NetBSD/sys/miscfs/kernfs/kernfs_subr.c
2003-09-10 03:42:07 +00:00

441 lines
12 KiB
C

/* $NetBSD: kernfs_subr.c,v 1.3 2003/09/10 03:42:07 itojun Exp $ */
/*
* Copyright (c) 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* Jan-Simon Pendry.
*
* 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. 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.
*
* @(#)kernfs_subr.c 8.6 (Berkeley) 5/14/95
*/
/*
* Copyright (c) 1994 Christopher G. Demetriou. All rights reserved.
* Copyright (c) 1993 Jan-Simon Pendry
*
* This code is derived from software contributed to Berkeley by
* Jan-Simon Pendry.
*
* 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 by the University of
* California, Berkeley and its contributors.
* 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.
*
* @(#)kernfs_subr.c 8.6 (Berkeley) 5/14/95
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: kernfs_subr.c,v 1.3 2003/09/10 03:42:07 itojun Exp $");
#ifdef _KERNEL_OPT
#include "opt_ipsec.h"
#endif
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/time.h>
#include <sys/kernel.h>
#include <sys/proc.h>
#include <sys/vnode.h>
#include <sys/malloc.h>
#include <sys/stat.h>
#include <sys/file.h>
#include <sys/filedesc.h>
#include <sys/mount.h>
#include <miscfs/kernfs/kernfs.h>
#ifdef IPSEC
#include <sys/mbuf.h>
#include <net/route.h>
#include <netinet/in.h>
#include <netinet6/ipsec.h>
#include <netkey/keydb.h>
#include <netkey/key.h>
#endif
void kernfs_hashins __P((struct kernfs_node *));
void kernfs_hashrem __P((struct kernfs_node *));
struct vnode *kernfs_hashget __P((kfstype, struct mount *,
const struct kern_target *, u_int32_t));
static LIST_HEAD(kfs_hashhead, kernfs_node) *kfs_hashtbl;
static u_long kfs_ihash; /* size of hash table - 1 */
#define KFSVALUEHASH(v) ((v) & kfs_ihash)
static struct lock kfs_hashlock;
static struct simplelock kfs_hash_slock;
#define ISSET(t, f) ((t) & (f))
/*
* allocate a kfsnode/vnode pair. the vnode is
* referenced, and locked.
*
* the kfs_type, kfs_value and mount point uniquely
* identify a kfsnode. the mount point is needed
* because someone might mount this filesystem
* twice.
*
* all kfsnodes are maintained on a singly-linked
* list. new nodes are only allocated when they cannot
* be found on this list. entries on the list are
* removed when the vfs reclaim entry is called.
*
* a single lock is kept for the entire list. this is
* needed because the getnewvnode() function can block
* waiting for a vnode to become free, in which case there
* may be more than one process trying to get the same
* vnode. this lock is only taken if we are going to
* call getnewvnode, since the kernel itself is single-threaded.
*
* if an entry is found on the list, then call vget() to
* take a reference. this is done because there may be
* zero references to it and so it needs to removed from
* the vnode free list.
*/
int
kernfs_allocvp(mp, vpp, kfs_type, kt, value)
struct mount *mp;
struct vnode **vpp;
kfstype kfs_type;
const struct kern_target *kt;
u_int32_t value;
{
struct kernfs_node *kfs = NULL, *kfsp;
