Skip to contentSkip to conversation

syscalls/kill01: Remove it

PATCH New Feiyu Zhu 4 participants 14 messages Aug 20, 2020 – Aug 21, 2020

lore
Share & tools
lore

Feiyu Zhu

author
#0 lore raw

PATCH 1/4 syscalls/kill01: Remove it

kill11 has covered this test, so remove it.

---
 runtest/syscalls                          |   1 -
 testcases/kernel/syscalls/kill/.gitignore |   1 -
 testcases/kernel/syscalls/kill/kill01.c   | 161 ------------------------------
 3 files changed, 163 deletions(-)
 delete mode 100644 testcases/kernel/syscalls/kill/kill01.c
runtest/syscalls +0 -1
index 860c5c3..5e2bdb1 100644
@@ -614,7 +614,6 @@ kcmp01 kcmp01
 kcmp02 kcmp02
 kcmp03 kcmp03
 
-kill01 kill01
 kill02 kill02
 kill03 kill03
 kill04 kill04
testcases/kernel/syscalls/kill/.gitignore +0 -1
index 490a1e2..b62662f 100644
@@ -1,4 +1,3 @@
-/kill01
 /kill02
 /kill03
 /kill04
testcases/kernel/syscalls/kill/kill01.c deleted +0 -161

161 lines not on this page · raw message

Signature · 1 line
1.8.3.1
author
#1 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

1) Remove uclinux code
2) Take use of safe macro
3) Merge kill04 into kill03
4) Add error test that pid is INT_MIN

---
 runtest/syscalls                          |   1 -
 testcases/kernel/syscalls/kill/.gitignore |   1 -
 testcases/kernel/syscalls/kill/kill03.c   | 203 +++++++++---------------------
 testcases/kernel/syscalls/kill/kill04.c   | 133 --------------------
 4 files changed, 59 insertions(+), 279 deletions(-)
 delete mode 100644 testcases/kernel/syscalls/kill/kill04.c
runtest/syscalls +0 -1
index 5e2bdb1..a6ab75b 100644
@@ -616,7 +616,6 @@ kcmp03 kcmp03
 
 kill02 kill02
 kill03 kill03
-kill04 kill04
 kill05 kill05
 kill06 kill06
 kill07 kill07
testcases/kernel/syscalls/kill/.gitignore +0 -1
index b62662f..75fdaa5 100644
@@ -1,6 +1,5 @@
 /kill02
 /kill03
-/kill04
 /kill05
 /kill06
 /kill07
testcases/kernel/syscalls/kill/kill03.c +59 -144

224 lines not on this page · raw message

testcases/kernel/syscalls/kill/kill04.c deleted +0 -133

133 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#2 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Feiyu Zhu <[email protected]> wrote:

...

testcases/kernel/syscalls/kill/kill03.c line 24

Quoted 11 lines
testcases/kernel/syscalls/kill/kill03.c +10 -0
+#include <unistd.h>
+#include "tst_test.h"
+
+static pid_t real_pid, fake_pid, int_min_pid;
+static void cleanup(void);
+
+static struct tcase {
+       int test_sig;
+       int exp_errno;
+       int child_flag;

The child_flag field is not necessary, we could prepare a real child in setup()
and reclaim it after testing via check the real_pid value, that will be more easily.

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200820/2115494a/attachment.htm>
#3 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Hi Li

Quoted 16 lines

Feiyu Zhu <[email protected] <mailto:[email protected]>> wrote:

    ...
    +#include <unistd.h>
    +#include "tst_test.h"
    +
    +static pid_t real_pid, fake_pid, int_min_pid;
    +static void cleanup(void);
    +
    +static struct tcase {
    +? ? ? ?int test_sig;
    +? ? ? ?int exp_errno;
    +? ? ? ?int child_flag;
Quoted 4 lines

The child_flag field is not necessary, we could prepare a real child in setup()
and?reclaim it after testing via check the real_pid value, that will be more easily.

