notifier_call_chain_robust+0x4c/0xa4
    [<000000005f95895f>] blocking_notifier_call_chain_robust+0x4c/0x74
    [<0000000048e36fa7>] load_module+0x1a2c/0x1c40
    [<0000000004eb8a91>] init_module_from_file+0x94/0xcc
    [<0000000037dbba28>] idempotent_init_module+0x184/0x278
    [<00000000161b75cb>] __arm64_sys_finit_module+0x68/0xa8
    [<000000006dc1669b>] invoke_syscall+0x44/0x100
    [<00000000fa87e304>] el0_svc_common.constprop.1+0x68/0xe0
    [<000000009d8ad866>] do_el0_svc+0x1c/0x28
    [<000000005b83c607>] el0_svc+0x3c/0xc4

Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
Fixes: b67abaad4d25 ("kunit: Allow kunit test modules to use test filtering")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
---
 lib/kunit/executor.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
index a6348489d45f..a037a46fae5e 100644
--- a/lib/kunit/executor.c
+++ b/lib/kunit/executor.c
@@ -137,8 +137,10 @@ void kunit_free_suite_set(struct kunit_suite_set suite_set)
 {
 	struct kunit_suite * const *suites;
 
-	for (suites = suite_set.start; suites < suite_set.end; suites++)
+	for (suites = suite_set.start; suites < suite_set.end; suites++) {
+		kfree((*suites)->test_cases);
 		kfree(*suites);
+	}
 	kfree(suite_set.start);
 }
 
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 13+ messages in thread

* [PATCH 2/4] kunit: Fix the wrong kfree of copy for kunit_filter_suites()
  2023-09-14 11:46 [PATCH 0/4] kunit: Fix some bugs in kunit Jinjie Ruan
  2023-09-14 11:46 ` [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set() Jinjie Ruan
@ 2023-09-14 11:46 ` Jinjie Ruan
  2023-09-19 21:18   ` Rae Moar
  2023-09-14 11:46 ` [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites() Jinjie Ruan
  2023-09-14 11:46 ` [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test Jinjie Ruan
  3 siblings, 1 reply; 13+ messages in thread
From: Jinjie Ruan @ 2023-09-14 11:46 UTC (permalink / raw)
  To: brendan.higgins, davidgow, skhan, dlatypov, rmoar,
	janusz.krzysztofik, linux-kselftest, kunit-dev, Ruan Jinjie

If the outer layer for loop is iterated more than once and it fails not
in the first iteration, the copy pointer has been moved. So it should free
the original copy's backup copy_start.

Fixes: abbf73816b6f ("kunit: fix possible memory leak in kunit_filter_suites()")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
---
 lib/kunit/executor.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
index a037a46fae5e..9358ed2df839 100644
--- a/lib/kunit/executor.c
+++ b/lib/kunit/executor.c
@@ -243,7 +243,7 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
 
 free_copy:
 	if (*err)
-		kfree(copy);
+		kfree(copy_start);
 
 	return filtered;
 }
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 13+ messages in thread

* [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites()
  2023-09-14 11:46 [PATCH 0/4] kunit: Fix some bugs in kunit Jinjie Ruan
  2023-09-14 11:46 ` [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set() Jinjie Ruan
  2023-09-14 11:46 ` [PATCH 2/4] kunit: Fix the wrong kfree of copy for kunit_filter_suites() Jinjie Ruan
@ 2023-09-14 11:46 ` Jinjie Ruan
  2023-09-19 21:18   ` Rae Moar
  2023-09-14 11:46 ` [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test Jinjie Ruan
  3 siblings, 1 reply; 13+ messages in thread
From: Jinjie Ruan @ 2023-09-14 11:46 UTC (permalink / raw)
  To: brendan.higgins, davidgow, skhan, dlatypov, rmoar,
	janusz.krzysztofik, linux-kselftest, kunit-dev
  Cc: ruanjinjie

If the outer layer for loop is iterated more than once and it fails not
in the first iteration, the filtered_suite and filtered_suite->test_cases
allocated in the last kunit_filter_attr_tests() in last inner for loop
is leaked.

So add a new free_filtered_suite err label and free the filtered_suite
and filtered_suite->test_cases so far. And change kmalloc_array of copy
to kcalloc to Clear the copy to make the kfree safe.

Fixes: 5d31f71efcb6 ("kunit: add kunit.filter_glob cmdline option to filter suites")
Fixes: 529534e8cba3 ("kunit: Add ability to filter attributes")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
---
 lib/kunit/executor.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
index 9358ed2df839..1236b3cd2fbb 100644
--- a/lib/kunit/executor.c
+++ b/lib/kunit/executor.c
@@ -157,10 +157,11 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
 	struct kunit_suite_set filtered = {NULL, NULL};
 	struct kunit_glob_filter parsed_glob;
 	struct kunit_attr_filter *parsed_filters = NULL;
+	struct kunit_suite * const *suites;
 
 	const size_t max = suite_set->end - suite_set->start;
 
-	copy = kmalloc_array(max, sizeof(*filtered.start), GFP_KERNEL);
+	copy = kcalloc(max, sizeof(*filtered.start), GFP_KERNEL);
 	if (!copy) { /* won't be able to run anything, return an empty set */
 		return filtered;
 	}
@@ -195,7 +196,7 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
 					parsed_glob.test_glob);
 			if (IS_ERR(filtered_suite)) {
 				*err = PTR_ERR(filtered_suite);
-				goto free_parsed_filters;
+				goto free_filtered_suite;
 			}
 		}
 		if (filter_count > 0 && parsed_filters != NULL) {
@@ -212,11 +213,11 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
 				filtered_suite = new_filtered_suite;
 
 				if (*err)
-					goto free_parsed_filters;
+					goto free_filtered_suite;
 
 				if (IS_ERR(filtered_suite)) {
 					*err = PTR_ERR(filtered_suite);
-					goto free_parsed_filters;
+					goto free_filtered_suite;
 				}
 				if (!filtered_suite)
 					break;
@@ -231,6 +232,14 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
 	filtered.start = copy_start;
 	filtered.end = copy;
 
+free_filtered_suite:
+	if (*err) {
+		for (suites = copy_start; suites < copy; suites++) {
+			kfree((*suites)->test_cases);
+			kfree(*suites);
+		}
+	}
+
 free_parsed_filters:
 	if (filter_count)
 		kfree(parsed_filters);
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 13+ messages in thread

* [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test
  2023-09-14 11:46 [PATCH 0/4] kunit: Fix some bugs in kunit Jinjie Ruan
                   ` (2 preceding siblings ...)
  2023-09-14 11:46 ` [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites() Jinjie Ruan
@ 2023-09-14 11:46 ` Jinjie Ruan
  2023-09-14 14:44   ` kernel test robot
  2023-09-19 21:19   ` Rae Moar
  3 siblings, 2 replies; 13+ messages in thread
From: Jinjie Ruan @ 2023-09-14 11:46 UTC (permalink / raw)
  To: brendan.higgins, davidgow, skhan, dlatypov, rmoar,
	janusz.krzysztofik, linux-kselftest, kunit-dev
  Cc: ruanjinjie

If kunit_filter_suites() succeeds, not only copy but also filtered_suite
and filtered_suite->test_cases should be freed.

