ppoll01: add signal() syscall to register the signal handler

The ppoll01 testcase has a signal handler in it but this signal handler is not registered with the kernel through the signal() syscall. The testcase fails when it sends a SIGINT to itself as part of one of the case of the test run. The following patch
1. Adds a signal() syscall to register the signal handler
2. Corrects the expected 'revents' for the case where ppoll() is called on a file which is only opened in read/write mode, i.e case00.

Signed-off-by Chandru S <[email protected]>

---
 testcases/kernel/syscalls/ppoll/ppoll01.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

kernel/syscalls/ppoll/ppoll01.c renamed from kernel/syscalls/ppoll/ppoll01.c.orig +5 -3

@@ -79,6 +79,8 @@ char *TCID = "ppoll01";  /* Test program
 int  testno;
 int  TST_TOTAL = 1;                   /* total number of tests in this file.   */
 
+void sighandler(int sig);       /* signals handler function for the test */
+
 /* Extern Global Functions */
 /******************************************************************************/
 /*                                                                            */

@@ -125,7 +127,8 @@ extern void cleanup() {
 /*                                                                            */
 /******************************************************************************/
 void setup() {
-        /* Capture signals if any */
+        /* Capture signals */
+	signal(SIGINT, &sighandler);
         /* Create temporary directories */
         TEST_PAUSE;
         tst_tmpdir();

@@ -191,7 +194,7 @@ struct test_case {
 static struct test_case tcase[] = {
 	{ // case00
                 .ttype          = NORMAL,
-                .expect_revents = POLLOUT,
+                .expect_revents = POLLOUT | POLLIN,
                 .ret            = 0,
                 .err            = 0,
         },
 
------------------------------------------------------------------------------

Come build with us! The BlackBerry&reg; Developer Conference in SF, CA

is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9&#45;12, 2009. Register now&#33;
http://p.sf.net/sfu/devconf
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Quoted 5 lines · Chandru

The ppoll01 testcase has a signal handler in it but this signal handler is not registered with the kernel through the signal() syscall. The testcase fails when it sends a SIGINT to itself as part of one of the case of the test run. The following patch
1. Adds a signal() syscall to register the signal handler
2. Corrects the expected 'revents' for the case where ppoll() is called on a file which is only opened in read/write mode, i.e case00.

Quoted 1 line

Signed-off-by Chandru S <[email protected]>

Hi Chandru,

Can you please resend this with a diff from the Latest CVS snapshot ?

Quoted 6 lines · On Tuesday 13 October 2009 17:43:37 Subrata Modak wrote:

Quoted 5 lines · Chandru

The ppoll01 testcase has a signal handler in it but this signal handler is not registered with the kernel through the signal() syscall. The testcase fails when it sends a SIGINT to itself as part of one of the case of the test run. The following patch
1. Adds a signal() syscall to register the signal handler
2. Corrects the expected 'revents' for the case where ppoll() is called on a file which is only opened in read/write mode, i.e case00.

Quoted 6 lines

Quoted 1 line

Signed-off-by Chandru S <[email protected]>

Hi Chandru,

Can you please resend this with a diff from the Latest CVS snapshot ?

Hi Subrata,

Thanks for letting me know. Here is the patch taken from the latest snapshot.

The following patch adds a signal() syscall to the ppoll01 testcase. The testcase has a signal handler in it but this signal handler is not registered with the kernel through the signal() syscall. Without the patch the testcase fails when it sends a SIGINT to itself as part of one of the case of the test run.

Signed-off-by Chandru S <[email protected]>

---

testcases/kernel/syscalls/ppoll/ppoll01.c renamed from testcases/kernel/syscalls/ppoll/ppoll01.c.orig +2 -0

@@ -106,6 +106,8 @@ extern void cleanup() {
         tst_exit();
 }
 
+void sighandler(int sig);       /* signals handler function for the test */
+
 /* Local  Functions */
 /******************************************************************************/
 /*                                                                            */

Quoted 14 lines · Chandru

Quoted 6 lines · On Tuesday 13 October 2009 17:43:37 Subrata Modak wrote:

Quoted 5 lines · Chandru

The ppoll01 testcase has a signal handler in it but this signal handler is not registered with the kernel through the signal() syscall. The testcase fails when it sends a SIGINT to itself as part of one of the case of the test run. The following patch
1. Adds a signal() syscall to register the signal handler
2. Corrects the expected 'revents' for the case where ppoll() is called on a file which is only opened in read/write mode, i.e case00.

Quoted 6 lines

Quoted 1 line

Signed-off-by Chandru S <[email protected]>

Hi Chandru,

Can you please resend this with a diff from the Latest CVS snapshot ?

