freebsd-nq/sys/xdr/xdr.c

837 lines
15 KiB
C
Raw Normal View History

Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
/* $NetBSD: xdr.c,v 1.22 2000/07/06 03:10:35 christos Exp $ */
/*
* Sun RPC is a product of Sun Microsystems, Inc. and is provided for
* unrestricted use provided that this legend is included on all tape
* media and as a part of the software program in whole or part. Users
* may copy or modify Sun RPC without charge, but are not authorized
* to license or distribute it to anyone else except as part of a product or
* program developed by the user.
*
* SUN RPC IS PROVIDED AS IS WITH NO WARRANTIES OF ANY KIND INCLUDING THE
* WARRANTIES OF DESIGN, MERCHANTIBILITY AND FITNESS FOR A PARTICULAR
* PURPOSE, OR ARISING FROM A COURSE OF DEALING, USAGE OR TRADE PRACTICE.
*
* Sun RPC is provided with no support and without any obligation on the
* part of Sun Microsystems, Inc. to assist in its use, correction,
* modification or enhancement.
*
* SUN MICROSYSTEMS, INC. SHALL HAVE NO LIABILITY WITH RESPECT TO THE
* INFRINGEMENT OF COPYRIGHTS, TRADE SECRETS OR ANY PATENTS BY SUN RPC
* OR ANY PART THEREOF.
*
* In no event will Sun Microsystems, Inc. be liable for any lost revenue
* or profits or other special, indirect and consequential damages, even if
* Sun has been advised of the possibility of such damages.
*
* Sun Microsystems, Inc.
* 2550 Garcia Avenue
* Mountain View, California 94043
*/
#if defined(LIBC_SCCS) && !defined(lint)
static char *sccsid2 = "@(#)xdr.c 1.35 87/08/12";
static char *sccsid = "@(#)xdr.c 2.1 88/07/29 4.0 RPCSRC";
#endif
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
/*
* xdr.c, Generic XDR routines implementation.
*
* Copyright (C) 1986, Sun Microsystems, Inc.
*
* These are the "generic" xdr routines used to serialize and de-serialize
* most common data items. See xdr.h for more info on the interface to
* xdr.
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <rpc/rpc.h>
#include <rpc/rpc_com.h>
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
#include <rpc/types.h>
#include <rpc/xdr.h>
typedef quad_t longlong_t; /* ANSI long long type */
typedef u_quad_t u_longlong_t; /* ANSI unsigned long long type */
/*
* constants specific to the xdr "protocol"
*/
#define XDR_FALSE ((long) 0)
#define XDR_TRUE ((long) 1)
/*
* for unit alignment
*/
static const char xdr_zero[BYTES_PER_XDR_UNIT] = { 0, 0, 0, 0 };
/*
* Free a data structure using XDR
* Not a filter, but a convenient utility nonetheless
*/
void
xdr_free(xdrproc_t proc, void *objp)
{
XDR x;
x.x_op = XDR_FREE;
(*proc)(&x, objp);
}
/*
* XDR nothing
*/
bool_t
xdr_void(void)
{
return (TRUE);
}
/*
* XDR integers
*/
bool_t
xdr_int(XDR *xdrs, int *ip)
{
long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (long) *ip;
return (XDR_PUTLONG(xdrs, &l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, &l)) {
return (FALSE);
}
*ip = (int) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR unsigned integers
*/
bool_t
xdr_u_int(XDR *xdrs, u_int *up)
{
u_long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (u_long) *up;
return (XDR_PUTLONG(xdrs, (long *)&l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, (long *)&l)) {
return (FALSE);
}
*up = (u_int) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR long integers
* same as xdr_u_long - open coded to save a proc call!
*/
bool_t
xdr_long(XDR *xdrs, long *lp)
{
switch (xdrs->x_op) {
case XDR_ENCODE:
return (XDR_PUTLONG(xdrs, lp));
case XDR_DECODE:
return (XDR_GETLONG(xdrs, lp));
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR unsigned long integers
* same as xdr_long - open coded to save a proc call!
