[LTP] [PATCH v2 0/2] Rewrite the gettid() testing suite

[LTP] [PATCH v2 0/2] Rewrite the gettid() testing suite

[Andrea Cervesato]

From: Andrea Cervesato <andrea.cervesato@suse.com>

Rewrite gettid01 and create a new gettid02 test in order to check all
gettid() basic functionalities, according with its manual.

Andrea Cervesato (2):
  Rewrite gettid01 test
  Add gettid02 test

 runtest/syscalls                            |   1 +
 testcases/kernel/syscalls/gettid/.gitignore |   1 +
 testcases/kernel/syscalls/gettid/Makefile   |   4 +-
 testcases/kernel/syscalls/gettid/gettid01.c | 105 ++++----------------
 testcases/kernel/syscalls/gettid/gettid02.c |  68 +++++++++++++
 5 files changed, 93 insertions(+), 86 deletions(-)
 create mode 100644 testcases/kernel/syscalls/gettid/gettid02.c

-- 
2.35.3


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

[LTP] [PATCH v2 1/2] Rewrite gettid01 test

[Andrea Cervesato]

From: Andrea Cervesato <andrea.cervesato@suse.com>

The old test wasn't doing anything meaningful, but just checking that
__NR_gettid syscall is working. In this test we check that
tid == pid, in the case we are not inside a thread.

Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
---
Check if we are inside a thread
Make use of tst_syscall

 testcases/kernel/syscalls/gettid/gettid01.c | 105 ++++----------------
 1 file changed, 20 insertions(+), 85 deletions(-)

diff --git a/testcases/kernel/syscalls/gettid/gettid01.c b/testcases/kernel/syscalls/gettid/gettid01.c
index 7e5b6b175..5f9623e9a 100644
--- a/testcases/kernel/syscalls/gettid/gettid01.c
+++ b/testcases/kernel/syscalls/gettid/gettid01.c
@@ -1,96 +1,31 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
 /*
- * Crackerjack Project
- *
- * Copyright (C) 2007-2008, Hitachi, Ltd.
- * Author(s): Takahiro Yasui <takahiro.yasui.mp@hitachi.com>,
- *            Yumiko Sugita <yumiko.sugita.yf@hitachi.com>,
- *            Satoshi Fujiwara <sa-fuji@sdl.hitachi.co.jp>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
- *
- * $Id: gettid01.c,v 1.5 2009/10/26 14:55:47 subrata_modak Exp $
+ * Copyright (C) 2023 SUSE LLC Andrea Cervesato <andrea.cervesato@suse.com>
+ */
+/*\
+ * [Description]
  *
+ * This test checks if parent pid is equal to tid in single-threaded
+ * application.
  */
 
-/* Porting from Crackerjack to LTP is done
-   by Masatake YAMATO <yamato@redhat.com> */
-
-#include <sys/types.h>
-#include <linux/unistd.h>
-#include <errno.h>
-
-#include "test.h"
-
-void setup();
-void cleanup();
-
-char *TCID = "gettid01";
-
-int TST_TOTAL = 1;
-
-pid_t my_gettid(void)
-{
-	return (pid_t) syscall(__NR_gettid);
-}
-
-int main(int ac, char **av)
-{
-	int lc;
-
-	tst_parse_opts(ac, av, NULL, NULL);
-
-	setup();
-
-	/*
-	 * The following loop checks looping state if -c option given
-	 */
-	for (lc = 0; TEST_LOOPING(lc); lc++) {
+#include "tst_test.h"
+#include "lapi/syscalls.h"
 
-		tst_count = 0;
-
-		TEST(my_gettid());
-
-		if (TEST_RETURN == -1) {
-			tst_resm(TFAIL, "gettid() Failed, errno=%d: %s",
-				 TEST_ERRNO, strerror(TEST_ERRNO));
-		} else {
-			tst_resm(TPASS, "gettid() returned %ld",
-				 TEST_RETURN);
-		}
-	}
-
-	cleanup();
-	tst_exit();
-}
-
-/*
- * setup() - performs all ONE TIME setup for this test.
- */
-void setup(void)
+static void run(void)
 {
+	long pid, tid;
 
