#endif /* CONFIG_NET_RADIO */
#include <asm/uaccess.h>
+#include <asm/unistd.h>
+
#include <net/compat.h>
#include <net/sock.h>
#include <linux/netfilter.h>
+#include <linux/vs_socket.h>
static int sock_no_open(struct inode *irrelevant, struct file *dontcare);
static ssize_t sock_aio_read(struct kiocb *iocb, char __user *buf,
static int sock_close(struct inode *inode, struct file *file);
static unsigned int sock_poll(struct file *file,
struct poll_table_struct *wait);
-static int sock_ioctl(struct inode *inode, struct file *file,
+static long sock_ioctl(struct file *file,
unsigned int cmd, unsigned long arg);
static int sock_fasync(int fd, struct file *filp, int on);
static ssize_t sock_readv(struct file *file, const struct iovec *vector,
.aio_read = sock_aio_read,
.aio_write = sock_aio_write,
.poll = sock_poll,
- .ioctl = sock_ioctl,
+ .unlocked_ioctl = sock_ioctl,
.mmap = sock_mmap,
.open = sock_no_open, /* special open code to disallow open via /proc */
.release = sock_close,
#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
static atomic_t net_family_lockct = ATOMIC_INIT(0);
-static spinlock_t net_family_lock = SPIN_LOCK_UNLOCKED;
+static DEFINE_SPINLOCK(net_family_lock);
/* The strategy is: modifications net_family vector are short, do not
sleep and veeery rare, but read access should be free of any exclusive
ei->socket.ops = NULL;
ei->socket.sk = NULL;
ei->socket.file = NULL;
- ei->socket.passcred = 0;
+ ei->socket.flags = 0;
return &ei->vfs_inode;
}
static int init_inodecache(void)
{
sock_inode_cachep = kmem_cache_create("sock_inode_cache",
- sizeof(struct socket_alloc),
- 0, SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT,
- init_once, NULL);
+ sizeof(struct socket_alloc),
+ 0, SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT,
+ init_once, NULL);
if (sock_inode_cachep == NULL)
return -ENOMEM;
return 0;
sock->file = file;
file->f_op = SOCK_INODE(sock)->i_fop = &socket_file_ops;
- file->f_mode = 3;
+ file->f_mode = FMODE_READ | FMODE_WRITE;
file->f_flags = O_RDWR;
file->f_pos = 0;
fd_install(fd, file);
* NULL is returned.
*/
-struct socket *sock_alloc(void)
+static struct socket *sock_alloc(void)
{
struct inode * inode;
struct socket * sock;
struct msghdr *msg, size_t size)
{
struct sock_iocb *si = kiocb_to_siocb(iocb);
- int err;
+ int err, len;
si->sock = sock;
si->scm = NULL;
if (err)
return err;
- return sock->ops->sendmsg(iocb, sock, msg, size);
+ len = sock->ops->sendmsg(iocb, sock, msg, size);
+ if (sock->sk) {
+ if (len == size)
+ vx_sock_send(sock->sk, size);
+ else
+ vx_sock_fail(sock->sk, size);
+ }
+ vxdprintk(VXD_CBIT(net, 7),
+ "__sock_sendmsg: %p[%p,%p,%p;%d]:%d/%d",
+ sock, sock->sk,
+ (sock->sk)?sock->sk->sk_nx_info:0,
+ (sock->sk)?sock->sk->sk_vx_info:0,
+ (sock->sk)?sock->sk->sk_xid:0,
+ (unsigned int)size, len);
+ return len;
}
int sock_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
{
struct kiocb iocb;
+ struct sock_iocb siocb;
int ret;
init_sync_kiocb(&iocb, NULL);
+ iocb.private = &siocb;
ret = __sock_sendmsg(&iocb, sock, msg, size);
if (-EIOCBQUEUED == ret)
ret = wait_on_sync_kiocb(&iocb);
return ret;
}
+int kernel_sendmsg(struct socket *sock, struct msghdr *msg,
+ struct kvec *vec, size_t num, size_t size)
+{
+ mm_segment_t oldfs = get_fs();
+ int result;
+
+ set_fs(KERNEL_DS);
+ /*
+ * the following is safe, since for compiler definitions of kvec and
