/*
- * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2004 Apple Computer, Inc. All rights reserved.
*
- * @APPLE_LICENSE_HEADER_START@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License"). You may not use this file except in compliance with the
- * License. Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.
*
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ *
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT. Please see the
- * License for the specific language governing rights and limitations
- * under the License.
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and
+ * limitations under the License.
*
- * @APPLE_LICENSE_HEADER_END@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
*/
/*
* Copyright 1997,1998 Julian Elischer. All rights reserved.
#include <sys/param.h>
#include <sys/systm.h>
-#include <sys/buf.h>
#include <sys/namei.h>
#include <sys/kernel.h>
#include <sys/fcntl.h>
#include <sys/disklabel.h>
#include <sys/lock.h>
#include <sys/stat.h>
-#include <sys/mount.h>
+#include <sys/mount_internal.h>
#include <sys/proc.h>
+#include <sys/kauth.h>
#include <sys/time.h>
-#include <sys/vnode.h>
+#include <sys/vnode_internal.h>
#include <miscfs/specfs/specdev.h>
#include <sys/dirent.h>
#include <sys/vmmeter.h>
#include <sys/vm.h>
+#include <sys/uio_internal.h>
#include "devfsdefs.h"
+static int devfs_update(struct vnode *vp, struct timeval *access,
+ struct timeval *modify);
+
+
/*
* Convert a component of a pathname into a pointer to a locked node.
* This is a very central and rather complicated routine.
* NOTE: (LOOKUP | LOCKPARENT) currently returns the parent node unlocked.
*/
static int
-devfs_lookup(struct vop_lookup_args *ap)
- /*struct vop_lookup_args {
+devfs_lookup(struct vnop_lookup_args *ap)
+ /*struct vnop_lookup_args {
struct vnode * a_dvp; directory vnode ptr
struct vnode ** a_vpp; where to put the result
struct componentname * a_cnp; the name we want
+ vfs_context_t a_context;
};*/
{
struct componentname *cnp = ap->a_cnp;
+ vfs_context_t ctx = cnp->cn_context;
+ struct proc *p = vfs_context_proc(ctx);
struct vnode *dir_vnode = ap->a_dvp;
struct vnode **result_vnode = ap->a_vpp;
devnode_t * dir_node; /* the directory we are searching */
devdirent_t * nodename;
int flags = cnp->cn_flags;
int op = cnp->cn_nameiop; /* LOOKUP, CREATE, RENAME, or DELETE */
- int lockparent = flags & LOCKPARENT;
int wantparent = flags & (LOCKPARENT|WANTPARENT);
int error = 0;
- struct proc *p = cnp->cn_proc;
char heldchar; /* the char at the end of the name componet */
+retry:
+
*result_vnode = NULL; /* safe not sorry */ /*XXX*/
- if (dir_vnode->v_usecount == 0)
- printf("devfs_lookup: dir had no refs ");
+ //if (dir_vnode->v_usecount == 0)
+ //printf("devfs_lookup: dir had no refs ");
dir_node = VTODN(dir_vnode);
/*
- * Check accessiblity of directory.
+ * Make sure that our node is a directory as well.
*/
if (dir_node->dn_type != DEV_DIR) {
return (ENOTDIR);
}
- if ((error = VOP_ACCESS(dir_vnode, VEXEC, cnp->cn_cred, p)) != 0) {
- return (error);
- }
-
- /* temporarily terminate string component */
+ DEVFS_LOCK();
+ /*
+ * temporarily terminate string component
+ */
heldchar = cnp->cn_nameptr[cnp->cn_namelen];
cnp->cn_nameptr[cnp->cn_namelen] = '\0';
- DEVFS_LOCK(p);
- nodename = dev_findname(dir_node,cnp->cn_nameptr);
- if (nodename) {
- /* entry exists */
- node = nodename->de_dnp;
- node->dn_last_lookup = nodename; /* for unlink */
- /* Do potential vnode allocation here inside the lock
- * to make sure that our device node has a non-NULL dn_vn
- * associated with it. The device node might otherwise
- * get deleted out from under us (see devfs_dn_free()).
- */
- error = devfs_dntovn(node, result_vnode, p);
- }
- DEVFS_UNLOCK(p);
- /* restore saved character */
+
+ nodename = dev_findname(dir_node, cnp->cn_nameptr);
+ /*
+ * restore saved character
+ */
cnp->cn_nameptr[cnp->cn_namelen] = heldchar;
- if (error)
- return (error);
+ if (nodename) {
+ /* entry exists */
+ node = nodename->de_dnp;
- if (!nodename) { /* no entry */
- /* If it doesn't exist and we're not the last component,
+ /* Do potential vnode allocation here inside the lock
+ * to make sure that our device node has a non-NULL dn_vn
+ * associated with it. The device node might otherwise
+ * get deleted out from under us (see devfs_dn_free()).
+ */
+ error = devfs_dntovn(node, result_vnode, p);
+ }
+ DEVFS_UNLOCK();
+
+ if (error) {
+ if (error == EAGAIN)
+ goto retry;
+ return error;
+ }
+ if (!nodename) {
+ /*
+ * we haven't called devfs_dntovn if we get here
+ * we have not taken a reference on the node.. no
+ * vnode_put is necessary on these error returns
+ *
+ * If it doesn't exist and we're not the last component,
* or we're at the last component, but we're not creating
* or renaming, return ENOENT.
*/
if (!(flags & ISLASTCN) || !(op == CREATE || op == RENAME)) {
return ENOENT;
}
- /*
- * Access for write is interpreted as allowing
- * creation of files in the directory.
- */
- if ((error = VOP_ACCESS(dir_vnode, VWRITE,
- cnp->cn_cred, p)) != 0)
- {
- return (error);
- }
/*
* We return with the directory locked, so that
* the parameters we set up above will still be
* We return ni_vp == NULL to indicate that the entry
* does not currently exist; we leave a pointer to
* the (locked) directory vnode in namei_data->ni_dvp.
- * The pathname buffer is saved so that the name
- * can be obtained later.
*
* NB - if the directory is unlocked, then this
* information cannot be used.
*/
- cnp->cn_flags |= SAVENAME;
- if (!lockparent)
- VOP_UNLOCK(dir_vnode, 0, p);
return (EJUSTRETURN);
}
+ /*
+ * from this point forward, we need to vnode_put the reference
+ * picked up in devfs_dntovn if we decide to return an error
+ */
/*
* If deleting, and at end of pathname, return
* on and lock the node, being careful with ".".
*/
if (op == DELETE && (flags & ISLASTCN)) {
- /*
- * Write access to directory required to delete files.
- */
- if ((error = VOP_ACCESS(dir_vnode, VWRITE,
- cnp->cn_cred, p)) != 0)
- return (error);
+
/*
* we are trying to delete '.'. What does this mean? XXX
*/
if (dir_node == node) {
- VREF(dir_vnode);
- *result_vnode = dir_vnode;
- return (0);
- }
-#ifdef NOTYET
- /*
- * If directory is "sticky", then user must own
- * the directory, or the file in it, else she
- * may not delete it (unless she's root). This
- * implements append-only directories.
- */
- if ((dir_node->mode & ISVTX) &&
- cnp->cn_cred->cr_uid != 0 &&
- cnp->cn_cred->cr_uid != dir_node->uid &&
- cnp->cn_cred->cr_uid != node->uid) {
- VOP_UNLOCK(*result_vnode, 0, p);
- return (EPERM);
+ if (*result_vnode) {
+ vnode_put(*result_vnode);
+ *result_vnode = NULL;
+ }
+ if ( ((error = vnode_get(dir_vnode)) == 0) ) {
+ *result_vnode = dir_vnode;
+ }
+ return (error);
}
-#endif
- if (!lockparent)
- VOP_UNLOCK(dir_vnode, 0, p);
return (0);
}
* regular file, or empty directory.
*/
if (op == RENAME && wantparent && (flags & ISLASTCN)) {
- /*
- * Are we allowed to change the holding directory?
- */
- if ((error = VOP_ACCESS(dir_vnode, VWRITE,
- cnp->cn_cred, p)) != 0)
- return (error);
+
/*
* Careful about locking second node.
* This can only occur if the target is ".".
*/
- if (dir_node == node)
- return (EISDIR);
- /* hmm save the 'from' name (we need to delete it) */
- cnp->cn_flags |= SAVENAME;
- if (!lockparent)
- VOP_UNLOCK(dir_vnode, 0, p);
+ if (dir_node == node) {
+ error = EISDIR;
+ goto drop_ref;
+ }
return (0);
}
* work if the file system has any hard links other than ".."
* that point backwards in the directory structure.