So use kunit_free_suite_set() to free the filtered_suite,
filtered_suite->test_cases and copy as kunit_module_exit() and
kunit_run_all_tests() do it.

Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
---
 lib/kunit/executor_test.c | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)

diff --git a/lib/kunit/executor_test.c b/lib/kunit/executor_test.c
index b4f6f96b2844..987b81dce01e 100644
--- a/lib/kunit/executor_test.c
+++ b/lib/kunit/executor_test.c
@@ -56,7 +56,6 @@ static void filter_suites_test(struct kunit *test)
 	got = kunit_filter_suites(&suite_set, "suite2", NULL, NULL, &err);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
 	KUNIT_ASSERT_EQ(test, err, 0);
-	kfree_at_end(test, got.start);
 
 	/* Validate we just have suite2 */
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
@@ -64,6 +63,9 @@ static void filter_suites_test(struct kunit *test)
 
 	/* Contains one element (end is 1 past end) */
 	KUNIT_ASSERT_EQ(test, got.end - got.start, 1);
+
+	if (!err)
+		kunit_free_suite_set(got);
 }
 
 static void filter_suites_test_glob_test(struct kunit *test)
@@ -82,7 +84,6 @@ static void filter_suites_test_glob_test(struct kunit *test)
 	got = kunit_filter_suites(&suite_set, "suite2.test2", NULL, NULL, &err);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
 	KUNIT_ASSERT_EQ(test, err, 0);
-	kfree_at_end(test, got.start);
 
 	/* Validate we just have suite2 */
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
@@ -93,6 +94,9 @@ static void filter_suites_test_glob_test(struct kunit *test)
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
 	KUNIT_EXPECT_STREQ(test, (const char *)got.start[0]->test_cases[0].name, "test2");
 	KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
+
+	if (!err)
+		kunit_free_suite_set(got);
 }
 
 static void filter_suites_to_empty_test(struct kunit *test)
@@ -109,10 +113,12 @@ static void filter_suites_to_empty_test(struct kunit *test)
 
 	got = kunit_filter_suites(&suite_set, "not_found", NULL, NULL, &err);
 	KUNIT_ASSERT_EQ(test, err, 0);
-	kfree_at_end(test, got.start); /* just in case */
 
 	KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
 				"should be empty to indicate no match");
+
+	if (!err)
+		kunit_free_suite_set(got);
 }
 
 static void parse_filter_attr_test(struct kunit *test)
@@ -172,7 +178,6 @@ static void filter_attr_test(struct kunit *test)
 	got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
 	KUNIT_ASSERT_EQ(test, err, 0);
-	kfree_at_end(test, got.start);
 
 	/* Validate we just have normal_suite */
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
@@ -183,6 +188,9 @@ static void filter_attr_test(struct kunit *test)
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
 	KUNIT_EXPECT_STREQ(test, got.start[0]->test_cases[0].name, "normal");
 	KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
+
+	if (!err)
+		kunit_free_suite_set(got);
 }
 
 static void filter_attr_empty_test(struct kunit *test)
@@ -200,10 +208,12 @@ static void filter_attr_empty_test(struct kunit *test)
 
 	got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
 	KUNIT_ASSERT_EQ(test, err, 0);
-	kfree_at_end(test, got.start); /* just in case */
 
 	KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
 				"should be empty to indicate no match");
+
+	if (!err)
+		kunit_free_suite_set(got);
 }
 
 static void filter_attr_skip_test(struct kunit *test)
@@ -222,7 +232,6 @@ static void filter_attr_skip_test(struct kunit *test)
 	got = kunit_filter_suites(&suite_set, NULL, filter, "skip", &err);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
 	KUNIT_ASSERT_EQ(test, err, 0);
-	kfree_at_end(test, got.start);
 
 	/* Validate we have both the slow and normal test */
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
@@ -233,6 +242,9 @@ static void filter_attr_skip_test(struct kunit *test)
 	/* Now ensure slow is skipped and normal is not */
 	KUNIT_EXPECT_EQ(test, got.start[0]->test_cases[0].status, KUNIT_SKIPPED);
 	KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].status);
+
+	if (!err)
+		kunit_free_suite_set(got);
 }
 
 static struct kunit_case executor_test_cases[] = {
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 13+ messages in thread

* Re: [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test
  2023-09-14 11:46 ` [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test Jinjie Ruan
@ 2023-09-14 14:44   ` kernel test robot
  2023-09-19 21:19   ` Rae Moar
  1 sibling, 0 replies; 13+ messages in thread
From: kernel test robot @ 2023-09-14 14:44 UTC (permalink / raw)
  To: Jinjie Ruan, brendan.higgins, davidgow, skhan, dlatypov, rmoar,
	janusz.krzysztofik, linux-kselftest, kunit-dev
  Cc: oe-kbuild-all, ruanjinjie

Hi Jinjie,

kernel test robot noticed the following build warnings:

[auto build test WARNING on linus/master]
[also build test WARNING on next-20230914]
[cannot apply to v6.6-rc1]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Jinjie-Ruan/kunit-Fix-missed-memory-release-in-kunit_free_suite_set/20230914-194915
base:   linus/master
patch link:    https://lore.kernel.org/r/20230914114629.1517650-5-ruanjinjie%40huawei.com
patch subject: [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test
config: m68k-allyesconfig (https://download.01.org/0day-ci/archive/20230914/202309142251.uJ8saAZv-lkp@intel.com/config)
compiler: m68k-linux-gcc (GCC) 13.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20230914/202309142251.uJ8saAZv-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202309142251.uJ8saAZv-lkp@intel.com/

All warnings (new ones prefixed by >>):

