Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/* $NetBSD: lock_proc.c,v 1.7 2000/10/11 20:23:56 is Exp $ */
|
|
|
|
/* $FreeBSD$ */
|
|
|
|
/*
|
|
|
|
* Copyright (c) 1995
|
|
|
|
* A.R. Gordon (andrew.gordon@net-tel.co.uk). 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. All advertising materials mentioning features or use of this software
|
|
|
|
* must display the following acknowledgement:
|
|
|
|
* This product includes software developed for the FreeBSD project
|
|
|
|
* 4. Neither the name of the author nor the names of any co-contributors
|
|
|
|
* may be used to endorse or promote products derived from this software
|
|
|
|
* without specific prior written permission.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY ANDREW GORDON AND CONTRIBUTORS ``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 OR CONTRIBUTORS 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.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <sys/cdefs.h>
|
|
|
|
#ifndef lint
|
|
|
|
__RCSID("$NetBSD: lock_proc.c,v 1.7 2000/10/11 20:23:56 is Exp $");
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#include <sys/param.h>
|
|
|
|
#include <sys/socket.h>
|
|
|
|
|
|
|
|
#include <netinet/in.h>
|
|
|
|
#include <arpa/inet.h>
|
|
|
|
|
|
|
|
#include <netdb.h>
|
|
|
|
#include <stdio.h>
|
|
|
|
#include <string.h>
|
|
|
|
#include <syslog.h>
|
|
|
|
#include <netconfig.h>
|
|
|
|
|
|
|
|
#include <rpc/rpc.h>
|
|
|
|
#include <rpcsvc/sm_inter.h>
|
|
|
|
|
|
|
|
#include "lockd.h"
|
|
|
|
#include <rpcsvc/nlm_prot.h>
|
|
|
|
#include "lockd_lock.h"
|
|
|
|
|
|
|
|
|
|
|
|
#define CLIENT_CACHE_SIZE 64 /* No. of client sockets cached */
|
|
|
|
#define CLIENT_CACHE_LIFETIME 120 /* In seconds */
|
|
|
|
|
2002-03-22 19:20:35 +00:00
|
|
|
static void log_from_addr(const char *, struct svc_req *);
|
2002-03-21 22:52:45 +00:00
|
|
|
static void log_netobj(netobj *obj);
|
|
|
|
static int addrcmp(struct sockaddr *, struct sockaddr *);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
|
|
|
|
/* log_from_addr ----------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Log name of function called and source address
|
|
|
|
* Returns: Nothing
|
|
|
|
* Notes: Extracts the source address from the transport handle
|
|
|
|
* passed in as part of the called procedure specification
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
log_from_addr(fun_name, req)
|
2002-03-22 19:20:35 +00:00
|
|
|
const char *fun_name;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *req;
|
|
|
|
{
|
|
|
|
struct sockaddr *addr;
|
|
|
|
char hostname_buf[NI_MAXHOST];
|
|
|
|
|
|
|
|
addr = svc_getrpccaller(req->rq_xprt)->buf;
|
|
|
|
if (getnameinfo(addr , addr->sa_len, hostname_buf, sizeof hostname_buf,
|
|
|
|
NULL, 0, 0) != 0)
|
|
|
|
return;
|
|
|
|
|
|
|
|
syslog(LOG_DEBUG, "%s from %s", fun_name, hostname_buf);
|
|
|
|
}
|
|
|
|
|
2001-10-14 18:36:35 +00:00
|
|
|
/* log_netobj ----------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Log a netobj
|
|
|
|
* Returns: Nothing
|
|
|
|
* Notes: This function should only really be called as part of
|
|
|
|
* a debug subsystem.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
log_netobj(obj)
|
|
|
|
netobj *obj;
|
|
|
|
{
|
|
|
|
char objvalbuffer[(sizeof(char)*2)*MAX_NETOBJ_SZ+2];
|
|
|
|
char objascbuffer[sizeof(char)*MAX_NETOBJ_SZ+1];
|
2001-11-13 11:24:23 +00:00
|
|
|
unsigned int i, maxlen;
|
2001-10-14 18:36:35 +00:00
|
|
|
char *tmp1, *tmp2;
|
|
|
|
|
|
|
|
/* Notify of potential security attacks */
|
|
|
|
if (obj->n_len > MAX_NETOBJ_SZ) {
|
|
|
|
syslog(LOG_DEBUG, "SOMEONE IS TRYING TO DO SOMETHING NASTY!\n");
|
|
|
|
syslog(LOG_DEBUG, "netobj too large! Should be %d was %d\n",
|
|
|
|
MAX_NETOBJ_SZ, obj->n_len);
|
|
|
|
}
|
|
|
|
/* Prevent the security hazard from the buffer overflow */
|
|
|
|
maxlen = (obj->n_len < MAX_NETOBJ_SZ ? obj->n_len : MAX_NETOBJ_SZ);
|
|
|
|
for (i=0, tmp1 = objvalbuffer, tmp2 = objascbuffer; i < obj->n_len;
|
|
|
|
i++, tmp1 +=2, tmp2 +=1) {
|
|
|
|
sprintf(tmp1,"%02X",*(obj->n_bytes+i));
|
|
|
|
sprintf(tmp2,"%c",*(obj->n_bytes+i));
|
|
|
|
}
|
|
|
|
*tmp1 = '\0';
|
|
|
|
*tmp2 = '\0';
|
|
|
|
syslog(LOG_DEBUG,"netobjvals: %s\n",objvalbuffer);
|
|
|
|
syslog(LOG_DEBUG,"netobjascs: %s\n",objascbuffer);
|
|
|
|
}
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/* get_client -------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Get a CLIENT* for making RPC calls to lockd on given host
|
|
|
|
* Returns: CLIENT* pointer, from clnt_udp_create, or NULL if error
|
|
|
|
* Notes: Creating a CLIENT* is quite expensive, involving a
|
|
|
|
* conversation with the remote portmapper to get the
|
|
|
|
* port number. Since a given client is quite likely
|
|
|
|
* to make several locking requests in succession, it is
|
|
|
|
* desirable to cache the created CLIENT*.