-	tst_sig(NOFORK, DEF_HANDLER, cleanup);
+	SAFE_FILE_LINES_SCANF("/proc/self/status", "Pid: %ld", &pid);
+	SAFE_FILE_LINES_SCANF("/proc/self/status", "Tgid: %ld", &tid);
 
-	TEST_PAUSE;
+	if (pid != tid)
+		tst_brk(TBROK, "Test function has been moved inside a thread?");
 
+	TST_EXP_EQ_LI(tst_syscall(__NR_gettid), tst_syscall(__NR_getpid));
+	TST_EXP_EQ_LI(tst_syscall(__NR_gettid), pid);
 }
 
-/*
- * cleanup() - performs all ONE TIME cleanup for this test at
- *		completion or premature exit.
- */
-void cleanup(void)
-{
-}
+static struct tst_test test = {
+	.test_all = run,
+};
-- 
2.35.3


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

[LTP] [PATCH v2 2/2] Add gettid02 test

[Andrea Cervesato]

From: Andrea Cervesato <andrea.cervesato@suse.com>

This new test is checking if __NR_gettid syscall is properly handling a
multi-threaded application, by assigning a different TID for each
thread.

Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
---
Make use of tst_syscall

 runtest/syscalls                            |  1 +
 testcases/kernel/syscalls/gettid/.gitignore |  1 +
 testcases/kernel/syscalls/gettid/Makefile   |  4 +-
 testcases/kernel/syscalls/gettid/gettid02.c | 68 +++++++++++++++++++++
 4 files changed, 73 insertions(+), 1 deletion(-)
 create mode 100644 testcases/kernel/syscalls/gettid/gettid02.c

diff --git a/runtest/syscalls b/runtest/syscalls
index 8652e0bd3..bba266ea3 100644
--- a/runtest/syscalls
+++ b/runtest/syscalls
@@ -538,6 +538,7 @@ getsockopt01 getsockopt01
 getsockopt02 getsockopt02
 
 gettid01 gettid01
+gettid02 gettid02
 
 gettimeofday01 gettimeofday01
 gettimeofday02 gettimeofday02
diff --git a/testcases/kernel/syscalls/gettid/.gitignore b/testcases/kernel/syscalls/gettid/.gitignore
index 78dce3499..9014f7c3a 100644
--- a/testcases/kernel/syscalls/gettid/.gitignore
+++ b/testcases/kernel/syscalls/gettid/.gitignore
@@ -1 +1,2 @@
 /gettid01
+/gettid02
diff --git a/testcases/kernel/syscalls/gettid/Makefile b/testcases/kernel/syscalls/gettid/Makefile
index 4e9982f76..5345eb0f5 100644
--- a/testcases/kernel/syscalls/gettid/Makefile
+++ b/testcases/kernel/syscalls/gettid/Makefile
@@ -10,7 +10,9 @@ top_srcdir		?= ../../../..
 include $(top_srcdir)/include/mk/testcases.mk
 
 ifeq ($(ANDROID), 1)
-FILTER_OUT_MAKE_TARGETS	+= gettid01
+FILTER_OUT_MAKE_TARGETS	+= gettid01 gettid02
 endif
 