   In file included from lib/kunit/executor.c:353:
>> lib/kunit/executor_test.c:274:13: warning: 'kfree_at_end' defined but not used [-Wunused-function]
     274 | static void kfree_at_end(struct kunit *test, const void *to_free)
         |             ^~~~~~~~~~~~


vim +/kfree_at_end +274 lib/kunit/executor_test.c

1d71307a6f94df Daniel Latypov 2021-04-20  270  
1d71307a6f94df Daniel Latypov 2021-04-20  271  /* Use the resource API to register a call to kfree(to_free).
1d71307a6f94df Daniel Latypov 2021-04-20  272   * Since we never actually use the resource, it's safe to use on const data.
1d71307a6f94df Daniel Latypov 2021-04-20  273   */
1d71307a6f94df Daniel Latypov 2021-04-20 @274  static void kfree_at_end(struct kunit *test, const void *to_free)
1d71307a6f94df Daniel Latypov 2021-04-20  275  {
1d71307a6f94df Daniel Latypov 2021-04-20  276  	/* kfree() handles NULL already, but avoid allocating a no-op cleanup. */
1d71307a6f94df Daniel Latypov 2021-04-20  277  	if (IS_ERR_OR_NULL(to_free))
1d71307a6f94df Daniel Latypov 2021-04-20  278  		return;
00e63f8afcfc6b David Gow      2023-05-25  279  
00e63f8afcfc6b David Gow      2023-05-25  280  	kunit_add_action(test,
00e63f8afcfc6b David Gow      2023-05-25  281  			(kunit_action_t *)kfree,
1d71307a6f94df Daniel Latypov 2021-04-20  282  			(void *)to_free);
1d71307a6f94df Daniel Latypov 2021-04-20  283  }
1d71307a6f94df Daniel Latypov 2021-04-20  284  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set()
  2023-09-14 11:46 ` [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set() Jinjie Ruan
@ 2023-09-19 21:18   ` Rae Moar
  0 siblings, 0 replies; 13+ messages in thread
From: Rae Moar @ 2023-09-19 21:18 UTC (permalink / raw)
  To: Jinjie Ruan
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev

On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
<kunit-dev@googlegroups.com> wrote:
>
> modprobe cpumask_kunit and rmmod cpumask_kunit, kmemleak detect
> a suspected memory leak as below.
>
> If kunit_filter_suites() in kunit_module_init() succeeds, the
> suite_set.start will not be NULL and the kunit_free_suite_set() in
> kunit_module_exit() should free all the memory which has not
> been freed. However the test_cases in suites is left out.
>
> unreferenced object 0xffff54ac47e83200 (size 512):
>   comm "modprobe", pid 592, jiffies 4294913238 (age 1367.612s)
>   hex dump (first 32 bytes):
>     84 13 1a f0 d3 b6 ff ff 30 68 1a f0 d3 b6 ff ff  ........0h......
>     00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
>   backtrace:
>     [<000000008dec63a2>] slab_post_alloc_hook+0xb8/0x368
>     [<00000000ec280d8e>] __kmem_cache_alloc_node+0x174/0x290
>     [<00000000896c7740>] __kmalloc+0x60/0x2c0
>     [<000000007a50fa06>] kunit_filter_suites+0x254/0x5b8
>     [<0000000078cc98e2>] kunit_module_notify+0xf4/0x240
>     [<0000000033cea952>] notifier_call_chain+0x98/0x17c
>     [<00000000973d05cc>] notifier_call_chain_robust+0x4c/0xa4
>     [<000000005f95895f>] blocking_notifier_call_chain_robust+0x4c/0x74
>     [<0000000048e36fa7>] load_module+0x1a2c/0x1c40
>     [<0000000004eb8a91>] init_module_from_file+0x94/0xcc
>     [<0000000037dbba28>] idempotent_init_module+0x184/0x278
>     [<00000000161b75cb>] __arm64_sys_finit_module+0x68/0xa8
>     [<000000006dc1669b>] invoke_syscall+0x44/0x100
>     [<00000000fa87e304>] el0_svc_common.constprop.1+0x68/0xe0
>     [<000000009d8ad866>] do_el0_svc+0x1c/0x28
>     [<000000005b83c607>] el0_svc+0x3c/0xc4
>
> Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
> Fixes: b67abaad4d25 ("kunit: Allow kunit test modules to use test filtering")
> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>

Hello!

This looks good to me.

Reviewed-by: Rae Moar <rmoar@google.com>

Thanks!

-Rae

> ---
>  lib/kunit/executor.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
> index a6348489d45f..a037a46fae5e 100644
> --- a/lib/kunit/executor.c
> +++ b/lib/kunit/executor.c
> @@ -137,8 +137,10 @@ void kunit_free_suite_set(struct kunit_suite_set suite_set)
>  {
>         struct kunit_suite * const *suites;
>
> -       for (suites = suite_set.start; suites < suite_set.end; suites++)
> +       for (suites = suite_set.start; suites < suite_set.end; suites++) {
> +               kfree((*suites)->test_cases);
>                 kfree(*suites);
> +       }
>         kfree(suite_set.start);
>  }
>
> --
> 2.34.1
>
> --
> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-2-ruanjinjie%40huawei.com.

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 2/4] kunit: Fix the wrong kfree of copy for kunit_filter_suites()
  2023-09-14 11:46 ` [PATCH 2/4] kunit: Fix the wrong kfree of copy for kunit_filter_suites() Jinjie Ruan
@ 2023-09-19 21:18   ` Rae Moar
  0 siblings, 0 replies; 13+ messages in thread
From: Rae Moar @ 2023-09-19 21:18 UTC (permalink / raw)
  To: Jinjie Ruan
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev

On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
<kunit-dev@googlegroups.com> wrote:
>
> If the outer layer for loop is iterated more than once and it fails not
> in the first iteration, the copy pointer has been moved. So it should free
> the original copy's backup copy_start.
>
> Fixes: abbf73816b6f ("kunit: fix possible memory leak in kunit_filter_suites()")
> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>

Hello!

This looks good to me as well.

Reviewed-by: Rae Moar <rmoar@google.com>

Thanks!

-Rae

> ---
>  lib/kunit/executor.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
> index a037a46fae5e..9358ed2df839 100644
> --- a/lib/kunit/executor.c
> +++ b/lib/kunit/executor.c
> @@ -243,7 +243,7 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>
>  free_copy:
>         if (*err)
> -               kfree(copy);
> +               kfree(copy_start);
>
>         return filtered;
>  }
> --
> 2.34.1
>
> --
> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-3-ruanjinjie%40huawei.com.

