/*-
 * Copyright (c) 2005-2009
 *      Hans Petter Selasky. All rights reserved.
 * Copyright (c) 2000-2004
 *	Poul-Henning Kamp.  All rights reserved.
 * Copyright (c) 1989, 1992-1993, 1995
 *	The Regents of the University of California.  All rights reserved.
 * Copyright (c)
 *       Jan-Simon Pendry. All rights reserved.
 *
 * 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. 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.
 *
 * This file is a lite version of "FreeBSD/src/sys/fs/devfs/devfs_vnops.c"
 *
 * On NetBSD the locking rules are defined in: /usr/src/sys/kern/vnode_if.src
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/conf.h>
#include <sys/dirent.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/proc.h>
#include <sys/sysctl.h>
#include <sys/vnode.h>
#include <sys/fcntl.h>

#include <sys/freebsd_compat.h>

typedef int (__vnodeop_t)(void *);

static __vnodeop_t **devfs_vnodeop_p;

#if (__NetBSD_Version__ >= 500000000)
#define	get_a_p(ap) (curthread)
#define	get_cn_proc(cnp) (curthread)
#define	BSD_VOP_ACCESS(a,b,c,d0,d1) \
  VOP_ACCESS(a,b,c)
#elif (__NetBSD_Version__ >= 400000000)
#define	get_a_p(ap) ((ap)->a_l->l_proc)
#define	get_cn_proc(cnp) ((cnp)->cn_lwp->l_proc)
#define	BSD_VOP_ACCESS(a,b,c,d0,d1) \
  VOP_ACCESS(a,b,c,d1)
#else
#define	get_a_p(ap) ((ap)->a_p)
#define	get_cn_proc(cnp) ((cnp)->cn_proc)
#define	BSD_VOP_ACCESS(a,b,c,d0,d1) \
  VOP_ACCESS(a,b,c,d0)
#endif

/*
 * helper functions
 */
int
devfs_allocv(struct devfs_dirent *de, struct mount *mp,
    struct vnode **vpp, struct proc *td)
{
	struct vnode *vp;
	struct cdev **pdev;
	struct cdev *dev;
	int error;

	while ((vp = de->de_vnode) != NULL) {

		error = vget(vp, LK_EXCLUSIVE | LK_RETRY);
		if (error == ENOENT) {
			/* vnode was reclaimed. */
			continue;
		}
		vpp[0] = vp;
		return (error);
	}
	if (de->de_dirent->d_type == DT_FIFO) {
		pdev = devfs_itod(de->de_inode);
		if (pdev == NULL)
			return (ENOENT);
		dev = *pdev;
		if (dev == NULL)
			return (ENOENT);
	} else {
		dev = NULL;
	}

	error = getnewvnode(VT_NON, mp, devfs_vnodeop_p, &vp);
	if (error) {
		printf("%s: ERROR: %d: failed to allocate new vnode!\n",
		    __FUNCTION__, error);
		return (error);
	}
	if (de->de_dirent->d_type == DT_FIFO) {
		vp->v_type = VFIFO;

		dev_ref(dev);

		dev_lock();
		LIST_INSERT_HEAD(&dev->si_alist, de, de_alias);
		dev_unlock();
	} else if (de->de_dirent->d_type == DT_DIR) {
		vp->v_type = VDIR;
	} else if (de->de_dirent->d_type == DT_LNK) {
		vp->v_type = VLNK;
	} else {
		vp->v_type = VBAD;
	}

	de->de_dev = dev;

	vp->v_data = de;
	de->de_vnode = vp;
	vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);
	uvm_vnp_setsize(vp, 0);		/* XXX not used */

	vpp[0] = vp;
	return (0);
}

static int
devfs_fp_check(struct vnode *vp, struct cdev **devp, struct __cdevsw **dswp)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_dirent *de = vp->v_data;
	struct cdev *cdev;
	struct __cdevsw *csw;

	if (de == NULL) {
		return (ENXIO);
	}
	cdev = de->de_dev;

	if (cdev == NULL) {
		return (ENXIO);
	}
	csw = dev_ref_cdevsw(cdev);
	if (csw == NULL) {
		return (ENXIO);
	}
	*devp = cdev;
	*dswp = csw;

	return (0);
}

static u_int32_t
devfs_random(void)
{
	/*
	 * make sure that people don't make assumptions about device
	 * major/minor numbers in userspace
	 */
	return ((boottime.tv_sec & 0x0f0f) | 0xFF000000UL);
}

dev_t
dev2udev(struct cdev *dev)
{
	if (dev == NULL) {
		return (NODEV);
	}
	return (dev->si_inode ^ devfs_random());
}

/*
 * vnode operations
 */

static int
devfs_open(struct vop_open_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;
	struct proc *td = get_a_p(ap);
	struct __cdevsw *dsw;
	struct cdev *dev;
	int error;

	if (vp->v_type == VDIR)
		return (0);

	if (vp->v_type != VFIFO)
		return (ENXIO);

	error = devfs_fp_check(vp, &dev, &dsw);
	if (error) {
		goto done;
	}
	dev->si_file_flags &= ~(O_NONBLOCK | O_DIRECT);