+gettid02: LDLIBS += -lpthread
+
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
diff --git a/testcases/kernel/syscalls/gettid/gettid02.c b/testcases/kernel/syscalls/gettid/gettid02.c
new file mode 100644
index 000000000..22e843bad
--- /dev/null
+++ b/testcases/kernel/syscalls/gettid/gettid02.c
@@ -0,0 +1,68 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C) 2023 SUSE LLC Andrea Cervesato <andrea.cervesato@suse.com>
+ */
+/*\
+ * [Description]
+ *
+ * This test spawns multiple threads, then check for each one of them if the
+ * parent ID is different AND if the thread ID is different from all the other
+ * spwaned threads.
+ */
+
+#include "tst_test.h"
+#include "lapi/syscalls.h"
+#include "tst_safe_pthread.h"
+
+#define THREADS_NUM 10
+
+static pid_t tids[THREADS_NUM];
+
+static void *threaded(void *arg)
+{
+	int i = *(int *)arg;
+	pid_t pid, tid;
+
+	pid = tst_syscall(__NR_getpid);
+	tid = tst_syscall(__NR_gettid);
+
+	TST_EXP_EXPR(pid != tid,
+		"parent ID (%d) differs from thread[%d] ID (%d)",
+		pid, i, tid);
+
+	return (void *)tst_syscall(__NR_gettid);
+}
+
+static void run(void)
+{
+	pthread_t thread;
+	int error = 0;
+
+	for (int i = 0; i < THREADS_NUM; i++) {
+		SAFE_PTHREAD_CREATE(&thread, NULL, threaded, &i);
+		SAFE_PTHREAD_JOIN(thread, (void **)&tids[i]);
+	}
+
+	for (int i = 0; i < THREADS_NUM; i++) {
+		for (int j = 0; j < THREADS_NUM; j++) {
+			if (i == j)
+				continue;
+
+			if (tids[i] == tids[j]) {
+				tst_res(TINFO, "thread[%d] and thread[%d] have the same ID", i, j);
+				error = 1;
+				goto end;
+			}
+		}
+	}
+
+end:
+	if (error)
+		tst_res(TFAIL, "Some threads have the same TID");
+	else
+		tst_res(TPASS, "All threads have a different TID");
+}
+
+static struct tst_test test = {
+	.test_all = run,
+};
-- 
2.35.3


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 0/2] Rewrite the gettid() testing suite

[Marius Kittler]

Am Donnerstag, 5. Oktober 2023, 15:49:09 CEST schrieb Andrea Cervesato:
> From: Andrea Cervesato <andrea.cervesato@suse.com>
> 
> Rewrite gettid01 and create a new gettid02 test in order to check all
> gettid() basic functionalities, according with its manual.
> 
> Andrea Cervesato (2):
>   Rewrite gettid01 test
>   Add gettid02 test
> 
>  runtest/syscalls                            |   1 +
>  testcases/kernel/syscalls/gettid/.gitignore |   1 +
>  testcases/kernel/syscalls/gettid/Makefile   |   4 +-
>  testcases/kernel/syscalls/gettid/gettid01.c | 105 ++++----------------
>  testcases/kernel/syscalls/gettid/gettid02.c |  68 +++++++++++++
>  5 files changed, 93 insertions(+), 86 deletions(-)
>  create mode 100644 testcases/kernel/syscalls/gettid/gettid02.c

Looks generally good to merge.

I'm only wondering whether the goto in the 2nd test could be avoided. After 
all it might even be a benefit to get an info for all the wrongly equivalent 
thread IDs (and not just the first).

Then it would also make sense to use `int j = i + 1;` in the inner loop to 
avoid duplicate/redundant commutative comparisons. And then you could actually 
also drop the `i == j` check.




-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH v2 0/2] Rewrite the gettid() testing suite

[Andrea Cervesato via ltp]

Hi!

On 10/5/23 16:24, Marius Kittler wrote:
> Am Donnerstag, 5. Oktober 2023, 15:49:09 CEST schrieb Andrea Cervesato:
>> From: Andrea Cervesato <andrea.cervesato@suse.com>
>>
>> Rewrite gettid01 and create a new gettid02 test in order to check all
>> gettid() basic functionalities, according with its manual.
>>
>> Andrea Cervesato (2):
>>    Rewrite gettid01 test
>>    Add gettid02 test
>>
>>   runtest/syscalls                            |   1 +
>>   testcases/kernel/syscalls/gettid/.gitignore |   1 +
>>   testcases/kernel/syscalls/gettid/Makefile   |   4 +-
>>   testcases/kernel/syscalls/gettid/gettid01.c | 105 ++++----------------
>>   testcases/kernel/syscalls/gettid/gettid02.c |  68 +++++++++++++
>>   5 files changed, 93 insertions(+), 86 deletions(-)
>>   create mode 100644 testcases/kernel/syscalls/gettid/gettid02.c
> Looks generally good to merge.
>
> I'm only wondering whether the goto in the 2nd test could be avoided. After
> all it might even be a benefit to get an info for all the wrongly equivalent
> thread IDs (and not just the first).
>
> Then it would also make sense to use `int j = i + 1;` in the inner loop to
> avoid duplicate/redundant commutative comparisons. And then you could actually
> also drop the `i == j` check.
>
>
>
>
It makes sense, I update the code.

Andrea


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp