mirror of
https://git.hardenedbsd.org/hardenedbsd/HardenedBSD.git
synced 2024-12-30 15:38:06 +01:00
26f9a76710
Reviewed by: Rodney W. Grimes Submitted by: John Dyson and David Greenman
400 lines
9.5 KiB
C
400 lines
9.5 KiB
C
/*
|
|
* Copyright (c) 1992, 1993
|
|
* The Regents of the University of California. All rights reserved.
|
|
*
|
|
* This code is derived from software donated 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.
|
|
*
|
|
* @(#)umap_subr.c 8.6 (Berkeley) 1/26/94
|
|
*
|
|
* $Id: lofs_subr.c, v 1.11 1992/05/30 10:05:43 jsp Exp jsp $
|
|
*/
|
|
|
|
#include <sys/param.h>
|
|
#include <sys/systm.h>
|
|
#include <sys/time.h>
|
|
#include <sys/types.h>
|
|
#include <sys/vnode.h>
|
|
#include <sys/mount.h>
|
|
#include <sys/namei.h>
|
|
#include <sys/malloc.h>
|
|
#include <miscfs/umapfs/umap.h>
|
|
|
|
#define LOG2_SIZEVNODE 7 /* log2(sizeof struct vnode) */
|
|
#define NUMAPNODECACHE 16
|
|
#define UMAP_NHASH(vp) ((((u_long) vp)>>LOG2_SIZEVNODE) & (NUMAPNODECACHE-1))
|
|
|
|
/*
|
|
* Null layer cache:
|
|
* Each cache entry holds a reference to the target vnode
|
|
* along with a pointer to the alias vnode. When an
|
|
* entry is added the target vnode is VREF'd. When the
|
|
* alias is removed the target vnode is vrele'd.
|
|
*/
|
|
|
|
/*
|
|
* Cache head
|
|
*/
|
|
struct umap_node_cache {
|
|
struct umap_node *ac_forw;
|
|
struct umap_node *ac_back;
|
|
};
|
|
|
|
static struct umap_node_cache umap_node_cache[NUMAPNODECACHE];
|
|
|
|
/*
|
|
* Initialise cache headers
|
|
*/
|
|
int
|
|
umapfs_init()
|
|
{
|
|
struct umap_node_cache *ac;
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
printf("umapfs_init\n"); /* printed during system boot */
|
|
#endif
|
|
|
|
for (ac = umap_node_cache; ac < umap_node_cache + NUMAPNODECACHE; ac++)
|
|
ac->ac_forw = ac->ac_back = (struct umap_node *) ac;
|
|
return (0);
|
|
}
|
|
|
|
/*
|
|
* Compute hash list for given target vnode
|
|
*/
|
|
static struct umap_node_cache *
|
|
umap_node_hash(targetvp)
|
|
struct vnode *targetvp;
|
|
{
|
|
|
|
return (&umap_node_cache[UMAP_NHASH(targetvp)]);
|
|
}
|
|
|
|
/*
|
|
* umap_findid is called by various routines in umap_vnodeops.c to
|
|
* find a user or group id in a map.
|
|
*/
|
|
static u_long
|
|
umap_findid(id, map, nentries)
|
|
u_long id;
|
|
u_long map[][2];
|
|
int nentries;
|
|
{
|
|
int i;
|
|
|
|
/* Find uid entry in map */
|
|
i = 0;
|
|
while ((i<nentries) && ((map[i][0]) != id))
|
|
i++;
|
|
|
|
if (i < nentries)
|
|
return (map[i][1]);
|
|
else
|
|
return (-1);
|
|
|
|
}
|
|
|
|
/*
|
|
* umap_reverse_findid is called by umap_getattr() in umap_vnodeops.c to
|
|
* find a user or group id in a map, in reverse.
|
|
*/
|
|
u_long
|
|
umap_reverse_findid(id, map, nentries)
|
|
u_long id;
|
|
u_long map[][2];
|
|
int nentries;
|
|
{
|
|
int i;
|
|
|
|
/* Find uid entry in map */
|
|
i = 0;
|
|
while ((i<nentries) && ((map[i][1]) != id))
|
|
i++;
|
|
|
|
if (i < nentries)
|
|
return (map[i][0]);
|
|
else
|
|
return (-1);
|
|
|
|
}
|
|
|
|
/*
|
|
* Return alias for target vnode if already exists, else 0.