|
|
|
|
*
|
|
|
|
* Since we are using UDP rather than TCP, there is no cost
|
|
|
|
* to the remote system in keeping these cached indefinitely.
|
|
|
|
* Unfortunately there is a snag: if the remote system
|
|
|
|
* reboots, the cached portmapper results will be invalid,
|
|
|
|
* and we will never detect this since all of the xxx_msg()
|
|
|
|
* calls return no result - we just fire off a udp packet
|
|
|
|
* and hope for the best.
|
|
|
|
*
|
|
|
|
* We solve this by discarding cached values after two
|
|
|
|
* minutes, regardless of whether they have been used
|
|
|
|
* in the meanwhile (since a bad one might have been used
|
|
|
|
* plenty of times, as the host keeps retrying the request
|
|
|
|
* and we keep sending the reply back to the wrong port).
|
|
|
|
*
|
|
|
|
* Given that the entries will always expire in the order
|
|
|
|
* that they were created, there is no point in a LRU
|
|
|
|
* algorithm for when the cache gets full - entries are
|
|
|
|
* always re-used in sequence.
|
|
|
|
*/
|
|
|
|
static CLIENT *clnt_cache_ptr[CLIENT_CACHE_SIZE];
|
|
|
|
static long clnt_cache_time[CLIENT_CACHE_SIZE]; /* time entry created */
|
|
|
|
static struct sockaddr_storage clnt_cache_addr[CLIENT_CACHE_SIZE];
|
2001-04-28 04:26:32 +00:00
|
|
|
static rpcvers_t clnt_cache_vers[CLIENT_CACHE_SIZE];
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
static int clnt_cache_next_to_use = 0;
|
|
|
|
|
|
|
|
static int
|
|
|
|
addrcmp(sa1, sa2)
|
|
|
|
struct sockaddr *sa1;
|
|
|
|
struct sockaddr *sa2;
|
|
|
|
{
|
|
|
|
int len;
|
|
|
|
void *p1, *p2;
|
|
|
|
|
|
|
|
if (sa1->sa_family != sa2->sa_family)
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
switch (sa1->sa_family) {
|
|
|
|
case AF_INET:
|
|
|
|
p1 = &((struct sockaddr_in *)sa1)->sin_addr;
|
|
|
|
p2 = &((struct sockaddr_in *)sa2)->sin_addr;
|
|
|
|
len = 4;
|
|
|
|
break;
|
|
|
|
case AF_INET6:
|
|
|
|
p1 = &((struct sockaddr_in6 *)sa1)->sin6_addr;
|
|
|
|
p2 = &((struct sockaddr_in6 *)sa2)->sin6_addr;
|
|
|
|
len = 16;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
return memcmp(p1, p2, len);
|
|
|
|
}
|
|
|
|
|
|
|
|
CLIENT *
|
|
|
|
get_client(host_addr, vers)
|
|
|
|
struct sockaddr *host_addr;
|
|
|
|
rpcvers_t vers;
|
|
|
|
{
|
|
|
|
CLIENT *client;
|
|
|
|
struct timeval retry_time, time_now;
|
|
|
|
int i;
|
2002-03-22 20:00:10 +00:00
|
|
|
const char *netid;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct netconfig *nconf;
|
|
|
|
char host[NI_MAXHOST];
|
|
|
|
|
|
|
|
gettimeofday(&time_now, NULL);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Search for the given client in the cache, zapping any expired
|
|
|
|
* entries that we happen to notice in passing.
|
|
|
|
*/
|
|
|
|
for (i = 0; i < CLIENT_CACHE_SIZE; i++) {
|
|
|
|
client = clnt_cache_ptr[i];
|
|
|
|
if (client && ((clnt_cache_time[i] + CLIENT_CACHE_LIFETIME)
|
|
|
|
< time_now.tv_sec)) {
|
|
|
|
/* Cache entry has expired. */
|
|
|
|
if (debug_level > 3)
|
|
|
|
syslog(LOG_DEBUG, "Expired CLIENT* in cache");
|
|
|
|
clnt_cache_time[i] = 0L;
|
|
|
|
clnt_destroy(client);
|
|
|
|
clnt_cache_ptr[i] = NULL;
|
|
|
|
client = NULL;
|
|
|
|
}
|
|
|
|
if (client && !addrcmp((struct sockaddr *)&clnt_cache_addr[i],
|
2001-04-28 04:26:32 +00:00
|
|
|
host_addr) && clnt_cache_vers[i] == vers) {
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/* Found it! */
|
|
|
|
if (debug_level > 3)
|
|
|
|
syslog(LOG_DEBUG, "Found CLIENT* in cache");
|
|
|
|
return (client);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2001-04-28 04:26:32 +00:00
|
|
|
if (debug_level > 3)
|
|
|
|
syslog(LOG_DEBUG, "CLIENT* not found in cache, creating");
|
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/* Not found in cache. Free the next entry if it is in use. */
|
|
|
|
if (clnt_cache_ptr[clnt_cache_next_to_use]) {
|
|
|
|
clnt_destroy(clnt_cache_ptr[clnt_cache_next_to_use]);
|
|
|
|
clnt_cache_ptr[clnt_cache_next_to_use] = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Need a host string for clnt_tp_create. Use NI_NUMERICHOST
|
|
|
|
* to avoid DNS lookups.