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites()
  2023-09-14 11:46 ` [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites() Jinjie Ruan
@ 2023-09-19 21:18   ` Rae Moar
  2023-09-20  2:34     ` Ruan Jinjie
  0 siblings, 1 reply; 13+ messages in thread
From: Rae Moar @ 2023-09-19 21:18 UTC (permalink / raw)
  To: Jinjie Ruan
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev

On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
<kunit-dev@googlegroups.com> wrote:
>
> If the outer layer for loop is iterated more than once and it fails not
> in the first iteration, the filtered_suite and filtered_suite->test_cases
> allocated in the last kunit_filter_attr_tests() in last inner for loop
> is leaked.
>
> So add a new free_filtered_suite err label and free the filtered_suite
> and filtered_suite->test_cases so far. And change kmalloc_array of copy
> to kcalloc to Clear the copy to make the kfree safe.
>
> Fixes: 5d31f71efcb6 ("kunit: add kunit.filter_glob cmdline option to filter suites")
> Fixes: 529534e8cba3 ("kunit: Add ability to filter attributes")
> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>

Hello!

This looks good to me. I just have one comment below.

Reviewed-by: Rae Moar <rmoar@google.com>

Thanks!
-Rae

> ---
>  lib/kunit/executor.c | 17 +++++++++++++----
>  1 file changed, 13 insertions(+), 4 deletions(-)
>
> diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
> index 9358ed2df839..1236b3cd2fbb 100644
> --- a/lib/kunit/executor.c
> +++ b/lib/kunit/executor.c
> @@ -157,10 +157,11 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>         struct kunit_suite_set filtered = {NULL, NULL};
>         struct kunit_glob_filter parsed_glob;
>         struct kunit_attr_filter *parsed_filters = NULL;
> +       struct kunit_suite * const *suites;
>
>         const size_t max = suite_set->end - suite_set->start;
>
> -       copy = kmalloc_array(max, sizeof(*filtered.start), GFP_KERNEL);
> +       copy = kcalloc(max, sizeof(*filtered.start), GFP_KERNEL);
>         if (!copy) { /* won't be able to run anything, return an empty set */
>                 return filtered;
>         }
> @@ -195,7 +196,7 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>                                         parsed_glob.test_glob);
>                         if (IS_ERR(filtered_suite)) {
>                                 *err = PTR_ERR(filtered_suite);
> -                               goto free_parsed_filters;
> +                               goto free_filtered_suite;
>                         }
>                 }
>                 if (filter_count > 0 && parsed_filters != NULL) {
> @@ -212,11 +213,11 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>                                 filtered_suite = new_filtered_suite;
>
>                                 if (*err)
> -                                       goto free_parsed_filters;
> +                                       goto free_filtered_suite;
>
>                                 if (IS_ERR(filtered_suite)) {
>                                         *err = PTR_ERR(filtered_suite);
> -                                       goto free_parsed_filters;
> +                                       goto free_filtered_suite;
>                                 }
>                                 if (!filtered_suite)
>                                         break;
> @@ -231,6 +232,14 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>         filtered.start = copy_start;
>         filtered.end = copy;
>
> +free_filtered_suite:
> +       if (*err) {
> +               for (suites = copy_start; suites < copy; suites++) {
> +                       kfree((*suites)->test_cases);
> +                       kfree(*suites);
> +               }
> +       }
> +

As this is pretty similar code to kunit_free_suite_set, I wish you
could use that method instead but I'm not actually sure it would be
cleaner.


>  free_parsed_filters:
>         if (filter_count)
>                 kfree(parsed_filters);
> --
> 2.34.1
>
> --
> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-4-ruanjinjie%40huawei.com.

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test
  2023-09-14 11:46 ` [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test Jinjie Ruan
  2023-09-14 14:44   ` kernel test robot
@ 2023-09-19 21:19   ` Rae Moar
  2023-09-20  2:57     ` Ruan Jinjie
  2023-09-20  7:00     ` Ruan Jinjie
  1 sibling, 2 replies; 13+ messages in thread
From: Rae Moar @ 2023-09-19 21:19 UTC (permalink / raw)
  To: Jinjie Ruan
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev

On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
<kunit-dev@googlegroups.com> wrote:
>
> If kunit_filter_suites() succeeds, not only copy but also filtered_suite
> and filtered_suite->test_cases should be freed.
>
> So use kunit_free_suite_set() to free the filtered_suite,
> filtered_suite->test_cases and copy as kunit_module_exit() and
> kunit_run_all_tests() do it.
>
> Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>

Hello!

This looks mostly good to me. But I have one notable comment. See below.

Thanks!
-Rae

> ---
>  lib/kunit/executor_test.c | 24 ++++++++++++++++++------
>  1 file changed, 18 insertions(+), 6 deletions(-)
>
> diff --git a/lib/kunit/executor_test.c b/lib/kunit/executor_test.c
> index b4f6f96b2844..987b81dce01e 100644
> --- a/lib/kunit/executor_test.c
> +++ b/lib/kunit/executor_test.c
> @@ -56,7 +56,6 @@ static void filter_suites_test(struct kunit *test)
>         got = kunit_filter_suites(&suite_set, "suite2", NULL, NULL, &err);
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>         KUNIT_ASSERT_EQ(test, err, 0);
> -       kfree_at_end(test, got.start);
>
>         /* Validate we just have suite2 */
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
> @@ -64,6 +63,9 @@ static void filter_suites_test(struct kunit *test)
>
>         /* Contains one element (end is 1 past end) */
>         KUNIT_ASSERT_EQ(test, got.end - got.start, 1);
> +
> +       if (!err)
> +               kunit_free_suite_set(got);

I definitely appreciate the change to free all of "got" rather than
just "got.start".

However, kfree_at_end used deferred actions to ensure the kfree would
occur at the end of the test. With this change, if the test fails the
suite set could not be freed.

Intead, is there any way to alter the function kfree_at_end (could be
renamed) to take in "got" and then use deferred actions to ensure
kunit_free_suite_set is called at the end of the test?

Let me know what you think about this.