*/
- if (flags & ISDOTDOT) {
- VOP_UNLOCK(dir_vnode, 0, p); /* race to get the node */
- if (lockparent && (flags & ISLASTCN))
- vn_lock(dir_vnode, LK_EXCLUSIVE | LK_RETRY, p);
- } else if (dir_node == node) {
-#if 0
- /*
- * this next statement is wrong: we already did a vget in
- * devfs_dntovn(); DWS 4/16/1999
- */
- VREF(dir_vnode); /* we want ourself, ie "." */
-#endif
+ if ((flags & ISDOTDOT) == 0 && dir_node == node) {
+ if (*result_vnode) {
+ vnode_put(*result_vnode);
+ *result_vnode = NULL;
+ }
+ if ( (error = vnode_get(dir_vnode)) ) {
+ return (error);
+ }
*result_vnode = dir_vnode;
- } else {
- if (!lockparent || (flags & ISLASTCN))
- VOP_UNLOCK(dir_vnode, 0, p);
}
-
return (0);
+
+drop_ref:
+ if (*result_vnode) {
+ vnode_put(*result_vnode);
+ *result_vnode = NULL;
+ }
+ return (error);
}
static int
-devfs_access(struct vop_access_args *ap)
- /*struct vop_access_args {
+devfs_getattr(struct vnop_getattr_args *ap)
+ /*struct vnop_getattr_args {
struct vnode *a_vp;
- int a_mode;
- struct ucred *a_cred;
+ struct vnode_attr *a_vap;
+ kauth_cred_t a_cred;
struct proc *a_p;
} */
{
- /*
- * mode is filled with a combination of VREAD, VWRITE,
- * and/or VEXEC bits turned on. In an octal number these
- * are the Y in 0Y00.
- */
struct vnode *vp = ap->a_vp;
- int mode = ap->a_mode;
- struct ucred *cred = ap->a_cred;
+ struct vnode_attr *vap = ap->a_vap;
devnode_t * file_node;
- gid_t *gp;
- int i;
- struct proc *p = ap->a_p;
+ struct timeval now;
file_node = VTODN(vp);
- /*
- * if we are not running as a process, we are in the
- * kernel and we DO have permission
- */
- if (p == NULL)
- return 0;
- /*
- * Access check is based on only one of owner, group, public.
- * If not owner, then check group. If not a member of the
- * group, then check public access.
- */
- if (cred->cr_uid != file_node->dn_uid)
- {
- /* failing that.. try groups */
- mode >>= 3;
- gp = cred->cr_groups;
- for (i = 0; i < cred->cr_ngroups; i++, gp++)
- {
- if (file_node->dn_gid == *gp)
- {
- goto found;
- }
- }
- /* failing that.. try general access */
- mode >>= 3;
-found:
- ;
- }
- if ((file_node->dn_mode & mode) == mode)
- return (0);
- /*
- * Root gets to do anything.
- * but only use suser prives as a last resort
- * (Use of super powers is recorded in ap->a_p->p_acflag)
- */
- if( suser(cred, &ap->a_p->p_acflag) == 0) /* XXX what if no proc? */
- return 0;
- return (EACCES);
-}
+ DEVFS_LOCK();
-static int
-devfs_getattr(struct vop_getattr_args *ap)
- /*struct vop_getattr_args {
- struct vnode *a_vp;
- struct vattr *a_vap;
- struct ucred *a_cred;
- struct proc *a_p;
- } */
-{
- struct vnode *vp = ap->a_vp;
- struct vattr *vap = ap->a_vap;
- devnode_t * file_node;
- struct timeval tv;
+ microtime(&now);
+ dn_times(file_node, &now, &now, &now);
+
+ VATTR_RETURN(vap, va_mode, file_node->dn_mode);
- file_node = VTODN(vp);
- tv = time;
- dn_times(file_node, tv, tv);
- vap->va_rdev = 0;/* default value only */
- vap->va_mode = file_node->dn_mode;
switch (file_node->dn_type)
{
case DEV_DIR:
- vap->va_rdev = (dev_t)file_node->dn_dvm;
+ VATTR_RETURN(vap, va_rdev, (dev_t)file_node->dn_dvm);
vap->va_mode |= (S_IFDIR);
break;
case DEV_CDEV:
- vap->va_rdev = file_node->dn_typeinfo.dev;
+ VATTR_RETURN(vap, va_rdev, file_node->dn_typeinfo.dev);
vap->va_mode |= (S_IFCHR);
break;
case DEV_BDEV:
- vap->va_rdev = file_node->dn_typeinfo.dev;
+ VATTR_RETURN(vap, va_rdev, file_node->dn_typeinfo.dev);
vap->va_mode |= (S_IFBLK);
break;
case DEV_SLNK:
+ VATTR_RETURN(vap, va_rdev, 0);
vap->va_mode |= (S_IFLNK);
break;
+ default:
+ VATTR_RETURN(vap, va_rdev, 0); /* default value only */
}
- vap->va_type = vp->v_type;
- vap->va_nlink = file_node->dn_links;
- vap->va_uid = file_node->dn_uid;
- vap->va_gid = file_node->dn_gid;
- vap->va_fsid = (int32_t)(void *)file_node->dn_dvm;
- vap->va_fileid = (int32_t)(void *)file_node;
- vap->va_size = file_node->dn_len; /* now a u_quad_t */
- /* this doesn't belong here */
+ VATTR_RETURN(vap, va_type, vp->v_type);
+ VATTR_RETURN(vap, va_nlink, file_node->dn_links);
+ VATTR_RETURN(vap, va_uid, file_node->dn_uid);
+ VATTR_RETURN(vap, va_gid, file_node->dn_gid);
+ VATTR_RETURN(vap, va_fsid, (uintptr_t)file_node->dn_dvm);
+ VATTR_RETURN(vap, va_fileid, (uintptr_t)file_node);
+ VATTR_RETURN(vap, va_data_size, file_node->dn_len);
+
+ /* return an override block size (advisory) */
if (vp->v_type == VBLK)
- vap->va_blocksize = BLKDEV_IOSIZE;
+ VATTR_RETURN(vap, va_iosize, BLKDEV_IOSIZE);
else if (vp->v_type == VCHR)
- vap->va_blocksize = MAXPHYSIO;
+ VATTR_RETURN(vap, va_iosize, MAXPHYSIO);
else
- vap->va_blocksize = vp->v_mount->mnt_stat.f_iosize;
+ VATTR_RETURN(vap, va_iosize, vp->v_mount->mnt_vfsstat.f_iosize);
/* if the time is bogus, set it to the boot time */
- if (file_node->dn_ctime.tv_sec == 0)
- file_node->dn_ctime.tv_sec = boottime.tv_sec;
+ if (file_node->dn_ctime.tv_sec == 0) {
+ file_node->dn_ctime.tv_sec = boottime_sec();
+ file_node->dn_ctime.tv_nsec = 0;
+ }
if (file_node->dn_mtime.tv_sec == 0)
- file_node->dn_mtime.tv_sec = boottime.tv_sec;
+ file_node->dn_mtime = file_node->dn_ctime;
if (file_node->dn_atime.tv_sec == 0)
- file_node->dn_atime.tv_sec = boottime.tv_sec;
- vap->va_ctime = file_node->dn_ctime;
- vap->va_mtime = file_node->dn_mtime;
- vap->va_atime = file_node->dn_atime;
- vap->va_gen = 0;
- vap->va_flags = 0;
- vap->va_bytes = file_node->dn_len; /* u_quad_t */
- vap->va_filerev = 0; /* XXX */ /* u_quad_t */
- vap->va_vaflags = 0; /* XXX */
+ file_node->dn_atime = file_node->dn_ctime;
+ VATTR_RETURN(vap, va_change_time, file_node->dn_ctime);
+ VATTR_RETURN(vap, va_modify_time, file_node->dn_mtime);
+ VATTR_RETURN(vap, va_access_time, file_node->dn_atime);
+ VATTR_RETURN(vap, va_gen, 0);
+ VATTR_RETURN(vap, va_flags, 0);
+ VATTR_RETURN(vap, va_filerev, 0);
+ VATTR_RETURN(vap, va_acl, NULL);
+
+ DEVFS_UNLOCK();
+
return 0;
}
static int
-devfs_setattr(struct vop_setattr_args *ap)
- /*struct vop_setattr_args {
- struct vnode *a_vp;
- struct vattr *a_vap;
- struct ucred *a_cred;
- struct proc *a_p;
- } */
+devfs_setattr(struct vnop_setattr_args *ap)
+ /*struct vnop_setattr_args {
+ struct vnode *a_vp;
+ struct vnode_attr *a_vap;
+ vfs_context_t a_context;
+ } */
{
- struct vnode *vp = ap->a_vp;
- struct vattr *vap = ap->a_vap;
- struct ucred *cred = ap->a_cred;
- struct proc *p = ap->a_p;
- int error = 0;
- gid_t *gp;
- int i;
- devnode_t * file_node;
- struct timeval atimeval, mtimeval;
-
- if (vap->va_flags != VNOVAL) /* XXX needs to be implemented */
- return (EOPNOTSUPP);
-
- file_node = VTODN(vp);
-
- if ((vap->va_type != VNON) ||
- (vap->va_nlink != VNOVAL) ||
- (vap->va_fsid != VNOVAL) ||
- (vap->va_fileid != VNOVAL) ||
- (vap->va_blocksize != VNOVAL) ||
- (vap->va_rdev != VNOVAL) ||
- (vap->va_bytes != VNOVAL) ||
- (vap->va_gen != VNOVAL ))
- {
- return EINVAL;
- }
-
- /*
- * Go through the fields and update iff not VNOVAL.
- */
- if (vap->va_atime.tv_sec != VNOVAL || vap->va_mtime.tv_sec != VNOVAL) {
- if (cred->cr_uid != file_node->dn_uid &&
- (error = suser(cred, &p->p_acflag)) &&
- ((vap->va_vaflags & VA_UTIMES_NULL) == 0 ||
- (error = VOP_ACCESS(vp, VWRITE, cred, p))))
- return (error);
- if (vap->va_atime.tv_sec != VNOVAL)
- file_node->dn_flags |= DN_ACCESS;
- if (vap->va_mtime.tv_sec != VNOVAL)
- file_node->dn_flags |= DN_CHANGE | DN_UPDATE;
- atimeval.tv_sec = vap->va_atime.tv_sec;
- atimeval.tv_usec = vap->va_atime.tv_nsec / 1000;
- mtimeval.tv_sec = vap->va_mtime.tv_sec;
- mtimeval.tv_usec = vap->va_mtime.tv_nsec / 1000;
- if (error = VOP_UPDATE(vp, &atimeval, &mtimeval, 1))
- return (error);
- }
-
- /*
- * Change the permissions.. must be root or owner to do this.
- */
- if (vap->va_mode != (u_short)VNOVAL) {
- if ((cred->cr_uid != file_node->dn_uid)
- && (error = suser(cred, &p->p_acflag)))
- return (error);
- file_node->dn_mode &= ~07777;
- file_node->dn_mode |= vap->va_mode & 07777;
- }
-
- /*
- * Change the owner.. must be root to do this.
- */
- if (vap->va_uid != (uid_t)VNOVAL) {
- if (error = suser(cred, &p->p_acflag))
- return (error);
- file_node->dn_uid = vap->va_uid;
- }
-
- /*
- * Change the group.. must be root or owner to do this.
- * If we are the owner, we must be in the target group too.
- * don't use suser() unless you have to as it reports
- * whether you needed suser powers or not.
- */
- if (vap->va_gid != (gid_t)VNOVAL) {
- if (cred->cr_uid == file_node->dn_uid){
- gp = cred->cr_groups;
- for (i = 0; i < cred->cr_ngroups; i++, gp++) {
- if (vap->va_gid == *gp)
- goto cando;
- }
- }
- /*
- * we can't do it with normal privs,
- * do we have an ace up our sleeve?
- */
- if (error = suser(cred, &p->p_acflag))
- return (error);
-cando:
- file_node->dn_gid = vap->va_gid;
- }
-#if 0
- /*
- * Copied from somewhere else
- * but only kept as a marker and reminder of the fact that
- * flags should be handled some day
- */
- if (vap->va_flags != VNOVAL) {
- if (error = suser(cred, &p->p_acflag))
- return error;
- if (cred->cr_uid == 0)
- ;
- else {
+ struct vnode *vp = ap->a_vp;
+ struct vnode_attr *vap = ap->a_vap;
+ kauth_cred_t cred = vfs_context_ucred(ap->a_context);
+ struct proc *p = vfs_context_proc(ap->a_context);
+ int error = 0;
+ devnode_t * file_node;
+ struct timeval atimeval, mtimeval;
+
+ file_node = VTODN(vp);
+
+ DEVFS_LOCK();
+ /*
+ * Go through the fields and update if set.
+ */
+ if (VATTR_IS_ACTIVE(vap, va_access_time) || VATTR_IS_ACTIVE(vap, va_modify_time)) {
+
+
+ if (VATTR_IS_ACTIVE(vap, va_access_time))
+ file_node->dn_access = 1;
+ if (VATTR_IS_ACTIVE(vap, va_modify_time)) {
+ file_node->dn_change = 1;
+ file_node->dn_update = 1;
}
- }
-#endif
+ atimeval.tv_sec = vap->va_access_time.tv_sec;
+ atimeval.tv_usec = vap->va_access_time.tv_nsec / 1000;
+ mtimeval.tv_sec = vap->va_modify_time.tv_sec;
+ mtimeval.tv_usec = vap->va_modify_time.tv_nsec / 1000;
+
+ if ( (error = devfs_update(vp, &atimeval, &mtimeval)) )
+ goto exit;
+ }
+ VATTR_SET_SUPPORTED(vap, va_access_time);
+ VATTR_SET_SUPPORTED(vap, va_change_time);
+
+ /*
+ * Change the permissions.
+ */
+ if (VATTR_IS_ACTIVE(vap, va_mode)) {
+ file_node->dn_mode &= ~07777;
+ file_node->dn_mode |= vap->va_mode & 07777;
+ }
+ VATTR_SET_SUPPORTED(vap, va_mode);
+
+ /*
+ * Change the owner.
+ */
+ if (VATTR_IS_ACTIVE(vap, va_uid))
+ file_node->dn_uid = vap->va_uid;
+ VATTR_SET_SUPPORTED(vap, va_uid);
+
+ /*
+ * Change the group.
+ */
+ if (VATTR_IS_ACTIVE(vap, va_gid))
+ file_node->dn_gid = vap->va_gid;
+ VATTR_SET_SUPPORTED(vap, va_gid);
+ exit:
+ DEVFS_UNLOCK();
+
return error;
}
static int
-devfs_read(struct vop_read_args *ap)
- /*struct vop_read_args {
+devfs_read(struct vnop_read_args *ap)
+ /* struct vnop_read_args {
struct vnode *a_vp;
struct uio *a_uio;
int a_ioflag;
- struct ucred *a_cred;
+ vfs_context_t a_context;
} */
{
devnode_t * dn_p = VTODN(ap->a_vp);
switch (ap->a_vp->v_type) {
case VDIR: {
- dn_p->dn_flags |= DN_ACCESS;
- return VOP_READDIR(ap->a_vp, ap->a_uio, ap->a_cred,
- NULL, NULL, NULL);
+ dn_p->dn_access = 1;
+
+ return VNOP_READDIR(ap->a_vp, ap->a_uio, 0, NULL, NULL, ap->a_context);
}
default: {
printf("devfs_read(): bad file type %d", ap->a_vp->v_type);
}
static int
-devfs_close(ap)
- struct vop_close_args /* {
+devfs_close(struct vnop_close_args *ap)
+ /* struct vnop_close_args {
struct vnode *a_vp;
int a_fflag;
- struct ucred *a_cred;
- struct proc *a_p;
- } */ *ap;
+ vfs_context_t a_context;
+ } */
{
struct vnode * vp = ap->a_vp;
register devnode_t * dnp = VTODN(vp);
+ struct timeval now;
- simple_lock(&vp->v_interlock);
- if (vp->v_usecount > 1)
- dn_times(dnp, time, time);
- simple_unlock(&vp->v_interlock);
+ if (vnode_isinuse(vp, 1)) {
+ DEVFS_LOCK();
+ microtime(&now);
+ dn_times(dnp, &now, &now, &now);
+ DEVFS_UNLOCK();
+ }
return (0);
}
static int
-devfsspec_close(ap)
- struct vop_close_args /* {
+devfsspec_close(struct vnop_close_args *ap)
+ /* struct vnop_close_args {
struct vnode *a_vp;
int a_fflag;
- struct ucred *a_cred;
- struct proc *a_p;
- } */ *ap;
+ vfs_context_t a_context;
+ } */
{
struct vnode * vp = ap->a_vp;
register devnode_t * dnp = VTODN(vp);
+ struct timeval now;
- simple_lock(&vp->v_interlock);
- if (vp->v_usecount > 1)
- dn_times(dnp, time, time);
- simple_unlock(&vp->v_interlock);
- return (VOCALL (spec_vnodeop_p, VOFFSET(vop_close), ap));
+ if (vnode_isinuse(vp, 1)) {
+ DEVFS_LOCK();
+ microtime(&now);
+ dn_times(dnp, &now, &now, &now);
+ DEVFS_UNLOCK();
+ }
+ return (VOCALL (spec_vnodeop_p, VOFFSET(vnop_close), ap));
}
static int
-devfsspec_read(struct vop_read_args *ap)
- /*struct vop_read_args {
+devfsspec_read(struct vnop_read_args *ap)
+ /* struct vnop_read_args {
struct vnode *a_vp;
struct uio *a_uio;
int a_ioflag;
- struct ucred *a_cred;
+ kauth_cred_t a_cred;
} */
{
- VTODN(ap->a_vp)->dn_flags |= DN_ACCESS;
- return (VOCALL (spec_vnodeop_p, VOFFSET(vop_read), ap));
+ register devnode_t * dnp = VTODN(ap->a_vp);
+
+ dnp->dn_access = 1;
+
+ return (VOCALL (spec_vnodeop_p, VOFFSET(vnop_read), ap));
}
static int
-devfsspec_write(struct vop_write_args *ap)
- /*struct vop_write_args {
+devfsspec_write(struct vnop_write_args *ap)
+ /* struct vnop_write_args {
struct vnode *a_vp;
struct uio *a_uio;
int a_ioflag;
- struct ucred *a_cred;
+ vfs_context_t a_context;
} */
{
- VTODN(ap->a_vp)->dn_flags |= DN_CHANGE | DN_UPDATE;
- return (VOCALL (spec_vnodeop_p, VOFFSET(vop_write), ap));
+ register devnode_t * dnp = VTODN(ap->a_vp);
+
+ dnp->dn_change = 1;
+ dnp->dn_update = 1;
+
+ return (VOCALL (spec_vnodeop_p, VOFFSET(vnop_write), ap));
}
/*
* Write data to a file or directory.
