Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-11-11 14:42:24 +00:00
Code Issues Packages Projects Releases Wiki Activity
cdb897e327
linux/net/atm/svc.c
David Howells cdfbabfb2f net: Work around lockdep limitation in sockets that use sockets 
Lockdep issues a circular dependency warning when AFS issues an operation
through AF_RXRPC from a context in which the VFS/VM holds the mmap_sem.

The theory lockdep comes up with is as follows:

 (1) If the pagefault handler decides it needs to read pages from AFS, it
     calls AFS with mmap_sem held and AFS begins an AF_RXRPC call, but
     creating a call requires the socket lock:

	mmap_sem must be taken before sk_lock-AF_RXRPC

 (2) afs_open_socket() opens an AF_RXRPC socket and binds it.  rxrpc_bind()
     binds the underlying UDP socket whilst holding its socket lock.
     inet_bind() takes its own socket lock:

	sk_lock-AF_RXRPC must be taken before sk_lock-AF_INET

 (3) Reading from a TCP socket into a userspace buffer might cause a fault
     and thus cause the kernel to take the mmap_sem, but the TCP socket is
     locked whilst doing this:

	sk_lock-AF_INET must be taken before mmap_sem

However, lockdep's theory is wrong in this instance because it deals only
with lock classes and not individual locks.  The AF_INET lock in (2) isn't
really equivalent to the AF_INET lock in (3) as the former deals with a
socket entirely internal to the kernel that never sees userspace.  This is
a limitation in the design of lockdep.

Fix the general case by:

 (1) Double up all the locking keys used in sockets so that one set are
     used if the socket is created by userspace and the other set is used
     if the socket is created by the kernel.

 (2) Store the kern parameter passed to sk_alloc() in a variable in the
     sock struct (sk_kern_sock).  This informs sock_lock_init(),
     sock_init_data() and sk_clone_lock() as to the lock keys to be used.

     Note that the child created by sk_clone_lock() inherits the parent's
     kern setting.

 (3) Add a 'kern' parameter to ->accept() that is analogous to the one
     passed in to ->create() that distinguishes whether kernel_accept() or
     sys_accept4() was the caller and can be passed to sk_alloc().

     Note that a lot of accept functions merely dequeue an already
     allocated socket.  I haven't touched these as the new socket already
     exists before we get the parameter.

     Note also that there are a couple of places where I've made the accepted
     socket unconditionally kernel-based:

	irda_accept()
	rds_rcp_accept_one()
	tcp_accept_from_sock()

     because they follow a sock_create_kern() and accept off of that.

Whilst creating this, I noticed that lustre and ocfs don't create sockets
through sock_create_kern() and thus they aren't marked as for-kernel,
though they appear to be internal.  I wonder if these should do that so
that they use the new set of lock keys.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-03-09 18:23:27 -08:00