|
|
|
|
*/
|
|
|
|
if (getnameinfo(host_addr, host_addr->sa_len, host, sizeof host,
|
|
|
|
NULL, 0, NI_NUMERICHOST) != 0) {
|
|
|
|
syslog(LOG_ERR, "unable to get name string for caller");
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
#if 1
|
|
|
|
if (host_addr->sa_family == AF_INET6)
|
|
|
|
netid = "udp6";
|
|
|
|
else
|
|
|
|
netid = "udp";
|
|
|
|
#else
|
|
|
|
if (host_addr->sa_family == AF_INET6)
|
|
|
|
netid = "tcp6";
|
|
|
|
else
|
|
|
|
netid = "tcp";
|
|
|
|
#endif
|
|
|
|
nconf = getnetconfigent(netid);
|
|
|
|
if (nconf == NULL) {
|
|
|
|
syslog(LOG_ERR, "could not get netconfig info for '%s': "
|
|
|
|
"no /etc/netconfig file?", netid);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
client = clnt_tp_create(host, NLM_PROG, vers, nconf);
|
|
|
|
freenetconfigent(nconf);
|
|
|
|
|
|
|
|
if (!client) {
|
|
|
|
syslog(LOG_ERR, "%s", clnt_spcreateerror("clntudp_create"));
|
|
|
|
syslog(LOG_ERR, "Unable to return result to %s", host);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Success - update the cache entry */
|
|
|
|
clnt_cache_ptr[clnt_cache_next_to_use] = client;
|
|
|
|
memcpy(&clnt_cache_addr[clnt_cache_next_to_use], host_addr,
|
|
|
|
host_addr->sa_len);
|
2001-04-28 04:26:32 +00:00
|
|
|
clnt_cache_vers[clnt_cache_next_to_use] = vers;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
clnt_cache_time[clnt_cache_next_to_use] = time_now.tv_sec;
|
|
|
|
if (++clnt_cache_next_to_use > CLIENT_CACHE_SIZE)
|
|
|
|
clnt_cache_next_to_use = 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Disable the default timeout, so we can specify our own in calls
|
|
|
|
* to clnt_call(). (Note that the timeout is a different concept
|
|
|
|
* from the retry period set in clnt_udp_create() above.)
|
|
|
|
*/
|
|
|
|
retry_time.tv_sec = -1;
|
|
|
|
retry_time.tv_usec = -1;
|
|
|
|
clnt_control(client, CLSET_TIMEOUT, (char *)&retry_time);
|
|
|
|
|
|
|
|
if (debug_level > 3)
|
|
|
|
syslog(LOG_DEBUG, "Created CLIENT* for %s", host);
|
|
|
|
return client;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* transmit_result --------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Transmit result for nlm_xxx_msg pseudo-RPCs
|
|
|
|
* Returns: Nothing - we have no idea if the datagram got there
|
|
|
|
* Notes: clnt_call() will always fail (with timeout) as we are
|
|
|
|
* calling it with timeout 0 as a hack to just issue a datagram
|
|
|
|
* without expecting a result
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
transmit_result(opcode, result, addr)
|
|
|
|
int opcode;
|
|
|
|
nlm_res *result;
|
|
|
|
struct sockaddr *addr;
|
|
|
|
{
|
|
|
|
static char dummy;
|
|
|
|
CLIENT *cli;
|
|
|
|
struct timeval timeo;
|
|
|
|
int success;
|
|
|
|
|
|
|
|
if ((cli = get_client(addr, NLM_VERS)) != NULL) {
|
|
|
|
timeo.tv_sec = 0; /* No timeout - not expecting response */
|
|
|
|
timeo.tv_usec = 0;
|
|
|
|
|
|
|
|
success = clnt_call(cli, opcode, xdr_nlm_res, result, xdr_void,
|
|
|
|
&dummy, timeo);
|
|
|
|
|
|
|
|
if (debug_level > 2)
|
|
|
|
syslog(LOG_DEBUG, "clnt_call returns %d(%s)",
|
|
|
|
success, clnt_sperrno(success));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
/* transmit4_result --------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Transmit result for nlm4_xxx_msg pseudo-RPCs
|
|
|
|
* Returns: Nothing - we have no idea if the datagram got there
|
|
|
|
* Notes: clnt_call() will always fail (with timeout) as we are
|
|
|
|
* calling it with timeout 0 as a hack to just issue a datagram
|
|
|
|
* without expecting a result
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
transmit4_result(opcode, result, addr)
|
|
|
|
int opcode;
|
|
|
|
nlm4_res *result;
|
|
|
|
struct sockaddr *addr;
|
|
|
|
{
|
|
|
|
static char dummy;
|
|
|
|
CLIENT *cli;
|
|
|
|
struct timeval timeo;
|
|
|
|
int success;
|
|
|
|
|
|
|
|
if ((cli = get_client(addr, NLM_VERS4)) != NULL) {
|
|
|
|
timeo.tv_sec = 0; /* No timeout - not expecting response */
|
|
|
|
timeo.tv_usec = 0;
|
|
|
|
|
|
|
|
success = clnt_call(cli, opcode, xdr_nlm4_res, result, xdr_void,
|
|
|
|
&dummy, timeo);
|
|
|
|
|
|
|
|
if (debug_level > 2)
|
|
|
|
syslog(LOG_DEBUG, "clnt_call returns %d(%s)",
|
|
|
|
success, clnt_sperrno(success));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* converts a struct nlm_lock to struct nlm4_lock
|
|
|
|
*/
|
2002-03-21 22:52:45 +00:00
|
|
|
static void nlmtonlm4(struct nlm_lock *, struct nlm4_lock *);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
static void
|
|
|
|
nlmtonlm4(arg, arg4)
|
|
|
|
struct nlm_lock *arg;
|
|
|
|
struct nlm4_lock *arg4;
|
|
|
|
{
|
|
|
|
memcpy(arg4, arg, sizeof(nlm_lock));
|
|
|
|
arg4->l_offset = arg->l_offset;
|
|
|
|
arg4->l_len = arg->l_len;
|
|
|
|
}
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Functions for Unix<->Unix locking (ie. monitored locking, with rpc.statd
|
|
|
|
* involved to ensure reclaim of locks after a crash of the "stateless"
|
|
|
|
* server.
|
|
|
|
*
|
|
|
|
* These all come in two flavours - nlm_xxx() and nlm_xxx_msg().
|
|
|
|
* The first are standard RPCs with argument and result.
|
|
|
|
* The nlm_xxx_msg() calls implement exactly the same functions, but
|
|
|
|
* use two pseudo-RPCs (one in each direction). These calls are NOT
|
|
|
|
* standard use of the RPC protocol in that they do not return a result
|
|
|
|
* at all (NB. this is quite different from returning a void result).
|
|
|
|
* The effect of this is to make the nlm_xxx_msg() calls simple unacknowledged
|
|
|
|
* datagrams, requiring higher-level code to perform retries.
|
|
|
|
*
|
|
|
|
* Despite the disadvantages of the nlm_xxx_msg() approach (some of which
|
|
|
|
* are documented in the comments to get_client() above), this is the
|
|
|
|
* interface used by all current commercial NFS implementations
|
|
|
|
* [Solaris, SCO, AIX etc.]. This is presumed to be because these allow
|
|
|
|
* implementations to continue using the standard RPC libraries, while
|
|
|
|
* avoiding the block-until-result nature of the library interface.
|
|
|
|
*
|
|
|
|
* No client implementations have been identified so far that make use
|
|
|
|
* of the true RPC version (early SunOS releases would be a likely candidate
|
|
|
|
* for testing).