*/
static int
-devfs_write(struct vop_write_args *ap)
- /*struct vop_write_args {
+devfs_write(struct vnop_write_args *ap)
+ /* struct vnop_write_args {
struct vnode *a_vp;
struct uio *a_uio;
int a_ioflag;
- struct ucred *a_cred;
+ kauth_cred_t a_cred;
} */
{
switch (ap->a_vp->v_type) {
}
static int
-devfs_remove(struct vop_remove_args *ap)
- /*struct vop_remove_args {
+devfs_remove(struct vnop_remove_args *ap)
+ /* struct vnop_remove_args {
struct vnode *a_dvp;
struct vnode *a_vp;
struct componentname *a_cnp;
struct vnode *vp = ap->a_vp;
struct vnode *dvp = ap->a_dvp;
struct componentname *cnp = ap->a_cnp;
+ vfs_context_t ctx = cnp->cn_context;
devnode_t * tp;
devnode_t * tdp;
devdirent_t * tnp;
int doingdirectory = 0;
int error = 0;
- uid_t ouruid = cnp->cn_cred->cr_uid;
- struct proc *p = cnp->cn_proc;
+ uid_t ouruid = kauth_cred_getuid(vfs_context_ucred(ctx));
/*
- * Lock our directories and get our name pointers
- * assume that the names are null terminated as they
+ * assume that the name is null terminated as they
* are the end of the path. Get pointers to all our
* devfs structures.
*/
tp = VTODN(vp);
tdp = VTODN(dvp);
- /*
- * Assuming we are atomic, dev_lookup left this for us
- */
- tnp = tp->dn_last_lookup;
- /*
- * Check we are doing legal things WRT the new flags
- */
- if ((tp->dn_flags & (IMMUTABLE | APPEND))
- || (tdp->dn_flags & APPEND) /*XXX eh?*/ ) {
- error = EPERM;
- goto abort;
+ DEVFS_LOCK();
+
+ tnp = dev_findname(tdp, cnp->cn_nameptr);
+
+ if (tnp == NULL) {
+ error = ENOENT;
+ goto abort;
}
/*
/***********************************
* Start actually doing things.... *
***********************************/
- tdp->dn_flags |= DN_CHANGE | DN_UPDATE;
+ tdp->dn_change = 1;
+ tdp->dn_update = 1;
- /*
- * own the parent directory, or the destination of the rename,
- * otherwise the destination may not be changed (except by
- * root). This implements append-only directories.
- * XXX shoudn't this be in generic code?
- */
- if ((tdp->dn_mode & S_ISTXT)
- && ouruid != 0
- && ouruid != tdp->dn_uid
- && ouruid != tp->dn_uid ) {
- error = EPERM;
- goto abort;
- }
/*
* Target must be empty if a directory and have no links
* to it. Also, ensure source and target are compatible
error = ENOTEMPTY;
goto abort;
}
- DEVFS_LOCK(p);
dev_free_name(tnp);
- DEVFS_UNLOCK(p);
- abort:
- if (dvp == vp)
- vrele(vp);
- else
- vput(vp);
- vput(dvp);
+abort:
+ DEVFS_UNLOCK();
+
return (error);
}
/*
*/
static int
-devfs_link(struct vop_link_args *ap)
- /*struct vop_link_args {
+devfs_link(struct vnop_link_args *ap)
+ /*struct vnop_link_args {
struct vnode *a_tdvp;
struct vnode *a_vp;
struct componentname *a_cnp;
+ vfs_context_t a_context;
} */
{
struct vnode *vp = ap->a_vp;
struct vnode *tdvp = ap->a_tdvp;
struct componentname *cnp = ap->a_cnp;
- struct proc *p = cnp->cn_proc;
devnode_t * fp;
devnode_t * tdp;
devdirent_t * tnp;
int error = 0;
- struct timeval tv;
+ struct timeval now;
/*
* First catch an arbitrary restriction for this FS
fp = VTODN(vp);
if (tdvp->v_mount != vp->v_mount) {
- error = EXDEV;
- VOP_ABORTOP(tdvp, cnp);
- goto out2;
- }
- if (tdvp != vp && (error = vn_lock(vp, LK_EXCLUSIVE, p))) {
- VOP_ABORTOP(tdvp, cnp);
- goto out2;
+ return (EXDEV);
}
+ DEVFS_LOCK();
- /*
- * Check we are doing legal things WRT the new flags
- */
- if (fp->dn_flags & (IMMUTABLE | APPEND)) {
- VOP_ABORTOP(tdvp, cnp);
- error = EPERM;
- goto out1;
- }
/***********************************
* Start actually doing things.... *
***********************************/
- fp->dn_flags |= DN_CHANGE;
- tv = time;
- error = VOP_UPDATE(vp, &tv, &tv, 1);
+ fp->dn_change = 1;
+
+ microtime(&now);
+ error = devfs_update(vp, &now, &now);
+
if (!error) {
- DEVFS_LOCK(p);
error = dev_add_name(cnp->cn_nameptr, tdp, NULL, fp, &tnp);
- DEVFS_UNLOCK(p);
}
out1:
- if (tdvp != vp)
- VOP_UNLOCK(vp, 0, p);
-out2:
- vput(tdvp);
- return (error);
-}
+ DEVFS_UNLOCK();
-/*
- * Check if source directory is in the path of the target directory.
- * Target is supplied locked, source is unlocked.
- * The target is always vput before returning.
- */
-int
-devfs_checkpath(source, target)
- devnode_t *source, *target;
-{
- int error = 0;
- devnode_t * ntmp;
- devnode_t * tmp;
- struct vnode *vp;
-
- vp = target->dn_vn;
- tmp = target;
-
- do {
- if (tmp == source) {
- error = EINVAL;
- break;
- }
- ntmp = tmp;
- } while ((tmp = tmp->dn_typeinfo.Dir.parent) != ntmp);
-
- if (vp != NULL)
- vput(vp);
- return (error);
+ return (error);
}
/*
* directory.
*/
static int
-devfs_rename(struct vop_rename_args *ap)
- /*struct vop_rename_args {
+devfs_rename(struct vnop_rename_args *ap)
+ /*struct vnop_rename_args {
struct vnode *a_fdvp;
struct vnode *a_fvp;
struct componentname *a_fcnp;
struct vnode *a_tdvp;
struct vnode *a_tvp;
struct componentname *a_tcnp;
+ vfs_context_t a_context;
} */
{
struct vnode *tvp = ap->a_tvp;
struct vnode *fdvp = ap->a_fdvp;
struct componentname *tcnp = ap->a_tcnp;
struct componentname *fcnp = ap->a_fcnp;
- struct proc *p = fcnp->cn_proc;
devnode_t *fp, *fdp, *tp, *tdp;
devdirent_t *fnp,*tnp;
int doingdirectory = 0;
int error = 0;
- struct timeval tv;
+ struct timeval now;
+ DEVFS_LOCK();
/*
* First catch an arbitrary restriction for this FS
*/
- if(tcnp->cn_namelen > DEVMAXNAMESIZE) {
+ if (tcnp->cn_namelen > DEVMAXNAMESIZE) {
error = ENAMETOOLONG;
- goto abortit;
+ goto out;
}
/*
- * Lock our directories and get our name pointers
* assume that the names are null terminated as they
* are the end of the path. Get pointers to all our
* devfs structures.
tdp = VTODN(tdvp);
fdp = VTODN(fdvp);
fp = VTODN(fvp);
- fnp = fp->dn_last_lookup;
+
+ fnp = dev_findname(fdp, fcnp->cn_nameptr);
+
+ if (fnp == NULL) {
+ error = ENOENT;
+ goto out;
+ }
tp = NULL;
tnp = NULL;
+
if (tvp) {
- tp = VTODN(tvp);
- tnp = tp->dn_last_lookup;
- }
-
- /*
- * trying to move it out of devfs?