|
|
*/
|
|
static struct vnode *
|
|
umap_node_find(mp, targetvp)
|
|
struct mount *mp;
|
|
struct vnode *targetvp;
|
|
{
|
|
struct umap_node_cache *hd;
|
|
struct umap_node *a;
|
|
struct vnode *vp;
|
|
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
printf("umap_node_find(mp = %x, target = %x)\n", mp, targetvp);
|
|
#endif
|
|
|
|
/*
|
|
* Find hash base, and then search the (two-way) linked
|
|
* list looking for a umap_node structure which is referencing
|
|
* the target vnode. If found, the increment the umap_node
|
|
* reference count (but NOT the target vnode's VREF counter).
|
|
*/
|
|
hd = umap_node_hash(targetvp);
|
|
|
|
loop:
|
|
for (a = hd->ac_forw; a != (struct umap_node *) hd; a = a->umap_forw) {
|
|
if (a->umap_lowervp == targetvp &&
|
|
a->umap_vnode->v_mount == mp) {
|
|
vp = UMAPTOV(a);
|
|
/*
|
|
* We need vget for the VXLOCK
|
|
* stuff, but we don't want to lock
|
|
* the lower node.
|
|
*/
|
|
if (vget(vp, 0)) {
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
printf ("umap_node_find: vget failed.\n");
|
|
#endif
|
|
goto loop;
|
|
}
|
|
return (vp);
|
|
}
|
|
}
|
|
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
printf("umap_node_find(%x, %x): NOT found\n", mp, targetvp);
|
|
#endif
|
|
|
|
return (0);
|
|
}
|
|
|
|
/*
|
|
* Make a new umap_node node.
|
|
* Vp is the alias vnode, lofsvp is the target vnode.
|
|
* Maintain a reference to (targetvp).
|
|
*/
|
|
static int
|
|
umap_node_alloc(mp, lowervp, vpp)
|
|
struct mount *mp;
|
|
struct vnode *lowervp;
|
|
struct vnode **vpp;
|
|
{
|
|
struct umap_node_cache *hd;
|
|
struct umap_node *xp;
|
|
struct vnode *othervp, *vp;
|
|
int error;
|
|
|
|
if (error = getnewvnode(VT_UMAP, mp, umap_vnodeop_p, vpp))
|
|
return (error);
|
|
vp = *vpp;
|
|
|
|
MALLOC(xp, struct umap_node *, sizeof(struct umap_node),
|
|
M_TEMP, M_WAITOK);
|
|
vp->v_type = lowervp->v_type;
|
|
xp->umap_vnode = vp;
|
|
vp->v_data = xp;
|
|
xp->umap_lowervp = lowervp;
|
|
/*
|
|
* Before we insert our new node onto the hash chains,
|
|
* check to see if someone else has beaten us to it.
|
|
* (We could have slept in MALLOC.)
|
|
*/
|
|
if (othervp = umap_node_find(lowervp)) {
|
|
FREE(xp, M_TEMP);
|
|
vp->v_type = VBAD; /* node is discarded */
|
|
vp->v_usecount = 0; /* XXX */
|
|
*vpp = othervp;
|
|
return (0);
|
|
}
|
|
VREF(lowervp); /* Extra VREF will be vrele'd in umap_node_create */
|
|
hd = umap_node_hash(lowervp);
|
|
insque(xp, hd);
|
|
return (0);
|
|
}
|
|
|
|
|
|
/*
|
|
* Try to find an existing umap_node vnode refering
|
|
* to it, otherwise make a new umap_node vnode which
|
|
* contains a reference to the target vnode.
|
|
*/
|
|
int
|
|
umap_node_create(mp, targetvp, newvpp)
|
|
struct mount *mp;
|
|
struct vnode *targetvp;
|
|
struct vnode **newvpp;
|
|
{
|
|
struct vnode *aliasvp;
|
|
|
|
if (aliasvp = umap_node_find(mp, targetvp)) {
|
|
/*
|
|
* Take another reference to the alias vnode
|
|
*/
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
vprint("umap_node_create: exists", ap->umap_vnode);
|
|
#endif
|
|
/* VREF(aliasvp); */
|
|
} else {
|
|
int error;
|
|
|
|
/*
|
|
* Get new vnode.
|
|
*/
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
printf("umap_node_create: create new alias vnode\n");
|
|
#endif
|
|
/*
|
|
* Make new vnode reference the umap_node.