692 lines
16 KiB
C
Raw Blame History

/* net/atm/svc.c - ATM SVC sockets */
/* Written 1995-2000 by Werner Almesberger, EPFL LRC/ICA */
#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__
#include <linux/string.h>
#include <linux/net.h> /* struct socket, struct proto_ops */
#include <linux/errno.h> /* error codes */
#include <linux/kernel.h> /* printk */
#include <linux/skbuff.h>
#include <linux/wait.h>
#include <linux/sched/signal.h>
#include <linux/fcntl.h> /* O_NONBLOCK */
#include <linux/init.h>
#include <linux/atm.h> /* ATM stuff */
#include <linux/atmsap.h>
#include <linux/atmsvc.h>
#include <linux/atmdev.h>
#include <linux/bitops.h>
#include <net/sock.h> /* for sock_no_* */
#include <linux/uaccess.h>
#include <linux/export.h>
#include "resources.h"
#include "common.h" /* common for PVCs and SVCs */
#include "signaling.h"
#include "addr.h"
static int svc_create(struct net *net, struct socket *sock, int protocol,
int kern);
/*
* Note: since all this is still nicely synchronized with the signaling demon,
* there's no need to protect sleep loops with clis. If signaling is
* moved into the kernel, that would change.
*/
static int svc_shutdown(struct socket *sock, int how)
{
return 0;
}
static void svc_disconnect(struct atm_vcc *vcc)
{
DEFINE_WAIT(wait);
struct sk_buff *skb;
struct sock *sk = sk_atm(vcc);
pr_debug("%p\n", vcc);
if (test_bit(ATM_VF_REGIS, &vcc->flags)) {
sigd_enq(vcc, as_close, NULL, NULL, NULL);
for (;;) {
prepare_to_wait(sk_sleep(sk), &wait, TASK_UNINTERRUPTIBLE);
if (test_bit(ATM_VF_RELEASED, &vcc->flags) || !sigd)
break;
schedule();
}
finish_wait(sk_sleep(sk), &wait);
}
/* beware - socket is still in use by atmsigd until the last
as_indicate has been answered */
while ((skb = skb_dequeue(&sk->sk_receive_queue)) != NULL) {
atm_return(vcc, skb->truesize);
pr_debug("LISTEN REL\n");
sigd_enq2(NULL, as_reject, vcc, NULL, NULL, &vcc->qos, 0);
dev_kfree_skb(skb);
}
clear_bit(ATM_VF_REGIS, &vcc->flags);
/* ... may retry later */
}
static int svc_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct atm_vcc *vcc;
if (sk) {
vcc = ATM_SD(sock);
pr_debug("%p\n", vcc);
clear_bit(ATM_VF_READY, &vcc->flags);
/*
* VCC pointer is used as a reference,
* so we must not free it (thereby subjecting it to re-use)
* before all pending connections are closed
*/
svc_disconnect(vcc);
vcc_release(sock);
}
return 0;
}
static int svc_bind(struct socket *sock, struct sockaddr *sockaddr,
int sockaddr_len)
{
DEFINE_WAIT(wait);
struct sock *sk = sock->sk;
struct sockaddr_atmsvc *addr;
struct atm_vcc *vcc;
int error;
if (sockaddr_len != sizeof(struct sockaddr_atmsvc))
return -EINVAL;
lock_sock(sk);
if (sock->state == SS_CONNECTED) {
error = -EISCONN;
goto out;
}
if (sock->state != SS_UNCONNECTED) {
error = -EINVAL;
goto out;
}
vcc = ATM_SD(sock);
addr = (struct sockaddr_atmsvc *) sockaddr;
if (addr->sas_family != AF_ATMSVC) {
error = -EAFNOSUPPORT;
goto out;
}
clear_bit(ATM_VF_BOUND, &vcc->flags);
/* failing rebind will kill old binding */
/* @@@ check memory (de)allocation on rebind */