>  }
>
>  static void filter_suites_test_glob_test(struct kunit *test)
> @@ -82,7 +84,6 @@ static void filter_suites_test_glob_test(struct kunit *test)
>         got = kunit_filter_suites(&suite_set, "suite2.test2", NULL, NULL, &err);
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>         KUNIT_ASSERT_EQ(test, err, 0);
> -       kfree_at_end(test, got.start);
>
>         /* Validate we just have suite2 */
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
> @@ -93,6 +94,9 @@ static void filter_suites_test_glob_test(struct kunit *test)
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>         KUNIT_EXPECT_STREQ(test, (const char *)got.start[0]->test_cases[0].name, "test2");
>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
> +
> +       if (!err)
> +               kunit_free_suite_set(got);
>  }
>
>  static void filter_suites_to_empty_test(struct kunit *test)
> @@ -109,10 +113,12 @@ static void filter_suites_to_empty_test(struct kunit *test)
>
>         got = kunit_filter_suites(&suite_set, "not_found", NULL, NULL, &err);
>         KUNIT_ASSERT_EQ(test, err, 0);
> -       kfree_at_end(test, got.start); /* just in case */
>
>         KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
>                                 "should be empty to indicate no match");
> +
> +       if (!err)
> +               kunit_free_suite_set(got);
>  }
>
>  static void parse_filter_attr_test(struct kunit *test)
> @@ -172,7 +178,6 @@ static void filter_attr_test(struct kunit *test)
>         got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>         KUNIT_ASSERT_EQ(test, err, 0);
> -       kfree_at_end(test, got.start);
>
>         /* Validate we just have normal_suite */
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
> @@ -183,6 +188,9 @@ static void filter_attr_test(struct kunit *test)
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>         KUNIT_EXPECT_STREQ(test, got.start[0]->test_cases[0].name, "normal");
>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
> +
> +       if (!err)
> +               kunit_free_suite_set(got);
>  }
>
>  static void filter_attr_empty_test(struct kunit *test)
> @@ -200,10 +208,12 @@ static void filter_attr_empty_test(struct kunit *test)
>
>         got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
>         KUNIT_ASSERT_EQ(test, err, 0);
> -       kfree_at_end(test, got.start); /* just in case */
>
>         KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
>                                 "should be empty to indicate no match");
> +
> +       if (!err)
> +               kunit_free_suite_set(got);
>  }
>
>  static void filter_attr_skip_test(struct kunit *test)
> @@ -222,7 +232,6 @@ static void filter_attr_skip_test(struct kunit *test)
>         got = kunit_filter_suites(&suite_set, NULL, filter, "skip", &err);
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>         KUNIT_ASSERT_EQ(test, err, 0);
> -       kfree_at_end(test, got.start);
>
>         /* Validate we have both the slow and normal test */
>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
> @@ -233,6 +242,9 @@ static void filter_attr_skip_test(struct kunit *test)
>         /* Now ensure slow is skipped and normal is not */
>         KUNIT_EXPECT_EQ(test, got.start[0]->test_cases[0].status, KUNIT_SKIPPED);
>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].status);
> +
> +       if (!err)
> +               kunit_free_suite_set(got);
>  }
>
>  static struct kunit_case executor_test_cases[] = {
> --
> 2.34.1
>
> --
> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-5-ruanjinjie%40huawei.com.

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites()
  2023-09-19 21:18   ` Rae Moar
@ 2023-09-20  2:34     ` Ruan Jinjie
  0 siblings, 0 replies; 13+ messages in thread
From: Ruan Jinjie @ 2023-09-20  2:34 UTC (permalink / raw)
  To: Rae Moar
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev



On 2023/9/20 5:18, Rae Moar wrote:
> On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
> <kunit-dev@googlegroups.com> wrote:
>>
>> If the outer layer for loop is iterated more than once and it fails not
>> in the first iteration, the filtered_suite and filtered_suite->test_cases
>> allocated in the last kunit_filter_attr_tests() in last inner for loop
>> is leaked.
>>
>> So add a new free_filtered_suite err label and free the filtered_suite
>> and filtered_suite->test_cases so far. And change kmalloc_array of copy
>> to kcalloc to Clear the copy to make the kfree safe.
>>
>> Fixes: 5d31f71efcb6 ("kunit: add kunit.filter_glob cmdline option to filter suites")
>> Fixes: 529534e8cba3 ("kunit: Add ability to filter attributes")
>> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
> 
> Hello!
> 
> This looks good to me. I just have one comment below.
> 
> Reviewed-by: Rae Moar <rmoar@google.com>
> 
> Thanks!
> -Rae
> 
>> ---
>>  lib/kunit/executor.c | 17 +++++++++++++----
>>  1 file changed, 13 insertions(+), 4 deletions(-)
>>
>> diff --git a/lib/kunit/executor.c b/lib/kunit/executor.c
>> index 9358ed2df839..1236b3cd2fbb 100644
>> --- a/lib/kunit/executor.c
>> +++ b/lib/kunit/executor.c
>> @@ -157,10 +157,11 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>>         struct kunit_suite_set filtered = {NULL, NULL};
>>         struct kunit_glob_filter parsed_glob;
>>         struct kunit_attr_filter *parsed_filters = NULL;
>> +       struct kunit_suite * const *suites;
>>
>>         const size_t max = suite_set->end - suite_set->start;
>>
>> -       copy = kmalloc_array(max, sizeof(*filtered.start), GFP_KERNEL);
>> +       copy = kcalloc(max, sizeof(*filtered.start), GFP_KERNEL);
>>         if (!copy) { /* won't be able to run anything, return an empty set */
>>                 return filtered;
>>         }
>> @@ -195,7 +196,7 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>>                                         parsed_glob.test_glob);
>>                         if (IS_ERR(filtered_suite)) {
>>                                 *err = PTR_ERR(filtered_suite);
>> -                               goto free_parsed_filters;
>> +                               goto free_filtered_suite;
>>                         }
>>                 }
>>                 if (filter_count > 0 && parsed_filters != NULL) {
>> @@ -212,11 +213,11 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>>                                 filtered_suite = new_filtered_suite;
>>
>>                                 if (*err)
>> -                                       goto free_parsed_filters;
>> +                                       goto free_filtered_suite;
>>
>>                                 if (IS_ERR(filtered_suite)) {
>>                                         *err = PTR_ERR(filtered_suite);
>> -                                       goto free_parsed_filters;
>> +                                       goto free_filtered_suite;
>>                                 }
>>                                 if (!filtered_suite)
>>                                         break;
>> @@ -231,6 +232,14 @@ kunit_filter_suites(const struct kunit_suite_set *suite_set,
>>         filtered.start = copy_start;
>>         filtered.end = copy;
>>
>> +free_filtered_suite:
>> +       if (*err) {
>> +               for (suites = copy_start; suites < copy; suites++) {
>> +                       kfree((*suites)->test_cases);
>> +                       kfree(*suites);
>> +               }
>> +       }
>> +
> 
> As this is pretty similar code to kunit_free_suite_set, I wish you
> could use that method instead but I'm not actually sure it would be
> cleaner.

There is a slight difference between here and kunit_free_suite_set(), it
do not kfree(suite_set.start) which is kfree(copy_start) here as it is
the first kcalloc.