	VOP_UNLOCK(vp, 0);

	if (dsw->d_fdopen) {
		error = EINVAL;		/* not supported */
	} else {
		error = dsw->d_open(dev, ap->a_mode, S_IFCHR, td);
	}

	vn_lock(vp, LK_EXCLUSIVE | LK_RETRY);

	dev_rel(dev);

done:
	return (error);
}

static int
devfs_read(struct vop_read_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	int error;

	if (ap->a_vp->v_type == VDIR) {
		error = VOP_READDIR(ap->a_vp, ap->a_uio, ap->a_cred,
		    NULL, NULL, NULL);
	} else if (ap->a_vp->v_type == VFIFO) {
		struct __cdevsw *dsw;
		struct cdev *dev;
		struct uio *uio;
		int ioflag;
		int resid;

		error = devfs_fp_check(ap->a_vp, &dev, &dsw);
		if (error) {
			goto done;
		}
		uio = ap->a_uio;
		resid = uio->uio_resid;

		ioflag = dev->si_file_flags & (O_NONBLOCK | O_DIRECT);
		if (ioflag & O_DIRECT) {
			ioflag |= IO_DIRECT;
		}
		if (ap->a_ioflag & FNONBLOCK) {
			ioflag |= O_NONBLOCK;
		}
		VOP_UNLOCK(ap->a_vp, 0);

		error = dsw->d_read(dev, uio, ioflag);

		vn_lock(ap->a_vp, LK_EXCLUSIVE | LK_RETRY);

		dev_rel(dev);

		if ((uio->uio_resid != resid) ||
		    ((error == 0) && (resid != 0))) {
			devfs_timestamp(&dev->si_atime);
		}
	} else {
		error = ENXIO;
	}

done:
	return error;
}

static int
devfs_write(struct vop_write_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	int error;

	if (ap->a_vp->v_type == VFIFO) {
		struct uio *uio;
		struct cdev *dev;
		struct __cdevsw *dsw;
		int ioflag;
		int resid;

		error = devfs_fp_check(ap->a_vp, &dev, &dsw);
		if (error) {
			goto done;
		}
		uio = ap->a_uio;
		resid = uio->uio_resid;

		ioflag = dev->si_file_flags & (O_NONBLOCK | O_DIRECT);
		if (ioflag & O_DIRECT) {
			ioflag |= IO_DIRECT;
		}
		if (ap->a_ioflag & FNONBLOCK) {
			ioflag |= O_NONBLOCK;
		}
		VOP_UNLOCK(ap->a_vp, 0);

		error = dsw->d_write(dev, uio, ioflag);

		vn_lock(ap->a_vp, LK_EXCLUSIVE | LK_RETRY);

		dev_rel(dev);

		if ((uio->uio_resid != resid) ||
		    ((error == 0) && (resid != 0))) {
			devfs_timestamp(&dev->si_ctime);
			dev->si_mtime = dev->si_ctime;
		}
	} else {
		error = ENXIO;
	}
done:
	return error;
}

static int
devfs_ioctl(struct vop_ioctl_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	int error;

	if (ap->a_vp->v_type == VFIFO) {
		struct cdev *dev;
		struct __cdevsw *dsw;
		u_int32_t cmd = ap->a_command;
		void *data = ap->a_data;

		error = devfs_fp_check(ap->a_vp, &dev, &dsw);
		if (error) {
			goto done;
		}
		if (cmd == FIONBIO) {
			if (*(int *)data) {
				dev->si_file_flags |= O_NONBLOCK;
			} else {
				dev->si_file_flags &= ~O_NONBLOCK;
			}

			/* this IOCTL is also forwarded to the driver */
		} else if (cmd == FIODTYPE) {
			*(int *)data = (dsw->d_flags & D_TYPEMASK);
			dev_rel(dev);
			error = 0;
			goto done;
		} else if (cmd == FIODGNAME) {
			struct fiodgname_arg *fgn = data;
			u_int8_t buf[128];

			error = snprintf(&buf[0], sizeof(buf), "%s", devtoname(dev));
			dev_rel(dev);

			if ((error < 0) ||
			    (error >= (sizeof(buf) - 1))) {
				error = EINVAL;
				goto done;
			}
			error++;
			if (error > fgn->len) {
				error = EINVAL;
				goto done;
			}
			error = copyout(&buf[0], fgn->buf, error);
			goto done;
		}
		error = dsw->d_ioctl(dev, cmd, data, ap->a_fflag, get_a_p(ap));

		dev_rel(dev);
#if 0
		if (error == ENOIOCTL) {
			error = ENOTTY;
		}
#endif
	} else {
		error = ENXIO;
	}
done:
	return error;
}

static int
devfs_poll(struct vop_poll_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	int error;

	if (ap->a_vp->v_type == VFIFO) {
		struct cdev *dev;
		struct __cdevsw *dsw;

		error = devfs_fp_check(ap->a_vp, &dev, &dsw);
		if (error) {
			goto done;
		}
		error = dsw->d_poll(dev, ap->a_events, get_a_p(ap));