if (!test_bit(ATM_VF_HASQOS, &vcc->flags)) {
error = -EBADFD;
goto out;
}
vcc->local = *addr;
set_bit(ATM_VF_WAITING, &vcc->flags);
sigd_enq(vcc, as_bind, NULL, NULL, &vcc->local);
for (;;) {
prepare_to_wait(sk_sleep(sk), &wait, TASK_UNINTERRUPTIBLE);
if (!test_bit(ATM_VF_WAITING, &vcc->flags) || !sigd)
break;
schedule();
}
finish_wait(sk_sleep(sk), &wait);
clear_bit(ATM_VF_REGIS, &vcc->flags); /* doesn't count */
if (!sigd) {
error = -EUNATCH;
goto out;
}
if (!sk->sk_err)
set_bit(ATM_VF_BOUND, &vcc->flags);
error = -sk->sk_err;
out:
release_sock(sk);
return error;
}
static int svc_connect(struct socket *sock, struct sockaddr *sockaddr,
int sockaddr_len, int flags)
{
DEFINE_WAIT(wait);
struct sock *sk = sock->sk;
struct sockaddr_atmsvc *addr;
struct atm_vcc *vcc = ATM_SD(sock);
int error;
pr_debug("%p\n", vcc);
lock_sock(sk);
if (sockaddr_len != sizeof(struct sockaddr_atmsvc)) {
error = -EINVAL;
goto out;
}
switch (sock->state) {
default:
error = -EINVAL;
goto out;
case SS_CONNECTED:
error = -EISCONN;
goto out;
case SS_CONNECTING:
if (test_bit(ATM_VF_WAITING, &vcc->flags)) {
error = -EALREADY;
goto out;
}
sock->state = SS_UNCONNECTED;
if (sk->sk_err) {
error = -sk->sk_err;
goto out;
}
break;
case SS_UNCONNECTED:
addr = (struct sockaddr_atmsvc *) sockaddr;
if (addr->sas_family != AF_ATMSVC) {
error = -EAFNOSUPPORT;
goto out;
}
if (!test_bit(ATM_VF_HASQOS, &vcc->flags)) {
error = -EBADFD;
goto out;
}
if (vcc->qos.txtp.traffic_class == ATM_ANYCLASS ||
vcc->qos.rxtp.traffic_class == ATM_ANYCLASS) {
error = -EINVAL;
goto out;
}
if (!vcc->qos.txtp.traffic_class &&
!vcc->qos.rxtp.traffic_class) {
error = -EINVAL;
goto out;
}
vcc->remote = *addr;
set_bit(ATM_VF_WAITING, &vcc->flags);
sigd_enq(vcc, as_connect, NULL, NULL, &vcc->remote);
if (flags & O_NONBLOCK) {
sock->state = SS_CONNECTING;
error = -EINPROGRESS;
goto out;
}
error = 0;
prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
while (test_bit(ATM_VF_WAITING, &vcc->flags) && sigd) {
schedule();
if (!signal_pending(current)) {
prepare_to_wait(sk_sleep(sk), &wait,
TASK_INTERRUPTIBLE);
continue;
}
pr_debug("*ABORT*\n");
/*
* This is tricky:
* Kernel ---close--> Demon
* Kernel <--close--- Demon
* or
* Kernel ---close--> Demon
* Kernel <--error--- Demon
* or
* Kernel ---close--> Demon
* Kernel <--okay---- Demon
* Kernel <--close--- Demon
*/
sigd_enq(vcc, as_close, NULL, NULL, NULL);
while (test_bit(ATM_VF_WAITING, &vcc->flags) && sigd) {
prepare_to_wait(sk_sleep(sk), &wait,
TASK_INTERRUPTIBLE);
schedule();
}
if (!sk->sk_err)
while (!test_bit(ATM_VF_RELEASED, &vcc->flags) &&
sigd) {
prepare_to_wait(sk_sleep(sk), &wait,
TASK_INTERRUPTIBLE);
schedule();
}
clear_bit(ATM_VF_REGIS, &vcc->flags);
clear_bit(ATM_VF_RELEASED, &vcc->flags);
clear_bit(ATM_VF_CLOSE, &vcc->flags);
/* we're gone now but may connect later */
error = -EINTR;
break;
}
finish_wait(sk_sleep(sk), &wait);
if (error)
goto out;
if (!sigd) {
error = -EUNATCH;
goto out;
}
if (sk->sk_err) {
error = -sk->sk_err;
goto out;
}
}
vcc->qos.txtp.max_pcr = SELECT_TOP_PCR(vcc->qos.txtp);
vcc->qos.txtp.pcr = 0;
