freebsd-nq/sys/dev/streams/streams.c
David Malone e1419c08e2 falloc allocates a file structure and adds it to the file descriptor
table, acquiring the necessary locks as it works. It usually returns
two references to the new descriptor: one in the descriptor table
and one via a pointer argument.

As falloc releases the FILEDESC lock before returning, there is a
potential for a process to close the reference in the file descriptor
table before falloc's caller gets to use the file. I don't think this
can happen in practice at the moment, because Giant indirectly protects
closes.

To stop the file being completly closed in this situation, this change
makes falloc set the refcount to two when both references are returned.
This makes life easier for several of falloc's callers, because the
first thing they previously did was grab an extra reference on the
file.

Reviewed by:	iedowse
Idea run past:	jhb
2003-10-19 20:41:07 +00:00

426 lines
10 KiB
C

/*
* Copyright (c) 1998 Mark Newton
* Copyright (c) 1994 Christos Zoulas
* Copyright (c) 1997 Todd Vierling
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. The names of the authors may not be used to endorse or promote products
* derived from this software without specific prior written permission
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* Stolen from NetBSD /sys/compat/svr4/svr4_net.c. Pseudo-device driver
* skeleton produced from /usr/share/examples/drivers/make_pseudo_driver.sh
* in 3.0-980524-SNAP then hacked a bit (but probably not enough :-).
*
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h> /* SYSINIT stuff */
#include <sys/conf.h> /* cdevsw stuff */
#include <sys/malloc.h> /* malloc region definitions */
#include <sys/file.h>
#include <sys/filedesc.h>
#include <sys/unistd.h>
#include <sys/fcntl.h>
#include <sys/socket.h>
#include <sys/protosw.h>
#include <sys/socketvar.h>
#include <sys/un.h>
#include <sys/domain.h>
#include <net/if.h>
#include <netinet/in.h>
#include <sys/proc.h>
#include <sys/uio.h>
#include <sys/sysproto.h>
#include <compat/svr4/svr4_types.h>
#include <compat/svr4/svr4_util.h>
#include <compat/svr4/svr4_signal.h>
#include <compat/svr4/svr4_ioctl.h>
#include <compat/svr4/svr4_stropts.h>
#include <compat/svr4/svr4_socket.h>
static int svr4_soo_close(struct file *, struct thread *);
static int svr4_ptm_alloc(struct thread *);
static d_open_t streamsopen;
struct svr4_sockcache_head svr4_head;
/* Initialization flag (set/queried by svr4_mod LKM) */
int svr4_str_initialized = 0;
/*
* Device minor numbers
*/
enum {
dev_ptm = 10,
dev_arp = 26,
dev_icmp = 27,
dev_ip = 28,
dev_tcp = 35,
dev_udp = 36,
dev_rawip = 37,
dev_unix_dgram = 38,
dev_unix_stream = 39,
dev_unix_ord_stream = 40
};
static dev_t dt_ptm, dt_arp, dt_icmp, dt_ip, dt_tcp, dt_udp, dt_rawip,
dt_unix_dgram, dt_unix_stream, dt_unix_ord_stream;
static struct fileops svr4_netops = {
.fo_read = soo_read,
.fo_write = soo_write,
.fo_ioctl = soo_ioctl,
.fo_poll = soo_poll,
.fo_kqfilter = soo_kqfilter,
.fo_stat = soo_stat,
.fo_close = svr4_soo_close
};
#define CDEV_MAJOR 103
static struct cdevsw streams_cdevsw = {
.d_open = streamsopen,
.d_name = "streams",
.d_maj = CDEV_MAJOR,
};
struct streams_softc {
struct isa_device *dev;
} ;