		dev_rel(dev);
	} else {
		error = ENXIO;
	}
done:
	return error;
}

static int
devfs_kqfilter(struct vop_kqfilter_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	int error;

	if (ap->a_vp->v_type == VFIFO) {
		struct cdev *dev;
		struct __cdevsw *dsw;

		error = devfs_fp_check(ap->a_vp, &dev, &dsw);
		if (error) {
			goto done;
		}
		error = dsw->d_kqfilter(dev, ap->a_kn);

		dev_rel(dev);
	} else {
		error = ENXIO;
	}
done:
	return error;
}

static int
devfs_close(struct vop_close_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	int error;

	if (ap->a_vp->v_type == VFIFO) {
		struct cdev *dev;
		struct __cdevsw *dsw;

		error = devfs_fp_check(ap->a_vp, &dev, &dsw);
		if (error) {
			goto done;
		}
		VOP_UNLOCK(ap->a_vp, 0);

		error = dsw->d_close(dev, ap->a_fflag, S_IFCHR, get_a_p(ap));

		vn_lock(ap->a_vp, LK_EXCLUSIVE | LK_RETRY);

		dev_rel(dev);
	} else {
		error = ENXIO;
	}
done:
	return error;
}

static int
devfs_access(struct vop_access_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;
	struct devfs_dirent *de;
	int error;

	de = vp->v_data;
	if (vp->v_type == VDIR) {
		de = de->de_dir;
	}
	error = vaccess
	    (vp->v_type, de->de_mode, de->de_uid, de->de_gid,
	    ap->a_mode, ap->a_cred);

	if (error == 0) {
		goto done;
	}
done:
	return (error);
}

static void
fix_time(struct timespec *tv)
{
	if (tv->tv_sec == 0) {
		tv->tv_sec = boottime.tv_sec;
		tv->tv_nsec = boottime.tv_usec * 1000;
	}
	return;
}

static int
devfs_getattr(struct vop_getattr_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;
	struct vattr *vap = ap->a_vap;
	struct devfs_dirent *de = vp->v_data;
	struct cdev *dev;

	if (vp->v_type == VDIR) {
		de = de->de_dir;
	}
	bzero(vap, sizeof(*vap));

	vattr_null(vap);
	vap->va_uid = de->de_uid;
	vap->va_gid = de->de_gid;
	vap->va_mode = de->de_mode;

	if (vp->v_type == VLNK)
		vap->va_size = strlen(de->de_symlink);
	else if (vp->v_type == VDIR)
		vap->va_size = vap->va_bytes = DEV_BSIZE;
	else
		vap->va_size = 0;

	if (vp->v_type != VDIR)
		vap->va_bytes = 0;

	vap->va_blocksize = DEV_BSIZE;
	vap->va_type = vp->v_type;

	if (vp->v_type != VFIFO) {
		fix_time(&de->de_atime);
		fix_time(&de->de_mtime);
		fix_time(&de->de_ctime);

		vap->va_atime = de->de_atime;
		vap->va_mtime = de->de_mtime;
		vap->va_ctime = de->de_ctime;
	} else {
		dev = de->de_dev;

		fix_time(&dev->si_atime);
		fix_time(&dev->si_mtime);
		fix_time(&dev->si_ctime);

		vap->va_atime = dev->si_atime;
		vap->va_mtime = dev->si_mtime;
		vap->va_ctime = dev->si_ctime;

		vap->va_rdev = dev->si_inode ^ devfs_random();
	}

	fix_time(&vap->va_birthtime);

	vap->va_gen = 0;
	vap->va_flags = 0;
	vap->va_nlink = de->de_links;
	vap->va_fileid = de->de_inode;
#if (__NetBSD_Version__ >= 300000000)
	vap->va_fsid = vp->v_mount->mnt_stat.f_fsidx.__fsid_val[0];
#else
	vap->va_fsid = vp->v_mount->mnt_stat.f_fsid.val[0];
#endif
	return (0);
}

static int
devfs_setattr(struct vop_setattr_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_dirent *de;
	struct cdev *dev;
	struct vattr *vap;
	struct vnode *vp;
	uid_t uid;
	gid_t gid;
	int error;
	u_int8_t c;

	vap = ap->a_vap;
	vp = ap->a_vp;
	if ((vap->va_type != VNON) ||
	    (vap->va_nlink != VNOVAL) ||
	    (vap->va_fsid != VNOVAL) ||
	    (vap->va_fileid != VNOVAL) ||
	    (vap->va_blocksize != VNOVAL) ||
	    ((vap->va_flags != VNOVAL) && (vap->va_flags != 0)) ||
	    (vap->va_rdev != VNOVAL) ||
#if 0
	    (vap->va_size != VNOVAL) ||
#endif
	    (vap->va_bytes != VNOVAL) ||
	    (vap->va_gen != VNOVAL)) {
		error = EINVAL;
		goto done;
	}
	de = vp->v_data;
	dev = de->de_dev;

	if (vp->v_type == VDIR) {
		de = de->de_dir;
	}
	error = 0;
	c = 0;