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* nlm_test ---------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Test whether a specified lock would be granted if requested
|
|
|
|
* Returns: nlm_granted (or error code)
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm_testres *
|
|
|
|
nlm_test_1_svc(arg, rqstp)
|
|
|
|
nlm_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_testres res;
|
|
|
|
struct nlm4_lock arg4;
|
|
|
|
struct nlm4_holder *holder;
|
|
|
|
nlmtonlm4(&arg->alock, &arg4);
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_test", rqstp);
|
|
|
|
|
2001-10-14 18:36:35 +00:00
|
|
|
holder = testlock(&arg4, arg->exclusive, 0);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/*
|
|
|
|
* Copy the cookie from the argument into the result. Note that this
|
|
|
|
* is slightly hazardous, as the structure contains a pointer to a
|
|
|
|
* malloc()ed buffer that will get freed by the caller. However, the
|
|
|
|
* main function transmits the result before freeing the argument
|
|
|
|
* so it is in fact safe.
|
|
|
|
*/
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
if (holder == NULL) {
|
|
|
|
res.stat.stat = nlm_granted;
|
|
|
|
} else {
|
|
|
|
res.stat.stat = nlm_denied;
|
|
|
|
memcpy(&res.stat.nlm_testrply_u.holder, holder,
|
|
|
|
sizeof(struct nlm_holder));
|
|
|
|
res.stat.nlm_testrply_u.holder.l_offset = holder->l_offset;
|
|
|
|
res.stat.nlm_testrply_u.holder.l_len = holder->l_len;
|
|
|
|
}
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm_test_msg_1_svc(arg, rqstp)
|
|
|
|
nlm_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
nlm_testres res;
|
|
|
|
static char dummy;
|
|
|
|
struct sockaddr *addr;
|
|
|
|
CLIENT *cli;
|
|
|
|
int success;
|
|
|
|
struct timeval timeo;
|
|
|
|
struct nlm4_lock arg4;
|
|
|
|
struct nlm4_holder *holder;
|
|
|
|
|
|
|
|
nlmtonlm4(&arg->alock, &arg4);
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_test_msg", rqstp);
|
|
|
|
|
2001-10-14 18:36:35 +00:00
|
|
|
holder = testlock(&arg4, arg->exclusive, 0);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
if (holder == NULL) {
|
|
|
|
res.stat.stat = nlm_granted;
|
|
|
|
} else {
|
|
|
|
res.stat.stat = nlm_denied;
|
|
|
|
memcpy(&res.stat.nlm_testrply_u.holder, holder,
|
|
|
|
sizeof(struct nlm_holder));
|
|
|
|
res.stat.nlm_testrply_u.holder.l_offset = holder->l_offset;
|
|
|
|
res.stat.nlm_testrply_u.holder.l_len = holder->l_len;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* nlm_test has different result type to the other operations, so
|
|
|
|
* can't use transmit_result() in this case
|
|
|
|
*/
|
|
|
|
addr = svc_getrpccaller(rqstp->rq_xprt)->buf;
|
|
|
|
if ((cli = get_client(addr, NLM_VERS)) != NULL) {
|
|
|
|
timeo.tv_sec = 0; /* No timeout - not expecting response */
|
|
|
|
timeo.tv_usec = 0;
|
|
|
|
|
|
|
|
success = clnt_call(cli, NLM_TEST_RES, xdr_nlm_testres,
|
|
|
|
&res, xdr_void, &dummy, timeo);
|
|
|
|
|
|
|
|
if (debug_level > 2)
|
|
|
|
syslog(LOG_DEBUG, "clnt_call returns %d", success);
|
|
|
|
}
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_lock ---------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purposes: Establish a lock
|
|
|
|
* Returns: granted, denied or blocked
|
|
|
|
* Notes: *** grace period support missing
|
|
|
|
*/
|
|
|
|
nlm_res *
|
|
|
|
nlm_lock_1_svc(arg, rqstp)
|
|
|
|
nlm_lockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
struct nlm4_lockargs arg4;
|
|
|
|
nlmtonlm4(&arg->alock, &arg4.alock);
|
|
|
|
arg4.cookie = arg->cookie;
|
|
|
|
arg4.block = arg->block;
|
|
|
|
arg4.exclusive = arg->exclusive;
|
|
|
|
arg4.reclaim = arg->reclaim;
|
|
|
|
arg4.state = arg->state;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_lock", rqstp);
|
|
|
|
|
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
res.stat.stat = getlock(&arg4, rqstp, LOCK_MON);
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm_lock_msg_1_svc(arg, rqstp)
|
|
|
|
nlm_lockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
struct nlm4_lockargs arg4;
|
|
|
|
|
|
|
|
nlmtonlm4(&arg->alock, &arg4.alock);
|
|
|
|
arg4.cookie = arg->cookie;
|
|
|
|
arg4.block = arg->block;
|
|
|
|
arg4.exclusive = arg->exclusive;
|
|
|
|
arg4.reclaim = arg->reclaim;
|
|
|
|
arg4.state = arg->state;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_lock_msg", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat.stat = getlock(&arg4, rqstp, LOCK_ASYNC | LOCK_MON);
|
|
|
|
transmit_result(NLM_LOCK_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_cancel -------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Cancel a blocked lock request
|
|
|
|
* Returns: granted or denied
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm_res *
|
|
|
|
nlm_cancel_1_svc(arg, rqstp)
|
|
|
|
nlm_cancargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
struct nlm4_lock arg4;
|
|
|
|
|
|
|
|
nlmtonlm4(&arg->alock, &arg4);
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_cancel", rqstp);
|
|
|
|
|
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Since at present we never return 'nlm_blocked', there can never be
|
|
|
|
* a lock to cancel, so this call always fails.