- * if we move a dir across mnt points. we need to fix all
- * the mountpoint pointers! XXX
- * so for now keep dirs within the same mount
- */
- if ((fvp->v_mount != tdvp->v_mount) ||
- (tvp && (fvp->v_mount != tvp->v_mount))) {
- error = EXDEV;
-abortit:
- VOP_ABORTOP(tdvp, tcnp);
- if (tdvp == tvp) /* eh? */
- vrele(tdvp);
- else
- vput(tdvp);
- if (tvp)
- vput(tvp);
- VOP_ABORTOP(fdvp, fcnp); /* XXX, why not in NFS? */
- vrele(fdvp);
- vrele(fvp);
- return (error);
- }
+ tnp = dev_findname(tdp, tcnp->cn_nameptr);
- /*
- * Check we are doing legal things WRT the new flags
- */
- if ((tp && (tp->dn_flags & (IMMUTABLE | APPEND)))
- || (fp->dn_flags & (IMMUTABLE | APPEND))
- || (fdp->dn_flags & APPEND)) {
- error = EPERM;
- goto abortit;
+ if (tnp == NULL) {
+ error = ENOENT;
+ goto out;
+ }
+ tp = VTODN(tvp);
}
-
+
/*
* Make sure that we don't try do something stupid
*/
|| (tcnp->cn_flags&ISDOTDOT)
|| (tdp == fp )) {
error = EINVAL;
- goto abortit;
+ goto out;
}
doingdirectory++;
}
*/
if (doingdirectory && (tdp != fdp)) {
devnode_t * tmp, *ntmp;
- error = VOP_ACCESS(fvp, VWRITE, tcnp->cn_cred, tcnp->cn_proc);
tmp = tdp;
do {
if(tmp == fp) {
/***********************************
* Start actually doing things.... *
***********************************/
- fp->dn_flags |= DN_CHANGE;
- tv = time;
- if (error = VOP_UPDATE(fvp, &tv, &tv, 1)) {
- VOP_UNLOCK(fvp, 0, p);
- goto bad;
+ fp->dn_change = 1;
+ microtime(&now);
+
+ if ( (error = devfs_update(fvp, &now, &now)) ) {
+ goto out;
}
/*
* Check if just deleting a link name.
if (fvp == tvp) {
if (fvp->v_type == VDIR) {
error = EINVAL;
- goto abortit;
+ goto out;
}
-
/* Release destination completely. */
- VOP_ABORTOP(tdvp, tcnp);
- vput(tdvp);
- vput(tvp);
-
- /* Delete source. */
- VOP_ABORTOP(fdvp, fcnp); /*XXX*/
- vrele(fdvp);
- vrele(fvp);
dev_free_name(fnp);
+
+ DEVFS_UNLOCK();
return 0;
}
-
- vrele(fdvp);
-
/*
* 1) Bump link count while we're moving stuff
* around. If we crash somewhere before
* We could do that as well but won't
*/
if (tp) {
- int ouruid = tcnp->cn_cred->cr_uid;
- /*
- * If the parent directory is "sticky", then the user must
- * own the parent directory, or the destination of the rename,
- * otherwise the destination may not be changed (except by
- * root). This implements append-only directories.
- * XXX shoudn't this be in generic code?
- */
- if ((tdp->dn_mode & S_ISTXT)
- && ouruid != 0
- && ouruid != tdp->dn_uid
- && ouruid != tp->dn_uid ) {
- error = EPERM;
- goto bad;
- }
+ int ouruid = kauth_cred_getuid(vfs_context_ucred(tcnp->cn_context));
/*
* Target must be empty if a directory and have no links
* to it. Also, ensure source and target are compatible
* (both directories, or both not directories).
*/
if (( doingdirectory) && (tp->dn_links > 2)) {
- error = ENOTEMPTY;
- goto bad;
+ error = ENOTEMPTY;
+ goto bad;
}
dev_free_name(tnp);
tp = NULL;
dev_add_name(tcnp->cn_nameptr,tdp,NULL,fp,&tnp);
fnp->de_dnp = NULL;
fp->dn_links--; /* one less link to it.. */
- dev_free_name(fnp);
- fp->dn_links--; /* we added one earlier*/
- if (tdp)
- vput(tdvp);
- if (tp)
- vput(fvp);
- vrele(fvp);
- return (error);
+ dev_free_name(fnp);
bad:
- if (tp)
- vput(tvp);
- vput(tdvp);
+ fp->dn_links--; /* we added one earlier*/
out:
- if (vn_lock(fvp, LK_EXCLUSIVE | LK_RETRY, p) == 0) {
- fp->dn_links--; /* we added one earlier*/
- vput(fvp);
- } else
- vrele(fvp);
+ DEVFS_UNLOCK();
return (error);
}
static int
-devfs_symlink(struct vop_symlink_args *ap)
- /*struct vop_symlink_args {
+devfs_symlink(struct vnop_symlink_args *ap)
+ /*struct vnop_symlink_args {
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
- struct vattr *a_vap;
+ struct vnode_attr *a_vap;
char *a_target;
+ vfs_context_t a_context;
} */
{
- struct componentname * cnp = ap->a_cnp;
- struct vnode *vp = NULL;
+ struct componentname * cnp = ap->a_cnp;
+ vfs_context_t ctx = cnp->cn_context;
+ struct proc *p = vfs_context_proc(ctx);
int error = 0;
devnode_t * dir_p;
devnode_type_t typeinfo;
devdirent_t * nm_p;
devnode_t * dev_p;
- struct vattr * vap = ap->a_vap;
+ struct vnode_attr * vap = ap->a_vap;
struct vnode * * vpp = ap->a_vpp;
- struct proc *p = cnp->cn_proc;
- struct timeval tv;
dir_p = VTODN(ap->a_dvp);
typeinfo.Slnk.name = ap->a_target;
typeinfo.Slnk.namelen = strlen(ap->a_target);
- DEVFS_LOCK(p);
+
+ DEVFS_LOCK();
error = dev_add_entry(cnp->cn_nameptr, dir_p, DEV_SLNK,
&typeinfo, NULL, NULL, &nm_p);
- DEVFS_UNLOCK(p);
if (error) {
goto failure;
}
-
dev_p = nm_p->de_dnp;
dev_p->dn_uid = dir_p->dn_uid;
dev_p->dn_gid = dir_p->dn_gid;
dev_p->dn_mode = vap->va_mode;
dn_copy_times(dev_p, dir_p);
+
error = devfs_dntovn(dev_p, vpp, p);
- if (error)
- goto failure;
- vp = *vpp;
- vput(vp);
- failure:
- if ((cnp->cn_flags & SAVESTART) == 0)
- FREE_ZONE(cnp->cn_pnbuf, cnp->cn_pnlen, M_NAMEI);
- vput(ap->a_dvp);
+failure:
+ DEVFS_UNLOCK();
+
return error;