	if (vap->va_uid == (uid_t)VNOVAL)
		uid = de->de_uid;
	else
		uid = vap->va_uid;

	if (vap->va_gid == (gid_t)VNOVAL)
		gid = de->de_gid;
	else
		gid = vap->va_gid;

	if ((uid != de->de_uid) ||
	    (gid != de->de_gid)) {
		if (((kauth_cred_getuid(ap->a_cred) != de->de_uid) ||
		    (uid != de->de_uid) ||
		    ((gid != de->de_gid) && !groupmember(gid, ap->a_cred))) &&
		    (error = suser(get_a_p(ap)))) {
			goto done;
		}
		de->de_uid = uid;
		de->de_gid = gid;
		c = 1;
	}
	if (vap->va_mode != (mode_t)VNOVAL) {
		if ((kauth_cred_getuid(ap->a_cred) != de->de_uid) &&
		    (error = suser(get_a_p(ap)))) {
			goto done;
		}
		de->de_mode = vap->va_mode;
		c = 1;
	}
	if ((vap->va_atime.tv_sec != VNOVAL) ||
	    (vap->va_mtime.tv_sec != VNOVAL)) {
		if (((vap->va_vaflags & VA_UTIMES_NULL) == 0) &&
		    (error = BSD_VOP_ACCESS(vp, VWRITE, ap->a_cred, get_a_p(ap), ap->a_l))) {
			goto done;
		}
		if (vap->va_atime.tv_sec != VNOVAL) {
			if (vp->v_type == VFIFO)
				dev->si_atime = vap->va_atime;
			else
				de->de_atime = vap->va_atime;
		}
		if (vap->va_mtime.tv_sec != VNOVAL) {
			if (vp->v_type == VFIFO)
				dev->si_mtime = vap->va_mtime;
			else
				de->de_mtime = vap->va_mtime;
		}
		c = 1;
	}
	if (c) {
		if (vp->v_type == VFIFO)
			devfs_timestamp(&dev->si_ctime);
		else
			devfs_timestamp(&de->de_mtime);
	}
done:
	return error;
}

/*
 * Construct the fully qualified path name relative to the mountpoint
 */
static char *
devfs_fqpn(char *buf, int len, struct vnode *dvp, struct componentname *cnp)
{
	struct devfs_dirent *de, *dd;
	struct devfs_mount *dmp;

	dmp = VFSTODEVFS(dvp->v_mount);
	dd = dvp->v_data;
	len--;
	buf[len] = '\0';
	len -= cnp->cn_namelen;
	if (len < 0) {
		return (NULL);
	}
	bcopy(cnp->cn_nameptr, buf + len, cnp->cn_namelen);
	de = dd;
	while (de != dmp->dm_basedir) {
		len--;
		if (len < 0) {
			return (NULL);
		}
		buf[len] = '/';
		len -= de->de_dirent->d_namlen;
		if (len < 0) {
			return (NULL);
		}
		bcopy(de->de_dirent->d_name, buf + len,
		    de->de_dirent->d_namlen);
		de = TAILQ_FIRST(&de->de_dlist);	/* "." */
		de = TAILQ_NEXT(de, de_list);	/* ".." */
		de = de->de_dir;
	}
	return (buf + len);
}

#undef DECONST
#define	DECONST(arg) ((void *)(((const char *)arg) - ((const char *)0)))

static int
devfs_lookup_sub(struct vop_lookup_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct componentname *cnp = ap->a_cnp;
	struct vnode **vpp = ap->a_vpp;
	struct vnode *dvp = ap->a_dvp;
	struct devfs_dirent **dde;
	struct devfs_dirent *de;
	struct devfs_dirent *dd;
	struct devfs_mount *dmp;
	struct cdev *cdev;
	struct proc *td = get_cn_proc(cnp);
	int error;
	int flags = cnp->cn_flags;
	int nameiop = cnp->cn_nameiop;
	u_int8_t specname[SPECNAMELEN + 1];
	u_int8_t *pname;
	u_int16_t name_len;

#if (__NetBSD_Version__ < 400000000)
	u_int8_t unlock = (!(flags & ISLASTCN)) || (!(flags & LOCKPARENT));

#endif

	pname = DECONST(cnp->cn_nameptr);
	dmp = VFSTODEVFS(dvp->v_mount);
	dd = dvp->v_data;
	vpp[0] = NULLVP;

	if ((flags & ISLASTCN) && (nameiop == RENAME)) {
		error = EOPNOTSUPP;
		goto done;
	}
	if (dvp->v_type != VDIR) {
		error = ENOTDIR;
		goto done;
	}
#if (__NetBSD_Version__ >= 500000000)
	if ((flags & ISDOTDOT) && (dvp->v_vflag & VV_ROOT))
#else
	if ((flags & ISDOTDOT) && (dvp->v_flag & VROOT))
#endif
	{
		error = EIO;
		goto done;
	}
	error = BSD_VOP_ACCESS(dvp, VEXEC, cnp->cn_cred, td, cnp->cn_lwp);
	if (error) {
		goto done;
	}
	if ((cnp->cn_namelen == 1) && (pname[0] == '.')) {
		if ((flags & ISLASTCN) && (nameiop != LOOKUP)) {
			error = EINVAL;
			goto done;
		}
		vpp[0] = dvp;
		VREF(dvp);
#if (__NetBSD_Version__ < 400000000)
		unlock = 0;
#endif
		goto done;
	}
	if (flags & ISDOTDOT) {
		if ((flags & ISLASTCN) && (nameiop != LOOKUP)) {
			error = EINVAL;
			goto done;
		}
		VOP_UNLOCK(dvp, 0);