|
|
|
|
*/
|
|
|
|
res.stat.stat = unlock(&arg4, LOCK_CANCEL);
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm_cancel_msg_1_svc(arg, rqstp)
|
|
|
|
nlm_cancargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
struct nlm4_lock arg4;
|
|
|
|
|
|
|
|
nlmtonlm4(&arg->alock, &arg4);
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_cancel_msg", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
/*
|
|
|
|
* Since at present we never return 'nlm_blocked', there can never be
|
|
|
|
* a lock to cancel, so this call always fails.
|
|
|
|
*/
|
|
|
|
res.stat.stat = unlock(&arg4, LOCK_CANCEL);
|
|
|
|
transmit_result(NLM_CANCEL_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_unlock -------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Release an existing lock
|
|
|
|
* Returns: Always granted, unless during grace period
|
|
|
|
* Notes: "no such lock" error condition is ignored, as the
|
|
|
|
* protocol uses unreliable UDP datagrams, and may well
|
|
|
|
* re-try an unlock that has already succeeded.
|
|
|
|
*/
|
|
|
|
nlm_res *
|
|
|
|
nlm_unlock_1_svc(arg, rqstp)
|
|
|
|
nlm_unlockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
struct nlm4_lock arg4;
|
|
|
|
|
|
|
|
nlmtonlm4(&arg->alock, &arg4);
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_unlock", rqstp);
|
|
|
|
|
|
|
|
res.stat.stat = unlock(&arg4, 0);
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm_unlock_msg_1_svc(arg, rqstp)
|
|
|
|
nlm_unlockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
struct nlm4_lock arg4;
|
|
|
|
|
|
|
|
nlmtonlm4(&arg->alock, &arg4);
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_unlock_msg", rqstp);
|
|
|
|
|
|
|
|
res.stat.stat = unlock(&arg4, 0);
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
transmit_result(NLM_UNLOCK_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Client-side pseudo-RPCs for results. Note that for the client there
|
|
|
|
* are only nlm_xxx_msg() versions of each call, since the 'real RPC'
|
|
|
|
* version returns the results in the RPC result, and so the client
|
|
|
|
* does not normally receive incoming RPCs.
|
|
|
|
*
|
|
|
|
* The exception to this is nlm_granted(), which is genuinely an RPC
|
|
|
|
* call from the server to the client - a 'call-back' in normal procedure
|
|
|
|
* call terms.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* nlm_granted ------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Receive notification that formerly blocked lock now granted
|
|
|
|
* Returns: always success ('granted')
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm_res *
|
|
|
|
nlm_granted_1_svc(arg, rqstp)
|
|
|
|
nlm_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_granted", rqstp);
|
|
|
|
|
2001-04-17 20:45:23 +00:00
|
|
|
res.stat.stat = lock_answer(arg->alock.svid, &arg->cookie,
|
|
|
|
nlm_granted, NULL, NLM_VERS) == 0 ?
|
|
|
|
nlm_granted : nlm_denied;
|
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm_granted_msg_1_svc(arg, rqstp)
|
|
|
|
nlm_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_granted_msg", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
2003-05-14 13:50:40 +00:00
|
|
|
res.stat.stat = lock_answer(arg->alock.svid, &arg->cookie,
|
|
|
|
nlm_granted, NULL, NLM_VERS) == 0 ?
|
|
|
|
nlm_granted : nlm_denied;
|
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
transmit_result(NLM_GRANTED_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_test_res ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_test_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_test_res_1_svc(arg, rqstp)
|
|
|
|
nlm_testres *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_test_res", rqstp);
|
2001-04-17 20:45:23 +00:00
|
|
|
(void)lock_answer(-1, &arg->cookie, arg->stat.stat,
|
|
|
|
&arg->stat.nlm_testrply_u.holder.svid, NLM_VERS);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_lock_res ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_lock_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_lock_res_1_svc(arg, rqstp)
|
|
|
|
nlm_res *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_lock_res", rqstp);
|
|
|
|
|
2001-04-17 20:45:23 +00:00
|
|
|
(void)lock_answer(-1, &arg->cookie, arg->stat.stat, NULL, NLM_VERS);
|
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_cancel_res ---------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_cancel_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_cancel_res_1_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
nlm_res *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_cancel_res", rqstp);
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_unlock_res ---------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_unlock_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_unlock_res_1_svc(arg, rqstp)
|
|
|
|
nlm_res *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_unlock_res", rqstp);
|
2001-04-17 20:45:23 +00:00
|
|
|
|
2001-11-12 16:34:59 +00:00
|
|
|
lock_answer(-1, &arg->cookie, arg->stat.stat, NULL, NLM_VERS);
|
2001-04-17 20:45:23 +00:00
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_granted_res --------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_granted_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_granted_res_1_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
nlm_res *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_granted_res", rqstp);
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Calls for PCNFS locking (aka non-monitored locking, no involvement
|
|
|
|
* of rpc.statd).
|
|
|
|
*
|
|
|
|
* These are all genuine RPCs - no nlm_xxx_msg() nonsense here.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* nlm_share --------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Establish a DOS-style lock
|
|
|
|
* Returns: success or failure
|
|
|
|
* Notes: Blocking locks are not supported - client is expected
|
|
|
|
* to retry if required.
|
|
|
|
*/
|
|
|
|
nlm_shareres *
|
|
|
|
nlm_share_3_svc(arg, rqstp)
|
|
|
|
nlm_shareargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_shareres res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_share", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat = nlm_granted;
|
|
|
|
res.sequence = 1234356; /* X/Open says this field is ignored? */
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_unshare ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Release a DOS-style lock
|
|
|
|
* Returns: nlm_granted, unless in grace period
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm_shareres *
|
|
|
|
nlm_unshare_3_svc(arg, rqstp)
|
|
|
|
nlm_shareargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_shareres res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_unshare", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat = nlm_granted;
|
|
|
|
res.sequence = 1234356; /* X/Open says this field is ignored? */
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_nm_lock ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: non-monitored version of nlm_lock()
|
|
|
|
* Returns: as for nlm_lock()
|
|
|
|
* Notes: These locks are in the same style as the standard nlm_lock,
|
|
|
|
* but the rpc.statd should not be called to establish a
|
|
|
|
* monitor for the client machine, since that machine is
|
|
|
|
* declared not to be running a rpc.statd, and so would not
|
|
|
|
* respond to the statd protocol.