*/
bool_t
xdr_u_long(XDR *xdrs, u_long *ulp)
{
switch (xdrs->x_op) {
case XDR_ENCODE:
return (XDR_PUTLONG(xdrs, (long *)ulp));
case XDR_DECODE:
return (XDR_GETLONG(xdrs, (long *)ulp));
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR 32-bit integers
* same as xdr_uint32_t - open coded to save a proc call!
*/
bool_t
xdr_int32_t(XDR *xdrs, int32_t *int32_p)
{
long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (long) *int32_p;
return (XDR_PUTLONG(xdrs, &l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, &l)) {
return (FALSE);
}
*int32_p = (int32_t) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR unsigned 32-bit integers
* same as xdr_int32_t - open coded to save a proc call!
*/
bool_t
xdr_uint32_t(XDR *xdrs, uint32_t *uint32_p)
{
u_long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (u_long) *uint32_p;
return (XDR_PUTLONG(xdrs, (long *)&l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, (long *)&l)) {
return (FALSE);
}
*uint32_p = (uint32_t) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR short integers
*/
bool_t
xdr_short(XDR *xdrs, short *sp)
{
long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (long) *sp;
return (XDR_PUTLONG(xdrs, &l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, &l)) {
return (FALSE);
}
*sp = (short) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR unsigned short integers
*/
bool_t
xdr_u_short(XDR *xdrs, u_short *usp)
{
u_long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (u_long) *usp;
return (XDR_PUTLONG(xdrs, (long *)&l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, (long *)&l)) {
return (FALSE);
}
*usp = (u_short) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR 16-bit integers
*/
bool_t
xdr_int16_t(XDR *xdrs, int16_t *int16_p)
{
long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (long) *int16_p;
return (XDR_PUTLONG(xdrs, &l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, &l)) {
return (FALSE);
}
*int16_p = (int16_t) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR unsigned 16-bit integers
*/
bool_t
xdr_uint16_t(XDR *xdrs, uint16_t *uint16_p)
{
u_long l;
switch (xdrs->x_op) {
case XDR_ENCODE:
l = (u_long) *uint16_p;
return (XDR_PUTLONG(xdrs, (long *)&l));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, (long *)&l)) {
return (FALSE);
}
*uint16_p = (uint16_t) l;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR a char
*/
bool_t
xdr_char(XDR *xdrs, char *cp)
{
int i;
i = (*cp);
if (!xdr_int(xdrs, &i)) {
return (FALSE);
}
*cp = i;
return (TRUE);
}
/*
* XDR an unsigned char
*/
bool_t
xdr_u_char(XDR *xdrs, u_char *cp)
{
u_int u;
u = (*cp);
if (!xdr_u_int(xdrs, &u)) {
return (FALSE);
}
*cp = u;
return (TRUE);
}
/*
* XDR booleans
*/
bool_t
xdr_bool(XDR *xdrs, bool_t *bp)
{
long lb;
switch (xdrs->x_op) {
case XDR_ENCODE:
lb = *bp ? XDR_TRUE : XDR_FALSE;
return (XDR_PUTLONG(xdrs, &lb));
case XDR_DECODE:
if (!XDR_GETLONG(xdrs, &lb)) {
return (FALSE);
}
*bp = (lb == XDR_FALSE) ? FALSE : TRUE;
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR enumerations
*/
bool_t
xdr_enum(XDR *xdrs, enum_t *ep)
{
enum sizecheck { SIZEVAL }; /* used to find the size of an enum */
/*
* enums are treated as ints
*/
/* LINTED */ if (sizeof (enum sizecheck) == sizeof (long)) {
return (xdr_long(xdrs, (long *)(void *)ep));
} else /* LINTED */ if (sizeof (enum sizecheck) == sizeof (int)) {
return (xdr_int(xdrs, (int *)(void *)ep));
} else /* LINTED */ if (sizeof (enum sizecheck) == sizeof (short)) {
return (xdr_short(xdrs, (short *)(void *)ep));
} else {
return (FALSE);
}
}
/*
* XDR opaque data
* Allows the specification of a fixed size sequence of opaque bytes.