> 
> 
>>  free_parsed_filters:
>>         if (filter_count)
>>                 kfree(parsed_filters);
>> --
>> 2.34.1
>>
>> --
>> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-4-ruanjinjie%40huawei.com.
> 

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test
  2023-09-19 21:19   ` Rae Moar
@ 2023-09-20  2:57     ` Ruan Jinjie
  2023-09-20  7:00     ` Ruan Jinjie
  1 sibling, 0 replies; 13+ messages in thread
From: Ruan Jinjie @ 2023-09-20  2:57 UTC (permalink / raw)
  To: Rae Moar
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev



On 2023/9/20 5:19, Rae Moar wrote:
> On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
> <kunit-dev@googlegroups.com> wrote:
>>
>> If kunit_filter_suites() succeeds, not only copy but also filtered_suite
>> and filtered_suite->test_cases should be freed.
>>
>> So use kunit_free_suite_set() to free the filtered_suite,
>> filtered_suite->test_cases and copy as kunit_module_exit() and
>> kunit_run_all_tests() do it.
>>
>> Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
>> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
> 
> Hello!
> 
> This looks mostly good to me. But I have one notable comment. See below.
> 
> Thanks!
> -Rae
> 
>> ---
>>  lib/kunit/executor_test.c | 24 ++++++++++++++++++------
>>  1 file changed, 18 insertions(+), 6 deletions(-)
>>
>> diff --git a/lib/kunit/executor_test.c b/lib/kunit/executor_test.c
>> index b4f6f96b2844..987b81dce01e 100644
>> --- a/lib/kunit/executor_test.c
>> +++ b/lib/kunit/executor_test.c
>> @@ -56,7 +56,6 @@ static void filter_suites_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, "suite2", NULL, NULL, &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we just have suite2 */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
>> @@ -64,6 +63,9 @@ static void filter_suites_test(struct kunit *test)
>>
>>         /* Contains one element (end is 1 past end) */
>>         KUNIT_ASSERT_EQ(test, got.end - got.start, 1);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
> 
> I definitely appreciate the change to free all of "got" rather than
> just "got.start".
> 
> However, kfree_at_end used deferred actions to ensure the kfree would
> occur at the end of the test. With this change, if the test fails the
> suite set could not be freed.

Only when suite_set.start != NULL which is equivalent to err == 0 the
suite set will be freed in kunit_module_exit(), and
kunit_free_suite_set() will be called only when kunit_filter_suites()
succeeds with err == 0. So judging from the use of kunit_filter_suites()
in kunit_module_exit() and kunit_run_all_tests(), it only wants to free
the suite set when kunit_filter_suites() succeeds with err == 0. So in
kunit_filter_suites() , we free
copy, filtered_suite and filtered_suite->test_cases if (*err) is true.

So if the test fails the suite set will be freed also. And there's a
double free problem in kfree_at_end(test, got.start) if the test fails.
So only free the suite set when err == 0.


737 static void kunit_module_init(struct module *mod)
738 {
739         struct kunit_suite_set suite_set = {
740                 mod->kunit_suites, mod->kunit_suites +
mod->num_kunit_suites,
741         };
742         const char *action = kunit_action();
743         int err = 0;
744
745         suite_set = kunit_filter_suites(&suite_set,
746                                         kunit_filter_glob() ?: "*.*",
747                                         kunit_filter(),
kunit_filter_action(),
748                                         &err);

765 static void kunit_module_exit(struct module *mod)
766 {
767         struct kunit_suite_set suite_set = {
768                 mod->kunit_suites, mod->kunit_suites +
mod->num_kunit_suites,
769         };
770         const char *action = kunit_action();
771
772         if (!action)
773                 __kunit_test_suites_exit(mod->kunit_suites,
774                                          mod->num_kunit_suites);
775
776         if (suite_set.start)
777                 kunit_free_suite_set(suite_set);
778 }



314 int kunit_run_all_tests(void)
315 {
           ......
325         if (filter_glob_param || filter_param) {
326                 suite_set = kunit_filter_suites(&suite_set,
filter_glob_param,
327                                 filter_param, filter_action_param,
&err);
328                 if (err) {
329                         pr_err("kunit executor: error filtering
suites: %d\n", err);
330                         goto out;
331                 }
332         }
......
342
343         if (filter_glob_param || filter_param) { /* a copy was made
of each suite */
344                 kunit_free_suite_set(suite_set);
345         }
346
347 out:
348         kunit_handle_shutdown();
349         return err;
350 }


> 
> Intead, is there any way to alter the function kfree_at_end (could be
> renamed) to take in "got" and then use deferred actions to ensure
> kunit_free_suite_set is called at the end of the test?

It is good iead. And it may be fine to call kfree_at_end(test, got) if
err == 0 to avoid double free issue.

> 
> Let me know what you think about this.
> 
> 
>>  }
>>
>>  static void filter_suites_test_glob_test(struct kunit *test)
>> @@ -82,7 +84,6 @@ static void filter_suites_test_glob_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, "suite2.test2", NULL, NULL, &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we just have suite2 */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
>> @@ -93,6 +94,9 @@ static void filter_suites_test_glob_test(struct kunit *test)
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>>         KUNIT_EXPECT_STREQ(test, (const char *)got.start[0]->test_cases[0].name, "test2");
>>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void filter_suites_to_empty_test(struct kunit *test)
>> @@ -109,10 +113,12 @@ static void filter_suites_to_empty_test(struct kunit *test)
>>
>>         got = kunit_filter_suites(&suite_set, "not_found", NULL, NULL, &err);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start); /* just in case */
>>
>>         KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
>>                                 "should be empty to indicate no match");
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void parse_filter_attr_test(struct kunit *test)
>> @@ -172,7 +178,6 @@ static void filter_attr_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we just have normal_suite */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
>> @@ -183,6 +188,9 @@ static void filter_attr_test(struct kunit *test)
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>>         KUNIT_EXPECT_STREQ(test, got.start[0]->test_cases[0].name, "normal");
>>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void filter_attr_empty_test(struct kunit *test)
>> @@ -200,10 +208,12 @@ static void filter_attr_empty_test(struct kunit *test)
>>
>>         got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start); /* just in case */
>>
>>         KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
>>                                 "should be empty to indicate no match");
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void filter_attr_skip_test(struct kunit *test)
>> @@ -222,7 +232,6 @@ static void filter_attr_skip_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, NULL, filter, "skip", &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we have both the slow and normal test */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>> @@ -233,6 +242,9 @@ static void filter_attr_skip_test(struct kunit *test)
>>         /* Now ensure slow is skipped and normal is not */
>>         KUNIT_EXPECT_EQ(test, got.start[0]->test_cases[0].status, KUNIT_SKIPPED);
>>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].status);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static struct kunit_case executor_test_cases[] = {
>> --
>> 2.34.1
>>
>> --
>> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-5-ruanjinjie%40huawei.com.
> 

