freebsd-dev/sys/fs/nfsclient/nfs_clcomsubs.c
Rick Macklem 9ec7b004d0 Add the experimental nfs subtree to the kernel, that includes
support for NFSv4 as well as NFSv2 and 3.
	It lives in 3 subdirs under sys/fs:
	nfs - functions that are common to the client and server
	nfsclient - a mutation of sys/nfsclient that call generic functions
	to do RPCs and handle state. As such, it retains the
	buffer cache handling characteristics and vnode semantics that
	are found in sys/nfsclient, for the most part.
	nfsserver - the server. It includes a DRC designed specifically for
	NFSv4, that is used instead of the generic DRC in sys/rpc.
	The build glue will be checked in later, so at this point, it
	consists of 3 new subdirs that should not affect kernel building.

Approved by:	kib (mentor)
2009-05-04 15:23:58 +00:00

522 lines
14 KiB
C

/*-
* Copyright (c) 1989, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* Rick Macklem at The University of Guelph.
*
* 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.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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>
__FBSDID("$FreeBSD$");
/*
* These functions support the macros and help fiddle mbuf chains for
* the nfs op functions. They do things like create the rpc header and
* copy data between mbuf chains and uio lists.
*/
#ifndef APPLEKEXT
#include <fs/nfs/nfsport.h>
extern struct nfsstats newnfsstats;
extern struct nfsv4_opflag nfsv4_opflag[NFSV4OP_NOPS];
extern int ncl_mbuf_mlen;
extern enum vtype newnv2tov_type[8];
extern enum vtype nv34tov_type[8];
NFSCLSTATEMUTEX;
#endif /* !APPLEKEXT */
static nfsuint64 nfs_nullcookie = {{ 0, 0 }};
static struct {
int op;
int opcnt;
const u_char *tag;
int taglen;
} nfsv4_opmap[NFS_NPROCS] = {
{ 0, 1, "Null", 4 },
{ NFSV4OP_GETATTR, 1, "Getattr", 7, },
{ NFSV4OP_SETATTR, 2, "Setattr", 7, },
{ NFSV4OP_LOOKUP, 3, "Lookup", 6, },
{ NFSV4OP_ACCESS, 2, "Access", 6, },
{ NFSV4OP_READLINK, 2, "Readlink", 8, },
{ NFSV4OP_READ, 1, "Read", 4, },
{ NFSV4OP_WRITE, 2, "Write", 5, },
{ NFSV4OP_OPEN, 3, "Open", 4, },
{ NFSV4OP_CREATE, 3, "Create", 6, },
{ NFSV4OP_CREATE, 1, "Create", 6, },
{ NFSV4OP_CREATE, 3, "Create", 6, },
{ NFSV4OP_REMOVE, 1, "Remove", 6, },
{ NFSV4OP_REMOVE, 1, "Remove", 6, },
{ NFSV4OP_SAVEFH, 5, "Rename", 6, },
{ NFSV4OP_SAVEFH, 4, "Link", 4, },
{ NFSV4OP_READDIR, 2, "Readdir", 7, },
{ NFSV4OP_READDIR, 2, "Readdir", 7, },
{ NFSV4OP_GETATTR, 1, "Getattr", 7, },
{ NFSV4OP_GETATTR, 1, "Getattr", 7, },
{ NFSV4OP_GETATTR, 1, "Getattr", 7, },
{ NFSV4OP_COMMIT, 2, "Commit", 6, },
{ NFSV4OP_LOOKUPP, 3, "Lookupp", 7, },
{ NFSV4OP_SETCLIENTID, 1, "SetClientID", 11, },
{ NFSV4OP_SETCLIENTIDCFRM, 1, "SetClientIDConfirm", 18, },
{ NFSV4OP_LOCK, 1, "Lock", 4, },
{ NFSV4OP_LOCKU, 1, "LockU", 5, },
{ NFSV4OP_OPEN, 2, "Open", 4, },
{ NFSV4OP_CLOSE, 1, "Close", 5, },
{ NFSV4OP_OPENCONFIRM, 1, "Openconfirm", 11, },
{ NFSV4OP_LOCKT, 1, "LockT", 5, },
{ NFSV4OP_OPENDOWNGRADE, 1, "Opendowngrade", 13, },
{ NFSV4OP_RENEW, 1, "Renew", 5, },
{ NFSV4OP_PUTROOTFH, 1, "Dirpath", 7, },
{ NFSV4OP_RELEASELCKOWN, 1, "Rellckown", 9, },
{ NFSV4OP_DELEGRETURN, 1, "Delegret", 8, },
{ NFSV4OP_DELEGRETURN, 3, "DelegRemove", 11, },
{ NFSV4OP_DELEGRETURN, 7, "DelegRename1", 12, },
{ NFSV4OP_DELEGRETURN, 9, "DelegRename2", 12, },
{ NFSV4OP_GETATTR, 1, "Getacl", 6, },
{ NFSV4OP_SETATTR, 1, "Setacl", 6, },
};
/*
* NFS RPCS that have large request message size.
