[patch 1/2][RFC] add socketat syscall
Subrata Modak
subrata at linux.vnet.ibm.com
Fri Nov 7 01:45:05 PST 2008
Thanks Daniel/Michael,
Adding Veerendra, Sudhir & Gowri as they are handling NS tests now and
they may include it in their Plan.
Regards--
Subrata
On Thu, 2008-11-06 at 17:18 +0100, Daniel Lezcano wrote:
> Michael Kerrisk wrote:
> >> On Fri, Oct 31, 2008 at 4:56 PM, Daniel Lezcano <dlezcano at fr.ibm.com> wrote:
> >>> This patch adds the socketat syscall which allows to specify in
> >>> which network namespace we want to create a socket. The network
> >>> namespace destination is referred by a socket fd previously opened
> >>> in the destination network namespace.
> >
> > Daniel,
> >
> > Is there any documentation for this system call, and/or test programs?
>
> Not yet.
>
> This small patch is a proposition to Andreas and Vivien to have a single
> process being able to manage several network namespaces.
>
> When a process unshares the network, it creates a socket which is used
> as a socket control (it belongs to the network namespace). Each time a
> network namespace is created, a socket control is created.
>
> When the process has to create a socket for a specific network
> namespace, it can use the socket control to specify it. This is the
> purpose of the socketat syscall.
>
> One example for a program in userspace:
>
> int main(int argc, char *argv[])
> {
> const int maxunshare = 128;
> int scs[maxunshare];
> int i, fd;
>
> for (i = 0; i < maxunshare; i++) {
> scs[i] = socket(PF_INET, SOCK_DGRAM, 0);
> unshare(CLONE_NEWNET);
> }
>
> ....
>
> /* I want to create a socket inside the network namespace #10 */
>
> fd = socketat(scs[10], PF_INET, SOCKET_STREAM, 0);
>
> ....
>
> bind, listen, etc ...
> }
>
> >>> Signed-off-by: Daniel Lezcano <dlezcano at fr.ibm.com>
> >>> ---
> >>> arch/x86/include/asm/unistd_32.h | 1
> >>> arch/x86/include/asm/unistd_64.h | 3 +-
> >>> arch/x86/kernel/syscall_table_32.S | 1
> >>> include/linux/syscalls.h | 1
> >>> kernel/sys_ni.c | 1
> >>> net/socket.c | 45 +++++++++++++++++++++++++++++++++++++
> >>> 6 files changed, 51 insertions(+), 1 deletion(-)
> >>>
> >>> Index: net-next-2.6/arch/x86/include/asm/unistd_32.h
> >>> ===================================================================
> >>> --- net-next-2.6.orig/arch/x86/include/asm/unistd_32.h
> >>> +++ net-next-2.6/arch/x86/include/asm/unistd_32.h
> >>> @@ -338,6 +338,7 @@
> >>> #define __NR_dup3 330
> >>> #define __NR_pipe2 331
> >>> #define __NR_inotify_init1 332
> >>> +#define __NR_socketat 333
> >>>
> >>> #ifdef __KERNEL__
> >>>
> >>> Index: net-next-2.6/arch/x86/include/asm/unistd_64.h
> >>> ===================================================================
> >>> --- net-next-2.6.orig/arch/x86/include/asm/unistd_64.h
> >>> +++ net-next-2.6/arch/x86/include/asm/unistd_64.h
> >>> @@ -653,7 +653,8 @@ __SYSCALL(__NR_dup3, sys_dup3)
> >>> __SYSCALL(__NR_pipe2, sys_pipe2)
> >>> #define __NR_inotify_init1 294
> >>> __SYSCALL(__NR_inotify_init1, sys_inotify_init1)
> >>> -
> >>> +#define __NR_socketat 295
> >>> +__SYSCALL(__NR_socketat, sys_socketat)
> >>>
> >>> #ifndef __NO_STUBS
> >>> #define __ARCH_WANT_OLD_READDIR
> >>> Index: net-next-2.6/arch/x86/kernel/syscall_table_32.S
> >>> ===================================================================
> >>> --- net-next-2.6.orig/arch/x86/kernel/syscall_table_32.S
> >>> +++ net-next-2.6/arch/x86/kernel/syscall_table_32.S
> >>> @@ -332,3 +332,4 @@ ENTRY(sys_call_table)
> >>> .long sys_dup3 /* 330 */
> >>> .long sys_pipe2
> >>> .long sys_inotify_init1
> >>> + .long sys_socketat
> >>> Index: net-next-2.6/net/socket.c