When I reviewed this patch in internal, I had the same idea. But when I try it and this case will hang because sub test will wait children finished by using tst_reap_childrens below:

lib/tst_test.c
static void run_tests(void)

{
	...
	 for (i = 0; i < tst_test->tcnt; i++) {
                 saved_results = *results;
                 tst_test->test(i);

                 if (getpid() != main_pid) {
                         exit(0);
                 }

                 tst_reap_children();

                 if (results_equal(&saved_results, results))
                         tst_brk(TBROK, "Test %i haven't reported 
results!", i);
         }

}

Also, we can use the current process id but it may has unexpected result when kill succeed. So fork a child to test maybe a better solution.

Quoted 6 lines
Signature · 3 lines
-- 
Regards,
Li Wang
#4 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Quoted 48 lines · Yang Xu

Hi Li

Quoted 2 lines

Feiyu Zhu <[email protected] <mailto:[email protected]>>

Quoted 14 lines · wrote:
    ...
    +#include <unistd.h>
    +#include "tst_test.h"
    +
    +static pid_t real_pid, fake_pid, int_min_pid;
    +static void cleanup(void);
    +
    +static struct tcase {
    +       int test_sig;
    +       int exp_errno;
    +       int child_flag;
Quoted 4 lines

The child_flag field is not necessary, we could prepare a real child in setup()
and reclaim it after testing via check the real_pid value, that will be more easily.

When I reviewed this patch in internal, I had the same idea. But when I try it and this case will hang because sub test will wait children finished by using tst_reap_childrens below:

lib/tst_test.c
static void run_tests(void)

{
        ...
         for (i = 0; i < tst_test->tcnt; i++) {
                 saved_results = *results;
                 tst_test->test(i);

                 if (getpid() != main_pid) {
                         exit(0);
                 }

                 tst_reap_children();

                 if (results_equal(&saved_results, results))
                         tst_brk(TBROK, "Test %i haven't reported
results!", i);
         }
Quoted 5 lines

}

Also, we can use the current process id but it may has unexpected result when kill succeed. So fork a child to test maybe a better solution.

Hmm, sorry for the uncleared description, actually I meant, to use real_pid instead of the tc->child_flag directly, then start to reclaim the child when the
real_pid test finishing.

Does this below diff work for you?

testcases/kernel/syscalls/kill/kill03.c +3 -10
@@ -21,24 +21,17 @@ static void cleanup(void);
 static struct tcase {
        int test_sig;
        int exp_errno;
-       int child_flag;
        pid_t *pid;
 } tcases[] = {
-       {2000, EINVAL, 1, &real_pid},
-       {SIGKILL, ESRCH, 0, &fake_pid},
-       {SIGKILL, ESRCH, 0, &int_min_pid}
+       {2000, EINVAL, &real_pid},
+       {SIGKILL, ESRCH, &fake_pid},
+       {SIGKILL, ESRCH, &int_min_pid}
 };
 
 static void verify_kill(unsigned int n)
 {
        struct tcase *tc = &tcases[n];
 
-       if (tc->child_flag) {
-               real_pid = SAFE_FORK();
-               if (!real_pid)
-                       pause();
-       }
-
        TEST(kill(*tc->pid, tc->test_sig));
        if (TST_RET != -1) {
                tst_res(TFAIL, "kill should fail but not, return %ld",

TST_RET);
@@ -51,14 +44,19 @@ static void verify_kill(unsigned int n)
tst_res(TFAIL | TTERRNO, "kill expected %s but got",
tst_strerrno(tc->exp_errno));

- if (tc->child_flag) {

+       if (real_pid) {
                cleanup();
                real_pid = 0;
        }

+
}

static void setup(void)
{
+ real_pid = SAFE_FORK();
+ if (!real_pid)
+ pause();
+

        fake_pid = tst_get_unused_pid();
        int_min_pid = INT_MIN;

}

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200821/9d88dd10/attachment-0001.htm>
#5 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Hi Li

Quoted 63 lines

On Fri, Aug 21, 2020 at 9:18 AM Yang Xu <[email protected] <mailto:[email protected]>> wrote:

    Hi Li


     >
     > Feiyu Zhu <[email protected]
    <mailto:[email protected]> <mailto:[email protected]
    <mailto:[email protected]>>> wrote:
     >
     >? ? ?...
     >? ? ?+#include <unistd.h>
     >? ? ?+#include "tst_test.h"
     >? ? ?+
     >? ? ?+static pid_t real_pid, fake_pid, int_min_pid;
     >? ? ?+static void cleanup(void);
     >? ? ?+
     >? ? ?+static struct tcase {
     >? ? ?+? ? ? ?int test_sig;
     >? ? ?+? ? ? ?int exp_errno;
     >? ? ?+? ? ? ?int child_flag;
     >
     >
     > The child_flag field is not necessary, we could prepare a real
    child in
     > setup()
     > and?reclaim it after testing via check the real_pid value, that
    will be
     > more easily.
    When I reviewed this patch in internal, I had the same idea. But when I
    try it and this case will hang because sub test will wait children
    finished by using tst_reap_childrens below:

    lib/tst_test.c
    static void run_tests(void)
    {
     ? ? ? ? ...
     ? ? ? ? ?for (i = 0; i < tst_test->tcnt; i++) {
     ? ? ? ? ? ? ? ? ?saved_results = *results;
     ? ? ? ? ? ? ? ? ?tst_test->test(i);

     ? ? ? ? ? ? ? ? ?if (getpid() != main_pid) {
     ? ? ? ? ? ? ? ? ? ? ? ? ?exit(0);
     ? ? ? ? ? ? ? ? ?}

     ? ? ? ? ? ? ? ? ?tst_reap_children();

     ? ? ? ? ? ? ? ? ?if (results_equal(&saved_results, results))
     ? ? ? ? ? ? ? ? ? ? ? ? ?tst_brk(TBROK, "Test %i haven't reported
    results!", i);
     ? ? ? ? ?}


    }

    Also, we can use the current process id but it may has unexpected
    result
    when kill succeed. So fork a child to test maybe a better solution.
Quoted 6 lines

Hmm, sorry for the uncleared description, actually I meant, to use real_pid instead of the?tc->child_flag directly, then start to reclaim the child when the
real_pid test finishing.

Does this below diff work for you?

It looks well. But the real_pid only valid when the first sub test and the real pid is equals to 0 when using -i parameters because we have killed this children.

For pid = 0, it means then sig is sent to every process in the process group of the calling process. So it looks like we test different thing when using -i parameters. What do you think about this?

Quoted 56 lines
testcases/kernel/syscalls/kill/kill03.c +3 -10
@@ -21,24 +21,17 @@ static void cleanup(void);
 ?static struct tcase {
 ? ? ? ? int test_sig;
 ? ? ? ? int exp_errno;
- ? ? ? int child_flag;
 ? ? ? ? pid_t *pid;
 ?} tcases[] = {
- ? ? ? {2000, EINVAL, 1, &real_pid},
- ? ? ? {SIGKILL, ESRCH, 0, &fake_pid},
- ? ? ? {SIGKILL, ESRCH, 0, &int_min_pid}
+ ? ? ? {2000, EINVAL, &real_pid},
+ ? ? ? {SIGKILL, ESRCH, &fake_pid},
+ ? ? ? {SIGKILL, ESRCH, &int_min_pid}
 ?};
 
 ?static void verify_kill(unsigned int n)
 ?{
 ? ? ? ? struct tcase *tc = &tcases[n];
 
- ? ? ? if (tc->child_flag) {
- ? ? ? ? ? ? ? real_pid = SAFE_FORK();
- ? ? ? ? ? ? ? if (!real_pid)
- ? ? ? ? ? ? ? ? ? ? ? pause();
- ? ? ? }
-
 ? ? ? ? TEST(kill(*tc->pid, tc->test_sig));
 ? ? ? ? if (TST_RET != -1) {
 ? ? ? ? ? ? ? ? tst_res(TFAIL, "kill should fail but not, return %ld", 

TST_RET);

testcases/kernel/syscalls/kill/kill03.c +6 -1
@@ -51,14 +44,19 @@ static void verify_kill(unsigned int n)
 ? ? ? ? ? ? ? ? tst_res(TFAIL | TTERRNO, "kill expected %s but got",
 ? ? ? ? ? ? ? ? ? ? ? ? tst_strerrno(tc->exp_errno));
 
- ? ? ? if (tc->child_flag) {
+ ? ? ? if (real_pid) {
 ? ? ? ? ? ? ? ? cleanup();
 ? ? ? ? ? ? ? ? real_pid = 0;
 ? ? ? ? }
+
 ?}
 
 ?static void setup(void)
 ?{
+ ? ? ? real_pid = SAFE_FORK();
+ ? ? ? if (!real_pid)
+ ? ? ? ? ? ? ? pause();
+
 ? ? ? ? fake_pid = tst_get_unused_pid();
 ? ? ? ? int_min_pid = INT_MIN;
 ?}
Signature · 2 lines
Regards,
Li Wang
#6 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Quoted 8 lines · Yang Xu

...