}
/*
* Mknod vnode call
*/
-/* ARGSUSED */
-int
-devfs_mknod(ap)
- struct vop_mknod_args /* {
+static int
+devfs_mknod(struct vnop_mknod_args *ap)
+ /* struct vnop_mknod_args {
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
- struct vattr *a_vap;
- } */ *ap;
+ struct vnode_attr *a_vap;
+ vfs_context_t a_context;
+ } */
{
struct componentname * cnp = ap->a_cnp;
+ vfs_context_t ctx = cnp->cn_context;
+ struct proc *p = vfs_context_proc(ctx);
devnode_t * dev_p;
devdirent_t * devent;
devnode_t * dir_p; /* devnode for parent directory */
struct vnode * dvp = ap->a_dvp;
int error = 0;
devnode_type_t typeinfo;
- struct vattr * vap = ap->a_vap;
+ struct vnode_attr * vap = ap->a_vap;
struct vnode ** vpp = ap->a_vpp;
- struct proc * p = cnp->cn_proc;
*vpp = NULL;
- if (!vap->va_type == VBLK && !vap->va_type == VCHR) {
- error = EINVAL; /* only support mknod of special files */
- goto failure;
+ if (!(vap->va_type == VBLK) && !(vap->va_type == VCHR)) {
+ return (EINVAL); /* only support mknod of special files */
}
dir_p = VTODN(dvp);
typeinfo.dev = vap->va_rdev;
- DEVFS_LOCK(p);
+
+ DEVFS_LOCK();
error = dev_add_entry(cnp->cn_nameptr, dir_p,
(vap->va_type == VBLK) ? DEV_BDEV : DEV_CDEV,
&typeinfo, NULL, NULL, &devent);
- DEVFS_UNLOCK(p);
if (error) {
- goto failure;
+ goto failure;
}
dev_p = devent->de_dnp;
error = devfs_dntovn(dev_p, vpp, p);
if (error)
- goto failure;
- dev_p->dn_uid = cnp->cn_cred->cr_uid;
- dev_p->dn_gid = dir_p->dn_gid;
+ goto failure;
+ dev_p->dn_uid = vap->va_uid;
+ dev_p->dn_gid = vap->va_gid;
dev_p->dn_mode = vap->va_mode;
- failure:
- if (*vpp) {
- vput(*vpp);
- *vpp = 0;
- }
- if ((cnp->cn_flags & SAVESTART) == 0)
- FREE_ZONE(cnp->cn_pnbuf, cnp->cn_pnlen, M_NAMEI);
- vput(dvp);
+ VATTR_SET_SUPPORTED(vap, va_uid);
+ VATTR_SET_SUPPORTED(vap, va_gid);
+ VATTR_SET_SUPPORTED(vap, va_mode);
+failure:
+ DEVFS_UNLOCK();
+
return (error);
}
* Vnode op for readdir
*/
static int
-devfs_readdir(struct vop_readdir_args *ap)
- /*struct vop_readdir_args {
+devfs_readdir(struct vnop_readdir_args *ap)
+ /*struct vnop_readdir_args {
struct vnode *a_vp;
struct uio *a_uio;
- struct ucred *a_cred;
- int *eofflag;
- int *ncookies;
- u_int **cookies;
+ int a_flags;
+ int *a_eofflag;
+ int *a_numdirent;
+ vfs_context_t a_context;
} */
{
struct vnode *vp = ap->a_vp;
int reclen;
int nodenumber;
int startpos,pos;
- struct proc * p = uio->uio_procp;
+
+ if (ap->a_flags & (VNODE_READDIR_EXTENDED | VNODE_READDIR_REQSEEKOFF))
+ return (EINVAL);
/* set up refs to dir */
dir_node = VTODN(vp);
- if(dir_node->dn_type != DEV_DIR)
+ if (dir_node->dn_type != DEV_DIR)
return(ENOTDIR);
-
pos = 0;
startpos = uio->uio_offset;
- DEVFS_LOCK(p);
+
+ DEVFS_LOCK();
+
name_node = dir_node->dn_typeinfo.Dir.dirlist;
nodenumber = 0;
- dir_node->dn_flags |= DN_ACCESS;
- while ((name_node || (nodenumber < 2)) && (uio->uio_resid > 0))
+ dir_node->dn_access = 1;
+
+ while ((name_node || (nodenumber < 2)) && (uio_resid(uio) > 0))
{
switch(nodenumber)
{
if(pos >= startpos) /* made it to the offset yet? */
{
- if (uio->uio_resid < reclen) /* will it fit? */
+ if (uio_resid(uio) < reclen) /* will it fit? */
break;
strcpy( dirent.d_name,name);
if ((error = uiomove ((caddr_t)&dirent,
name_node = name_node->de_next;
nodenumber++;
}
- DEVFS_UNLOCK(p);
+ DEVFS_UNLOCK();
uio->uio_offset = pos;
return (error);
/*
*/
static int
-devfs_readlink(struct vop_readlink_args *ap)
- /*struct vop_readlink_args {
+devfs_readlink(struct vnop_readlink_args *ap)
+ /*struct vnop_readlink_args {
struct vnode *a_vp;
struct uio *a_uio;
- struct ucred *a_cred;
+ vfs_context_t a_context;
} */
{
struct vnode *vp = ap->a_vp;
/* set up refs to dir */
lnk_node = VTODN(vp);
- if(lnk_node->dn_type != DEV_SLNK)
- return(EINVAL);
- if ((error = VOP_ACCESS(vp, VREAD, ap->a_cred, NULL)) != 0) { /* XXX */
- return error;
+
+ if (lnk_node->dn_type != DEV_SLNK) {
+ error = EINVAL;
+ goto out;
}
error = uiomove(lnk_node->dn_typeinfo.Slnk.name,
lnk_node->dn_typeinfo.Slnk.namelen, uio);
+out:
return error;
}
static int
-devfs_abortop(struct vop_abortop_args *ap)
- /*struct vop_abortop_args {
- struct vnode *a_dvp;
- struct componentname *a_cnp;
- } */
-{
- if ((ap->a_cnp->cn_flags & (HASBUF | SAVESTART)) == HASBUF) {
- FREE_ZONE(ap->a_cnp->cn_pnbuf, ap->a_cnp->cn_pnlen, M_NAMEI);
- }
- return 0;
-}
-
-
-static int
-devfs_reclaim(struct vop_reclaim_args *ap)
- /*struct vop_reclaim_args {
+devfs_reclaim(struct vnop_reclaim_args *ap)
+ /*struct vnop_reclaim_args {
struct vnode *a_vp;
} */
{
struct vnode * vp = ap->a_vp;
devnode_t * dnp = VTODN(vp);
+ DEVFS_LOCK();
+
if (dnp) {
/*
* do the same as devfs_inactive in case it is not called
*/
dnp->dn_vn = NULL;
vp->v_data = NULL;
+
if (dnp->dn_delete) {
devnode_free(dnp);
}
}
+ DEVFS_UNLOCK();
+
return(0);
}
+
/*
- * Print out the contents of a /devfs vnode.
+ * Get configurable pathname variables.