*/
static int nfs_bigrequest[NFS_NPROCS] = {
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
};
/*
* Start building a request. Mostly just put the first file handle in
* place.
*/
APPLESTATIC void
nfscl_reqstart(struct nfsrv_descript *nd, int procnum, struct nfsmount *nmp,
u_int8_t *nfhp, int fhlen, u_int32_t **opcntpp)
{
struct mbuf *mb;
u_int32_t *tl;
int opcnt;
nfsattrbit_t attrbits;
/*
* First, fill in some of the fields of nd.
*/
if (NFSHASNFSV4(nmp))
nd->nd_flag = ND_NFSV4;
else if (NFSHASNFSV3(nmp))
nd->nd_flag = ND_NFSV3;
else
nd->nd_flag = ND_NFSV2;
nd->nd_procnum = procnum;
nd->nd_repstat = 0;
/*
* Get the first mbuf for the request.
*/
if (nfs_bigrequest[procnum])
NFSMCLGET(mb, M_WAIT);
else
NFSMGET(mb);
mbuf_setlen(mb, 0);
nd->nd_mreq = nd->nd_mb = mb;
nd->nd_bpos = NFSMTOD(mb, caddr_t);
/*
* And fill the first file handle into the request.
*/
if (nd->nd_flag & ND_NFSV4) {
opcnt = nfsv4_opmap[procnum].opcnt +
nfsv4_opflag[nfsv4_opmap[procnum].op].needscfh;
/*
* What should the tag really be?
*/
(void) nfsm_strtom(nd, nfsv4_opmap[procnum].tag,
nfsv4_opmap[procnum].taglen);
NFSM_BUILD(tl, u_int32_t *, 3 * NFSX_UNSIGNED);
*tl++ = txdr_unsigned(NFSV4_MINORVERSION);
if (opcntpp != NULL)
*opcntpp = tl;
*tl++ = txdr_unsigned(opcnt);
if (nfsv4_opflag[nfsv4_opmap[procnum].op].needscfh > 0) {
*tl = txdr_unsigned(NFSV4OP_PUTFH);
(void) nfsm_fhtom(nd, nfhp, fhlen, 0);
if (nfsv4_opflag[nfsv4_opmap[procnum].op].needscfh==2){
NFSM_BUILD(tl, u_int32_t *, NFSX_UNSIGNED);
*tl = txdr_unsigned(NFSV4OP_GETATTR);
NFSWCCATTR_ATTRBIT(&attrbits);
(void) nfsrv_putattrbit(nd, &attrbits);
nd->nd_flag |= ND_V4WCCATTR;
}
NFSM_BUILD(tl, u_int32_t *, NFSX_UNSIGNED);
}
*tl = txdr_unsigned(nfsv4_opmap[procnum].op);
} else {
(void) nfsm_fhtom(nd, nfhp, fhlen, 0);
}
NFSINCRGLOBAL(newnfsstats.rpccnt[procnum]);
}
#ifndef APPLE
/*
* copies a uio scatter/gather list to an mbuf chain.