		de = TAILQ_FIRST(&dd->de_dlist);	/* "." */
		de = TAILQ_NEXT(de, de_list);	/* ".." */
		de = de->de_dir;
		error = devfs_allocv(de, dvp->v_mount, vpp, td);

#if (__NetBSD_Version__ < 400000000)
		if ((error == 0) && unlock) {
			unlock = 0;
			cnp->cn_flags |= PDIRUNLOCK;
			goto done;
		}
#endif
		vn_lock(dvp, LK_EXCLUSIVE | LK_RETRY);
		goto done;
	}
	devfs_populate(dmp);
	dd = dvp->v_data;
	TAILQ_FOREACH(de, &dd->de_dlist, de_list) {
		if (cnp->cn_namelen != de->de_dirent->d_namlen) {
			continue;
		}
		if (bcmp(cnp->cn_nameptr, de->de_dirent->d_name,
		    de->de_dirent->d_namlen)) {
			continue;
		}
		if (de->de_flags & DE_WHITEOUT) {
			goto notfound;
		}
		goto found;
	}

	if (nameiop == DELETE) {
		goto notfound;
	}
	/*
	 * OK, we didn't have an entry for the name we were asked for
	 * so we try to see if anybody can create it on demand.
	 */
	pname = devfs_fqpn(specname, sizeof(specname), dvp, cnp);
	if (pname == NULL) {
		goto notfound;
	}
	cdev = NULL;
	name_len = strlen(pname);
#if (__NetBSD_Version__ < 400000000)
	EVENTHANDLER_INVOKE(dev_clone, td->p_ucred, pname,
	    name_len, &cdev);
#else
	EVENTHANDLER_INVOKE(dev_clone, NULL /* XXX not supported */ , pname,
	    name_len, &cdev);
#endif
	if (cdev == NULL) {
		goto notfound;
	}
	devfs_populate(dmp);

	dde = devfs_itode(dmp, cdev->si_inode);
	dev_rel(cdev);

	if ((dde == NULL) ||
	    (dde[0] == NULL)) {
		goto notfound;
	}
	if ((dde[0])->de_flags & DE_WHITEOUT) {
		goto notfound;
	}
	de = dde[0];
	goto found;

notfound:

	if (((nameiop == CREATE) || (nameiop == RENAME)) &&
#if (__NetBSD_Version__ < 400000000)
	    (flags & (LOCKPARENT | WANTPARENT)) &&
#endif
	    (flags & ISLASTCN)) {
		error = BSD_VOP_ACCESS(dvp, VWRITE,
		    cnp->cn_cred, td, cnp->cn_lwp);
		if (error) {
			goto done;
		}
		cnp->cn_flags |= SAVENAME;
		error = EJUSTRETURN;
		goto done;
	}
	error = ENOENT;
	goto done;

found:

	if ((cnp->cn_nameiop == DELETE) && (flags & ISLASTCN)) {
		error = BSD_VOP_ACCESS(dvp, VWRITE, cnp->cn_cred, td, cnp->cn_lwp);
		if (error) {
			goto done;
		}
		if (vpp[0] == dvp) {
			VREF(dvp);
#if (__NetBSD_Version__ < 400000000)
			unlock = 0;
#endif
			goto done;
		}
	}
	error = devfs_allocv(de, dvp->v_mount, vpp, td);

done:
#if (__NetBSD_Version__ < 400000000)
	if ((error == 0) && unlock) {
		VOP_UNLOCK(dvp, 0);
		cnp->cn_flags |= PDIRUNLOCK;
	}
#endif
	return error;
}

static int
devfs_lookup(struct vop_lookup_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_mount *dmp;
	int error;

	dmp = VFSTODEVFS(ap->a_dvp->v_mount);
	__lockmgr(&dmp->dm_lock, LK_EXCLUSIVE, NULL, curthread);
	error = devfs_lookup_sub(ap);
	__lockmgr(&dmp->dm_lock, LK_RELEASE, NULL, curthread);
	return error;
}

static int
devfs_mknod(struct vop_mknod_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *dvp;

	dvp = ap->a_dvp;
	VOP_UNLOCK(dvp, 0);