+ * iovec are identical, yielding the same in-core layout and alignment
+ */
+ msg->msg_iov = (struct iovec *)vec,
+ msg->msg_iovlen = num;
+ result = sock_sendmsg(sock, msg, size);
+ set_fs(oldfs);
+ return result;
+}
static inline int __sock_recvmsg(struct kiocb *iocb, struct socket *sock,
struct msghdr *msg, size_t size, int flags)
{
- int err;
+ int err, len;
struct sock_iocb *si = kiocb_to_siocb(iocb);
si->sock = sock;
if (err)
return err;
- return sock->ops->recvmsg(iocb, sock, msg, size, flags);
+ len = sock->ops->recvmsg(iocb, sock, msg, size, flags);
+ if ((len >= 0) && sock->sk)
+ vx_sock_recv(sock->sk, len);
+ vxdprintk(VXD_CBIT(net, 7),
+ "__sock_recvmsg: %p[%p,%p,%p;%d]:%d/%d",
+ sock, sock->sk,
+ (sock->sk)?sock->sk->sk_nx_info:0,
+ (sock->sk)?sock->sk->sk_vx_info:0,
+ (sock->sk)?sock->sk->sk_xid:0,
+ (unsigned int)size, len);
+ return len;
}
int sock_recvmsg(struct socket *sock, struct msghdr *msg,
size_t size, int flags)
{
struct kiocb iocb;
+ struct sock_iocb siocb;
int ret;
init_sync_kiocb(&iocb, NULL);
+ iocb.private = &siocb;
ret = __sock_recvmsg(&iocb, sock, msg, size, flags);
if (-EIOCBQUEUED == ret)
ret = wait_on_sync_kiocb(&iocb);
return ret;
}
+int kernel_recvmsg(struct socket *sock, struct msghdr *msg,
+ struct kvec *vec, size_t num,
+ size_t size, int flags)
+{
+ mm_segment_t oldfs = get_fs();
+ int result;
+
+ set_fs(KERNEL_DS);
+ /*
+ * the following is safe, since for compiler definitions of kvec and
+ * iovec are identical, yielding the same in-core layout and alignment
+ */
+ msg->msg_iov = (struct iovec *)vec,
+ msg->msg_iovlen = num;
+ result = sock_recvmsg(sock, msg, size, flags);
+ set_fs(oldfs);
+ return result;
+}
+
+static void sock_aio_dtor(struct kiocb *iocb)
+{
+ kfree(iocb->private);
+}
+
/*
* Read data from a socket. ubuf is a user mode pointer. We make sure the user
* area ubuf...ubuf+size-1 is writable before asking the protocol.
static ssize_t sock_aio_read(struct kiocb *iocb, char __user *ubuf,
size_t size, loff_t pos)
{
- struct sock_iocb *x = kiocb_to_siocb(iocb);
+ struct sock_iocb *x, siocb;
struct socket *sock;
int flags;
if (size==0) /* Match SYS5 behaviour */
return 0;
+ if (is_sync_kiocb(iocb))
+ x = &siocb;
+ else {
+ x = kmalloc(sizeof(struct sock_iocb), GFP_KERNEL);
+ if (!x)
+ return -ENOMEM;
+ iocb->ki_dtor = sock_aio_dtor;
+ }
+ iocb->private = x;
+ x->kiocb = iocb;
sock = SOCKET_I(iocb->ki_filp->f_dentry->d_inode);
x->async_msg.msg_name = NULL;
static ssize_t sock_aio_write(struct kiocb *iocb, const char __user *ubuf,
size_t size, loff_t pos)
{
- struct sock_iocb *x = kiocb_to_siocb(iocb);
+ struct sock_iocb *x, siocb;
struct socket *sock;
if (pos != 0)
if(size==0) /* Match SYS5 behaviour */
return 0;
+ if (is_sync_kiocb(iocb))
+ x = &siocb;
+ else {
+ x = kmalloc(sizeof(struct sock_iocb), GFP_KERNEL);
+ if (!x)
+ return -ENOMEM;
+ iocb->ki_dtor = sock_aio_dtor;
+ }
+ iocb->private = x;
+ x->kiocb = iocb;
sock = SOCKET_I(iocb->ki_filp->f_dentry->d_inode);
x->async_msg.msg_name = NULL;
struct socket *sock;
int flags;
- if (ppos != &file->f_pos)
- return -ESPIPE;
-
sock = SOCKET_I(file->f_dentry->d_inode);
flags = !(file->f_flags & O_NONBLOCK) ? 0 : MSG_DONTWAIT;
return sock->ops->sendpage(sock, page, offset, size, flags);
}
-int sock_readv_writev(int type, struct inode * inode, struct file * file,
- const struct iovec * iov, long count, size_t size)
+static int sock_readv_writev(int type, struct inode * inode,