struct vnode *vp = NULL;
int error;
long *cookie;
do {
if ((*vpp = kernfs_hashget(kfs_type, mp, kt, value)) != NULL)
return (0);
} while (lockmgr(&kfs_hashlock, LK_EXCLUSIVE|LK_SLEEPFAIL, 0));
if (kfs_type == Pdevice) {
/* /kern/rootdev = look for device and obey */
/* /kern/rrootdev = look for device and obey */
dev_t *dp;
struct vnode *fvp;
#ifdef DIAGNOSTIC
if (!kt)
panic("kernfs: kt == NULL for Pdevice");
#endif
dp = kt->kt_data;
loop:
if (*dp == NODEV || !vfinddev(*dp, kt->kt_vtype, &fvp)) {
lockmgr(&kfs_hashlock, LK_RELEASE, NULL);
return (ENOENT);
}
vp = fvp;
if (vget(fvp, LK_EXCLUSIVE))
goto loop;
*vpp = vp;
lockmgr(&kfs_hashlock, LK_RELEASE, NULL);
return (0);
}
if ((error = getnewvnode(VT_KERNFS, mp, kernfs_vnodeop_p, &vp)) != 0) {
*vpp = NULL;
lockmgr(&kfs_hashlock, LK_RELEASE, NULL);
return (error);
}
MALLOC(kfs, void *, sizeof(struct kernfs_node), M_TEMP, M_WAITOK);
memset(kfs, 0, sizeof(*kfs));
vp->v_data = kfs;
cookie = &(VFSTOKERNFS(mp)->fileno_cookie);
again:
TAILQ_FOREACH(kfsp, &VFSTOKERNFS(mp)->nodelist, kfs_list) {
if (kfsp->kfs_cookie == *cookie) {
(*cookie) ++;
goto again;
}
if (TAILQ_NEXT(kfsp, kfs_list)) {
if (kfsp->kfs_cookie < *cookie &&
*cookie < TAILQ_NEXT(kfsp, kfs_list)->kfs_cookie)
break;
if (kfsp->kfs_cookie + 1 <
TAILQ_NEXT(kfsp, kfs_list)->kfs_cookie) {
*cookie = kfsp->kfs_cookie + 1;
break;
}
}
}
kfs->kfs_cookie = *cookie;
if (kfsp)
TAILQ_INSERT_AFTER(&VFSTOKERNFS(mp)->nodelist, kfsp, kfs,
kfs_list);
else
TAILQ_INSERT_TAIL(&VFSTOKERNFS(mp)->nodelist, kfs, kfs_list);
kfs->kfs_type = kfs_type;
kfs->kfs_vnode = vp;
kfs->kfs_fileno = KERNFS_FILENO(kt, kfs_type, kfs->kfs_cookie);
kfs->kfs_value = value;
kfs->kfs_kt = kt;
switch (kfs_type) {
case Pkern: /* /kern = dr-xr-xr-x */
kfs->kfs_mode = S_IRUSR|S_IXUSR|S_IRGRP|S_IXGRP|S_IROTH|S_IXOTH;
vp->v_type = VDIR;
vp->v_flag = VROOT;
break;
case Pnull: /* /kern/?? = -r--r--r-- */
case Ptime: /* /kern/time = -r--r--r-- */
case Pint: /* /kern/?? = -r--r--r-- */
case Pstring: /* /kern/?? = -r--r--r-- */
case Pavenrun: /* /kern/loadavg = -r--r--r-- */
case Pmsgbuf: /* /kern/msgbuf = -r--r--r-- */
kfs->kfs_mode = S_IRUSR|S_IRGRP|S_IROTH;
vp->v_type = VREG;
break;
case Pipsecsadir: /* /kern/ipsecsa = dr-x------ */
case Pipsecspdir: /* /kern/ipsecsp = dr-x------ */
kfs->kfs_mode = S_IRUSR|S_IXUSR;
vp->v_type = VDIR;
break;
case Pipsecsa: /* /kern/ipsecsa/N = -r-------- */
case Pipsecsp: /* /kern/ipsecsp/N = -r-------- */
kfs->kfs_mode = S_IRUSR;
vp->v_type = VREG;
break;
case Phostname: /* /kern/hostname = -rw-r-r--- */
kfs->kfs_mode = S_IRUSR|S_IWUSR|S_IRGRP|S_IROTH;
vp->v_type = VREG;
break;
default:
panic("kernfs_allocvp");
}
kernfs_hashins(kfs);
uvm_vnp_setsize(vp, 0);
lockmgr(&kfs_hashlock, LK_RELEASE, NULL);
*vpp = vp;
return (0);
}
int
kernfs_freevp(vp)
struct vnode *vp;
{
struct kernfs_node *kfs = VTOKERN(vp);
kernfs_hashrem(kfs);
TAILQ_REMOVE(&VFSTOKERNFS(vp->v_mount)->nodelist, kfs, kfs_list);
FREE(vp->v_data, M_TEMP);
vp->v_data = 0;
return (0);
}
/*
* Initialize kfsnode hash table.
