[LTP] [PATCH v1 0/2] Rewrite the gettid() testing suite

[LTP] [PATCH v1 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 | 102 +++-----------------
 testcases/kernel/syscalls/gettid/gettid02.c |  68 +++++++++++++
 5 files changed, 87 insertions(+), 89 deletions(-)
 create mode 100644 testcases/kernel/syscalls/gettid/gettid02.c

-- 
2.35.3


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

[LTP] [PATCH v1 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
gettid() syscall was working. In this test we checks if gettid() is
working and if PID differs from its return value.

Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
---
 testcases/kernel/syscalls/gettid/gettid01.c | 102 +++-----------------
 1 file changed, 14 insertions(+), 88 deletions(-)

diff --git a/testcases/kernel/syscalls/gettid/gettid01.c b/testcases/kernel/syscalls/gettid/gettid01.c
index 7e5b6b175..3ee139d5f 100644
--- a/testcases/kernel/syscalls/gettid/gettid01.c
+++ b/testcases/kernel/syscalls/gettid/gettid01.c
@@ -1,96 +1,22 @@
+// 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();
+#include "tst_test.h"
+#include "lapi/syscalls.h"
 
-char *TCID = "gettid01";
-
-int TST_TOTAL = 1;
-
-pid_t my_gettid(void)
+static void run(void)
 {
-	return (pid_t) syscall(__NR_gettid);
+	TST_EXP_EQ_LI(getpid(), tst_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++) {
-
-		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)
-{
-
-	tst_sig(NOFORK, DEF_HANDLER, cleanup);
-
-	TEST_PAUSE;
-
-}
-
-/*
- * 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 v1 2/2] Add gettid02 test

[Andrea Cervesato]

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

This new test is checking if gettid() is properly handling a
multi-threaded application, by assigning a different TID for each thread
which differs from the parent ID.

Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
---
 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 4fb76584f..d50f5a3e9 100644
--- a/runtest/syscalls
+++ b/runtest/syscalls
@@ -537,6 +537,7 @@ 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..c981d2b79
--- /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 = 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 v1 1/2] Rewrite gettid01 test

[Richard Palethorpe]

Hello,

Andrea Cervesato <andrea.cervesato@suse.de> writes:

> From: Andrea Cervesato <andrea.cervesato@suse.com>
>
> The old test wasn't doing anything meaningful, but just checking
> gettid() syscall was working. In this test we checks if gettid() is
> working and if PID differs from its return value.
>
> Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
> ---
>  testcases/kernel/syscalls/gettid/gettid01.c | 102 +++-----------------
>  1 file changed, 14 insertions(+), 88 deletions(-)
>
> diff --git a/testcases/kernel/syscalls/gettid/gettid01.c b/testcases/kernel/syscalls/gettid/gettid01.c
> index 7e5b6b175..3ee139d5f 100644
> --- a/testcases/kernel/syscalls/gettid/gettid01.c
> +++ b/testcases/kernel/syscalls/gettid/gettid01.c
> @@ -1,96 +1,22 @@
> +// 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();
> +#include "tst_test.h"
> +#include "lapi/syscalls.h"
>  
> -char *TCID = "gettid01";
> -
> -int TST_TOTAL = 1;
> -
> -pid_t my_gettid(void)
> +static void run(void)
>  {
> -	return (pid_t) syscall(__NR_gettid);
> +	TST_EXP_EQ_LI(getpid(), tst_syscall(__NR_gettid));

Perhaps this is nit picking, but this assumes libc didn't put us in a
thread or just caches the wrong value in getpid. So it is more a test of
libc than the kernel.

Is there some other way we could check that the main test process is not
a thread? There could be some file in proc I guess. Then we are
comparing information from multiple sources and it should all align.

Also getpid could be called directly avoiding libc.

>  }
>  
> -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++) {
> -
> -		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)
> -{
> -
> -	tst_sig(NOFORK, DEF_HANDLER, cleanup);
> -
> -	TEST_PAUSE;
> -
> -}
> -
> -/*
> - * 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


-- 
Thank you,
Richard.

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

Re: [LTP] [PATCH v1 2/2] Add gettid02 test

[Richard Palethorpe]

Hello,

Andrea Cervesato <andrea.cervesato@suse.de> writes:

> From: Andrea Cervesato <andrea.cervesato@suse.com>
>
> This new test is checking if gettid() is properly handling a
> multi-threaded application, by assigning a different TID for each thread
> which differs from the parent ID.
>
> Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
> ---
>  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 4fb76584f..d50f5a3e9 100644
> --- a/runtest/syscalls
> +++ b/runtest/syscalls
> @@ -537,6 +537,7 @@ 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..c981d2b79
> --- /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 = getpid();

Maybe we could call __NR_getpid directly as well and compare it?

Same reason as with other test; libc's like to be clever and cache this
value. So maybe we are not testing the kernel here.

Otherwise LGTM.

-- 
Thank you,
Richard.

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

Re: [LTP] [PATCH v1 1/2] Rewrite gettid01 test

[Andrea Cervesato via ltp]

Hi!