	return (EOPNOTSUPP);
}

static int
devfs_pathconf(struct vop_pathconf_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	switch (ap->a_name) {
	case _PC_NAME_MAX:
		ap->a_retval[0] = NAME_MAX;
		break;
	case _PC_PATH_MAX:
		ap->a_retval[0] = PATH_MAX;
		break;
#if 0
	case _PC_MAC_PRESENT:
		ap->a_retval[0] = 0;
		break;
#endif
	case _PC_LINK_MAX:
		ap->a_retval[0] = LINK_MAX;
		break;
	case _PC_PIPE_BUF:
		ap->a_retval[0] = PIPE_BUF;
		break;
	case _PC_CHOWN_RESTRICTED:
		ap->a_retval[0] = 1;
		break;
	case _PC_NO_TRUNC:
		ap->a_retval[0] = 1;
		break;
	case _PC_SYNC_IO:
		ap->a_retval[0] = 1;
		break;
	default:
		return (EINVAL);
	}
	return (0);
}

static int
devfs_readdir(struct vop_readdir_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_mount *dmp;
	struct devfs_dirent *dd;
	struct devfs_dirent *de;
	struct dirent *dps;
	struct dirent *dpe;
	struct dirent *dp;
	struct uio *uio;
	off_t *cookiebuf;
	off_t *cookiep;
	off_t oldoff;
	off_t off;
	int error;
	int ncookies = 0;

	if (ap->a_vp->v_type != VDIR) {
		return (ENOTDIR);
	}
	uio = ap->a_uio;
	if (uio->uio_offset < 0) {
		return EINVAL;
	}
	dmp = VFSTODEVFS(ap->a_vp->v_mount);
	__lockmgr(&dmp->dm_lock, LK_EXCLUSIVE, 0, curthread);
	devfs_populate(dmp);
	error = 0;
	de = ap->a_vp->v_data;
	off = 0;
	oldoff = uio->uio_offset;
	TAILQ_FOREACH(dd, &de->de_dlist, de_list) {
		if (dd->de_flags & DE_WHITEOUT) {
			continue;
		}
		if (dd->de_dirent->d_type == DT_DIR) {
			de = dd->de_dir;
		} else {
			de = dd;
		}
		dp = dd->de_dirent;
		if (dp->d_reclen > uio->uio_resid) {
			break;
		}
		dp->d_fileno = de->de_inode;
		if (off >= uio->uio_offset) {
			ncookies++;
			error = uiomove(dp, dp->d_reclen, uio);
			if (error) {
				break;
			}
		}
		off += dp->d_reclen;
	}

	if ((!error) && (ap->a_ncookies != NULL) && (ap->a_cookies != NULL)) {
		MALLOC(cookiebuf, off_t *, ncookies * sizeof(u_long),
		    M_DEVFS, M_WAITOK);

		cookiep = cookiebuf;
		dps = (struct dirent *)((char *)uio->uio_iov->iov_base -
		    (uio->uio_offset - oldoff));
		dpe = (struct dirent *)uio->uio_iov->iov_base;
		for (dp = dps;
		    dp < dpe;
		    dp = (struct dirent *)((caddr_t)dp + dp->d_reclen)) {
			oldoff += dp->d_reclen;
			*cookiep++ = (off_t)oldoff;
		}
		*ap->a_ncookies = ncookies;
		*ap->a_cookies = cookiebuf;
	}
	__lockmgr(&dmp->dm_lock, LK_RELEASE, 0, curthread);
	uio->uio_offset = off;
	return error;
}

static int
devfs_readlink(struct vop_readlink_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_dirent *de;
	int error;

	de = ap->a_vp->v_data;
	error = uiomove(de->de_symlink, strlen(de->de_symlink), ap->a_uio);
	return error;
}

static int
devfs_reclaim(struct vop_reclaim_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;
	struct devfs_dirent *de;
	struct cdev *dev;

#if (__NetBSD_Version__ < 500000000)
	cache_purge(vp);
#endif

	de = vp->v_data;
	vp->v_data = NULL;

	if (de == NULL) {
		goto done;
	}
	de->de_vnode = NULL;
	dev = de->de_dev;
	de->de_dev = NULL;

	if (dev == NULL)
		goto done;

	dev_lock();
	LIST_REMOVE(de, de_alias);
	dev_unlock();

	dev_rel(dev);
done:
	return (0);
}

static int
devfs_remove(struct vop_remove_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;
	struct devfs_dirent *dd;
	struct devfs_dirent *de;
	struct devfs_mount *dmp = VFSTODEVFS(vp->v_mount);

	__lockmgr(&dmp->dm_lock, LK_EXCLUSIVE, 0, curthread);
	dd = ap->a_dvp->v_data;
	de = vp->v_data;
	if (de->de_dirent->d_type == DT_LNK) {
		TAILQ_REMOVE(&dd->de_dlist, de, de_list);
		if (de->de_vnode) {
			de->de_vnode->v_data = NULL;
		}
		FREE(de, M_DEVFS);
	} else {
		de->de_flags |= DE_WHITEOUT;
	}
	__lockmgr(&dmp->dm_lock, LK_RELEASE, 0, curthread);

	VOP_UNLOCK(vp, 0);
	VOP_UNLOCK(ap->a_dvp, 0);
	return (0);
}

/*
 * Revoke is called on a tty when a terminal session ends.  The vnode
 * is orphaned by setting v_op to deadfs so we need to let go of it
 * as well so that we create a new one next time around.
 */
static int
devfs_revoke(struct vop_revoke_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;
	struct cdev *dev;
	struct devfs_dirent *de;