Quoted 6 lines

Also, we can use the current process id but it may has unexpected result
when kill succeed. So fork a child to test maybe a better solution.

Hmm, sorry for the uncleared description, actually I meant, to use

real_pid

Quoted 9 lines
Quoted 5 lines

instead of the tc->child_flag directly, then start to reclaim the child when the
real_pid test finishing.

Does this below diff work for you?

It looks well. But the real_pid only valid when the first sub test and the real pid is equals to 0 when using -i parameters because we have killed this children.

Yes, you're right.
How about moving up the real_pid creator to the main process? does it work for you?

static void verify_kill(unsigned int n)

{
        if (!real_pid) {
                real_pid = SAFE_FORK();
                if (!real_pid)
                        pause();
        }

        TEST(kill(*tc->pid, tc->test_sig));
        ...

        if (real_pid) {
                cleanup();
                real_pid = 0;
        }

}

Quoted 5 lines

For pid = 0, it means then sig is sent to every process in the process group of the calling process. So it looks like we test different thing when using -i parameters. What do you think about this?

I even think it is a good idea for code simplification:). In this case, the first
the subtest is just to verify invalid signal for kill(), it doesn't matter to go with
kill(0, invalid_signal), that only tries to kill the current main process. isn't it?

If we go this way, not only the tc->child_flag is no needed, but also not necessary
to fork a new child to be killed.

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200821/57245fb5/attachment-0001.htm>
#7 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Hi Li

Quoted 40 lines

On Fri, Aug 21, 2020 at 10:22 AM Yang Xu <[email protected] <mailto:[email protected]>> wrote:

...
>? ? ?Also, we can use the current process id but it may has unexpected >? ? ?result
>? ? ?when kill succeed. So fork a child to test maybe a better
solution.
>
>
> Hmm, sorry for the uncleared description, actually I meant, to
use real_pid
> instead of the?tc->child_flag directly, then start to reclaim the
child
> when the
> real_pid test finishing.
>
> Does this below diff work for you?
It looks well. But the real_pid only valid when the first sub test and the real pid is equals to 0 when using -i parameters because we have killed this children.

Yes, you're right.
How about moving up the real_pid creator to the main process? does it work for you?

static void verify_kill(unsigned int n)
{
? ? ? ? if (!real_pid) {
? ? ? ? ? ? ? ? real_pid = SAFE_FORK();
? ? ? ? ? ? ? ? if (!real_pid)
? ? ? ? ? ? ? ? ? ? ? ? pause();
? ? ? ? }

? ? ? ? TEST(kill(*tc->pid, tc->test_sig));
...

Quoted 7 lines

? ? ? ? if (real_pid) {
? ? ? ? ? ? ? ? cleanup();
? ? ? ? ? ? ? ? real_pid = 0;
? ? ? ? }

}

Yes, it looks ok,

Quoted 6 lines

For pid = 0, it means? then sig is sent to every process in the process group of the calling process.? So it looks like we test different thing when using -i parameters.? What do you think about this?

Quoted 6 lines

I even think it is a good idea for code simplification:). In this case, the first
the subtest is just to verify invalid signal for kill(), it doesn't matter to go with
kill(0, invalid_signal), that only tries to kill the current main process. isn't it?

Yes. I think using current process id instead of 0 is more easier because the invalid signal can't kill process forever.

static void setup(void)

{
         real_pid = getpid();
         fake_pid = tst_get_unused_pid();
         int_min_pid = INT_MIN;
}
Quoted 8 lines

If we go this way, not only the?tc->child_flag is no needed, but also not necessary
to fork a new child to be killed.