Quoted 7 lines

Hi Subrata,

Thanks for letting me know. Here is the patch taken from the latest snapshot.

The following patch adds a signal() syscall to the ppoll01 testcase. The testcase has a signal handler in it but this signal handler is not registered with the kernel through the signal() syscall.  Without the patch the testcase fails when it sends a SIGINT to itself as part of one of the case of the test run.

Signed-off-by Chandru S <[email protected]>

Thanks.

Regards--
Subrata

Quoted 8 lines

---

testcases/kernel/syscalls/ppoll/ppoll01.c     2009-10-14 16:53:42.828486675 +0530 renamed from testcases/kernel/syscalls/ppoll/ppoll01.c.orig        2009-10-14 16:58:45.547235527 +0530 +0 -0

@@ -106,6 +106,8 @@ extern void cleanup() {
          tst_exit();
  }
 

Quoted 6 lines

testcases/kernel/syscalls/ppoll/ppoll01.c     2009-10-14 16:53:42.828486675 +0530 renamed from testcases/kernel/syscalls/ppoll/ppoll01.c.orig        2009-10-14 16:58:45.547235527 +0530 +2 -0

+void sighandler(int sig);       /* signals handler function for the test */
+
  /* Local  Functions */
  /******************************************************************************/
  /*                                                                            */

testcases/kernel/syscalls/ppoll/ppoll01.c     2009-10-14 16:53:42.828486675 +0530 extern void cleanup() { · line 135

Quoted 8 lines

testcases/kernel/syscalls/ppoll/ppoll01.c     2009-10-14 16:53:42.828486675 +0530 renamed from testcases/kernel/syscalls/ppoll/ppoll01.c.orig        2009-10-14 16:58:45.547235527 +0530 +2 -0

@@ -126,6 +128,8 @@ extern void cleanup() {
  /******************************************************************************/
  void setup() {
          /* Capture signals if any */
+     signal(SIGINT, &sighandler);
+
          /* Create temporary directories */
          TEST_PAUSE;
          tst_tmpdir();

Just curious -- was the issue seen at the following point in the code previously?

361 TEST(sys_ret = syscall(__NR_ppoll, p_fds, nfds, p_ts,
p_sigmask, sigsetsize)); /* <-- Blocks here indefinitely */
362 sys_errno = errno;
363 if (sys_ret <= 0 || tc->ret < 0)
364 goto TEST_END;
365
366 cmp_ok = fds[0].revents == tc->expect_revents;

I've seen this issue 1 out of the 8 or 9 times I've run this test.

Quoted 11 lines · On Thursday 15 October 2009 06:55:29 Garrett Cooper wrote:

Just curious -- was the issue seen at the following point in the code previously?

361             TEST(sys_ret = syscall(__NR_ppoll, p_fds, nfds, p_ts, p_sigmask, sigsetsize)); /* <-- Blocks here indefinitely */ 362             sys_errno = errno;
363             if (sys_ret <= 0 || tc->ret < 0)
364                     goto TEST_END;
365
366             cmp_ok = fds[0].revents == tc->expect_revents;

Quoted 1 line

I've seen this issue 1 out of the 8 or 9 times I've run this test.

Nope, without the patch we only see case00 completing and the test exits while in the middle of case01

# ./ppoll01
ppoll01     0  TINFO  :  (case00) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       1 errno=0 (Success)
ppoll01     0  TINFO  :  (case00) END => OK
ppoll01     0  TINFO  :  (case01) START
#

With the patch we see the output as follows...

# ./ppoll01
ppoll01     0  TINFO  :  (case00) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       1 errno=0 (Success)
ppoll01     0  TINFO  :  (case00) END => OK
ppoll01     0  TINFO  :  (case01) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       0 errno=0 (Success)
ppoll01     0  TINFO  :  (case01) END => OK
ppoll01     0  TINFO  :  (case02) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       0 errno=0 (Success)
ppoll01     0  TINFO  :  (case02) END => OK
ppoll01     0  TINFO  :  (case03) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       1 errno=0 (Success)
ppoll01     0  TINFO  :  (case03) END => OK
ppoll01     0  TINFO  :  (case04) START
EXPECT: return value(ret)=-1 errno=4 (Interrupted system call)
RESULT: return value(ret)=-1 errno=4 (Interrupted system call)
ppoll01     0  TINFO  :  (case04) END => OK
ppoll01     0  TINFO  :  (case05) START
EXPECT: return value(ret)=-1 errno=4 (Interrupted system call)
RESULT: return value(ret)=-1 errno=4 (Interrupted system call)
ppoll01     0  TINFO  :  (case05) END => OK
ppoll01     0  TINFO  :  (case06) START