	__KASSERT((ap->a_flags & REVOKEALL) != 0,
	    ("%s: not REVOKEALL !", __FUNCTION__));

	de = vp->v_data;
	dev = de->de_dev;

	while (1) {
		dev_lock();
		de = LIST_FIRST(&dev->si_alist);
		dev_unlock();
		if (de == NULL) {
			break;
		}
		vgone(de->de_vnode);
	}
	return (0);
}

static int
devfs_symlink(struct vop_symlink_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_dirent *dd;
	struct devfs_dirent *de;
	struct devfs_mount *dmp;
	struct proc *td;
	int i, error;

	td = get_cn_proc(ap->a_cnp);
	__KASSERT(td == curthread, ("%s: td != curthread", __FUNCTION__));
	error = suser(td);
	if (error) {
		goto done;
	}
	dmp = VFSTODEVFS(ap->a_dvp->v_mount);
	dd = ap->a_dvp->v_data;
	de = devfs_newdirent(DECONST(ap->a_cnp->cn_nameptr),
	    ap->a_cnp->cn_namelen);
	de->de_uid = 0;
	de->de_gid = 0;
	de->de_mode = 0755;
	de->de_inode = dmp->dm_inode++;
	de->de_dirent->d_type = DT_LNK;
	i = strlen(ap->a_target) + 1;
	MALLOC(de->de_symlink, char *, i, M_DEVFS, M_WAITOK);
	bcopy(ap->a_target, de->de_symlink, i);
	__lockmgr(&dmp->dm_lock, LK_EXCLUSIVE, 0, td);
	TAILQ_INSERT_TAIL(&dd->de_dlist, de, de_list);
	error = devfs_allocv(de, ap->a_dvp->v_mount, ap->a_vpp, td);
	__lockmgr(&dmp->dm_lock, LK_RELEASE, 0, td);

done:
	VOP_UNLOCK(ap->a_dvp, 0);

	return (error);
}

/*
 * FreeBSD is moving from IO_XXX to O_XXX for device
 * drivers. To not break old drivers, the following
 * assertions must be met:
 */

#if ((O_NONBLOCK != IO_NDELAY) || !defined(O_NONBLOCK) || !defined(IO_NDELAY))
#error "Assertion not met!"
#endif

#if ((O_FSYNC != IO_SYNC) || !defined(O_FSYNC) || !defined(IO_SYNC))
#error "Assertion not met!"
#endif

static int
devfs_fsync(struct vop_fsync_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (0);
}

static int
devfs_print(struct vop_print_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct devfs_dirent *de = ap->a_vp->v_data;
	struct cdev *dev = de->de_dev;

	printf("\tdev %s\n", devtoname(dev));
	return (0);
}

static int
devfs_advlock(struct vop_advlock_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return ((ap->a_flags & F_FLOCK) ? EOPNOTSUPP : EINVAL);
}

extern int genfs_nolock(void *);
extern int genfs_nounlock(void *);
extern int genfs_noislocked(void *);

static int
devfs_lock(struct vop_lock_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (genfs_nolock(ap));
}

static int
devfs_unlock(struct vop_unlock_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (genfs_nounlock(ap));
}

static int
devfs_islocked(struct vop_islocked_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (genfs_noislocked(ap));
}

static int
devfs_inactive(struct vop_inactive_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	struct vnode *vp = ap->a_vp;

#if (__NetBSD_Version__ >= 500000000)
	/* XXX node is gone */
	*ap->a_recycle = 1;
#endif

	VOP_UNLOCK(vp, 0);

#if (__NetBSD_Version__ < 500000000)
	vgone(vp);
#endif
	return (0);
}

static int
devfs_create_vnop(struct vop_create_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	VOP_UNLOCK(ap->a_dvp, 0);
	return (EOPNOTSUPP);
}

#if (__NetBSD_Version__ < 500000000)
static int
devfs_lease(struct vop_lease_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

#endif

static int
devfs_fcntl(struct vop_fcntl_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

static int
devfs_mmap(struct vop_mmap_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

static int
devfs_seek(struct vop_seek_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (ap->a_newoff < 0) ? EINVAL : 0;
}

static int
devfs_link(struct vop_link_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	VOP_UNLOCK(ap->a_dvp, 0);
	return (EOPNOTSUPP);
}

static int
devfs_rename(struct vop_rename_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	VOP_UNLOCK(ap->a_tdvp, 0);
	if (ap->a_tvp) {
		VOP_UNLOCK(ap->a_tvp, 0);
	}
	return (EOPNOTSUPP);
}

static int
devfs_mkdir(struct vop_mkdir_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	VOP_UNLOCK(ap->a_dvp, 0);
	return (EOPNOTSUPP);
}

static int
devfs_rmdir(struct vop_rmdir_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	VOP_UNLOCK(ap->a_dvp, 0);
	VOP_UNLOCK(ap->a_vp, 0);
	return (EOPNOTSUPP);
}

static int
devfs_abortop(struct vop_abortop_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (0);
}

static int
devfs_bmap(struct vop_bmap_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

static int
devfs_strategy(struct vop_strategy_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