*/
static int
-devfs_print(struct vop_print_args *ap)
- /*struct vop_print_args {
+devs_vnop_pathconf(
+ struct vnop_pathconf_args /* {
struct vnode *a_vp;
- } */
+ int a_name;
+ int *a_retval;
+ vfs_context_t a_context;
+ } */ *ap)
{
+ switch (ap->a_name) {
+ case _PC_LINK_MAX:
+ /* arbitrary limit matching HFS; devfs has no hard limit */
+ *ap->a_retval = 32767;
+ break;
+ case _PC_NAME_MAX:
+ *ap->a_retval = DEVMAXNAMESIZE - 1; /* includes NUL */
+ break;
+ case _PC_PATH_MAX:
+ *ap->a_retval = DEVMAXPATHSIZE - 1; /* XXX nonconformant */
+ break;
+ case _PC_CHOWN_RESTRICTED:
+ *ap->a_retval = 1;
+ break;
+ case _PC_NO_TRUNC:
+ *ap->a_retval = 0;
+ break;
+ case _PC_CASE_SENSITIVE:
+ *ap->a_retval = 1;
+ break;
+ case _PC_CASE_PRESERVING:
+ *ap->a_retval = 1;
+ break;
+ default:
+ return (EINVAL);
+ }
return (0);
}
+
+
/**************************************************************************\
* pseudo ops *
\**************************************************************************/
/*
*
- * struct vop_inactive_args {
+ * struct vnop_inactive_args {
* struct vnode *a_vp;
- * struct proc *a_p;
+ * vfs_context_t a_context;
* }
*/
static int
-devfs_inactive(struct vop_inactive_args *ap)
+devfs_inactive(__unused struct vnop_inactive_args *ap)
{
- struct vnode * vp = ap->a_vp;
- devnode_t * dnp = VTODN(vp);
-
- if (dnp) {
- dnp->dn_vn = NULL;
- vp->v_data = NULL;
- if (dnp->dn_delete) {
- devnode_free(dnp);
- }
- }
- VOP_UNLOCK(vp, 0, ap->a_p);
return (0);
}
-int
-devfs_update(ap)
- struct vop_update_args /* {
- struct vnode *a_vp;
- struct timeval *a_access;
- struct timeval *a_modify;
- int a_waitfor;
- } */ *ap;
+/*
+ * called with DEVFS_LOCK held
+ */
+static int
+devfs_update(struct vnode *vp, struct timeval *access, struct timeval *modify)
{
- register struct fs *fs;
- int error;
devnode_t * ip;
+ struct timeval now;
+
+ ip = VTODN(vp);
+ if (vp->v_mount->mnt_flag & MNT_RDONLY) {
+ ip->dn_access = 0;
+ ip->dn_change = 0;
+ ip->dn_update = 0;
- ip = VTODN(ap->a_vp);
- if (ap->a_vp->v_mount->mnt_flag & MNT_RDONLY) {
- ip->dn_flags &=
- ~(DN_ACCESS | DN_CHANGE | DN_MODIFIED | DN_UPDATE);
return (0);
}
- if ((ip->dn_flags &
- (DN_ACCESS | DN_CHANGE | DN_MODIFIED | DN_UPDATE)) == 0)
- return (0);
- dn_times(ip, time, time);
+ microtime(&now);
+ dn_times(ip, access, modify, &now);
+
return (0);
}
/* The following ops are used by directories and symlinks */
int (**devfs_vnodeop_p)(void *);
static struct vnodeopv_entry_desc devfs_vnodeop_entries[] = {
- { &vop_default_desc, (VOPFUNC)vn_default_error },
- { &vop_lookup_desc, (VOPFUNC)devfs_lookup }, /* lookup */
- { &vop_create_desc, (VOPFUNC)err_create }, /* create */
- { &vop_whiteout_desc, (VOPFUNC)err_whiteout }, /* whiteout */
- { &vop_mknod_desc, (VOPFUNC)devfs_mknod }, /* mknod */
- { &vop_open_desc, (VOPFUNC)nop_open }, /* open */
- { &vop_close_desc, (VOPFUNC)devfs_close }, /* close */
- { &vop_access_desc, (VOPFUNC)devfs_access }, /* access */
- { &vop_getattr_desc, (VOPFUNC)devfs_getattr }, /* getattr */
- { &vop_setattr_desc, (VOPFUNC)devfs_setattr }, /* setattr */
- { &vop_read_desc, (VOPFUNC)devfs_read }, /* read */
- { &vop_write_desc, (VOPFUNC)devfs_write }, /* write */
- { &vop_lease_desc, (VOPFUNC)nop_lease }, /* lease */
- { &vop_ioctl_desc, (VOPFUNC)err_ioctl }, /* ioctl */
- { &vop_select_desc, (VOPFUNC)err_select }, /* select */
- { &vop_revoke_desc, (VOPFUNC)err_revoke }, /* revoke */
- { &vop_mmap_desc, (VOPFUNC)err_mmap }, /* mmap */
- { &vop_fsync_desc, (VOPFUNC)nop_fsync }, /* fsync */
- { &vop_seek_desc, (VOPFUNC)err_seek }, /* seek */
- { &vop_remove_desc, (VOPFUNC)devfs_remove }, /* remove */
- { &vop_link_desc, (VOPFUNC)devfs_link }, /* link */
- { &vop_rename_desc, (VOPFUNC)devfs_rename }, /* rename */
- { &vop_mkdir_desc, (VOPFUNC)err_mkdir }, /* mkdir */
- { &vop_rmdir_desc, (VOPFUNC)err_rmdir }, /* rmdir */
- { &vop_symlink_desc, (VOPFUNC)devfs_symlink }, /* symlink */
- { &vop_readdir_desc, (VOPFUNC)devfs_readdir }, /* readdir */
- { &vop_readlink_desc, (VOPFUNC)devfs_readlink }, /* readlink */
- { &vop_abortop_desc, (VOPFUNC)devfs_abortop }, /* abortop */
- { &vop_inactive_desc, (VOPFUNC)devfs_inactive }, /* inactive */
- { &vop_reclaim_desc, (VOPFUNC)devfs_reclaim }, /* reclaim */
- { &vop_lock_desc, (VOPFUNC)nop_lock }, /* lock */
- { &vop_unlock_desc, (VOPFUNC)nop_unlock }, /* unlock */
- { &vop_bmap_desc, (VOPFUNC)err_bmap }, /* bmap */
- { &vop_strategy_desc, (VOPFUNC)err_strategy }, /* strategy */
- { &vop_print_desc, (VOPFUNC)err_print }, /* print */
- { &vop_islocked_desc, (VOPFUNC)nop_islocked }, /* islocked */
- { &vop_pathconf_desc, (VOPFUNC)err_pathconf }, /* pathconf */
- { &vop_advlock_desc, (VOPFUNC)err_advlock }, /* advlock */
- { &vop_blkatoff_desc, (VOPFUNC)err_blkatoff }, /* blkatoff */
- { &vop_valloc_desc, (VOPFUNC)err_valloc }, /* valloc */
- { &vop_reallocblks_desc, (VOPFUNC)err_reallocblks }, /* reallocblks */
- { &vop_vfree_desc, (VOPFUNC)err_vfree }, /* vfree */
- { &vop_truncate_desc, (VOPFUNC)err_truncate }, /* truncate */
- { &vop_update_desc, (VOPFUNC)devfs_update }, /* update */
- { &vop_bwrite_desc, (VOPFUNC)err_bwrite },
- { &vop_pagein_desc, (VOPFUNC)err_pagein }, /* Pagein */
- { &vop_pageout_desc, (VOPFUNC)err_pageout }, /* Pageout */
- { &vop_copyfile_desc, (VOPFUNC)err_copyfile }, /* Copyfile */
- { &vop_blktooff_desc, (VOPFUNC)err_blktooff }, /* blktooff */
- { &vop_offtoblk_desc, (VOPFUNC)err_offtoblk }, /* offtoblk */
- { &vop_cmap_desc, (VOPFUNC)err_cmap }, /* cmap */
+ { &vnop_default_desc, (VOPFUNC)vn_default_error },
+ { &vnop_lookup_desc, (VOPFUNC)devfs_lookup }, /* lookup */
+ { &vnop_create_desc, (VOPFUNC)err_create }, /* create */
+ { &vnop_whiteout_desc, (VOPFUNC)err_whiteout }, /* whiteout */
+ { &vnop_mknod_desc, (VOPFUNC)devfs_mknod }, /* mknod */
+ { &vnop_open_desc, (VOPFUNC)nop_open }, /* open */
+ { &vnop_close_desc, (VOPFUNC)devfs_close }, /* close */
+ { &vnop_getattr_desc, (VOPFUNC)devfs_getattr }, /* getattr */
+ { &vnop_setattr_desc, (VOPFUNC)devfs_setattr }, /* setattr */
+ { &vnop_read_desc, (VOPFUNC)devfs_read }, /* read */
+ { &vnop_write_desc, (VOPFUNC)devfs_write }, /* write */
+ { &vnop_ioctl_desc, (VOPFUNC)err_ioctl }, /* ioctl */
+ { &vnop_select_desc, (VOPFUNC)err_select }, /* select */
+ { &vnop_revoke_desc, (VOPFUNC)err_revoke }, /* revoke */
+ { &vnop_mmap_desc, (VOPFUNC)err_mmap }, /* mmap */
+ { &vnop_fsync_desc, (VOPFUNC)nop_fsync }, /* fsync */
+ { &vnop_remove_desc, (VOPFUNC)devfs_remove }, /* remove */
+ { &vnop_link_desc, (VOPFUNC)devfs_link }, /* link */
+ { &vnop_rename_desc, (VOPFUNC)devfs_rename }, /* rename */