> >>> ===================================================================
> >>> --- net-next-2.6.orig/net/socket.c
> >>> +++ net-next-2.6/net/socket.c
> >>> @@ -1253,6 +1253,51 @@ out_release:
> >>> return retval;
> >>> }
> >>>
> >>> +asmlinkage long sys_socketat(int fd, int family, int type, int protocol)
> >>> +{
> >>> + int retval, fput_needed;
> >>> + struct socket *sock;
> >>> + struct socket *sockat;
> >>> + struct net *net;
> >>> + int flags;
> >>> +
> >>> + /* Check the SOCK_* constants for consistency. */
> >>> + BUILD_BUG_ON(SOCK_CLOEXEC != O_CLOEXEC);
> >>> + BUILD_BUG_ON((SOCK_MAX | SOCK_TYPE_MASK) != SOCK_TYPE_MASK);
> >>> + BUILD_BUG_ON(SOCK_CLOEXEC & SOCK_TYPE_MASK);
> >>> + BUILD_BUG_ON(SOCK_NONBLOCK & SOCK_TYPE_MASK);
> >>> +
> >>> + flags = type & ~SOCK_TYPE_MASK;
> >>> + if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
> >>> + return -EINVAL;
> >>> + type &= SOCK_TYPE_MASK;
> >>> +
> >>> + if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
> >>> + flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;
> >>> +
> >>> + sock = sockfd_lookup_light(fd, &retval, &fput_needed);
> >>> + if (!sock)
> >>> + goto out;
> >>> +
> >>> + net = sock_net(sock->sk);
> >>> +
> >>> + retval = __sock_create(net, family, type, protocol, &sockat, 0);
> >>> + if (retval)
> >>> + goto out_fput;
> >>> +
> >>> + retval = sock_map_fd(sock, flags & (O_CLOEXEC | O_NONBLOCK));
> >>> + if (retval < 0)
> >>> + goto out_release;
> >>> +out_fput:
> >>> + fput_light(sock->file, fput_needed);
> >>> +out:
> >>> + return retval;
> >>> +
> >>> +out_release:
> >>> + sock_release(sockat);
> >>> + goto out;
> >>> +}
> >>> +
> >>> /*
> >>> * Create a pair of connected sockets.
> >>> */
> >>> Index: net-next-2.6/include/linux/syscalls.h
> >>> ===================================================================
> >>> --- net-next-2.6.orig/include/linux/syscalls.h
> >>> +++ net-next-2.6/include/linux/syscalls.h
> >>> @@ -423,6 +423,7 @@ asmlinkage long sys_recvfrom(int, void _
> >>> struct sockaddr __user *, int __user *);
> >>> asmlinkage long sys_recvmsg(int fd, struct msghdr __user *msg, unsigned flags);
> >>> asmlinkage long sys_socket(int, int, int);
> >>> +asmlinkage long sys_socketat(int, int, int, int);
> >>> asmlinkage long sys_socketpair(int, int, int, int __user *);
> >>> asmlinkage long sys_socketcall(int call, unsigned long __user *args);
> >>> asmlinkage long sys_listen(int, int);
> >>> Index: net-next-2.6/kernel/sys_ni.c
> >>> ===================================================================
> >>> --- net-next-2.6.orig/kernel/sys_ni.c
> >>> +++ net-next-2.6/kernel/sys_ni.c
> >>> @@ -40,6 +40,7 @@ cond_syscall(sys_send);
> >>> cond_syscall(sys_recvfrom);
> >>> cond_syscall(sys_recv);
> >>> cond_syscall(sys_socket);
> >>> +cond_syscall(sys_socketat);
> >>> cond_syscall(sys_setsockopt);
> >>> cond_syscall(compat_sys_setsockopt);
> >>> cond_syscall(sys_getsockopt);
> >>>
> >>> --
> >>> _______________________________________________
> >>> Containers mailing list
> >>> Containers at lists.linux-foundation.org
> >>> https://lists.linux-foundation.org/mailman/listinfo/containers
> >>>
> >>
> >>
> >> --
> >> Michael Kerrisk Linux man-pages maintainer;
> >> http://www.kernel.org/doc/man-pages/ Found a documentation bug?
> >> http://www.kernel.org/doc/man-pages/reporting_bugs.html
> >> --
> >> To unsubscribe from this list: send the line "unsubscribe linux-api" in
> >> the body of a message to majordomo at vger.kernel.org
> >> More majordomo info at http://vger.kernel.org/majordomo-info.html
> >>
> >
> >
> >
>
>
More information about the Containers
mailing list