|
|
|
|
*/
|
|
|
|
nlm_res *
|
|
|
|
nlm_nm_lock_3_svc(arg, rqstp)
|
|
|
|
nlm_lockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_nm_lock", rqstp);
|
|
|
|
|
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat.stat = nlm_granted;
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_free_all ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Release all locks held by a named client
|
|
|
|
* Returns: Nothing
|
|
|
|
* Notes: Potential denial of service security problem here - the
|
|
|
|
* locks to be released are specified by a host name, independent
|
|
|
|
* of the address from which the request has arrived.
|
|
|
|
* Should probably be rejected if the named host has been
|
|
|
|
* using monitored locks.
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_free_all_3_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
nlm_notify *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static char dummy;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_free_all", rqstp);
|
|
|
|
return (&dummy);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* calls for nlm version 4 (NFSv3) */
|
|
|
|
/* nlm_test ---------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Test whether a specified lock would be granted if requested
|
|
|
|
* Returns: nlm_granted (or error code)
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm4_testres *
|
|
|
|
nlm4_test_4_svc(arg, rqstp)
|
|
|
|
nlm4_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_testres res;
|
|
|
|
struct nlm4_holder *holder;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_test", rqstp);
|
2001-10-14 18:36:35 +00:00
|
|
|
if (debug_level > 5) {
|
|
|
|
syslog(LOG_DEBUG, "Locking arguments:\n");
|
|
|
|
log_netobj(&(arg->cookie));
|
|
|
|
syslog(LOG_DEBUG, "Alock arguments:\n");
|
|
|
|
syslog(LOG_DEBUG, "Caller Name: %s\n",arg->alock.caller_name);
|
|
|
|
syslog(LOG_DEBUG, "File Handle:\n");
|
|
|
|
log_netobj(&(arg->alock.fh));
|
|
|
|
syslog(LOG_DEBUG, "Owner Handle:\n");
|
|
|
|
log_netobj(&(arg->alock.oh));
|
|
|
|
syslog(LOG_DEBUG, "SVID: %d\n", arg->alock.svid);
|
2001-11-13 11:24:23 +00:00
|
|
|
syslog(LOG_DEBUG, "Lock Offset: %llu\n",
|
|
|
|
(unsigned long long)arg->alock.l_offset);
|
|
|
|
syslog(LOG_DEBUG, "Lock Length: %llu\n",
|
|
|
|
(unsigned long long)arg->alock.l_len);
|
2001-10-14 18:36:35 +00:00
|
|
|
syslog(LOG_DEBUG, "Exclusive: %s\n",
|
|
|
|
(arg->exclusive ? "true" : "false"));
|
|
|
|
}
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
|
2001-10-14 18:36:35 +00:00
|
|
|
holder = testlock(&arg->alock, arg->exclusive, LOCK_V4);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Copy the cookie from the argument into the result. Note that this
|
|
|
|
* is slightly hazardous, as the structure contains a pointer to a
|
|
|
|
* malloc()ed buffer that will get freed by the caller. However, the
|
|
|
|
* main function transmits the result before freeing the argument
|
|
|
|
* so it is in fact safe.
|
|
|
|
*/
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
if (holder == NULL) {
|
|
|
|
res.stat.stat = nlm4_granted;
|
|
|
|
} else {
|
|
|
|
res.stat.stat = nlm4_denied;
|
|
|
|
memcpy(&res.stat.nlm4_testrply_u.holder, holder,
|
|
|
|
sizeof(struct nlm4_holder));
|
|
|
|
}
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm4_test_msg_4_svc(arg, rqstp)
|
|
|
|
nlm4_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
nlm4_testres res;
|
|
|
|
static char dummy;
|
|
|
|
struct sockaddr *addr;
|
|
|
|
CLIENT *cli;
|
|
|
|
int success;
|
|
|
|
struct timeval timeo;
|
|
|
|
struct nlm4_holder *holder;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_test_msg", rqstp);
|
|
|
|
|
2001-10-14 18:36:35 +00:00
|
|
|
holder = testlock(&arg->alock, arg->exclusive, LOCK_V4);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
if (holder == NULL) {
|
|
|
|
res.stat.stat = nlm4_granted;
|
|
|
|
} else {
|
|
|
|
res.stat.stat = nlm4_denied;
|
|
|
|
memcpy(&res.stat.nlm4_testrply_u.holder, holder,
|
|
|
|
sizeof(struct nlm4_holder));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* nlm_test has different result type to the other operations, so
|
|
|
|
* can't use transmit4_result() in this case
|
|
|
|
*/
|
|
|
|
addr = svc_getrpccaller(rqstp->rq_xprt)->buf;
|
|
|
|
if ((cli = get_client(addr, NLM_VERS4)) != NULL) {
|
|
|
|
timeo.tv_sec = 0; /* No timeout - not expecting response */
|
|
|
|
timeo.tv_usec = 0;
|
|
|
|
|
|
|
|
success = clnt_call(cli, NLM4_TEST_RES, xdr_nlm4_testres,
|
|
|
|
&res, xdr_void, &dummy, timeo);
|
|
|
|
|
|
|
|
if (debug_level > 2)
|
|
|
|
syslog(LOG_DEBUG, "clnt_call returns %d", success);
|
|
|
|
}
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_lock ---------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purposes: Establish a lock
|
|
|
|
* Returns: granted, denied or blocked
|
|
|
|
* Notes: *** grace period support missing
|
|
|
|
*/
|
|
|
|
nlm4_res *
|
|
|
|
nlm4_lock_4_svc(arg, rqstp)
|
|
|
|
nlm4_lockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_lock", rqstp);
|
2001-10-14 18:36:35 +00:00
|
|
|
if (debug_level > 5) {
|
|
|
|
syslog(LOG_DEBUG, "Locking arguments:\n");
|
|
|
|
log_netobj(&(arg->cookie));
|