* NOTE: can ony handle iovcnt == 1
*/
APPLESTATIC void
nfsm_uiombuf(struct nfsrv_descript *nd, struct uio *uiop, int siz)
{
char *uiocp;
struct mbuf *mp, *mp2;
int xfer, left, mlen;
int uiosiz, clflg, rem;
char *cp, *tcp;
#ifdef DIAGNOSTIC
if (uiop->uio_iovcnt != 1)
panic("nfsm_uiotombuf: iovcnt != 1");
#endif
if (siz > ncl_mbuf_mlen) /* or should it >= MCLBYTES ?? */
clflg = 1;
else
clflg = 0;
rem = NFSM_RNDUP(siz) - siz;
mp = mp2 = nd->nd_mb;
while (siz > 0) {
left = uiop->uio_iov->iov_len;
uiocp = uiop->uio_iov->iov_base;
if (left > siz)
left = siz;
uiosiz = left;
while (left > 0) {
mlen = M_TRAILINGSPACE(mp);
if (mlen == 0) {
if (clflg)
NFSMCLGET(mp, M_WAIT);
else
NFSMGET(mp);
mbuf_setlen(mp, 0);
mbuf_setnext(mp2, mp);
mp2 = mp;
mlen = M_TRAILINGSPACE(mp);
}
xfer = (left > mlen) ? mlen : left;
#ifdef notdef
/* Not Yet.. */
if (uiop->uio_iov->iov_op != NULL)
(*(uiop->uio_iov->iov_op))
(uiocp, NFSMTOD(mp, caddr_t) + mbuf_len(mp),
xfer);
else
#endif
if (uiop->uio_segflg == UIO_SYSSPACE)
NFSBCOPY(uiocp, NFSMTOD(mp, caddr_t) + mbuf_len(mp),
xfer);
else
copyin(CAST_USER_ADDR_T(uiocp), NFSMTOD(mp, caddr_t)
+ mbuf_len(mp), xfer);
mbuf_setlen(mp, mbuf_len(mp) + xfer);
left -= xfer;
uiocp += xfer;
uiop->uio_offset += xfer;
uiop->uio_resid -= xfer;
}
tcp = (char *)uiop->uio_iov->iov_base;
tcp += uiosiz;
uiop->uio_iov->iov_base = (void *)tcp;
uiop->uio_iov->iov_len -= uiosiz;
siz -= uiosiz;
}
if (rem > 0) {
if (rem > M_TRAILINGSPACE(mp)) {
NFSMGET(mp);
mbuf_setlen(mp, 0);
mbuf_setnext(mp2, mp);
}
cp = NFSMTOD(mp, caddr_t) + mbuf_len(mp);
for (left = 0; left < rem; left++)
*cp++ = '\0';
mbuf_setlen(mp, mbuf_len(mp) + rem);
nd->nd_bpos = cp;
} else
nd->nd_bpos = NFSMTOD(mp, caddr_t) + mbuf_len(mp);
nd->nd_mb = mp;
}
#endif /* !APPLE */
/*
* Load vnode attributes from the xdr file attributes.
* Returns EBADRPC if they can't be parsed, 0 otherwise.
*/
APPLESTATIC int
nfsm_loadattr(struct nfsrv_descript *nd, struct nfsvattr *nap)
{
struct nfs_fattr *fp;
int error = 0;
if (nd->nd_flag & ND_NFSV4) {
error = nfsv4_loadattr(nd, NULL, nap, NULL, NULL, 0, NULL,
NULL, NULL, NULL, NULL, 0, NULL, NULL, NULL, NULL, NULL);
} else if (nd->nd_flag & ND_NFSV3) {
NFSM_DISSECT(fp, struct nfs_fattr *, NFSX_V3FATTR);
nap->na_type = nfsv34tov_type(fp->fa_type);
nap->na_mode = fxdr_unsigned(u_short, fp->fa_mode);
nap->na_rdev = makedev(fxdr_unsigned(u_char, fp->fa3_rdev.specdata1),
fxdr_unsigned(u_char, fp->fa3_rdev.specdata2));
nap->na_nlink = fxdr_unsigned(u_short, fp->fa_nlink);
nap->na_uid = fxdr_unsigned(uid_t, fp->fa_uid);
nap->na_gid = fxdr_unsigned(gid_t, fp->fa_gid);
nap->na_size = fxdr_hyper(&fp->fa3_size);
nap->na_blocksize = NFS_FABLKSIZE;
nap->na_bytes = fxdr_hyper(&fp->fa3_used);
nap->na_fileid = fxdr_unsigned(int32_t,
fp->fa3_fileid.nfsuquad[1]);
fxdr_nfsv3time(&fp->fa3_atime, &nap->na_atime);
fxdr_nfsv3time(&fp->fa3_ctime, &nap->na_ctime);
fxdr_nfsv3time(&fp->fa3_mtime, &nap->na_mtime);
nap->na_flags = 0;
nap->na_filerev = 0;
} else {
NFSM_DISSECT(fp, struct nfs_fattr *, NFSX_V2FATTR);
nap->na_type = nfsv2tov_type(fp->fa_type);
nap->na_mode = fxdr_unsigned(u_short, fp->fa_mode);
if (nap->na_type == VNON || nap->na_type == VREG)
nap->na_type = IFTOVT(nap->na_mode);
nap->na_rdev = fxdr_unsigned(dev_t, fp->fa2_rdev);
/*
* Really ugly NFSv2 kludge.