* cp points to the opaque object and cnt gives the byte length.
*/
bool_t
xdr_opaque(XDR *xdrs, caddr_t cp, u_int cnt)
{
u_int rndup;
static int crud[BYTES_PER_XDR_UNIT];
/*
* if no data we are done
*/
if (cnt == 0)
return (TRUE);
/*
* round byte count to full xdr units
*/
rndup = cnt % BYTES_PER_XDR_UNIT;
if (rndup > 0)
rndup = BYTES_PER_XDR_UNIT - rndup;
if (xdrs->x_op == XDR_DECODE) {
if (!XDR_GETBYTES(xdrs, cp, cnt)) {
return (FALSE);
}
if (rndup == 0)
return (TRUE);
return (XDR_GETBYTES(xdrs, (caddr_t)(void *)crud, rndup));
}
if (xdrs->x_op == XDR_ENCODE) {
if (!XDR_PUTBYTES(xdrs, cp, cnt)) {
return (FALSE);
}
if (rndup == 0)
return (TRUE);
return (XDR_PUTBYTES(xdrs, xdr_zero, rndup));
}
if (xdrs->x_op == XDR_FREE) {
return (TRUE);
}
return (FALSE);
}
/*
* XDR counted bytes
* *cpp is a pointer to the bytes, *sizep is the count.
* If *cpp is NULL maxsize bytes are allocated
*/
bool_t
xdr_bytes(XDR *xdrs, char **cpp, u_int *sizep, u_int maxsize)
{
char *sp = *cpp; /* sp is the actual string pointer */
u_int nodesize;
bool_t ret, allocated = FALSE;
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
/*
* first deal with the length since xdr bytes are counted
*/
if (! xdr_u_int(xdrs, sizep)) {
return (FALSE);
}
nodesize = *sizep;
if ((nodesize > maxsize) && (xdrs->x_op != XDR_FREE)) {
return (FALSE);
}
/*
* now deal with the actual bytes
*/
switch (xdrs->x_op) {
case XDR_DECODE:
if (nodesize == 0) {
return (TRUE);
}
if (sp == NULL) {
*cpp = sp = mem_alloc(nodesize);
allocated = TRUE;
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
}
if (sp == NULL) {
printf("xdr_bytes: out of memory");
return (FALSE);
}
/* FALLTHROUGH */
case XDR_ENCODE:
ret = xdr_opaque(xdrs, sp, nodesize);
if ((xdrs->x_op == XDR_DECODE) && (ret == FALSE)) {
if (allocated == TRUE) {
mem_free(sp, nodesize);
*cpp = NULL;
}
}
return (ret);
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
case XDR_FREE:
if (sp != NULL) {
mem_free(sp, nodesize);
*cpp = NULL;
}
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* Implemented here due to commonality of the object.
*/
bool_t
xdr_netobj(XDR *xdrs, struct netobj *np)
{
return (xdr_bytes(xdrs, &np->n_bytes, &np->n_len, MAX_NETOBJ_SZ));
}
/*
* XDR a descriminated union
* Support routine for discriminated unions.
* You create an array of xdrdiscrim structures, terminated with
* an entry with a null procedure pointer. The routine gets
* the discriminant value and then searches the array of xdrdiscrims
* looking for that value. It calls the procedure given in the xdrdiscrim
* to handle the discriminant. If there is no specific routine a default
* routine may be called.
* If there is no specific or default routine an error is returned.
*/
bool_t
xdr_union(XDR *xdrs,
enum_t *dscmp, /* enum to decide which arm to work on */
char *unp, /* the union itself */
const struct xdr_discrim *choices, /* [value, xdr proc] for each arm */
xdrproc_t dfault) /* default xdr routine */
{
enum_t dscm;
/*
* we deal with the discriminator; it's an enum
*/
if (! xdr_enum(xdrs, dscmp)) {
return (FALSE);
}
dscm = *dscmp;
/*
* search choices for a value that matches the discriminator.
* if we find one, execute the xdr routine for that value.