Signature · 2 lines
Regards,
Li Wang
#8 lore raw

PATCH 2/4 syscalls/kill03, 04: Cleanup && Convert to new library

Quoted 8 lines · Yang Xu

...
Yes. I think using current process id instead of 0 is more easier because the invalid signal can't kill process forever.

static void setup(void)

{
         real_pid = getpid();

Agreed!

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200821/1cb95f23/attachment.htm>
author
#9 lore raw

PATCH 3/4 syscalls/kill05:Cleanup && Convert to new library

In reply to Feiyu Zhu

1) Take use of some safe macros

---
 testcases/kernel/syscalls/kill/Makefile |   4 +-
 testcases/kernel/syscalls/kill/kill05.c | 243 ++++++++------------------------
 2 files changed, 58 insertions(+), 189 deletions(-)
testcases/kernel/syscalls/kill/Makefile +2 -2
index 3d1b146..0cc064b 100644
@@ -3,11 +3,11 @@
 
 top_srcdir		?= ../../../..
 
-LTPLIBS = ltpipc
+LTPLIBS = ltpipc ltpnewipc
 
 include $(top_srcdir)/include/mk/testcases.mk
 
 kill07: LTPLDLIBS  = -lltpipc
-kill05: LTPLDLIBS  = -lltpipc
+kill05: LTPLDLIBS  = -lltpnewipc
 
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
testcases/kernel/syscalls/kill/kill05.c +56 -187

286 lines not on this page · raw message

Signature · 1 line
1.8.3.1
#10 lore raw

PATCH 3/4 syscalls/kill05:Cleanup && Convert to new library

Quoted 3 lines · Feiyu Zhu

...
* Looping with the -i option does not work correctly.

This line should be removed since it works well with -i.

Quoted 5 lines

+static void wait_for_flag(int value)

 {
        while (1) {
                if (*flag == value)

Sleep 1sec is too long here, we can use usleep(100) for a replacement.

+static struct tst_test test = {

testcases/kernel/syscalls/kill/kill05.c void wait_for_flag(int value) · line 108

Quoted 5 lines
testcases/kernel/syscalls/kill/kill05.c +4 -0
+       .setup = setup,
+       .cleanup = cleanup,
+       .test_all = verify_kill,
+       .needs_tmpdir = 1,

'.needs_tmdir' is redundant.

testcases/kernel/syscalls/kill/kill05.c void wait_for_flag(int value) · line 111

Quoted 4 lines
testcases/kernel/syscalls/kill/kill05.c +3 -0
+       .needs_root = 1,
+       .forks_child = 1,
+};

The remaining part looks good.

Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200821/d8fb8c0e/attachment.htm>
author
#11 lore raw

PATCH 4/4 syscalls/kill06:Cleanup && Convert to new library

In reply to Feiyu Zhu

1) Take use of some safe macros

---
 testcases/kernel/syscalls/kill/kill06.c | 208 ++++++--------------------------
 1 file changed, 37 insertions(+), 171 deletions(-)
testcases/kernel/syscalls/kill/kill06.c +37 -171
index 7a1e8bf..8fcc8bc 100644
@@ -1,187 +1,53 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
 /*
+ * Copyright (c) International Business Machines  Corp., 2001
  *
- *   Copyright (c) International Business Machines  Corp., 2001
- *
- *   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
- */
-
-/*
- * NAME
- *	kill06.c
- *
- * DESCRIPTION
- *	Test case to check the basic functionality of kill() when killing an
- *	entire process group with a negative pid.
- *
- * ALGORITHM
- *	call setup
- *	loop if the -i option was given
- *	fork 5 children
- *	execute the kill system call
- *	check the return value
- *	if return value is -1
- *		issue a FAIL message, break remaining tests and cleanup
- *	if we are doing functional testing
- *		if the processes were terminated with the expected signal.
- *			issue a PASS message
- *		otherwise
- *			issue a FAIL message
- *	call cleanup
- *
- * USAGE
- *  kill06 [-c n] [-f] [-i n] [-I x] [-P x] [-t]
- *     where,  -c n : Run n copies concurrently.
- *             -f   : Turn off functionality Testing.
- *             -i n : Execute test n times.
- *             -I x : Execute test for x seconds.
- *             -P x : Pause for x seconds between iterations.
- *             -t   : Turn on syscall timing.
+ * Test case to check the basic functionality of kill() when killing an
+ * entire process group with a negative pid.
  *
  * HISTORY
  *	07/2001 Ported by Wayne Boyer
- *
- * RESTRICTIONS
- *	This test should be run as a non-root user.
  */
 