|
|
*/
|
|
if (error = umap_node_alloc(mp, targetvp, &aliasvp))
|
|
return (error);
|
|
|
|
/*
|
|
* aliasvp is already VREF'd by getnewvnode()
|
|
*/
|
|
}
|
|
|
|
vrele(targetvp);
|
|
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
vprint("umap_node_create: alias", aliasvp);
|
|
vprint("umap_node_create: target", targetvp);
|
|
#endif
|
|
|
|
*newvpp = aliasvp;
|
|
return (0);
|
|
}
|
|
|
|
#ifdef UMAPFS_DIAGNOSTIC
|
|
int umap_checkvp_barrier = 1;
|
|
struct vnode *
|
|
umap_checkvp(vp, fil, lno)
|
|
struct vnode *vp;
|
|
char *fil;
|
|
int lno;
|
|
{
|
|
struct umap_node *a = VTOUMAP(vp);
|
|
#if 0
|
|
/*
|
|
* Can't do this check because vop_reclaim runs
|
|
* with funny vop vector.
|
|
*/
|
|
if (vp->v_op != umap_vnodeop_p) {
|
|
printf ("umap_checkvp: on non-umap-node\n");
|
|
while (umap_checkvp_barrier) /*WAIT*/ ;
|
|
panic("umap_checkvp");
|
|
}
|
|
#endif
|
|
if (a->umap_lowervp == NULL) {
|
|
/* Should never happen */
|
|
int i; u_long *p;
|
|
printf("vp = %x, ZERO ptr\n", vp);
|
|
for (p = (u_long *) a, i = 0; i < 8; i++)
|
|
printf(" %x", p[i]);
|
|
printf("\n");
|
|
/* wait for debugger */
|
|
while (umap_checkvp_barrier) /*WAIT*/ ;
|
|
panic("umap_checkvp");
|
|
}
|
|
if (a->umap_lowervp->v_usecount < 1) {
|
|
int i; u_long *p;
|
|
printf("vp = %x, unref'ed lowervp\n", vp);
|
|
for (p = (u_long *) a, i = 0; i < 8; i++)
|
|
printf(" %x", p[i]);
|
|
printf("\n");
|
|
/* wait for debugger */
|
|
while (umap_checkvp_barrier) /*WAIT*/ ;
|
|
panic ("umap with unref'ed lowervp");
|
|
}
|
|
#if 0
|
|
printf("umap %x/%d -> %x/%d [%s, %d]\n",
|
|
a->umap_vnode, a->umap_vnode->v_usecount,
|
|
a->umap_lowervp, a->umap_lowervp->v_usecount,
|
|
fil, lno);
|
|
#endif
|
|
return (a->umap_lowervp);
|
|
}
|
|
#endif
|
|
|
|
/* umap_mapids maps all of the ids in a credential, both user and group. */
|
|
|
|
void
|
|
umap_mapids(v_mount, credp)
|
|
struct mount *v_mount;
|
|
struct ucred *credp;
|
|
{
|
|
int i, unentries, gnentries;
|
|
u_long *groupmap, *usermap;
|
|
uid_t uid;
|
|
gid_t gid;
|
|
|
|
unentries = MOUNTTOUMAPMOUNT(v_mount)->info_nentries;
|
|
usermap = &(MOUNTTOUMAPMOUNT(v_mount)->info_mapdata[0][0]);
|
|
gnentries = MOUNTTOUMAPMOUNT(v_mount)->info_gnentries;
|
|
groupmap = &(MOUNTTOUMAPMOUNT(v_mount)->info_gmapdata[0][0]);
|
|
|
|
/* Find uid entry in map */
|
|
|
|
uid = (uid_t) umap_findid(credp->cr_uid, usermap, unentries);
|
|
|
|
if (uid != -1)
|
|
credp->cr_uid = uid;
|
|
else
|
|
credp->cr_uid = (uid_t) NOBODY;
|
|
|
|
#ifdef notdef
|
|
/* cr_gid is the same as cr_groups[0] in 4BSD */
|
|
|
|
/* Find gid entry in map */
|
|
|
|
gid = (gid_t) umap_findid(credp->cr_gid, groupmap, gnentries);
|
|
|
|
if (gid != -1)
|
|
credp->cr_gid = gid;
|
|
else
|
|
credp->cr_gid = NULLGROUP;
|
|
#endif
|
|
|
|
/* Now we must map each of the set of groups in the cr_groups
|
|
structure. */
|
|
|
|
i = 0;
|
|
while (credp->cr_groups[i] != 0) {
|
|
gid = (gid_t) umap_findid(credp->cr_groups[i],
|
|
groupmap, gnentries);
|
|
|
|
if (gid != -1)
|
|
credp->cr_groups[i++] = gid;
|
|
else
|
|
credp->cr_groups[i++] = NULLGROUP;
|
|
}
|
|
}
|