|
|
|
syslog(LOG_DEBUG, "Alock arguments:\n");
|
|
|
|
syslog(LOG_DEBUG, "Caller Name: %s\n",arg->alock.caller_name);
|
|
|
|
syslog(LOG_DEBUG, "File Handle:\n");
|
|
|
|
log_netobj(&(arg->alock.fh));
|
|
|
|
syslog(LOG_DEBUG, "Owner Handle:\n");
|
|
|
|
log_netobj(&(arg->alock.oh));
|
|
|
|
syslog(LOG_DEBUG, "SVID: %d\n", arg->alock.svid);
|
2001-11-13 11:24:23 +00:00
|
|
|
syslog(LOG_DEBUG, "Lock Offset: %llu\n",
|
|
|
|
(unsigned long long)arg->alock.l_offset);
|
|
|
|
syslog(LOG_DEBUG, "Lock Length: %llu\n",
|
|
|
|
(unsigned long long)arg->alock.l_len);
|
2001-10-14 18:36:35 +00:00
|
|
|
syslog(LOG_DEBUG, "Block: %s\n", (arg->block ? "true" : "false"));
|
|
|
|
syslog(LOG_DEBUG, "Exclusive: %s\n", (arg->exclusive ? "true" : "false"));
|
|
|
|
syslog(LOG_DEBUG, "Reclaim: %s\n", (arg->reclaim ? "true" : "false"));
|
|
|
|
syslog(LOG_DEBUG, "State num: %d\n", arg->state);
|
|
|
|
}
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
|
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_4() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
res.stat.stat = getlock(arg, rqstp, LOCK_MON | LOCK_V4);
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm4_lock_msg_4_svc(arg, rqstp)
|
|
|
|
nlm4_lockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_lock_msg", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat.stat = getlock(arg, rqstp, LOCK_MON | LOCK_ASYNC | LOCK_V4);
|
|
|
|
transmit4_result(NLM4_LOCK_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_cancel -------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Cancel a blocked lock request
|
|
|
|
* Returns: granted or denied
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm4_res *
|
|
|
|
nlm4_cancel_4_svc(arg, rqstp)
|
|
|
|
nlm4_cancargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_cancel", rqstp);
|
|
|
|
|
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Since at present we never return 'nlm_blocked', there can never be
|
|
|
|
* a lock to cancel, so this call always fails.
|
|
|
|
*/
|
|
|
|
res.stat.stat = unlock(&arg->alock, LOCK_CANCEL);
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm4_cancel_msg_4_svc(arg, rqstp)
|
|
|
|
nlm4_cancargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_cancel_msg", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
/*
|
|
|
|
* Since at present we never return 'nlm_blocked', there can never be
|
|
|
|
* a lock to cancel, so this call always fails.
|
|
|
|
*/
|
|
|
|
res.stat.stat = unlock(&arg->alock, LOCK_CANCEL | LOCK_V4);
|
|
|
|
transmit4_result(NLM4_CANCEL_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_unlock -------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Release an existing lock
|
|
|
|
* Returns: Always granted, unless during grace period
|
|
|
|
* Notes: "no such lock" error condition is ignored, as the
|
|
|
|
* protocol uses unreliable UDP datagrams, and may well
|
|
|
|
* re-try an unlock that has already succeeded.
|
|
|
|
*/
|
|
|
|
nlm4_res *
|
|
|
|
nlm4_unlock_4_svc(arg, rqstp)
|
|
|
|
nlm4_unlockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_unlock", rqstp);
|
|
|
|
|
|
|
|
res.stat.stat = unlock(&arg->alock, LOCK_V4);
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm4_unlock_msg_4_svc(arg, rqstp)
|
|
|
|
nlm4_unlockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_unlock_msg", rqstp);
|
|
|
|
|
|
|
|
res.stat.stat = unlock(&arg->alock, LOCK_V4);
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
transmit4_result(NLM4_UNLOCK_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getcaller(rqstp->rq_xprt));
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Client-side pseudo-RPCs for results. Note that for the client there
|
|
|
|
* are only nlm_xxx_msg() versions of each call, since the 'real RPC'
|
|
|
|
* version returns the results in the RPC result, and so the client
|
|
|
|
* does not normally receive incoming RPCs.
|
|
|
|
*
|
|
|
|
* The exception to this is nlm_granted(), which is genuinely an RPC
|
|
|
|
* call from the server to the client - a 'call-back' in normal procedure
|
|
|
|
* call terms.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* nlm_granted ------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Receive notification that formerly blocked lock now granted
|
|
|
|
* Returns: always success ('granted')
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm4_res *
|
|
|
|
nlm4_granted_4_svc(arg, rqstp)
|
|
|
|
nlm4_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_granted", rqstp);
|
|
|
|
|
2001-04-17 20:45:23 +00:00
|
|
|
res.stat.stat = lock_answer(arg->alock.svid, &arg->cookie,
|
|
|
|
nlm4_granted, NULL, NLM_VERS4) == 0 ?
|
|
|
|
nlm4_granted : nlm4_denied;
|
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
/* copy cookie from arg to result. See comment in nlm_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
nlm4_granted_msg_4_svc(arg, rqstp)
|
|
|
|
nlm4_testargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_granted_msg", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
2003-05-14 13:50:40 +00:00
|
|
|
res.stat.stat = lock_answer(arg->alock.svid, &arg->cookie,
|
|
|
|
nlm4_granted, NULL, NLM_VERS4) == 0 ?