+ { &vnop_mkdir_desc, (VOPFUNC)err_mkdir }, /* mkdir */
+ { &vnop_rmdir_desc, (VOPFUNC)err_rmdir }, /* rmdir */
+ { &vnop_symlink_desc, (VOPFUNC)devfs_symlink }, /* symlink */
+ { &vnop_readdir_desc, (VOPFUNC)devfs_readdir }, /* readdir */
+ { &vnop_readlink_desc, (VOPFUNC)devfs_readlink }, /* readlink */
+ { &vnop_inactive_desc, (VOPFUNC)devfs_inactive }, /* inactive */
+ { &vnop_reclaim_desc, (VOPFUNC)devfs_reclaim }, /* reclaim */
+ { &vnop_strategy_desc, (VOPFUNC)err_strategy }, /* strategy */
+ { &vnop_pathconf_desc, (VOPFUNC)devs_vnop_pathconf }, /* pathconf */
+ { &vnop_advlock_desc, (VOPFUNC)err_advlock }, /* advlock */
+ { &vnop_bwrite_desc, (VOPFUNC)err_bwrite },
+ { &vnop_pagein_desc, (VOPFUNC)err_pagein }, /* Pagein */
+ { &vnop_pageout_desc, (VOPFUNC)err_pageout }, /* Pageout */
+ { &vnop_copyfile_desc, (VOPFUNC)err_copyfile }, /* Copyfile */
+ { &vnop_blktooff_desc, (VOPFUNC)err_blktooff }, /* blktooff */
+ { &vnop_offtoblk_desc, (VOPFUNC)err_offtoblk }, /* offtoblk */
+ { &vnop_blockmap_desc, (VOPFUNC)err_blockmap }, /* blockmap */
{ (struct vnodeop_desc*)NULL, (int(*)())NULL }
};
struct vnodeopv_desc devfs_vnodeop_opv_desc =
/* The following ops are used by the device nodes */
int (**devfs_spec_vnodeop_p)(void *);
static struct vnodeopv_entry_desc devfs_spec_vnodeop_entries[] = {
- { &vop_default_desc, (VOPFUNC)vn_default_error },
- { &vop_lookup_desc, (VOPFUNC)spec_lookup }, /* lookup */
- { &vop_create_desc, (VOPFUNC)spec_create }, /* create */
- { &vop_mknod_desc, (VOPFUNC)spec_mknod }, /* mknod */
- { &vop_open_desc, (VOPFUNC)spec_open }, /* open */
- { &vop_close_desc, (VOPFUNC)devfsspec_close }, /* close */
- { &vop_access_desc, (VOPFUNC)devfs_access }, /* access */
- { &vop_getattr_desc, (VOPFUNC)devfs_getattr }, /* getattr */
- { &vop_setattr_desc, (VOPFUNC)devfs_setattr }, /* setattr */
- { &vop_read_desc, (VOPFUNC)devfsspec_read }, /* read */
- { &vop_write_desc, (VOPFUNC)devfsspec_write }, /* write */
- { &vop_lease_desc, (VOPFUNC)spec_lease_check }, /* lease */
- { &vop_ioctl_desc, (VOPFUNC)spec_ioctl }, /* ioctl */
- { &vop_select_desc, (VOPFUNC)spec_select }, /* select */
- { &vop_revoke_desc, (VOPFUNC)spec_revoke }, /* revoke */
- { &vop_mmap_desc, (VOPFUNC)spec_mmap }, /* mmap */
- { &vop_fsync_desc, (VOPFUNC)spec_fsync }, /* fsync */
- { &vop_seek_desc, (VOPFUNC)spec_seek }, /* seek */
- { &vop_remove_desc, (VOPFUNC)devfs_remove }, /* remove */
- { &vop_link_desc, (VOPFUNC)devfs_link }, /* link */
- { &vop_rename_desc, (VOPFUNC)spec_rename }, /* rename */
- { &vop_mkdir_desc, (VOPFUNC)spec_mkdir }, /* mkdir */
- { &vop_rmdir_desc, (VOPFUNC)spec_rmdir }, /* rmdir */
- { &vop_symlink_desc, (VOPFUNC)spec_symlink }, /* symlink */
- { &vop_readdir_desc, (VOPFUNC)spec_readdir }, /* readdir */
- { &vop_readlink_desc, (VOPFUNC)spec_readlink }, /* readlink */
- { &vop_abortop_desc, (VOPFUNC)spec_abortop }, /* abortop */
- { &vop_inactive_desc, (VOPFUNC)devfs_inactive }, /* inactive */
- { &vop_reclaim_desc, (VOPFUNC)devfs_reclaim }, /* reclaim */
- { &vop_lock_desc, (VOPFUNC)nop_lock }, /* lock */
- { &vop_unlock_desc, (VOPFUNC)nop_unlock }, /* unlock */
- { &vop_bmap_desc, (VOPFUNC)spec_bmap }, /* bmap */
- { &vop_strategy_desc, (VOPFUNC)spec_strategy }, /* strategy */
- { &vop_print_desc, (VOPFUNC)devfs_print }, /* print */
- { &vop_islocked_desc, (VOPFUNC)nop_islocked }, /* islocked */
- { &vop_pathconf_desc, (VOPFUNC)spec_pathconf }, /* pathconf */
- { &vop_advlock_desc, (VOPFUNC)spec_advlock }, /* advlock */
- { &vop_blkatoff_desc, (VOPFUNC)spec_blkatoff }, /* blkatoff */
- { &vop_valloc_desc, (VOPFUNC)spec_valloc }, /* valloc */
- { &vop_reallocblks_desc, (VOPFUNC)spec_reallocblks }, /* reallocblks */
- { &vop_vfree_desc, (VOPFUNC)nop_vfree }, /* vfree */
- { &vop_truncate_desc, (VOPFUNC)spec_truncate }, /* truncate */
- { &vop_update_desc, (VOPFUNC)devfs_update }, /* update */
- { &vop_bwrite_desc, (VOPFUNC)vn_bwrite },
- { &vop_devblocksize_desc, (VOPFUNC)spec_devblocksize }, /* devblocksize */
- { &vop_pagein_desc, (VOPFUNC)err_pagein }, /* Pagein */
- { &vop_pageout_desc, (VOPFUNC)err_pageout }, /* Pageout */
- { &vop_copyfile_desc, (VOPFUNC)err_copyfile }, /* Copyfile */
- { &vop_blktooff_desc, (VOPFUNC)spec_blktooff }, /* blktooff */
- { &vop_blktooff_desc, (VOPFUNC)spec_offtoblk }, /* blkofftoblk */
- { &vop_cmap_desc, (VOPFUNC)spec_cmap }, /* cmap */
+ { &vnop_default_desc, (VOPFUNC)vn_default_error },
+ { &vnop_lookup_desc, (VOPFUNC)spec_lookup }, /* lookup */
+ { &vnop_create_desc, (VOPFUNC)spec_create }, /* create */
+ { &vnop_mknod_desc, (VOPFUNC)spec_mknod }, /* mknod */
+ { &vnop_open_desc, (VOPFUNC)spec_open }, /* open */
+ { &vnop_close_desc, (VOPFUNC)devfsspec_close }, /* close */
+ { &vnop_getattr_desc, (VOPFUNC)devfs_getattr }, /* getattr */
+ { &vnop_setattr_desc, (VOPFUNC)devfs_setattr }, /* setattr */
+ { &vnop_read_desc, (VOPFUNC)devfsspec_read }, /* read */
+ { &vnop_write_desc, (VOPFUNC)devfsspec_write }, /* write */
+ { &vnop_ioctl_desc, (VOPFUNC)spec_ioctl }, /* ioctl */
+ { &vnop_select_desc, (VOPFUNC)spec_select }, /* select */
+ { &vnop_revoke_desc, (VOPFUNC)spec_revoke }, /* revoke */
+ { &vnop_mmap_desc, (VOPFUNC)spec_mmap }, /* mmap */
+ { &vnop_fsync_desc, (VOPFUNC)spec_fsync }, /* fsync */
+ { &vnop_remove_desc, (VOPFUNC)devfs_remove }, /* remove */
+ { &vnop_link_desc, (VOPFUNC)devfs_link }, /* link */
+ { &vnop_rename_desc, (VOPFUNC)spec_rename }, /* rename */
+ { &vnop_mkdir_desc, (VOPFUNC)spec_mkdir }, /* mkdir */
+ { &vnop_rmdir_desc, (VOPFUNC)spec_rmdir }, /* rmdir */
+ { &vnop_symlink_desc, (VOPFUNC)spec_symlink }, /* symlink */
+ { &vnop_readdir_desc, (VOPFUNC)spec_readdir }, /* readdir */
+ { &vnop_readlink_desc, (VOPFUNC)spec_readlink }, /* readlink */
+ { &vnop_inactive_desc, (VOPFUNC)devfs_inactive }, /* inactive */
+ { &vnop_reclaim_desc, (VOPFUNC)devfs_reclaim }, /* reclaim */
+ { &vnop_strategy_desc, (VOPFUNC)spec_strategy }, /* strategy */
+ { &vnop_pathconf_desc, (VOPFUNC)spec_pathconf }, /* pathconf */
+ { &vnop_advlock_desc, (VOPFUNC)spec_advlock }, /* advlock */
+ { &vnop_bwrite_desc, (VOPFUNC)vn_bwrite },
+ { &vnop_pagein_desc, (VOPFUNC)err_pagein }, /* Pagein */
+ { &vnop_pageout_desc, (VOPFUNC)err_pageout }, /* Pageout */
+ { &vnop_copyfile_desc, (VOPFUNC)err_copyfile }, /* Copyfile */
+ { &vnop_blktooff_desc, (VOPFUNC)spec_blktooff }, /* blktooff */
+ { &vnop_blktooff_desc, (VOPFUNC)spec_offtoblk }, /* blkofftoblk */
+ { &vnop_blockmap_desc, (VOPFUNC)spec_blockmap }, /* blockmap */
{ (struct vnodeop_desc*)NULL, (int(*)())NULL }
};
struct vnodeopv_desc devfs_spec_vnodeop_opv_desc =