#if (__NetBSD_Version__ < 400000000)
static int
devfs_blkatoff(struct vop_blkatoff_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

static int
devfs_valloc(struct vop_valloc_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (0);
}

static int
devfs_vfree(struct vop_vfree_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

static int
devfs_truncate(struct vop_truncate_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

static int
devfs_update(struct vop_update_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (0);
}

#endif

static int
devfs_bwrite(struct vop_bwrite_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (EOPNOTSUPP);
}

extern int genfs_getpages(void *);
extern int genfs_null_putpages(void *);

static int
devfs_getpages(struct vop_getpages_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (genfs_getpages(ap));
}

static int
devfs_putpages(struct vop_putpages_args *ap)
{
#ifdef DEVFS_DEBUG
	printf("%s\n", __FUNCTION__);
#endif
	return (genfs_null_putpages(ap));
}

/*
 * devfs export table
 */

typedef int (*const vop_t)(void *);

static const struct vnodeopv_entry_desc
	devfs_vnodeop_entries[] =
{
	{&vop_default_desc, (vop_t)&vn_default_error},

	/*
         * basic vnode operations:
         */

	{&vop_open_desc, (vop_t)&devfs_open},
	{&vop_read_desc, (vop_t)&devfs_read},
	{&vop_write_desc, (vop_t)&devfs_write},
	{&vop_ioctl_desc, (vop_t)&devfs_ioctl},
	{&vop_poll_desc, (vop_t)&devfs_poll},
	{&vop_kqfilter_desc, (vop_t)&devfs_kqfilter},
	{&vop_close_desc, (vop_t)&devfs_close},

	/*
         * extended vnode operations:
         */

	{&vop_access_desc, (vop_t)&devfs_access},
	{&vop_getattr_desc, (vop_t)&devfs_getattr},
	{&vop_setattr_desc, (vop_t)&devfs_setattr},
	{&vop_lookup_desc, (vop_t)&devfs_lookup},
	{&vop_mknod_desc, (vop_t)&devfs_mknod},
	{&vop_pathconf_desc, (vop_t)&devfs_pathconf},
	{&vop_readdir_desc, (vop_t)&devfs_readdir},
	{&vop_readlink_desc, (vop_t)&devfs_readlink},
	{&vop_reclaim_desc, (vop_t)&devfs_reclaim},
	{&vop_remove_desc, (vop_t)&devfs_remove},
	{&vop_revoke_desc, (vop_t)&devfs_revoke},
	{&vop_symlink_desc, (vop_t)&devfs_symlink},
	{&vop_fsync_desc, (vop_t)&devfs_fsync},
	{&vop_print_desc, (vop_t)&devfs_print},
	{&vop_advlock_desc, (vop_t)&devfs_advlock},

	{&vop_lock_desc, (vop_t)&devfs_lock},
	{&vop_unlock_desc, (vop_t)&devfs_unlock},
	{&vop_islocked_desc, (vop_t)&devfs_islocked},
	{&vop_inactive_desc, (vop_t)&devfs_inactive},

	/*
         * various vnode operations:
         */

	{&vop_create_desc, (vop_t)&devfs_create_vnop},
#if (__NetBSD_Version__ < 500000000)
	{&vop_lease_desc, (vop_t)&devfs_lease},
#endif
	{&vop_fcntl_desc, (vop_t)&devfs_fcntl},
	{&vop_mmap_desc, (vop_t)&devfs_mmap},
	{&vop_seek_desc, (vop_t)&devfs_seek},
	{&vop_link_desc, (vop_t)&devfs_link},
	{&vop_rename_desc, (vop_t)&devfs_rename},
	{&vop_mkdir_desc, (vop_t)&devfs_mkdir},
	{&vop_rmdir_desc, (vop_t)&devfs_rmdir},
	{&vop_abortop_desc, (vop_t)&devfs_abortop},
	{&vop_bmap_desc, (vop_t)&devfs_bmap},
	{&vop_strategy_desc, (vop_t)&devfs_strategy},
#if (__NetBSD_Version__ < 400000000)
	{&vop_blkatoff_desc, (vop_t)&devfs_blkatoff},
	{&vop_valloc_desc, (vop_t)&devfs_valloc},
	{&vop_vfree_desc, (vop_t)&devfs_vfree},
	{&vop_truncate_desc, (vop_t)&devfs_truncate},
	{&vop_update_desc, (vop_t)&devfs_update},
#endif
	{&vop_bwrite_desc, (vop_t)&devfs_bwrite},
	{&vop_getpages_desc, (vop_t)&devfs_getpages},
	{&vop_putpages_desc, (vop_t)&devfs_putpages},

	{NULL, NULL}
};

const struct vnodeopv_desc devfs_vnodeop_opv_desc = {
	&devfs_vnodeop_p,
	&devfs_vnodeop_entries[0],
};

/*
 * Not implemented:
 *
 * vop_stat_desc
 * vop_mmap_desc
 */