|
|
|
|
nlm4_granted : nlm4_denied;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
transmit4_result(NLM4_GRANTED_RES, &res,
|
|
|
|
(struct sockaddr *)svc_getrpccaller(rqstp->rq_xprt)->buf);
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_test_res ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_test_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm4_test_res_4_svc(arg, rqstp)
|
|
|
|
nlm4_testres *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_test_res", rqstp);
|
2001-04-17 20:45:23 +00:00
|
|
|
|
|
|
|
(void)lock_answer(-1, &arg->cookie, arg->stat.stat,
|
|
|
|
(int *)&arg->stat.nlm4_testrply_u.holder.svid,
|
|
|
|
NLM_VERS4);
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_lock_res ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_lock_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm4_lock_res_4_svc(arg, rqstp)
|
|
|
|
nlm4_res *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_lock_res", rqstp);
|
|
|
|
|
2001-04-17 20:45:23 +00:00
|
|
|
(void)lock_answer(-1, &arg->cookie, arg->stat.stat, NULL, NLM_VERS4);
|
|
|
|
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_cancel_res ---------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_cancel_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm4_cancel_res_4_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
nlm4_res *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_cancel_res", rqstp);
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_unlock_res ---------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_unlock_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm4_unlock_res_4_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
nlm4_res *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_unlock_res", rqstp);
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_granted_res --------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Accept result from earlier nlm_granted_msg() call
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm4_granted_res_4_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
nlm4_res *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_granted_res", rqstp);
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ------------------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Calls for PCNFS locking (aka non-monitored locking, no involvement
|
|
|
|
* of rpc.statd).
|
|
|
|
*
|
|
|
|
* These are all genuine RPCs - no nlm_xxx_msg() nonsense here.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* nlm_share --------------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: Establish a DOS-style lock
|
|
|
|
* Returns: success or failure
|
|
|
|
* Notes: Blocking locks are not supported - client is expected
|
|
|
|
* to retry if required.
|
|
|
|
*/
|
|
|
|
nlm4_shareres *
|
|
|
|
nlm4_share_4_svc(arg, rqstp)
|
|
|
|
nlm4_shareargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_shareres res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_share", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat = nlm4_granted;
|
|
|
|
res.sequence = 1234356; /* X/Open says this field is ignored? */
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm4_unshare ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Release a DOS-style lock
|
|
|
|
* Returns: nlm_granted, unless in grace period
|
|
|
|
* Notes:
|
|
|
|
*/
|
|
|
|
nlm4_shareres *
|
|
|
|
nlm4_unshare_4_svc(arg, rqstp)
|
|
|
|
nlm4_shareargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_shareres res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm_unshare", rqstp);
|
|
|
|
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat = nlm4_granted;
|
|
|
|
res.sequence = 1234356; /* X/Open says this field is ignored? */
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm4_nm_lock ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: non-monitored version of nlm4_lock()
|
|
|
|
* Returns: as for nlm4_lock()
|
|
|
|
* Notes: These locks are in the same style as the standard nlm4_lock,
|
|
|
|
* but the rpc.statd should not be called to establish a
|
|
|
|
* monitor for the client machine, since that machine is
|
|
|
|
* declared not to be running a rpc.statd, and so would not
|
|
|
|
* respond to the statd protocol.
|
|
|
|
*/
|
|
|
|
nlm4_res *
|
|
|
|
nlm4_nm_lock_4_svc(arg, rqstp)
|
|
|
|
nlm4_lockargs *arg;
|
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static nlm4_res res;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_nm_lock", rqstp);
|
|
|
|
|
|
|
|
/* copy cookie from arg to result. See comment in nlm4_test_1() */
|
|
|
|
res.cookie = arg->cookie;
|
|
|
|
res.stat.stat = nlm4_granted;
|
|
|
|
return (&res);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm4_free_all ------------------------------------------------------------ */
|
|
|
|
/*
|
|
|
|
* Purpose: Release all locks held by a named client
|
|
|
|
* Returns: Nothing
|
|
|
|
* Notes: Potential denial of service security problem here - the
|
|
|
|
* locks to be released are specified by a host name, independent
|
|
|
|
* of the address from which the request has arrived.
|
|
|
|
* Should probably be rejected if the named host has been
|
|
|
|
* using monitored locks.
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm4_free_all_4_svc(arg, rqstp)
|
2002-03-21 23:05:13 +00:00
|
|
|
struct nlm4_notify *arg __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
struct svc_req *rqstp;
|
|
|
|
{
|
|
|
|
static char dummy;
|
|
|
|
|
|
|
|
if (debug_level)
|
|
|
|
log_from_addr("nlm4_free_all", rqstp);
|
|
|
|
return (&dummy);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* nlm_sm_notify --------------------------------------------------------- */
|
|
|
|
/*
|
|
|
|
* Purpose: called by rpc.statd when a monitored host state changes.
|
|
|
|
* Returns: Nothing
|
|
|
|
*/
|
|
|
|
void *
|
|
|
|
nlm_sm_notify_0_svc(arg, rqstp)
|
|
|
|
struct nlm_sm_status *arg;
|
2002-03-21 23:05:13 +00:00
|
|
|
struct svc_req *rqstp __unused;
|
Bring in a hybrid of SunSoft's transport-independent RPC (TI-RPC) and
associated changes that had to happen to make this possible as well as
bugs fixed along the way.
Bring in required TLI library routines to support this.
Since we don't support TLI we've essentially copied what NetBSD
has done, adding a thin layer to emulate direct the TLI calls
into BSD socket calls.
This is mostly from Sun's tirpc release that was made in 1994,
however some fixes were backported from the 1999 release (supposedly
only made available after this porting effort was underway).
The submitter has agreed to continue on and bring us up to the
1999 release.
Several key features are introduced with this update:
Client calls are thread safe. (1999 code has server side thread
safe)
Updated, a more modern interface.
Many userland updates were done to bring the code up to par with
the recent RPC API.
There is an update to the pthreads library, a function
pthread_main_np() was added to emulate a function of Sun's threads
library.
While we're at it, bring in NetBSD's lockd, it's been far too
long of a wait.
New rpcbind(8) replaces portmap(8) (supporting communication over
an authenticated Unix-domain socket, and by default only allowing
set and unset requests over that channel). It's much more secure
than the old portmapper.
Umount(8), mountd(8), mount_nfs(8), nfsd(8) have also been upgraded
to support TI-RPC and to support IPV6.
Umount(8) is also fixed to unmount pathnames longer than 80 chars,
which are currently truncated by the Kernel statfs structure.
Submitted by: Martin Blapp <mb@imp.ch>
Manpage review: ru
Secure RPC implemented by: wpaul
2001-03-19 12:50:13 +00:00
|
|
|
{
|
|
|
|
static char dummy;
|
|
|
|
notify(arg->mon_name, arg->state);
|
|
|
|
return (&dummy);
|
|
|
|
}
|