"Serge E. Hallyn" <[EMAIL PROTECTED]> wrote on 27/12/2007 20:28:36:
> Quoting Rishikesh K. Rajak ([EMAIL PROTECTED]):
> >
>
> > This patch implements a basic test for cloning pid namespace. It
> > verifies that the newly created child process has pid == 1 and ppid ==
0.
> >
> > <Signed-off By>: Rishikesh K Rajak <[EMAIL PROTECTED]>
> > ---
> > Index: ltp-full-20071130/testcases/kernel/containers/pidns/pidns01.c
> > ===================================================================
> > --- /dev/null 1970-01-01 00:00:00.000000000 +0000
> > +++ ltp-full-20071130/testcases/kernel/containers/pidns/pidns01.c
> 2007-12-27 16:28:09.000000000 +0530
> > @@ -0,0 +1,162 @@
> > +/*
> > +* Copyright (c) International Business Machines Corp., 2007
> > +* 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307
USA
> > +*
> >
+***************************************************************************
> > +
> > +* File: pidns01.c
> > +*
> > +* Description:
> > +* The pidns01.c testcase builds into the ltp framework to verify
> > +* the basic functionality of PID Namespace.
> > +*
> > +* Verify that:
> > +* 1. When parent clone a process with flag CLONE_NEWPID, the process
ID of
> > +* child should be always one.
> > +*
> > +* 2. When parent clone a process with flag CLONE_NEWPID, the
> parent process ID of
> > +* should be always zero.
> > +*
> > +* Total Tests: 6 assertions
> > +*
> > +* Test Name: pidns01
> > +*
> > +* Test Assertion & Strategy:
> > +*
> > +* From main() clone a new child process with passing the
> clone_flag as CLONE_NEWPID,
> > +* Inside the cloned pid check for the getpid() and getppid()
> > +* Verify with global macro defined value for parent pid & child pid.
> > +*
> > +* Usage: <for command-line>
> > +* pidns01
> > +*
> > +* History:
> > +*
> > +* FLAG DATE NAME DESCRIPTION
> > +* 27/12/07 RISHIKESH K RAJAK <[EMAIL PROTECTED]> Created this test
> > +*
> >
>
+*******************************************************************************************/
> > +#define _GNU_SOURCE 1
> > +#include <sys/wait.h>
> > +#include <assert.h>
> > +#include <stdio.h>
> > +#include <stdlib.h>
> > +#include <unistd.h>
> > +#include <string.h>
> > +#include <errno.h>
> > +#ifndef NO_LTP
> > +#include <usctest.h>
> > +#include <test.h>
> > +#include <libclone.h>
> > +#else
> > +#include "../../../../include/usctest.h"
> > +#include "../libclone/libclone.h"
> > +#endif
> > +
> > +char *TCID = "pid_namespace1";
> > +int TST_TOTAL=1;
> > +
> > +void cleanup(void);
> > +
> > +#ifdef NO_LTP
> > +#define TFAIL "FAILURE: "
> > +#define TPASS "PASS: "
> > +#define TINFO "INFO: "
> > +#define TWARN "WARN: "
> > +#define tst_resm(x, format, arg...) printf("%s:" format, x, ## arg)
> > +#define tst_exit() exit(1)
> > +#endif
> > +
> > +#define CHILD_PID 1
> > +#define PARENT_PID 0
> > +
> > +
> > +/*
> > + * child_fn1() - Inside container
> > + */
> > +int child_fn1(void *ttype)
> > +{
> > + pid_t cpid, ppid;
> > + cpid = getpid();
> > + ppid = getppid();
> > +
> > + tst_resm(TINFO, "PIDNS test is running inside container\n");
> > + if(( cpid == CHILD_PID) &&
> > + ( ppid == PARENT_PID ) )
> > + {
> > + tst_resm(TPASS, "Success:" );
> > + }
> > + else
> > + {
> > + tst_resm(TFAIL, "FAIL: Got unexpected result of"
> > + " cpid=%d ppid=%d\n", cpid, ppid);
> > + }
>
> I'm still dubious - if this testcase fails, won't LTP claim all tests
> pssed?
>
> Yes the logs will have a FAIL message but one generally doesn't look at
> those if all tests reported as passed.
>
> So can someone confirm that LTP should report a failed test case the way
> this is done? (My own quick test suggests no)
Hi serge,
You are right, I investigated more on this and i think this is a library
and pan related problem of LTP.
Discussed with LTP maintainer, look like all the testcases following same
as i followed.
Will say more on this in near future.