-#include "test.h"
-
+#include <sys/types.h>
 #include <signal.h>
-#include <errno.h>
 #include <sys/wait.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include "tst_test.h"
 
-void cleanup(void);
-void setup(void);
-void do_child(void);
-
-char *TCID = "kill06";
-int TST_TOTAL = 1;
-
-#define TEST_SIG SIGKILL
-
-int main(int ac, char **av)
+static void verify_kill(void)
 {
-	int lc;
-	pid_t pid1, pid2;
-	int exno, status, nsig, i;
-
-	tst_parse_opts(ac, av, NULL, NULL);
-#ifdef UCLINUX
-	maybe_run_child(&do_child, "");
-#endif
-
-	setup();		/* global setup */
-
-	/* The following loop checks looping state if -i option given */
-	for (lc = 0; TEST_LOOPING(lc); lc++) {
-
-		/* reset tst_count in case we are looping */
-		tst_count = 0;
-		status = 1;
-		exno = 1;
-
-		/* Fork a process and set the process group so that */
-		/* it is different from this one.  Fork 5 more children. */
-
-		pid1 = FORK_OR_VFORK();
-		if (pid1 < 0) {
-			tst_brkm(TBROK, cleanup, "Fork of first child failed");
-		} else if (pid1 == 0) {
-			setpgrp();
-			for (i = 0; i < 5; i++) {
-				pid2 = FORK_OR_VFORK();
-				if (pid2 < 0) {
-					tst_brkm(TBROK, cleanup, "Fork failed");
-				} else if (pid2 == 0) {
-#ifdef UCLINUX
-					if (self_exec(av[0], "") < 0) {
-						tst_brkm(TBROK, cleanup,
-							 "self_exec of "
-							 "child failed");
-					}
-#else
-					do_child();
-#endif
-				}
-			}
-			/* Kill all processes in this process group */
-			TEST(kill(-getpgrp(), TEST_SIG));
-			sleep(300);
-
-			tst_resm(TINFO, "%d never received a"
-				 " signal", getpid());
-			exit(exno);
-		} else {
-			waitpid(pid1, &status, 0);
-			if (TEST_RETURN != 0) {
-				tst_brkm(TFAIL, cleanup, "%s failed - errno = "
-					 "%d : %s", TCID, TEST_ERRNO,
-					 strerror(TEST_ERRNO));
-			}
+	pid_t pid, child_pid[5];
+	int nsig, i, status;
+
+	pid = SAFE_FORK();
+	if (pid == 0) {
+		setpgrp();
+		for (i = 0; i < 5; i++) {
+			child_pid[i] = SAFE_FORK();
+			if (!child_pid[i])
+				pause();
 		}
 
-		/*
-		 * Check to see if the process was terminated with the
-		 * expected signal.
-		 */
-		nsig = WTERMSIG(status);
-		if (!nsig) {
-			tst_resm(TFAIL, "Did not receive any signal");
-		} else if (nsig == TEST_SIG) {
-			tst_resm(TPASS, "received expected signal %d",
-				 nsig);
-		} else {
-			tst_resm(TFAIL,
-				 "expected signal %d received %d",
-				 TEST_SIG, nsig);
-		}
+		TEST(kill(-getpgrp(), SIGKILL));
+		if (TST_RET != 0)
+			tst_res(TFAIL | TTERRNO, "kill failed");
+		tst_res(TINFO, "%d never received a signal", getpid());
+		exit(0);
 	}
-
-	cleanup();
-	tst_exit();
-}
-
-/*
- * do_child()
- */
-void do_child(void)
-{
-	int exno = 1;
-
-	sleep(299);
-
-	tst_resm(TINFO, "%d never received a" " signal", getpid());
-	exit(exno);
-}
-
-/*
- * setup() - performs all ONE TIME setup for this test
- */
-void setup(void)
-{
-	/* Setup default signal handling */
-	tst_sig(FORK, DEF_HANDLER, cleanup);
-
-	TEST_PAUSE;
+
+	SAFE_WAITPID(pid, &status, 0);
+	nsig = WTERMSIG(status);
+	if (nsig != SIGKILL) {
+		tst_res(TFAIL, "wait: unexpected signal %d returned, "
+			"expected SIGKILL(9)", nsig);
+		return;
+	}
+	tst_res(TPASS, "receive expected signal SIGKILL(9)");
 }
 