vcc->qos.txtp.min_pcr = 0;
error = vcc_connect(sock, vcc->itf, vcc->vpi, vcc->vci);
if (!error)
sock->state = SS_CONNECTED;
else
(void)svc_disconnect(vcc);
out:
release_sock(sk);
return error;
}
static int svc_listen(struct socket *sock, int backlog)
{
DEFINE_WAIT(wait);
struct sock *sk = sock->sk;
struct atm_vcc *vcc = ATM_SD(sock);
int error;
pr_debug("%p\n", vcc);
lock_sock(sk);
/* let server handle listen on unbound sockets */
if (test_bit(ATM_VF_SESSION, &vcc->flags)) {
error = -EINVAL;
goto out;
}
if (test_bit(ATM_VF_LISTEN, &vcc->flags)) {
error = -EADDRINUSE;
goto out;
}
set_bit(ATM_VF_WAITING, &vcc->flags);
sigd_enq(vcc, as_listen, NULL, NULL, &vcc->local);
for (;;) {
prepare_to_wait(sk_sleep(sk), &wait, TASK_UNINTERRUPTIBLE);
if (!test_bit(ATM_VF_WAITING, &vcc->flags) || !sigd)
break;
schedule();
}
finish_wait(sk_sleep(sk), &wait);
if (!sigd) {
error = -EUNATCH;
goto out;
}
set_bit(ATM_VF_LISTEN, &vcc->flags);
vcc_insert_socket(sk);
sk->sk_max_ack_backlog = backlog > 0 ? backlog : ATM_BACKLOG_DEFAULT;
error = -sk->sk_err;
out:
release_sock(sk);
return error;
}
static int svc_accept(struct socket *sock, struct socket *newsock, int flags,
bool kern)
{
struct sock *sk = sock->sk;
struct sk_buff *skb;
struct atmsvc_msg *msg;
struct atm_vcc *old_vcc = ATM_SD(sock);
struct atm_vcc *new_vcc;
int error;
lock_sock(sk);
error = svc_create(sock_net(sk), newsock, 0, kern);
if (error)
goto out;
new_vcc = ATM_SD(newsock);
pr_debug("%p -> %p\n", old_vcc, new_vcc);
while (1) {
DEFINE_WAIT(wait);
prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
while (!(skb = skb_dequeue(&sk->sk_receive_queue)) &&
sigd) {
if (test_bit(ATM_VF_RELEASED, &old_vcc->flags))
break;
if (test_bit(ATM_VF_CLOSE, &old_vcc->flags)) {
error = -sk->sk_err;
break;
}
if (flags & O_NONBLOCK) {
error = -EAGAIN;
break;
}
release_sock(sk);
schedule();
lock_sock(sk);
if (signal_pending(current)) {
error = -ERESTARTSYS;
break;
}
prepare_to_wait(sk_sleep(sk), &wait,
TASK_INTERRUPTIBLE);
}
finish_wait(sk_sleep(sk), &wait);
if (error)
goto out;
if (!skb) {
error = -EUNATCH;
goto out;
}
msg = (struct atmsvc_msg *)skb->data;
new_vcc->qos = msg->qos;
set_bit(ATM_VF_HASQOS, &new_vcc->flags);
new_vcc->remote = msg->svc;
new_vcc->local = msg->local;
new_vcc->sap = msg->sap;
error = vcc_connect(newsock, msg->pvc.sap_addr.itf,
msg->pvc.sap_addr.vpi,
msg->pvc.sap_addr.vci);
dev_kfree_skb(skb);
sk->sk_ack_backlog--;
if (error) {
sigd_enq2(NULL, as_reject, old_vcc, NULL, NULL,
&old_vcc->qos, error);
error = error == -EAGAIN ? -EBUSY : error;
goto out;
}
/* wait should be short, so we ignore the non-blocking flag */
set_bit(ATM_VF_WAITING, &new_vcc->flags);
sigd_enq(new_vcc, as_accept, old_vcc, NULL, NULL);
for (;;) {
prepare_to_wait(sk_sleep(sk_atm(new_vcc)), &wait,
TASK_UNINTERRUPTIBLE);
if (!test_bit(ATM_VF_WAITING, &new_vcc->flags) || !sigd)
break;
release_sock(sk);
schedule();
lock_sock(sk);
}
finish_wait(sk_sleep(sk_atm(new_vcc)), &wait);
if (!sigd) {
error = -EUNATCH;
goto out;
}
if (!sk_atm(new_vcc)->sk_err)
break;