^ permalink raw reply	[flat|nested] 13+ messages in thread

* Re: [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test
  2023-09-19 21:19   ` Rae Moar
  2023-09-20  2:57     ` Ruan Jinjie
@ 2023-09-20  7:00     ` Ruan Jinjie
  1 sibling, 0 replies; 13+ messages in thread
From: Ruan Jinjie @ 2023-09-20  7:00 UTC (permalink / raw)
  To: Rae Moar
  Cc: brendan.higgins, davidgow, skhan, dlatypov, janusz.krzysztofik,
	linux-kselftest, kunit-dev



On 2023/9/20 5:19, Rae Moar wrote:
> On Thu, Sep 14, 2023 at 7:47 AM 'Jinjie Ruan' via KUnit Development
> <kunit-dev@googlegroups.com> wrote:
>>
>> If kunit_filter_suites() succeeds, not only copy but also filtered_suite
>> and filtered_suite->test_cases should be freed.
>>
>> So use kunit_free_suite_set() to free the filtered_suite,
>> filtered_suite->test_cases and copy as kunit_module_exit() and
>> kunit_run_all_tests() do it.
>>
>> Fixes: e5857d396f35 ("kunit: flatten kunit_suite*** to kunit_suite** in .kunit_test_suites")
>> Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
> 
> Hello!
> 
> This looks mostly good to me. But I have one notable comment. See below.
> 
> Thanks!
> -Rae
> 
>> ---
>>  lib/kunit/executor_test.c | 24 ++++++++++++++++++------
>>  1 file changed, 18 insertions(+), 6 deletions(-)
>>
>> diff --git a/lib/kunit/executor_test.c b/lib/kunit/executor_test.c
>> index b4f6f96b2844..987b81dce01e 100644
>> --- a/lib/kunit/executor_test.c
>> +++ b/lib/kunit/executor_test.c
>> @@ -56,7 +56,6 @@ static void filter_suites_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, "suite2", NULL, NULL, &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we just have suite2 */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
>> @@ -64,6 +63,9 @@ static void filter_suites_test(struct kunit *test)
>>
>>         /* Contains one element (end is 1 past end) */
>>         KUNIT_ASSERT_EQ(test, got.end - got.start, 1);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
> 
> I definitely appreciate the change to free all of "got" rather than
> just "got.start".
> 
> However, kfree_at_end used deferred actions to ensure the kfree would
> occur at the end of the test. With this change, if the test fails the
> suite set could not be freed.
> 
> Intead, is there any way to alter the function kfree_at_end (could be
> renamed) to take in "got" and then use deferred actions to ensure
> kunit_free_suite_set is called at the end of the test?

I try it and it seems unfeasible because the got is a local struct
kunit_suite_set and kunit_free_suite_set use it in another func will
cause wild-memory-access as the struct kunit_suite_set has been freed
already.

[   49.490158] general protection fault, probably for non-canonical
address 0xe006fbfff71d514d: 0000 [#1] PREEMPT SMP KASAN
[   49.493858] KASAN: maybe wild-memory-access in range
[0x0037ffffb8ea8a68-0x0037ffffb8ea8a6f]
[   49.495391] CPU: 2 PID: 1439 Comm: kunit_try_catch Tainted: G    B
        N 6.6.0-rc2+ #29
[   49.496578] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
BIOS 1.15.0-1 04/01/2014
[   49.497715] RIP: 0010:kunit_free_suite_set+0x8e/0x150
[   49.498419] Code: 4d 89 fe 49 c1 ee 03 49 01 de 48 89 e8 48 c1 e8 03
80 3c 18 00 75 7d 4c 8b 65 00 49 8d bc 24 20 01 00 00 48 89 f8 48 c1 e8
03 <80> 3c 18 00 0f 85 81 00 00 00 49 8b bc 24 20 01 00 00 e8 2b 90 e0
[   49.500918] RSP: 0000:ffff8881047f7e18 EFLAGS: 00010207
[   49.501627] RAX: 0006fffff71d514d RBX: dffffc0000000000 RCX:
1ffff11020814d9e
[   49.502597] RDX: 1ffff110214f6fbc RSI: 0000000000000004 RDI:
0037ffffb8ea8a6c
[   49.503553] RBP: ffffffff811d098f R08: 0000000000000001 R09:
ffffed1020814d99
[   49.504526] R10: ffff8881040a6ccb R11: 0000000000000400 R12:
0037ffffb8ea894c
[   49.505489] R13: ffff88810a7b7dd8 R14: ffffed10214f6fbc R15:
ffff88810a7b7de0
[   49.506470] FS:  0000000000000000(0000) GS:ffff888119d00000(0000)
knlGS:0000000000000000
[   49.507497] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   49.508200] CR2: ffff88811948dfff CR3: 0000000005286001 CR4:
0000000000770ee0
[   49.509078] DR0: ffffffff8fdd6ce8 DR1: ffffffff8fdd6ce9 DR2:
ffffffff8fdd6cea
[   49.509909] DR3: ffffffff8fdd6ceb DR6: 00000000fffe0ff0 DR7:
0000000000000600
[   49.510790] PKRU: 55555554
[   49.511127] Call Trace:
[   49.511448]  <TASK>
[   49.511715]  ? die_addr+0x3d/0xa0
[   49.512133]  ? exc_general_protection+0x144/0x220
[   49.512725]  ? asm_exc_general_protection+0x22/0x30
[   49.513323]  ? do_exit+0x125f/0x2240
[   49.513785]  ? kunit_free_suite_set+0x8e/0x150
[   49.514330]  ? _raw_spin_lock_irqsave+0x8d/0xe0
[   49.514901]  kunit_remove_resource+0x191/0x2a0
[   49.515464]  ? __sched_text_end+0xa/0xa
[   49.515949]  ? __sched_text_end+0xa/0xa
[   49.516432]  kunit_cleanup+0x6f/0x110
[   49.516885]  ? kunit_cleanup+0x110/0x110
[   49.517374]  kunit_generic_run_threadfn_adapter+0x4a/0x90
[   49.518045]  ? kunit_try_catch_throw+0x80/0x80
[   49.518623]  kthread+0x2b5/0x380
[   49.519032]  ? kthread_complete_and_exit+0x20/0x20
[   49.519626]  ret_from_fork+0x2d/0x70
[   49.520078]  ? kthread_complete_and_exit+0x20/0x20
[   49.520679]  ret_from_fork_asm+0x11/0x20
[   49.521179]  </TASK>
[   49.521459] Modules linked in:
[   49.521849] Dumping ftrace buffer:
[   49.522273]    (ftrace buffer empty)
[   49.522764] ---[ end trace 0000000000000000 ]---
[   49.523416] RIP: 0010:kunit_free_suite_set+0x8e/0x150
[   49.524049] Code: 4d 89 fe 49 c1 ee 03 49 01 de 48 89 e8 48 c1 e8 03
80 3c 18 00 75 7d 4c 8b 65 00 49 8d bc 24 20 01 00 00 48 89 f8 48 c1 e8
03 <80> 3c 18 00 0f 85 81 00 00 00 49 8b bc 24 20 01 00 00 e8 2b 90 e0
[   49.526357] RSP: 0000:ffff8881047f7e18 EFLAGS: 00010207
[   49.527026] RAX: 0006fffff71d514d RBX: dffffc0000000000 RCX:
1ffff11020814d9e
[   49.527898] RDX: 1ffff110214f6fbc RSI: 0000000000000004 RDI:
0037ffffb8ea8a6c
[   49.528765] RBP: ffffffff811d098f R08: 0000000000000001 R09:
ffffed1020814d99
[   49.529633] R10: ffff8881040a6ccb R11: 0000000000000400 R12:
0037ffffb8ea894c
[   49.530524] R13: ffff88810a7b7dd8 R14: ffffed10214f6fbc R15:
ffff88810a7b7de0
[   49.531405] FS:  0000000000000000(0000) GS:ffff888119d00000(0000)
knlGS:0000000000000000
[   49.532386] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   49.533094] CR2: ffff88811948dfff CR3: 0000000005286001 CR4:
0000000000770ee0
[   49.533971] DR0: ffffffff8fdd6ce8 DR1: ffffffff8fdd6ce9 DR2:
ffffffff8fdd6cea
[   49.534879] DR3: ffffffff8fdd6ceb DR6: 00000000fffe0ff0 DR7:
0000000000000600
[   49.535774] PKRU: 55555554
[   49.536108] Kernel panic - not syncing: Fatal exception
[   49.537673] Dumping ftrace buffer:
[   49.538076]    (ftrace buffer empty)
[   49.538531] Kernel Offset: disabled
[   49.538963] Rebooting in 1 seconds..