*/
for (; choices->proc != NULL_xdrproc_t; choices++) {
if (choices->value == dscm)
return ((*(choices->proc))(xdrs, unp));
}
/*
* no match - execute the default xdr routine if there is one
*/
return ((dfault == NULL_xdrproc_t) ? FALSE :
(*dfault)(xdrs, unp));
}
/*
* Non-portable xdr primitives.
* Care should be taken when moving these routines to new architectures.
*/
/*
* XDR null terminated ASCII strings
* xdr_string deals with "C strings" - arrays of bytes that are
* terminated by a NULL character. The parameter cpp references a
* pointer to storage; If the pointer is null, then the necessary
* storage is allocated. The last parameter is the max allowed length
* of the string as specified by a protocol.
*/
bool_t
xdr_string(XDR *xdrs, char **cpp, u_int maxsize)
{
char *sp = *cpp; /* sp is the actual string pointer */
u_int size;
u_int nodesize;
bool_t ret, allocated = FALSE;
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
/*
* first deal with the length since xdr strings are counted-strings
*/
switch (xdrs->x_op) {
case XDR_FREE:
if (sp == NULL) {
return(TRUE); /* already free */
}
/* FALLTHROUGH */
case XDR_ENCODE:
size = strlen(sp);
break;
case XDR_DECODE:
break;
}
if (! xdr_u_int(xdrs, &size)) {
return (FALSE);
}
if (size > maxsize) {
return (FALSE);
}
nodesize = size + 1;
/*
* now deal with the actual bytes
*/
switch (xdrs->x_op) {
case XDR_DECODE:
if (nodesize == 0) {
return (TRUE);
}
if (sp == NULL) {
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
*cpp = sp = mem_alloc(nodesize);
allocated = TRUE;
}
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
if (sp == NULL) {
printf("xdr_string: out of memory");
return (FALSE);
}
sp[size] = 0;
/* FALLTHROUGH */
case XDR_ENCODE:
ret = xdr_opaque(xdrs, sp, size);
if ((xdrs->x_op == XDR_DECODE) && (ret == FALSE)) {
if (allocated == TRUE) {
mem_free(sp, nodesize);
*cpp = NULL;
}
}
return (ret);
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
case XDR_FREE:
mem_free(sp, nodesize);
*cpp = NULL;
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* Wrapper for xdr_string that can be called directly from
* routines like clnt_call
*/
bool_t
xdr_wrapstring(XDR *xdrs, char **cpp)
{
return xdr_string(xdrs, cpp, RPC_MAXDATASIZE);
Add the new kernel-mode NFS Lock Manager. To use it instead of the user-mode lock manager, build a kernel with the NFSLOCKD option and add '-k' to 'rpc_lockd_flags' in rc.conf. Highlights include: * Thread-safe kernel RPC client - many threads can use the same RPC client handle safely with replies being de-multiplexed at the socket upcall (typically driven directly by the NIC interrupt) and handed off to whichever thread matches the reply. For UDP sockets, many RPC clients can share the same socket. This allows the use of a single privileged UDP port number to talk to an arbitrary number of remote hosts. * Single-threaded kernel RPC server. Adding support for multi-threaded server would be relatively straightforward and would follow approximately the Solaris KPI. A single thread should be sufficient for the NLM since it should rarely block in normal operation. * Kernel mode NLM server supporting cancel requests and granted callbacks. I've tested the NLM server reasonably extensively - it passes both my own tests and the NFS Connectathon locking tests running on Solaris, Mac OS X and Ubuntu Linux. * Userland NLM client supported. While the NLM server doesn't have support for the local NFS client's locking needs, it does have to field async replies and granted callbacks from remote NLMs that the local client has contacted. We relay these replies to the userland rpc.lockd over a local domain RPC socket. * Robust deadlock detection for the local lock manager. In particular it will detect deadlocks caused by a lock request that covers more than one blocking request. As required by the NLM protocol, all deadlock detection happens synchronously - a user is guaranteed that if a lock request isn't rejected immediately, the lock will eventually be granted. The old system allowed for a 'deferred deadlock' condition where a blocked lock request could wake up and find that some other deadlock-causing lock owner had beaten them to the lock. * Since both local and remote locks are managed by the same kernel locking code, local and remote processes can safely use file locks for mutual exclusion. Local processes have no fairness advantage compared to remote processes when contending to lock a region that has just been unlocked - the local lock manager enforces a strict first-come first-served model for both local and remote lockers. Sponsored by: Isilon Systems PR: 95247 107555 115524 116679 MFC after: 2 weeks
2008-03-26 15:23:12 +00:00
}
/*
* NOTE: xdr_hyper(), xdr_u_hyper(), xdr_longlong_t(), and xdr_u_longlong_t()
* are in the "non-portable" section because they require that a `long long'
* be a 64-bit type.
