Re: [LTP] [lxc-dev] Re: [PATCH 5/9] Network NS patch , lib to create the netns

[Serge E. H. <se...@us...>]

Quoting Veerendra (ve...@li...):
> Veerendra wrote:
> > Signed-off-by: Veerendra C <vec...@in...>
> >
> > This patch consists of the files..
> >
> > containers/libclone/libclone.h
> > containers/libclone/libnetns.c
> > containers/libclone/Makefile
> >
> >
> > Regards
> > Veerendra C
> >
> > >
> >
> >   
> Attaching the patch..
> 
> 
> --~--~---------~--~----~------------~-------~--~----~
> You received this message because you are subscribed to the Google Groups "lxc-dev" group.
> To post to this group, send email to lx...@go...
> To unsubscribe from this group, send email to lxc...@go...
> For more options, visit this group at http://groups.google.com/group/lxc-dev?hl=en
> -~----------~----~----~----~------~----~------~--~---
> 

> Index: containers/libclone/libclone.h
> ===================================================================
> --- containers.orig/libclone/libclone.h
> +++ containers/libclone/libclone.h
> @@ -58,6 +58,46 @@ extern int  __clone2(int (*fn) (void *ar
>                  pid_t *parent_tid, void *tls, pid_t *child_tid);
>  #endif
> 
> +#ifndef HAVE_UNSHARE
> +
> +#if __i386__
> +#    define __NR_unshare 310
> +#elif __x86_64__
> +#    define __NR_unshare 272
> +#elif __ia64__
> +#    define __NR_unshare 1296
> +#elif __s390x__
> +#    define __NR_unshare 303
> +#elif __powerpc__
> +#    define __NR_unshare 282
> +#else
> +#    error "Architecture not supported"
> +#endif

The existing libclone.h already tries to define these for arches that
need them.  I certainly believe that there are distributions which
are still funky, but let's understand why.

Do you have a machine of your own on which you needed this?  Can you
remove this bit and let me know what happens?

> +
> +static inline int unshare(unsigned long flags)
> +{
> +        return syscall(__NR_unshare, flags);
> +}
> +
> +#endif /* HAVE_UNSHARE */
> +
> +#if __i386__
> +#    define __NR_clone64       333
> +#    define __NR_unshare64     334

What on earth are you using clone64 and unshare64 for?  They don't
actually exist, and you certainly don't need them for network
namespaces.

> +#elif __powerpc__
> +#    define __NR_clone64       313
> +#    define __NR_unshare64     314
> +#elif __s390x__ || __s390__
> +#    define __NR_clone64       322
> +#    define __NR_unshare64     323
> +#elif __x86_64__
> +#    define __NR_clone64       295
> +#    define __NR_unshare64     296
> +#else
> +#    error "Architecture not supported"
> +#endif
> +
> +
>  #ifndef CLONE_NEWUTS
>  #define CLONE_NEWUTS		0x04000000
>  #endif
> @@ -74,6 +114,10 @@ extern int  __clone2(int (*fn) (void *ar
>  #define CLONE_NEWPID            0x20000000
>  #endif
> 
> +#ifndef CLONE_NEWNET
> +#define CLONE_NEWNET		0x40000000
> +#endif
> +
>  /*
>   * Run fn1 in a unshared environmnent, and fn2 in the original context
>   * Fn2 may be NULL.
> @@ -97,4 +141,10 @@ int do_clone_unshare_tests(int use_clone
>  			int (*fn1)(void *arg), void *arg1,
>  			int (*fn2)(void *arg), void *arg2);
> 
> +static inline int unshare64(unsigned long long int flags)
> +{
> +       return syscall(__NR_unshare64, (unsigned long) (flags >> 32), 
> +                      (unsigned long) (flags & 0xffffffff));
> +}
> +

Again, you shouldn't need the above.

>  #endif
> Index: containers/libclone/libnetns.c
> ===================================================================
> --- /dev/null
> +++ containers/libclone/libnetns.c
> @@ -0,0 +1,105 @@
> +/*
> +* 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
> +*
> +***************************************************************************/
> +/*=========================================================================
> +* This testcase creates the network namespace. 
> +* It creates veth pair veth8 & veth9. Also assigns IP addresses to the childNS.
> +* Also it starts the sshd daemon @ port 7890
> +*
> +* Scripts Used: parentns.sh childns.sh
> +=========================================================================*/
> +
> +#include <sys/utsname.h>
> +#include <sched.h>
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include "libclone.h"
> +#include <sched.h>
> +#include <sys/syscall.h>
> +#include <unistd.h>
> +#include <signal.h>
> +#include <string.h>
> +#include <errno.h>
> +#include <libgen.h>
> +#include <fcntl.h>
> +#include <sys/types.h>
> +#include <sys/wait.h>
> + 
> +extern pid_t getpgid(pid_t pid);
> +extern pid_t getsid(pid_t pid);
> +
> +int crtchild(char *s1 , char *s2)
> +{
> +    char *cmd[] = { "/bin/bash", s1, s2, (char *)0 };
> +    execve("/bin/bash", cmd, __environ);
> +    printf("The code would not reach here on success\n");
> +    perror("execve");
> +    return 1;
> +}
> +
> +int create_net_namespace(char *p1, char *c1)
> +{
> +	int pid, status, ret;
> +    long long flags = 0;
> +    char par[256], child[256];
> +    char path[128];
> +
> +	if (tst_kvercmp(2,6,19) < 0)
> +		return 1;
> +
> +    if (ret = getenv("LTPROOT")) {
> +        sprintf(path,"%s%s",getenv("LTPROOT"),"/testcases/kernel/containers/netns/");
> +    } 
> +    else{
> +        printf("LTPROOT env variable is not set\n");
> +        printf("Please set LTPROOT and re-run the test.. Thankyou\n");
> +        return -1;
> +    } 
> +
> +    flags |= CLONE_NEWNS;
> +    flags |= CLONE_NEWNET;
> +
> +    sprintf(par , "%s%s %s", path, "parentns.sh", p1);
> +    sprintf(child , "%s%s", path,"childns.sh");
> +
> +    if ((pid = fork()) == 0) {
> +
> +        // Child.
> +        ret = unshare64(flags);

This can't be succeeding on a stock kernel.

Please test with
git://git.kernel.org/pub/scm/linux/kernel/git/daveh/linux-2.6-lxc.git
right now.

I suspect what happened was you cut and pasted code from an experimental
but obsolete ns_exec, but you need to use standard unshare() here.

> +        if (ret < 0) {
> +            perror("unshare");
> +            return 1;
> +        }
> +    return crtchild(child,c1);
> +    }
> +    else{
> +
> +        //parent
> +        ret = system(par);
> +        if (ret != 0) {
> +            printf("Error not able to assign the pid to the net\n");
> +            fflush(stdout);
> +            exit(1);
> +        }
> +    fflush(stdout);
> +
> +    if ((ret = waitpid(pid, &status, __WALL)) < 0){
> +        printf("waitpid() returns %d, errno %d\n", ret, errno);
> +        return errno;
> +    }
> +    return 0;
> +    }
> +}
> Index: containers/libclone/Makefile
> ===================================================================
> --- containers.orig/libclone/Makefile
> +++ containers/libclone/Makefile
> @@ -18,7 +18,7 @@
>  ##                                                                            ##
>  ################################################################################
> 
> -TARGET=libclone.a
> +TARGET=libclone.a libnetns.a
>  SRCS=$(wildcard *.c)
>  OBJS=$(patsubst %.c,%.o,$(SRCS))
>