Thanks
Rishi
>
> > + tst_exit();
> > +}
> > +
> >
+/***********************************************************************
> > +* M A I N
> >
+***********************************************************************/
> > +
> > +int main(int argc, char *argv[])
> > +{
> > + int ret, status;
> > +
> > + ret = do_clone_unshare_test(T_CLONE,
> > + CLONE_NEWPID, child_fn1, NULL);
> > +
> > + /* check return code */
> > + if (ret == -1) {
> > + tst_resm(TFAIL, "clone() Failed, errno = %d :"
> > + " %s", ret, strerror(ret));
> > + /* Cleanup & continue with next test case */
> > + cleanup();
> > + }
> > +
> > + /* Wait for child to finish */
> > + if ((wait(&status)) < 0) {
> > + tst_resm(TWARN, "wait() failed, skipping this"
> > + " test case");
> > + /* Cleanup & continue with next test case */
> > + cleanup();
> > + }
> > +
> > + if (WTERMSIG(status)) {
> > + tst_resm(TWARN, "child exited with signal %d",
> > + WTERMSIG(status));
> > + }
> > +
> > + /* cleanup and exit */
> > + cleanup();
> > +
> > + /*NOTREACHED*/
> > + return 0;
> > +
> > +} /* End main */
> > +
> > +/*
> > + * cleanup() - performs all ONE TIME cleanup for this test at
> > + * completion or premature exit.
> > + */
> > +void
> > +cleanup()
> > +{
> > +
> > +#ifndef NO_LTP
> > + /* Clean the test testcase as LTP wants*/
> > + TEST_CLEANUP;
> > +#endif
> > +
> > + /* exit with return code appropriate for results */
> > + tst_exit();
> > +}
> > Index:
ltp-full-20071130/testcases/kernel/containers/pidns/runpidnstest.sh
> > ===================================================================
> > --- /dev/null 1970-01-01 00:00:00.000000000 +0000
> > +++ ltp-
> full-20071130/testcases/kernel/containers/pidns/runpidnstest.sh
> 2007-12-27 16:27:39.000000000 +0530
> > @@ -0,0 +1,8 @@
> > +#!/bin/sh
> > +
> > +exit_code=0
> > +pidns01
> > +if [ $? -ne 0 ]; then
> > + exit_code="$?"
> > +fi
> > +exit $exit_code
> > Index: ltp-full-20071130/testcases/kernel/containers/pidns/Makefile
> > ===================================================================
> > --- ltp-full-20071130.
> orig/testcases/kernel/containers/pidns/Makefile 2007-12-27 16:26:
> 18.000000000 +0530
> > +++ ltp-full-20071130/testcases/kernel/containers/pidns/Makefile
> 2007-12-27 16:27:39.000000000 +0530
> > @@ -3,7 +3,7 @@
> > LDLIBS += -L../../../../lib -L../libclone ../libclone/libclone.a -lltp
> >
> > SRCS = $(wildcard *.c)
> > -NOLTPSRCS =
> > +NOLTPSRCS = pidns01.c
> > TARGETS = $(patsubst %.c,%,$(SRCS))
> > NOLTP_TARGETS = $(patsubst %.c,%_noltp,$(NOLTPSRCS))
> >
> > @@ -18,4 +18,7 @@
> > rm -f $(TARGETS) *.o $(NOLTP_TARGETS)
> >
> > install:
> > - @set -e; for i in $(TARGETS) check_pidns_enabled; do ln -f $$i
> ../../../bin/$$i; done
> > + @set -e; for i in $(TARGETS) runpidnstest.sh
> check_pidns_enabled; do ln -f $$i ../../../bin/$$i ; chmod +x
> runpidnstest.sh ; done
> > +
> > +noltp_check: noltp
> > + ./runtests_noltp.sh
> > Index: ltp-full-20071130/testcases/kernel/containers/container_test.sh
> > ===================================================================
> > --- ltp-full-20071130.
> orig/testcases/kernel/containers/container_test.sh 2007-12-27 16:
> 26:18.000000000 +0530
> > +++ ltp-full-20071130/testcases/kernel/containers/container_test.
> sh 2007-12-27 16:27:39.000000000 +0530
> > @@ -57,6 +57,7 @@
> > check_pidns_enabled
> > if [ $? -eq 0 ]; then
> > echo "Running pidns tests."
> > + runpidnstest.sh
> > else
> > echo "pid namespaces not enabled in kernel. Not running pidns
tests."
> > fi
>
> >
-------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2005.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > Ltp-list mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/ltp-list
>
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list