if (sk_atm(new_vcc)->sk_err != ERESTARTSYS) {
error = -sk_atm(new_vcc)->sk_err;
goto out;
}
}
newsock->state = SS_CONNECTED;
out:
release_sock(sk);
return error;
}
static int svc_getname(struct socket *sock, struct sockaddr *sockaddr,
int *sockaddr_len, int peer)
{
struct sockaddr_atmsvc *addr;
*sockaddr_len = sizeof(struct sockaddr_atmsvc);
addr = (struct sockaddr_atmsvc *) sockaddr;
memcpy(addr, peer ? &ATM_SD(sock)->remote : &ATM_SD(sock)->local,
sizeof(struct sockaddr_atmsvc));
return 0;
}
int svc_change_qos(struct atm_vcc *vcc, struct atm_qos *qos)
{
struct sock *sk = sk_atm(vcc);
DEFINE_WAIT(wait);
set_bit(ATM_VF_WAITING, &vcc->flags);
sigd_enq2(vcc, as_modify, NULL, NULL, &vcc->local, qos, 0);
for (;;) {
prepare_to_wait(sk_sleep(sk), &wait, TASK_UNINTERRUPTIBLE);
if (!test_bit(ATM_VF_WAITING, &vcc->flags) ||
test_bit(ATM_VF_RELEASED, &vcc->flags) || !sigd) {
break;
}
schedule();
}
finish_wait(sk_sleep(sk), &wait);
if (!sigd)
return -EUNATCH;
return -sk->sk_err;
}
static int svc_setsockopt(struct socket *sock, int level, int optname,
char __user *optval, unsigned int optlen)
{
struct sock *sk = sock->sk;
struct atm_vcc *vcc = ATM_SD(sock);
int value, error = 0;
lock_sock(sk);
switch (optname) {
case SO_ATMSAP:
if (level != SOL_ATM || optlen != sizeof(struct atm_sap)) {
error = -EINVAL;
goto out;
}
if (copy_from_user(&vcc->sap, optval, optlen)) {
error = -EFAULT;
goto out;
}
set_bit(ATM_VF_HASSAP, &vcc->flags);
break;
case SO_MULTIPOINT:
if (level != SOL_ATM || optlen != sizeof(int)) {
error = -EINVAL;
goto out;
}
if (get_user(value, (int __user *)optval)) {
error = -EFAULT;
goto out;
}
if (value == 1)
set_bit(ATM_VF_SESSION, &vcc->flags);
else if (value == 0)
clear_bit(ATM_VF_SESSION, &vcc->flags);
else
error = -EINVAL;
break;
default:
error = vcc_setsockopt(sock, level, optname, optval, optlen);
}
out:
release_sock(sk);
return error;
}
static int svc_getsockopt(struct socket *sock, int level, int optname,
char __user *optval, int __user *optlen)
{
struct sock *sk = sock->sk;
int error = 0, len;
lock_sock(sk);
if (!__SO_LEVEL_MATCH(optname, level) || optname != SO_ATMSAP) {
error = vcc_getsockopt(sock, level, optname, optval, optlen);
goto out;
}
if (get_user(len, optlen)) {
error = -EFAULT;
goto out;
}
if (len != sizeof(struct atm_sap)) {
error = -EINVAL;
goto out;
}
if (copy_to_user(optval, &ATM_SD(sock)->sap, sizeof(struct atm_sap))) {
error = -EFAULT;
goto out;
}
out:
release_sock(sk);
return error;
}
static int svc_addparty(struct socket *sock, struct sockaddr *sockaddr,
int sockaddr_len, int flags)
{
DEFINE_WAIT(wait);
struct sock *sk = sock->sk;
struct atm_vcc *vcc = ATM_SD(sock);
int error;
lock_sock(sk);
set_bit(ATM_VF_WAITING, &vcc->flags);
sigd_enq(vcc, as_addparty, NULL, NULL,
(struct sockaddr_atmsvc *) sockaddr);
if (flags & O_NONBLOCK) {
error = -EINPROGRESS;
goto out;
}
pr_debug("added wait queue\n");
for (;;) {
prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
if (!test_bit(ATM_VF_WAITING, &vcc->flags) || !sigd)
break;
schedule();
}
finish_wait(sk_sleep(sk), &wait);
error = -xchg(&sk->sk_err_soft, 0);