+ struct file * file, const struct iovec * iov,
+ long count, size_t size)
{
struct msghdr msg;
struct socket *sock;
*/
static DECLARE_MUTEX(br_ioctl_mutex);
-static int (*br_ioctl_hook)(unsigned long arg) = NULL;
+static int (*br_ioctl_hook)(unsigned int cmd, void __user *arg) = NULL;
-void brioctl_set(int (*hook)(unsigned long))
+void brioctl_set(int (*hook)(unsigned int, void __user *))
{
down(&br_ioctl_mutex);
br_ioctl_hook = hook;
EXPORT_SYMBOL(brioctl_set);
static DECLARE_MUTEX(vlan_ioctl_mutex);
-static int (*vlan_ioctl_hook)(unsigned long arg);
+static int (*vlan_ioctl_hook)(void __user *arg);
-void vlan_ioctl_set(int (*hook)(unsigned long))
+void vlan_ioctl_set(int (*hook)(void __user *))
{
down(&vlan_ioctl_mutex);
vlan_ioctl_hook = hook;
EXPORT_SYMBOL(vlan_ioctl_set);
static DECLARE_MUTEX(dlci_ioctl_mutex);
-static int (*dlci_ioctl_hook)(unsigned int, void *);
+static int (*dlci_ioctl_hook)(unsigned int, void __user *);
-void dlci_ioctl_set(int (*hook)(unsigned int, void *))
+void dlci_ioctl_set(int (*hook)(unsigned int, void __user *))
{
down(&dlci_ioctl_mutex);
dlci_ioctl_hook = hook;
* what to do with it - that's up to the protocol still.
*/
-static int sock_ioctl(struct inode *inode, struct file *file, unsigned int cmd,
- unsigned long arg)
+static long sock_ioctl(struct file *file, unsigned cmd, unsigned long arg)
{
struct socket *sock;
+ void __user *argp = (void __user *)arg;
int pid, err;
- unlock_kernel();
- sock = SOCKET_I(inode);
+ sock = SOCKET_I(file->f_dentry->d_inode);
if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) {
- err = dev_ioctl(cmd, (void *)arg);
+ err = dev_ioctl(cmd, argp);
} else
#ifdef WIRELESS_EXT
if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST) {
- err = dev_ioctl(cmd, (void *)arg);
+ err = dev_ioctl(cmd, argp);
} else
#endif /* WIRELESS_EXT */
switch (cmd) {
case FIOSETOWN:
case SIOCSPGRP:
err = -EFAULT;
- if (get_user(pid, (int *)arg))
+ if (get_user(pid, (int __user *)argp))
break;
err = f_setown(sock->file, pid, 1);
break;
case FIOGETOWN:
case SIOCGPGRP:
- err = put_user(sock->file->f_owner.pid, (int *)arg);
+ err = put_user(sock->file->f_owner.pid, (int __user *)argp);
break;
case SIOCGIFBR:
case SIOCSIFBR:
+ case SIOCBRADDBR:
+ case SIOCBRDELBR:
err = -ENOPKG;
if (!br_ioctl_hook)
request_module("bridge");
down(&br_ioctl_mutex);
if (br_ioctl_hook)
- err = br_ioctl_hook(arg);
+ err = br_ioctl_hook(cmd, argp);
up(&br_ioctl_mutex);
break;
case SIOCGIFVLAN:
down(&vlan_ioctl_mutex);
if (vlan_ioctl_hook)
- err = vlan_ioctl_hook(arg);
+ err = vlan_ioctl_hook(argp);
up(&vlan_ioctl_mutex);
break;
case SIOCGIFDIVERT:
case SIOCSIFDIVERT:
/* Convert this to call through a hook */
- err = divert_ioctl(cmd, (struct divert_cf *)arg);
+ err = divert_ioctl(cmd, argp);
break;
case SIOCADDDLCI:
case SIOCDELDLCI:
if (dlci_ioctl_hook) {
down(&dlci_ioctl_mutex);
- err = dlci_ioctl_hook(cmd, (void *)arg);
+ err = dlci_ioctl_hook(cmd, argp);
up(&dlci_ioctl_mutex);
}
break;
err = sock->ops->ioctl(sock, cmd, arg);
break;
}
- lock_kernel();
-
return err;
}
static int __sock_create(int family, int type, int protocol, struct socket **res, int kern)
{
- int i;
int err;
struct socket *sock;
if (type < 0 || type >= SOCK_MAX)
return -EINVAL;
+ /* disable IPv6 inside vservers for now */
+ if (family == PF_INET6 && !vx_check(0, VX_ADMIN))
+ return -EAFNOSUPPORT;
+
/* Compatibility.