*
* --thorpej@netbsd.org, November 30, 1999
*/
/*
* XDR 64-bit integers
*/
bool_t
xdr_int64_t(XDR *xdrs, int64_t *llp)
{
u_long ul[2];
switch (xdrs->x_op) {
case XDR_ENCODE:
ul[0] = (u_long)((uint64_t)*llp >> 32) & 0xffffffff;
ul[1] = (u_long)((uint64_t)*llp) & 0xffffffff;
if (XDR_PUTLONG(xdrs, (long *)&ul[0]) == FALSE)
return (FALSE);
return (XDR_PUTLONG(xdrs, (long *)&ul[1]));
case XDR_DECODE:
if (XDR_GETLONG(xdrs, (long *)&ul[0]) == FALSE)
return (FALSE);
if (XDR_GETLONG(xdrs, (long *)&ul[1]) == FALSE)
return (FALSE);
*llp = (int64_t)
(((uint64_t)ul[0] << 32) | ((uint64_t)ul[1]));
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR unsigned 64-bit integers
*/
bool_t
xdr_uint64_t(XDR *xdrs, uint64_t *ullp)
{
u_long ul[2];
switch (xdrs->x_op) {
case XDR_ENCODE:
ul[0] = (u_long)(*ullp >> 32) & 0xffffffff;
ul[1] = (u_long)(*ullp) & 0xffffffff;
if (XDR_PUTLONG(xdrs, (long *)&ul[0]) == FALSE)
return (FALSE);
return (XDR_PUTLONG(xdrs, (long *)&ul[1]));
case XDR_DECODE:
if (XDR_GETLONG(xdrs, (long *)&ul[0]) == FALSE)
return (FALSE);
if (XDR_GETLONG(xdrs, (long *)&ul[1]) == FALSE)
return (FALSE);
*ullp = (uint64_t)
(((uint64_t)ul[0] << 32) | ((uint64_t)ul[1]));
return (TRUE);
case XDR_FREE:
return (TRUE);
}
/* NOTREACHED */
return (FALSE);
}
/*
* XDR hypers
*/
bool_t
xdr_hyper(XDR *xdrs, longlong_t *llp)
{
/*
* Don't bother open-coding this; it's a fair amount of code. Just
* call xdr_int64_t().
*/
return (xdr_int64_t(xdrs, (int64_t *)llp));
}
/*
* XDR unsigned hypers
*/
bool_t
xdr_u_hyper(XDR *xdrs, u_longlong_t *ullp)
{
/*
* Don't bother open-coding this; it's a fair amount of code. Just
* call xdr_uint64_t().
*/
return (xdr_uint64_t(xdrs, (uint64_t *)ullp));
}
/*
* XDR longlong_t's
*/
bool_t
xdr_longlong_t(XDR *xdrs, longlong_t *llp)
{
/*
* Don't bother open-coding this; it's a fair amount of code. Just
* call xdr_int64_t().
*/
return (xdr_int64_t(xdrs, (int64_t *)llp));
}
/*
* XDR u_longlong_t's
*/
bool_t
xdr_u_longlong_t(XDR *xdrs, u_longlong_t *ullp)
{
/*
* Don't bother open-coding this; it's a fair amount of code. Just
* call xdr_uint64_t().
*/
return (xdr_uint64_t(xdrs, (uint64_t *)ullp));
}