*/
if (nap->na_type == VCHR && nap->na_rdev == ((dev_t)-1))
nap->na_type = VFIFO;
nap->na_nlink = fxdr_unsigned(u_short, fp->fa_nlink);
nap->na_uid = fxdr_unsigned(uid_t, fp->fa_uid);
nap->na_gid = fxdr_unsigned(gid_t, fp->fa_gid);
nap->na_size = fxdr_unsigned(u_int32_t, fp->fa2_size);
nap->na_blocksize = fxdr_unsigned(int32_t, fp->fa2_blocksize);
nap->na_bytes =
(u_quad_t)fxdr_unsigned(int32_t, fp->fa2_blocks) *
NFS_FABLKSIZE;
nap->na_fileid = fxdr_unsigned(int32_t, fp->fa2_fileid);
fxdr_nfsv2time(&fp->fa2_atime, &nap->na_atime);
fxdr_nfsv2time(&fp->fa2_mtime, &nap->na_mtime);
nap->na_flags = 0;
nap->na_ctime.tv_sec = fxdr_unsigned(u_int32_t,
fp->fa2_ctime.nfsv2_sec);
nap->na_ctime.tv_nsec = 0;
nap->na_gen = fxdr_unsigned(u_int32_t,fp->fa2_ctime.nfsv2_usec);
nap->na_filerev = 0;
}
nfsmout:
return (error);
}
/*
* This function finds the directory cookie that corresponds to the
* logical byte offset given.
*/
APPLESTATIC nfsuint64 *
nfscl_getcookie(struct nfsnode *np, off_t off, int add)
{
struct nfsdmap *dp, *dp2;
int pos;
pos = off / NFS_DIRBLKSIZ;
if (pos == 0) {
#ifdef DIAGNOSTIC
if (add)
panic("nfs getcookie add at 0");
#endif
return (&nfs_nullcookie);
}
pos--;
dp = LIST_FIRST(&np->n_cookies);
if (!dp) {
if (add) {
MALLOC(dp, struct nfsdmap *, sizeof (struct nfsdmap),
M_NFSDIROFF, M_WAITOK);
dp->ndm_eocookie = 0;
LIST_INSERT_HEAD(&np->n_cookies, dp, ndm_list);
} else
return (NULL);
}
while (pos >= NFSNUMCOOKIES) {
pos -= NFSNUMCOOKIES;
if (LIST_NEXT(dp, ndm_list) != NULL) {
if (!add && dp->ndm_eocookie < NFSNUMCOOKIES &&
pos >= dp->ndm_eocookie)
return (NULL);
dp = LIST_NEXT(dp, ndm_list);
} else if (add) {
MALLOC(dp2, struct nfsdmap *, sizeof (struct nfsdmap),
M_NFSDIROFF, M_WAITOK);
dp2->ndm_eocookie = 0;
LIST_INSERT_AFTER(dp, dp2, ndm_list);
dp = dp2;
} else
return (NULL);
}
if (pos >= dp->ndm_eocookie) {
if (add)
dp->ndm_eocookie = pos + 1;
else
return (NULL);
}
return (&dp->ndm_cookies[pos]);
}
/*
* Gets a file handle out of an nfs reply sent to the client and returns
* the file handle and the file's attributes.