On 10/5/23 10:50, Richard Palethorpe wrote:
> Hello,
>
> Andrea Cervesato <andrea.cervesato@suse.de> writes:
>
>> From: Andrea Cervesato <andrea.cervesato@suse.com>
>>
>> The old test wasn't doing anything meaningful, but just checking
>> gettid() syscall was working. In this test we checks if gettid() is
>> working and if PID differs from its return value.
>>
>> Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
>> ---
>>   testcases/kernel/syscalls/gettid/gettid01.c | 102 +++-----------------
>>   1 file changed, 14 insertions(+), 88 deletions(-)
>>
>> diff --git a/testcases/kernel/syscalls/gettid/gettid01.c b/testcases/kernel/syscalls/gettid/gettid01.c
>> index 7e5b6b175..3ee139d5f 100644
>> --- a/testcases/kernel/syscalls/gettid/gettid01.c
>> +++ b/testcases/kernel/syscalls/gettid/gettid01.c
>> @@ -1,96 +1,22 @@
>> +// 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();
>> +#include "tst_test.h"
>> +#include "lapi/syscalls.h"
>>   
>> -char *TCID = "gettid01";
>> -
>> -int TST_TOTAL = 1;
>> -
>> -pid_t my_gettid(void)
>> +static void run(void)
>>   {
>> -	return (pid_t) syscall(__NR_gettid);
>> +	TST_EXP_EQ_LI(getpid(), tst_syscall(__NR_gettid));
> Perhaps this is nit picking, but this assumes libc didn't put us in a
> thread or just caches the wrong value in getpid. So it is more a test of
> libc than the kernel.
I understand the cache part. Probably we need to call 
tst_syscall(__NR_getpid) instead of getpid(), but I don't get the thread 
part. Is it possible that libc moves test function inside a thread?
> Is there some other way we could check that the main test process is not
> a thread? There could be some file in proc I guess. Then we are
> comparing information from multiple sources and it should all align.
>
> Also getpid could be called directly avoiding libc.
>
>>   }
>>   
>> -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++) {
>> -
>> -		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)
>> -{
>> -
>> -	tst_sig(NOFORK, DEF_HANDLER, cleanup);
>> -
>> -	TEST_PAUSE;
>> -
>> -}
>> -
>> -/*
>> - * 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
>
Andrea


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

Re: [LTP] [PATCH v1 1/2] Rewrite gettid01 test

[Richard Palethorpe]

Hello,

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

> Hi!
>
> On 10/5/23 10:50, Richard Palethorpe wrote:
>> Hello,
>>
>> Andrea Cervesato <andrea.cervesato@suse.de> writes:
>>
>>> From: Andrea Cervesato <andrea.cervesato@suse.com>
>>>
>>> The old test wasn't doing anything meaningful, but just checking
>>> gettid() syscall was working. In this test we checks if gettid() is
>>> working and if PID differs from its return value.
>>>
>>> Signed-off-by: Andrea Cervesato <andrea.cervesato@suse.com>
>>> ---
>>>   testcases/kernel/syscalls/gettid/gettid01.c | 102 +++-----------------
>>>   1 file changed, 14 insertions(+), 88 deletions(-)
>>>
>>> diff --git a/testcases/kernel/syscalls/gettid/gettid01.c b/testcases/kernel/syscalls/gettid/gettid01.c
>>> index 7e5b6b175..3ee139d5f 100644
>>> --- a/testcases/kernel/syscalls/gettid/gettid01.c
>>> +++ b/testcases/kernel/syscalls/gettid/gettid01.c
>>> @@ -1,96 +1,22 @@
>>> +// 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();
>>> +#include "tst_test.h"
>>> +#include "lapi/syscalls.h"
>>>   -char *TCID = "gettid01";
>>> -
>>> -int TST_TOTAL = 1;
>>> -
>>> -pid_t my_gettid(void)
>>> +static void run(void)
>>>   {
>>> -	return (pid_t) syscall(__NR_gettid);
>>> +	TST_EXP_EQ_LI(getpid(), tst_syscall(__NR_gettid));
>> Perhaps this is nit picking, but this assumes libc didn't put us in a
>> thread or just caches the wrong value in getpid. So it is more a test of
>> libc than the kernel.
> I understand the cache part. Probably we need to call
> tst_syscall(__NR_getpid) instead of getpid(), but I don't get the
> thread part. Is it possible that libc moves test function inside a
> thread?

I suppose it is possible, but very unlikely. I haven't seen anything
like that; I'm just using my wild imagination.

Perhaps we could check 'grep Threads /proc/self/status'? It should be
equal to 1. If not then something may have instrumented the process or
the system's libc is insane or something else. It's just some extra
debug info.

>> Is there some other way we could check that the main test process is not
>> a thread? There could be some file in proc I guess. Then we are
>> comparing information from multiple sources and it should all align.
>>
>> Also getpid could be called directly avoiding libc.
>>
>>>   }
>>>   -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++) {
>>> -
>>> -		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)
>>> -{
>>> -
>>> -	tst_sig(NOFORK, DEF_HANDLER, cleanup);
>>> -
>>> -	TEST_PAUSE;
>>> -
>>> -}
>>> -
>>> -/*
>>> - * 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
>>
> Andrea


-- 
Thank you,
Richard.

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