> 
> Let me know what you think about this.
> 
> 
>>  }
>>
>>  static void filter_suites_test_glob_test(struct kunit *test)
>> @@ -82,7 +84,6 @@ static void filter_suites_test_glob_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, "suite2.test2", NULL, NULL, &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we just have suite2 */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
>> @@ -93,6 +94,9 @@ static void filter_suites_test_glob_test(struct kunit *test)
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>>         KUNIT_EXPECT_STREQ(test, (const char *)got.start[0]->test_cases[0].name, "test2");
>>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void filter_suites_to_empty_test(struct kunit *test)
>> @@ -109,10 +113,12 @@ static void filter_suites_to_empty_test(struct kunit *test)
>>
>>         got = kunit_filter_suites(&suite_set, "not_found", NULL, NULL, &err);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start); /* just in case */
>>
>>         KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
>>                                 "should be empty to indicate no match");
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void parse_filter_attr_test(struct kunit *test)
>> @@ -172,7 +178,6 @@ static void filter_attr_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we just have normal_suite */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]);
>> @@ -183,6 +188,9 @@ static void filter_attr_test(struct kunit *test)
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>>         KUNIT_EXPECT_STREQ(test, got.start[0]->test_cases[0].name, "normal");
>>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].name);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void filter_attr_empty_test(struct kunit *test)
>> @@ -200,10 +208,12 @@ static void filter_attr_empty_test(struct kunit *test)
>>
>>         got = kunit_filter_suites(&suite_set, NULL, filter, NULL, &err);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start); /* just in case */
>>
>>         KUNIT_EXPECT_PTR_EQ_MSG(test, got.start, got.end,
>>                                 "should be empty to indicate no match");
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static void filter_attr_skip_test(struct kunit *test)
>> @@ -222,7 +232,6 @@ static void filter_attr_skip_test(struct kunit *test)
>>         got = kunit_filter_suites(&suite_set, NULL, filter, "skip", &err);
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start);
>>         KUNIT_ASSERT_EQ(test, err, 0);
>> -       kfree_at_end(test, got.start);
>>
>>         /* Validate we have both the slow and normal test */
>>         KUNIT_ASSERT_NOT_ERR_OR_NULL(test, got.start[0]->test_cases);
>> @@ -233,6 +242,9 @@ static void filter_attr_skip_test(struct kunit *test)
>>         /* Now ensure slow is skipped and normal is not */
>>         KUNIT_EXPECT_EQ(test, got.start[0]->test_cases[0].status, KUNIT_SKIPPED);
>>         KUNIT_EXPECT_FALSE(test, got.start[0]->test_cases[1].status);
>> +
>> +       if (!err)
>> +               kunit_free_suite_set(got);
>>  }
>>
>>  static struct kunit_case executor_test_cases[] = {
>> --
>> 2.34.1
>>
>> --
>> You received this message because you are subscribed to the Google Groups "KUnit Development" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to kunit-dev+unsubscribe@googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/msgid/kunit-dev/20230914114629.1517650-5-ruanjinjie%40huawei.com.
> 

^ permalink raw reply	[flat|nested] 13+ messages in thread

end of thread, other threads:[~2023-09-20  7:00 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-09-14 11:46 [PATCH 0/4] kunit: Fix some bugs in kunit Jinjie Ruan
2023-09-14 11:46 ` [PATCH 1/4] kunit: Fix missed memory release in kunit_free_suite_set() Jinjie Ruan
2023-09-19 21:18   ` Rae Moar
2023-09-14 11:46 ` [PATCH 2/4] kunit: Fix the wrong kfree of copy for kunit_filter_suites() Jinjie Ruan
2023-09-19 21:18   ` Rae Moar
2023-09-14 11:46 ` [PATCH 3/4] kunit: Fix possible memory leak in kunit_filter_suites() Jinjie Ruan
2023-09-19 21:18   ` Rae Moar
2023-09-20  2:34     ` Ruan Jinjie
2023-09-14 11:46 ` [PATCH 4/4] kunit: test: Fix the possible memory leak in executor_test Jinjie Ruan
2023-09-14 14:44   ` kernel test robot
2023-09-19 21:19   ` Rae Moar
2023-09-20  2:57     ` Ruan Jinjie
2023-09-20  7:00     ` Ruan Jinjie

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.