X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=net%2Fsocket.c;h=1afd1123400620fcb554170ac0877c80de90d643;hb=6a77f38946aaee1cd85eeec6cf4229b204c15071;hp=57cbf98d3d82bbb02ed31d8b9bb20557547beaa4;hpb=9213980e6a70d8473e0ffd4b39ab5b6caaba9ff5;p=linux-2.6.git diff --git a/net/socket.c b/net/socket.c index 57cbf98d3..1afd11234 100644 --- a/net/socket.c +++ b/net/socket.c @@ -93,7 +93,6 @@ #include #include -#include #include static int sock_no_open(struct inode *irrelevant, struct file *dontcare); @@ -106,7 +105,7 @@ static int sock_mmap(struct file *file, struct vm_area_struct * vma); 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, @@ -128,7 +127,7 @@ static struct file_operations socket_file_ops = { .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, @@ -146,7 +145,7 @@ static struct net_proto_family *net_families[NPROTO]; #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 @@ -403,7 +402,7 @@ int sock_map_fd(struct socket *sock) 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); @@ -461,7 +460,7 @@ struct socket *sockfd_lookup(int fd, int *err) * NULL is returned. */ -struct socket *sock_alloc(void) +static struct socket *sock_alloc(void) { struct inode * inode; struct socket * sock; @@ -551,27 +550,47 @@ static inline int __sock_sendmsg(struct kiocb *iocb, struct socket *sock, else vx_sock_fail(sock->sk, size); } - vxdprintk("__sock_sendmsg: %p[%p,%p,%p;%d]:%d/%d\n", + 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, - size, len); + (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) @@ -592,12 +611,13 @@ static inline int __sock_recvmsg(struct kiocb *iocb, struct socket *sock, len = sock->ops->recvmsg(iocb, sock, msg, size, flags); if ((len >= 0) && sock->sk) vx_sock_recv(sock->sk, len); - vxdprintk("__sock_recvmsg: %p[%p,%p,%p;%d]:%d/%d\n", + 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, - size, len); + (unsigned int)size, len); return len; } @@ -605,15 +625,41 @@ 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. @@ -622,7 +668,7 @@ int sock_recvmsg(struct socket *sock, struct msghdr *msg, 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; @@ -631,6 +677,16 @@ static ssize_t sock_aio_read(struct kiocb *iocb, char __user *ubuf, 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; @@ -655,7 +711,7 @@ static ssize_t sock_aio_read(struct kiocb *iocb, char __user *ubuf, 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) @@ -663,6 +719,16 @@ static ssize_t sock_aio_write(struct kiocb *iocb, const char __user *ubuf, 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; @@ -686,9 +752,6 @@ ssize_t sock_sendpage(struct file *file, struct page *page, 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; @@ -698,8 +761,9 @@ ssize_t sock_sendpage(struct file *file, struct page *page, 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; @@ -790,15 +854,13 @@ EXPORT_SYMBOL(dlci_ioctl_set); * 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, argp); } else @@ -864,8 +926,6 @@ static int sock_ioctl(struct inode *inode, struct file *file, unsigned int cmd, err = sock->ops->ioctl(sock, cmd, arg); break; } - lock_kernel(); - return err; } @@ -1035,7 +1095,6 @@ int sock_wake_async(struct socket *sock, int how, int band) static int __sock_create(int family, int type, int protocol, struct socket **res, int kern) { - int i; int err; struct socket *sock; @@ -1084,7 +1143,7 @@ static int __sock_create(int family, int type, int protocol, struct socket **res net_family_read_lock(); if (net_families[family] == NULL) { - i = -EAFNOSUPPORT; + err = -EAFNOSUPPORT; goto out; } @@ -1094,10 +1153,9 @@ static int __sock_create(int family, int type, int protocol, struct socket **res * 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; } @@ -1108,11 +1166,11 @@ static int __sock_create(int family, int type, int protocol, struct socket **res * 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 @@ -1132,7 +1190,7 @@ static int __sock_create(int family, int type, int protocol, struct socket **res out: net_family_read_unlock(); - return i; + return err; out_module_put: module_put(net_families[family]->owner); out_release: @@ -1323,7 +1381,7 @@ asmlinkage long sys_accept(int fd, struct sockaddr __user *upeer_sockaddr, int _ if (!sock) goto out; - err = -EMFILE; + err = -ENFILE; if (!(newsock = sock_alloc())) goto out_put; @@ -2000,15 +2058,6 @@ extern void sk_init(void); 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. */ @@ -2058,8 +2107,6 @@ void socket_seq_show(struct seq_file *seq) /* 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); @@ -2071,3 +2118,5 @@ EXPORT_SYMBOL(sock_sendmsg); EXPORT_SYMBOL(sock_unregister); EXPORT_SYMBOL(sock_wake_async); EXPORT_SYMBOL(sockfd_lookup); +EXPORT_SYMBOL(kernel_sendmsg); +EXPORT_SYMBOL(kernel_recvmsg);