#define UNIT(dev) minor(dev) /* assume one minor number per unit */
typedef struct streams_softc *sc_p;
static int
streams_modevent(module_t mod, int type, void *unused)
{
switch (type) {
case MOD_LOAD:
/* XXX should make sure it isn't already loaded first */
dt_ptm = make_dev(&streams_cdevsw, dev_ptm, 0, 0, 0666,
"ptm");
dt_arp = make_dev(&streams_cdevsw, dev_arp, 0, 0, 0666,
"arp");
dt_icmp = make_dev(&streams_cdevsw, dev_icmp, 0, 0, 0666,
"icmp");
dt_ip = make_dev(&streams_cdevsw, dev_ip, 0, 0, 0666,
"ip");
dt_tcp = make_dev(&streams_cdevsw, dev_tcp, 0, 0, 0666,
"tcp");
dt_udp = make_dev(&streams_cdevsw, dev_udp, 0, 0, 0666,
"udp");
dt_rawip = make_dev(&streams_cdevsw, dev_rawip, 0, 0, 0666,
"rawip");
dt_unix_dgram = make_dev(&streams_cdevsw, dev_unix_dgram,
0, 0, 0666, "ticlts");
dt_unix_stream = make_dev(&streams_cdevsw, dev_unix_stream,
0, 0, 0666, "ticots");
dt_unix_ord_stream = make_dev(&streams_cdevsw,
dev_unix_ord_stream, 0, 0, 0666, "ticotsord");
if (! (dt_ptm && dt_arp && dt_icmp && dt_ip && dt_tcp &&
dt_udp && dt_rawip && dt_unix_dgram &&
dt_unix_stream && dt_unix_ord_stream)) {
printf("WARNING: device config for STREAMS failed\n");
printf("Suggest unloading streams KLD\n");
}
return 0;
case MOD_UNLOAD:
/* XXX should check to see if it's busy first */
destroy_dev(dt_ptm);
destroy_dev(dt_arp);
destroy_dev(dt_icmp);
destroy_dev(dt_ip);
destroy_dev(dt_tcp);
destroy_dev(dt_udp);
destroy_dev(dt_rawip);
destroy_dev(dt_unix_dgram);
destroy_dev(dt_unix_stream);
destroy_dev(dt_unix_ord_stream);
return 0;
default:
break;
}
return 0;
}
static moduledata_t streams_mod = {
"streams",
streams_modevent,
0
};
DECLARE_MODULE(streams, streams_mod, SI_SUB_DRIVERS, SI_ORDER_ANY);
MODULE_VERSION(streams, 1);
/*
* We only need open() and close() routines. open() calls socreate()
* to allocate a "real" object behind the stream and mallocs some state
* info for use by the svr4 emulator; close() deallocates the state
* information and passes the underlying object to the normal socket close
* routine.
*/
static int
streamsopen(dev_t dev, int oflags, int devtype, struct thread *td)
{
int type, protocol;
int fd, extraref;
struct file *fp;
struct socket *so;
int error;
int family;
struct proc *p = td->td_proc;
PROC_LOCK(p);
if (td->td_dupfd >= 0) {
PROC_UNLOCK(p);
return ENODEV;
}
PROC_UNLOCK(p);
switch (minor(dev)) {
case dev_udp:
family = AF_INET;
type = SOCK_DGRAM;
protocol = IPPROTO_UDP;
break;
case dev_tcp:
family = AF_INET;
type = SOCK_STREAM;
protocol = IPPROTO_TCP;
break;
case dev_ip:
case dev_rawip:
family = AF_INET;
type = SOCK_RAW;
protocol = IPPROTO_IP;
break;
case dev_icmp:
family = AF_INET;
type = SOCK_RAW;
protocol = IPPROTO_ICMP;
break;
case dev_unix_dgram:
family = AF_LOCAL;
type = SOCK_DGRAM;
protocol = 0;
break;
case dev_unix_stream:
case dev_unix_ord_stream:
family = AF_LOCAL;
type = SOCK_STREAM;
protocol = 0;
break;
case dev_ptm:
return svr4_ptm_alloc(td);
default:
return EOPNOTSUPP;
}
if ((error = falloc(td, &fp, &fd)) != 0)
return error;
/* An extra reference on `fp' has been held for us by falloc(). */
if ((error = socreate(family, &so, type, protocol,
td->td_ucred, td)) != 0) {
FILEDESC_LOCK(p->p_fd);
/* Check the fd table entry hasn't changed since we made it. */
extraref = 0;
if (p->p_fd->fd_ofiles[fd] == fp) {