*/
void
kernfs_hashinit()
{
lockinit(&kfs_hashlock, PINOD, "kfs_hashlock", 0, 0);
kfs_hashtbl = hashinit(desiredvnodes / 4, HASH_LIST, M_UFSMNT,
M_WAITOK, &kfs_ihash);
simple_lock_init(&kfs_hash_slock);
}
void
kernfs_hashreinit()
{
struct kernfs_node *pp;
struct kfs_hashhead *oldhash, *hash;
u_long i, oldmask, mask, val;
hash = hashinit(desiredvnodes / 4, HASH_LIST, M_UFSMNT, M_WAITOK,
&mask);
simple_lock(&kfs_hash_slock);
oldhash = kfs_hashtbl;
oldmask = kfs_ihash;
kfs_hashtbl = hash;
kfs_ihash = mask;
for (i = 0; i <= oldmask; i++) {
while ((pp = LIST_FIRST(&oldhash[i])) != NULL) {
LIST_REMOVE(pp, kfs_hash);
val = KFSVALUEHASH(pp->kfs_value);
LIST_INSERT_HEAD(&hash[val], pp, kfs_hash);
}
}
simple_unlock(&kfs_hash_slock);
hashdone(oldhash, M_UFSMNT);
}
/*
* Free kfsnode hash table.
*/
void
kernfs_hashdone()
{
hashdone(kfs_hashtbl, M_UFSMNT);
}
struct vnode *
kernfs_hashget(type, mp, kt, value)
kfstype type;
struct mount *mp;
const struct kern_target *kt;
u_int32_t value;
{
struct kfs_hashhead *ppp;
struct kernfs_node *pp;
struct vnode *vp;
loop:
simple_lock(&kfs_hash_slock);
ppp = &kfs_hashtbl[KFSVALUEHASH(value)];
LIST_FOREACH(pp, ppp, kfs_hash) {
vp = KERNFSTOV(pp);
if (pp->kfs_type == type && vp->v_mount == mp &&
pp->kfs_kt == kt && pp->kfs_value == value) {
simple_lock(&vp->v_interlock);
simple_unlock(&kfs_hash_slock);
if (vget(vp, LK_EXCLUSIVE | LK_INTERLOCK))
goto loop;
return (vp);
}
}
simple_unlock(&kfs_hash_slock);
return (NULL);
}
/*
* Insert the kfsnode into the hash table and lock it.
*/
void
kernfs_hashins(pp)
struct kernfs_node *pp;
{
struct kfs_hashhead *ppp;
/* lock the kfsnode, then put it on the appropriate hash list */
lockmgr(&pp->kfs_vnode->v_lock, LK_EXCLUSIVE, (struct simplelock *)0);
simple_lock(&kfs_hash_slock);
ppp = &kfs_hashtbl[KFSVALUEHASH(pp->kfs_value)];
LIST_INSERT_HEAD(ppp, pp, kfs_hash);
simple_unlock(&kfs_hash_slock);
}
/*
* Remove the kfsnode from the hash table.
*/
void
kernfs_hashrem(pp)
struct kernfs_node *pp;
{
simple_lock(&kfs_hash_slock);
LIST_REMOVE(pp, kfs_hash);
simple_unlock(&kfs_hash_slock);
}
#ifdef IPSEC
void
kernfs_revoke_sa(sav)
struct secasvar *sav;
{
struct kernfs_node *kfs, *pnext;
struct vnode *vp;
struct kfs_hashhead *ppp;
struct mbuf *m;
ppp = &kfs_hashtbl[KFSVALUEHASH(ntohl(sav->spi))];
for (kfs = LIST_FIRST(ppp); kfs; kfs = pnext) {
vp = KERNFSTOV(kfs);
pnext = LIST_NEXT(kfs, kfs_hash);
if (vp->v_usecount > 0 && kfs->kfs_type == Pipsecsa &&
kfs->kfs_value == ntohl(sav->spi)) {
m = key_setdumpsa_spi(sav->spi);
if (!m)
VOP_REVOKE(vp, REVOKEALL);
else
m_freem(m);
break;
}
}
}
void
kernfs_revoke_sp(sp)
struct secpolicy *sp;
{
struct kernfs_node *kfs, *pnext;
struct vnode *vp;
struct kfs_hashhead *ppp;
ppp = &kfs_hashtbl[KFSVALUEHASH(sp->id)];
for (kfs = LIST_FIRST(ppp); kfs; kfs = pnext) {
vp = KERNFSTOV(kfs);
pnext = LIST_NEXT(kfs, kfs_hash);
if (vp->v_usecount > 0 && kfs->kfs_type == Pipsecsa &&
kfs->kfs_value == sp->id)
VOP_REVOKE(vp, REVOKEALL);
}
}
#endif