* For V4, it assumes that Getfh and Getattr Op's results are here.
*/
APPLESTATIC int
nfscl_mtofh(struct nfsrv_descript *nd, struct nfsfh **nfhpp,
struct nfsvattr *nap, int *attrflagp)
{
u_int32_t *tl;
int error = 0, flag = 1;
*nfhpp = NULL;
*attrflagp = 0;
/*
* First get the file handle and vnode.
*/
if (nd->nd_flag & ND_NFSV3) {
NFSM_DISSECT(tl, u_int32_t *, NFSX_UNSIGNED);
flag = fxdr_unsigned(int, *tl);
} else if (nd->nd_flag & ND_NFSV4) {
NFSM_DISSECT(tl, u_int32_t *, 2 * NFSX_UNSIGNED);
}
if (flag) {
error = nfsm_getfh(nd, nfhpp);
if (error)
return (error);
}
/*
* Now, get the attributes.
*/
if (nd->nd_flag & ND_NFSV4) {
NFSM_DISSECT(tl, u_int32_t *, 2 * NFSX_UNSIGNED);
} else if (nd->nd_flag & ND_NFSV3) {
NFSM_DISSECT(tl, u_int32_t *, NFSX_UNSIGNED);
if (flag) {
flag = fxdr_unsigned(int, *tl);
} else if (fxdr_unsigned(int, *tl)) {
error = nfsm_advance(nd, NFSX_V3FATTR, -1);
if (error)
return (error);
}
}
if (flag) {
error = nfsm_loadattr(nd, nap);
if (!error)
*attrflagp = 1;
}
nfsmout:
return (error);
}
/*
* Put a state Id in the mbuf list.
*/
APPLESTATIC void
nfsm_stateidtom(struct nfsrv_descript *nd, nfsv4stateid_t *stateidp, int flag)
{
nfsv4stateid_t *st;
NFSM_BUILD(st, nfsv4stateid_t *, NFSX_STATEID);
if (flag == NFSSTATEID_PUTALLZERO) {
st->seqid = 0;
st->other[0] = 0;
st->other[1] = 0;
st->other[2] = 0;
} else if (flag == NFSSTATEID_PUTALLONE) {
st->seqid = 0xffffffff;
st->other[0] = 0xffffffff;
st->other[1] = 0xffffffff;
st->other[2] = 0xffffffff;
} else {
st->seqid = stateidp->seqid;
st->other[0] = stateidp->other[0];
st->other[1] = stateidp->other[1];
st->other[2] = stateidp->other[2];
}
}
/*
* Initialize the owner/delegation sleep lock.
*/
APPLESTATIC void
nfscl_lockinit(struct nfsv4lock *lckp)
{
lckp->nfslock_usecnt = 0;
lckp->nfslock_lock = 0;
}
/*
* Get an exclusive lock. (Not needed for OpenBSD4, since there is only one
* thread for each posix process in the kernel.)
*/
APPLESTATIC void
nfscl_lockexcl(struct nfsv4lock *lckp, void *mutex)
{
int igotlock;
do {
igotlock = nfsv4_lock(lckp, 1, NULL, mutex);
} while (!igotlock);
}
/*
* Release an exclusive lock.
*/
APPLESTATIC void
nfscl_lockunlock(struct nfsv4lock *lckp)
{
nfsv4_unlock(lckp, 0);
}
/*
* Called to derefernce a lock on a stateid (delegation or open owner).
*/
APPLESTATIC void
nfscl_lockderef(struct nfsv4lock *lckp)
{
NFSLOCKCLSTATE();
lckp->nfslock_usecnt--;
if (lckp->nfslock_usecnt == 0 && (lckp->nfslock_lock & NFSV4LOCK_WANTED)) {
lckp->nfslock_lock &= ~NFSV4LOCK_WANTED;
wakeup((caddr_t)lckp);
}
NFSUNLOCKCLSTATE();
}