This uglymoron is moved from INET layer to here to avoid
net_family_read_lock();
if (net_families[family] == NULL) {
- i = -EAFNOSUPPORT;
+ err = -EAFNOSUPPORT;
goto out;
}
* default.
*/
- if (!(sock = sock_alloc()))
- {
+ if (!(sock = sock_alloc())) {
printk(KERN_WARNING "socket: no more sockets\n");
- i = -ENFILE; /* Not exactly a match, but its the
+ err = -ENFILE; /* Not exactly a match, but its the
closest posix thing */
goto out;
}
* We will call the ->create function, that possibly is in a loadable
* module, so we have to bump that loadable module refcnt first.
*/
- i = -EAFNOSUPPORT;
+ err = -EAFNOSUPPORT;
if (!try_module_get(net_families[family]->owner))
goto out_release;
- if ((i = net_families[family]->create(sock, protocol)) < 0)
+ if ((err = net_families[family]->create(sock, protocol)) < 0)
goto out_module_put;
/*
* Now to bump the refcnt of the [loadable] module that owns this
out:
net_family_read_unlock();
- return i;
+ return err;
out_module_put:
module_put(net_families[family]->owner);
out_release:
if (retval < 0)
goto out;
+ set_bit(SOCK_USER_SOCKET, &sock->flags);
retval = sock_map_fd(sock);
if (retval < 0)
goto out_release;
err = sock_create(family, type, protocol, &sock1);
if (err < 0)
goto out;
+ set_bit(SOCK_USER_SOCKET, &sock1->flags);
err = sock_create(family, type, protocol, &sock2);
if (err < 0)
goto out_release_1;
+ set_bit(SOCK_USER_SOCKET, &sock2->flags);
err = sock1->ops->socketpair(sock1, sock2);
if (err < 0)
if (!sock)
goto out;
- err = -EMFILE;
+ err = -ENFILE;
if (!(newsock = sock_alloc()))
goto out_put;
return err;
}
+#ifdef __ARCH_WANT_SYS_SOCKETCALL
+
/* Argument list sizes for sys_socketcall */
#define AL(x) ((x) * sizeof(unsigned long))
static unsigned char nargs[18]={AL(0),AL(3),AL(3),AL(3),AL(2),AL(3),
return err;
}
+#endif /* __ARCH_WANT_SYS_SOCKETCALL */
+
/*
* This function is called by a protocol handler that wants to
* advertise its address family, and have it linked into the
void __init sock_init(void)
{
- int i;
-
- /*
- * Initialize all address (protocol) families.
- */
-
- for (i = 0; i < NPROTO; i++)
- net_families[i] = NULL;
-
/*
* Initialize sock SLAB cache.
*/
/* ABI emulation layers need these two */
EXPORT_SYMBOL(move_addr_to_kernel);
EXPORT_SYMBOL(move_addr_to_user);
-EXPORT_SYMBOL(sock_alloc);
-EXPORT_SYMBOL(sock_alloc_inode);
EXPORT_SYMBOL(sock_create);
EXPORT_SYMBOL(sock_create_kern);
EXPORT_SYMBOL(sock_create_lite);
EXPORT_SYMBOL(sock_unregister);
EXPORT_SYMBOL(sock_wake_async);
EXPORT_SYMBOL(sockfd_lookup);
+EXPORT_SYMBOL(kernel_sendmsg);
+EXPORT_SYMBOL(kernel_recvmsg);
git://git.onelab.eu / linux-2.6.git / blobdiff