EXPECT: return value(ret)=-1 errno=22 (Invalid argument)
RESULT: return value(ret)=-1 errno=22 (Invalid argument)
ppoll01     0  TINFO  :  (case06) END => OK
ppoll01     0  TINFO  :  (case07) START
EXPECT: return value(ret)=-1 errno=14 (Bad address)
RESULT: return value(ret)=-1 errno=14 (Bad address)
ppoll01     0  TINFO  :  (case07) END => OK
ppoll01     1  TPASS  :  ppoll01 call succeeded

Quoted 53 lines · Chandru

Quoted 11 lines · On Thursday 15 October 2009 06:55:29 Garrett Cooper wrote:

Just curious -- was the issue seen at the following point in the code previously?

361             TEST(sys_ret = syscall(__NR_ppoll, p_fds, nfds, p_ts, p_sigmask, sigsetsize)); /* <-- Blocks here indefinitely */ 362             sys_errno = errno;
363             if (sys_ret <= 0 || tc->ret < 0)
364                     goto TEST_END;
365
366             cmp_ok = fds[0].revents == tc->expect_revents;

Quoted 1 line

I've seen this issue 1 out of the 8 or 9 times I've run this test.

Nope, without the patch we only see case00 completing and the test exits while in the middle of case01

# ./ppoll01
ppoll01     0  TINFO  :  (case00) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       1 errno=0 (Success)
ppoll01     0  TINFO  :  (case00) END => OK
ppoll01     0  TINFO  :  (case01) START
#

With the patch we see the output as follows...

# ./ppoll01
ppoll01     0  TINFO  :  (case00) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       1 errno=0 (Success)
ppoll01     0  TINFO  :  (case00) END => OK
ppoll01     0  TINFO  :  (case01) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       0 errno=0 (Success)
ppoll01     0  TINFO  :  (case01) END => OK
ppoll01     0  TINFO  :  (case02) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       0 errno=0 (Success)
ppoll01     0  TINFO  :  (case02) END => OK
ppoll01     0  TINFO  :  (case03) START
EXPECT: return value(ret)=(N >= 0) errno=0 (Success)
RESULT: return value(ret)=       1 errno=0 (Success)
ppoll01     0  TINFO  :  (case03) END => OK
ppoll01     0  TINFO  :  (case04) START
EXPECT: return value(ret)=-1 errno=4 (Interrupted system call)
RESULT: return value(ret)=-1 errno=4 (Interrupted system call)
ppoll01     0  TINFO  :  (case04) END => OK
ppoll01     0  TINFO  :  (case05) START
EXPECT: return value(ret)=-1 errno=4 (Interrupted system call)
RESULT: return value(ret)=-1 errno=4 (Interrupted system call)
ppoll01     0  TINFO  :  (case05) END => OK
ppoll01     0  TINFO  :  (case06) START

Quoted 8 lines

EXPECT: return value(ret)=-1 errno=22 (Invalid argument)
RESULT: return value(ret)=-1 errno=22 (Invalid argument)
ppoll01     0  TINFO  :  (case06) END => OK
ppoll01     0  TINFO  :  (case07) START
EXPECT: return value(ret)=-1 errno=14 (Bad address)
RESULT: return value(ret)=-1 errno=14 (Bad address)
ppoll01     0  TINFO  :  (case07) END => OK
ppoll01     1  TPASS  :  ppoll01 call succeeded

   Interesting. The cause may be the same, but the effect may be different, as this is blocking with this section of the test:

       case MINUS_NSEC:
               ts.tv_sec = 0;
               ts.tv_nsec = -1;
               p_ts = &ts;
               break;

   Which is fine, but depending on whether or not a previous testcase fails with EOK and a return code of 0. Example:

#define _GNU_SOURCE
#include <signal.h>
#include <poll.h>
#include <errno.h>
#include <string.h>

int main (void) {

       struct pollfd fds;
       struct timespec timeout;
       sigset_t sigmask;

       sigemptyset(&sigmask);

       timeout.tv_sec = 0;
       timeout.tv_nsec = 5e8; /* .5 sec */

       if (ppoll(&fds, 1, &timeout, &sigmask) <= 0) { perror("ppoll");
               return 0;
       }
       return 1;
}

gcooper@orangebox ~ $ ./ppoll_test
ppoll: Success

   The manpage doesn't say that it shouldn't set errno to ETIME, etc, so this is a legitimate return value (despite the fact that the testcase fails claiming that the subtestcase failed and the call succeeded, IIRC -- or maybe it was another test...).
   I'll definitely keep my eyes peeled for this particular issue though in the future.