out:
release_sock(sk);
return error;
}
static int svc_dropparty(struct socket *sock, int ep_ref)
{
DEFINE_WAIT(wait);
struct sock *sk = sock->sk;
struct atm_vcc *vcc = ATM_SD(sock);
int error;
lock_sock(sk);
set_bit(ATM_VF_WAITING, &vcc->flags);
sigd_enq2(vcc, as_dropparty, NULL, NULL, NULL, NULL, ep_ref);
for (;;) {
prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
if (!test_bit(ATM_VF_WAITING, &vcc->flags) || !sigd)
break;
schedule();
}
finish_wait(sk_sleep(sk), &wait);
if (!sigd) {
error = -EUNATCH;
goto out;
}
error = -xchg(&sk->sk_err_soft, 0);
out:
release_sock(sk);
return error;
}
static int svc_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
int error, ep_ref;
struct sockaddr_atmsvc sa;
struct atm_vcc *vcc = ATM_SD(sock);
switch (cmd) {
case ATM_ADDPARTY:
if (!test_bit(ATM_VF_SESSION, &vcc->flags))
return -EINVAL;
if (copy_from_user(&sa, (void __user *) arg, sizeof(sa)))
return -EFAULT;
error = svc_addparty(sock, (struct sockaddr *)&sa, sizeof(sa),
0);
break;
case ATM_DROPPARTY:
if (!test_bit(ATM_VF_SESSION, &vcc->flags))
return -EINVAL;
if (copy_from_user(&ep_ref, (void __user *) arg, sizeof(int)))
return -EFAULT;
error = svc_dropparty(sock, ep_ref);
break;
default:
error = vcc_ioctl(sock, cmd, arg);
}
return error;
}
#ifdef CONFIG_COMPAT
static int svc_compat_ioctl(struct socket *sock, unsigned int cmd,
unsigned long arg)
{
/* The definition of ATM_ADDPARTY uses the size of struct atm_iobuf.
But actually it takes a struct sockaddr_atmsvc, which doesn't need
compat handling. So all we have to do is fix up cmd... */
if (cmd == COMPAT_ATM_ADDPARTY)
cmd = ATM_ADDPARTY;
if (cmd == ATM_ADDPARTY || cmd == ATM_DROPPARTY)
return svc_ioctl(sock, cmd, arg);
else
return vcc_compat_ioctl(sock, cmd, arg);
}
#endif /* CONFIG_COMPAT */
static const struct proto_ops svc_proto_ops = {
.family = PF_ATMSVC,
.owner = THIS_MODULE,
.release = svc_release,
.bind = svc_bind,
.connect = svc_connect,
.socketpair = sock_no_socketpair,
.accept = svc_accept,
.getname = svc_getname,
.poll = vcc_poll,
.ioctl = svc_ioctl,
#ifdef CONFIG_COMPAT
.compat_ioctl = svc_compat_ioctl,
#endif
.listen = svc_listen,
.shutdown = svc_shutdown,
.setsockopt = svc_setsockopt,
.getsockopt = svc_getsockopt,
.sendmsg = vcc_sendmsg,
.recvmsg = vcc_recvmsg,
.mmap = sock_no_mmap,
.sendpage = sock_no_sendpage,
};
static int svc_create(struct net *net, struct socket *sock, int protocol,
int kern)
{
int error;
if (!net_eq(net, &init_net))
return -EAFNOSUPPORT;
sock->ops = &svc_proto_ops;
error = vcc_create(net, sock, protocol, AF_ATMSVC, kern);
if (error)
return error;
ATM_SD(sock)->local.sas_family = AF_ATMSVC;
ATM_SD(sock)->remote.sas_family = AF_ATMSVC;
return 0;
}
static const struct net_proto_family svc_family_ops = {
.family = PF_ATMSVC,
.create = svc_create,
.owner = THIS_MODULE,
};
/*
* Initialize the ATM SVC protocol family
*/
int __init atmsvc_init(void)
{
return sock_register(&svc_family_ops);
}
void atmsvc_exit(void)
{
sock_unregister(PF_ATMSVC);
}
Reference in New Issue View Git Blame Copy Permalink