-/*
- * cleanup() - performs all the ONE TIME cleanup for this test@completion
- * or premature exit.
- */
-void cleanup(void)
-{
-
-}
+static struct tst_test test = {
+	.forks_child = 1,
+	.test_all = verify_kill,
+};
Signature · 1 line
1.8.3.1
#12 lore raw

PATCH 4/4 syscalls/kill06:Cleanup && Convert to new library

Feiyu Zhu <[email protected]> wrote:

...

Quoted 2 lines

+ pid_t pid, child_pid[5];

I'd suggest dropping these two variables since it makes no sense to save the children PIDs in the following testing.

testcases/kernel/syscalls/kill/kill06.c line 28

Quoted 8 lines
testcases/kernel/syscalls/kill/kill06.c +7 -0
+       int nsig, i, status;
+
+       pid = SAFE_FORK();
+       if (pid == 0) {
+               setpgrp();
+               for (i = 0; i < 5; i++) {
+                       child_pid[i] = SAFE_FORK();

if we drop the arrays, that loop would be like:

                for (i = 0; i < 5; i++) {
                        if (!SAFE_FORK())
                                pause();
                }
Quoted 14 lines
testcases/kernel/syscalls/kill/kill06.c +2 -9
+                       if (!child_pid[i])
+                               pause();
                }
 
-               /*
-                * Check to see if the process was terminated with the
-                * expected signal.
-                */
-               nsig = WTERMSIG(status);
-               if (!nsig) {
-                       tst_resm(TFAIL, "Did not receive any signal");
-               } else if (nsig == TEST_SIG) {
-                       tst_resm(TPASS, "received expected signal %d",

testcases/kernel/syscalls/kill/kill06.c line 36

Quoted 10 lines
testcases/kernel/syscalls/kill/kill06.c +4 -6
-                                nsig);
-               } else {
-                       tst_resm(TFAIL,
-                                "expected signal %d received %d",
-                                TEST_SIG, nsig);
-               }
+               TEST(kill(-getpgrp(), SIGKILL));
+               if (TST_RET != 0)
+                       tst_res(TFAIL | TTERRNO, "kill failed");
+               tst_res(TINFO, "%d never received a signal", getpid());

This TINFO is a misleading message to people. If the process receives a signal
but not been killed, how do you say that never received a signal? So I suggest deleting this line.

Quoted 22 lines
testcases/kernel/syscalls/kill/kill06.c +1 -19
+               exit(0);
        }
-
-       cleanup();
-       tst_exit();
-}
-
-/*
- * do_child()
- */
-void do_child(void)
-{
-       int exno = 1;
-
-       sleep(299);
-
-       tst_resm(TINFO, "%d never received a" " signal", getpid());
-       exit(exno);
-}
-
-/*

testcases/kernel/syscalls/kill/kill06.c line 40

Quoted 10 lines
testcases/kernel/syscalls/kill/kill06.c +2 -8
- * setup() - performs all ONE TIME setup for this test
- */
-void setup(void)
-{
-       /* Setup default signal handling */
-       tst_sig(FORK, DEF_HANDLER, cleanup);
-
-       TEST_PAUSE;
+
+       SAFE_WAITPID(pid, &status, 0);

What about waiting for any child process here?
SAFE_WAITPID(-1, &status, 0);

Quoted 9 lines
testcases/kernel/syscalls/kill/kill06.c +7 -0
+       nsig = WTERMSIG(status);
+       if (nsig != SIGKILL) {
+               tst_res(TFAIL, "wait: unexpected signal %d returned, "
+                       "expected SIGKILL(9)", nsig);
+               return;
+       }
+       tst_res(TPASS, "receive expected signal SIGKILL(9)");
 }
Signature · 5 lines
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20200821/43a364ab/attachment-0001.htm>