p->p_fd->fd_ofiles[fd] = NULL;
extraref = 1;
}
FILEDESC_UNLOCK(p->p_fd);
if (extraref)
fdrop(fp, td);
fdrop(fp, td);
return error;
}
FILEDESC_LOCK(p->p_fd);
fp->f_data = so;
fp->f_flag = FREAD|FWRITE;
fp->f_ops = &svr4_netops;
fp->f_type = DTYPE_SOCKET;
FILEDESC_UNLOCK(p->p_fd);
(void)svr4_stream_get(fp);
fdrop(fp, td);
PROC_LOCK(p);
td->td_dupfd = fd;
PROC_UNLOCK(p);
return ENXIO;
}
static int
svr4_ptm_alloc(td)
struct thread *td;
{
struct proc *p = td->td_proc;
/*
* XXX this is very, very ugly. But I can't find a better
* way that won't duplicate a big amount of code from
* sys_open(). Ho hum...
*
* Fortunately for us, Solaris (at least 2.5.1) makes the
* /dev/ptmx open automatically just open a pty, that (after
* STREAMS I_PUSHes), is just a plain pty. fstat() is used
* to get the minor device number to map to a tty.
*
* Cycle through the names. If sys_open() returns ENOENT (or
* ENXIO), short circuit the cycle and exit.
*/
static char ptyname[] = "/dev/ptyXX";
static char ttyletters[] = "pqrstuwxyzPQRST";
static char ttynumbers[] = "0123456789abcdef";
caddr_t sg = stackgap_init();
char *path = stackgap_alloc(&sg, sizeof(ptyname));
struct open_args oa;
int l = 0, n = 0;
register_t fd = -1;
int error;
oa.path = path;
oa.flags = O_RDWR;
oa.mode = 0;
while (fd == -1) {
ptyname[8] = ttyletters[l];
ptyname[9] = ttynumbers[n];
if ((error = copyout(ptyname, path, sizeof(ptyname))) != 0)
return error;
switch (error = open(td, &oa)) {
case ENOENT:
case ENXIO:
return error;
case 0:
PROC_LOCK(p);
td->td_dupfd = td->td_retval[0];
PROC_UNLOCK(p);
return ENXIO;
default:
if (ttynumbers[++n] == '\0') {
if (ttyletters[++l] == '\0')
break;
n = 0;
}
}
}
return ENOENT;
}
struct svr4_strm *
svr4_stream_get(fp)
struct file *fp;
{
struct socket *so;
struct svr4_strm *st;
if (fp == NULL || fp->f_type != DTYPE_SOCKET)
return NULL;
so = fp->f_data;
/*
* mpfixme: lock socketbuffer here
*/
if (so->so_emuldata) {
return so->so_emuldata;
}
/* Allocate a new one. */
st = malloc(sizeof(struct svr4_strm), M_TEMP, M_WAITOK);
st->s_family = so->so_proto->pr_domain->dom_family;
st->s_cmd = ~0;
st->s_afd = -1;
st->s_eventmask = 0;
/*
* avoid a race where we loose due to concurrancy issues
* of two threads trying to allocate the so_emuldata.
*/
if (so->so_emuldata) {
/* lost the race, use the existing emuldata */
FREE(st, M_TEMP);
st = so->so_emuldata;
} else {
/* we won, or there was no race, use our copy */
so->so_emuldata = st;
fp->f_ops = &svr4_netops;
}
return st;
}
void
svr4_delete_socket(p, fp)
struct proc *p;
struct file *fp;
{
struct svr4_sockcache_entry *e;
void *cookie = ((struct socket *)fp->f_data)->so_emuldata;
while (svr4_str_initialized != 2) {
if (atomic_cmpset_acq_int(&svr4_str_initialized, 0, 1)) {
TAILQ_INIT(&svr4_head);
atomic_store_rel_int(&svr4_str_initialized, 2);
}
return;
}
TAILQ_FOREACH(e, &svr4_head, entries)
if (e->p == p && e->cookie == cookie) {
TAILQ_REMOVE(&svr4_head, e, entries);
DPRINTF(("svr4_delete_socket: %s [%p,%d,%d]\n",
e->sock.sun_path, p, (int)e->dev, e->ino));
free(e, M_TEMP);
return;
}
}
static int
svr4_soo_close(struct file *fp, struct thread *td)
{
struct socket *so = fp->f_data;
/* CHECKUNIT_DIAG(ENXIO);*/
svr4_delete_socket(td->td_proc, fp);
free(so->so_emuldata, M_TEMP);
return soo_close(fp, td);
}