freebsd-dev/sys/kern/vfs_vnops.c

536 lines
12 KiB
C
Raw Normal View History

1994-05-24 10:09:53 +00:00
/*
* Copyright (c) 1982, 1986, 1989, 1993
* The Regents of the University of California. All rights reserved.
* (c) UNIX System Laboratories, Inc.
* All or some portions of this file are derived from material licensed
* to the University of California by American Telephone and Telegraph
* Co. or Unix System Laboratories, Inc. and are reproduced herein with
* the permission of UNIX System Laboratories, Inc.
*
* 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.
*
* @(#)vfs_vnops.c 8.2 (Berkeley) 1/21/94
* $Id: vfs_vnops.c,v 1.38 1997/09/14 02:51:16 peter Exp $
1994-05-24 10:09:53 +00:00
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/fcntl.h>
1994-05-24 10:09:53 +00:00
#include <sys/file.h>
#include <sys/stat.h>
#include <sys/proc.h>
#include <sys/mount.h>
#include <sys/namei.h>
#include <sys/vnode.h>
#include <sys/filio.h>
#include <sys/ttycom.h>
1994-05-24 10:09:53 +00:00
static int vn_closefile __P((struct file *fp, struct proc *p));
static int vn_ioctl __P((struct file *fp, int com, caddr_t data,
struct proc *p));
static int vn_read __P((struct file *fp, struct uio *uio,
struct ucred *cred));
1997-09-14 02:51:16 +00:00
static int vn_poll __P((struct file *fp, int events, struct ucred *cred,
struct proc *p));
static int vn_write __P((struct file *fp, struct uio *uio,
struct ucred *cred));
1994-05-24 10:09:53 +00:00
struct fileops vnops =
1997-09-14 02:51:16 +00:00
{ vn_read, vn_write, vn_ioctl, vn_poll, vn_closefile };
1994-05-24 10:09:53 +00:00
/*
* Common code for vnode open operations.
* Check permissions, and call the VOP_OPEN or VOP_CREATE routine.
*/
int
1994-05-24 10:09:53 +00:00
vn_open(ndp, fmode, cmode)
register struct nameidata *ndp;
int fmode, cmode;
{
register struct vnode *vp;
register struct proc *p = ndp->ni_cnd.cn_proc;
register struct ucred *cred = p->p_ucred;
struct vattr vat;
struct vattr *vap = &vat;
int error;
if (fmode & O_CREAT) {
ndp->ni_cnd.cn_nameiop = CREATE;
ndp->ni_cnd.cn_flags = LOCKPARENT | LOCKLEAF;
if ((fmode & O_EXCL) == 0)
ndp->ni_cnd.cn_flags |= FOLLOW;
error = namei(ndp);
if (error)
1994-05-24 10:09:53 +00:00
return (error);
if (ndp->ni_vp == NULL) {
VATTR_NULL(vap);
vap->va_type = VREG;
vap->va_mode = cmode;
if (fmode & O_EXCL)
vap->va_vaflags |= VA_EXCLUSIVE;
VOP_LEASE(ndp->ni_dvp, p, cred, LEASE_WRITE);
if (error = VOP_CREATE(ndp->ni_dvp, &ndp->ni_vp,
&ndp->ni_cnd, vap))
1994-05-24 10:09:53 +00:00
return (error);
ASSERT_VOP_UNLOCKED(ndp->ni_dvp, "create");
ASSERT_VOP_LOCKED(ndp->ni_vp, "create");
1994-05-24 10:09:53 +00:00
fmode &= ~O_TRUNC;
vp = ndp->ni_vp;
} else {
VOP_ABORTOP(ndp->ni_dvp, &ndp->ni_cnd);
if (ndp->ni_dvp == ndp->ni_vp)
vrele(ndp->ni_dvp);
else
vput(ndp->ni_dvp);
ndp->ni_dvp = NULL;
vp = ndp->ni_vp;
if (fmode & O_EXCL) {
error = EEXIST;
goto bad;
}
fmode &= ~O_CREAT;
}
} else {
ndp->ni_cnd.cn_nameiop = LOOKUP;
ndp->ni_cnd.cn_flags = FOLLOW | LOCKLEAF;
error = namei(ndp);
if (error)
1994-05-24 10:09:53 +00:00
return (error);
vp = ndp->ni_vp;
}
if (vp->v_type == VSOCK) {
error = EOPNOTSUPP;
goto bad;
}
if ((fmode & O_CREAT) == 0) {
if (fmode & FREAD) {
error = VOP_ACCESS(vp, VREAD, cred, p);
if (error)
1994-05-24 10:09:53 +00:00
goto bad;
}
if (fmode & (FWRITE | O_TRUNC)) {
if (vp->v_type == VDIR) {
error = EISDIR;
goto bad;
}
error = vn_writechk(vp);
if (error)
goto bad;
error = VOP_ACCESS(vp, VWRITE, cred, p);
if (error)
1994-05-24 10:09:53 +00:00
goto bad;
}
}
if (fmode & O_TRUNC) {
VOP_UNLOCK(vp, 0, p); /* XXX */
VOP_LEASE(vp, p, cred, LEASE_WRITE);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p); /* XXX */
1994-05-24 10:09:53 +00:00
VATTR_NULL(vap);
vap->va_size = 0;
error = VOP_SETATTR(vp, vap, cred, p);
if (error)
1994-05-24 10:09:53 +00:00
goto bad;
}
error = VOP_OPEN(vp, fmode, cred, p);
if (error)
1994-05-24 10:09:53 +00:00
goto bad;
/*
* Make sure that a VM object is created for VMIO support.
*/
if (vp->v_type == VREG) {
if ((error = vfs_object_create(vp, p, cred, 1)) != 0)
goto bad;
}
NOTE: libkvm, w, ps, 'top', and any other utility which depends on struct proc or any VM system structure will have to be rebuilt!!! Much needed overhaul of the VM system. Included in this first round of changes: 1) Improved pager interfaces: init, alloc, dealloc, getpages, putpages, haspage, and sync operations are supported. The haspage interface now provides information about clusterability. All pager routines now take struct vm_object's instead of "pagers". 2) Improved data structures. In the previous paradigm, there is constant confusion caused by pagers being both a data structure ("allocate a pager") and a collection of routines. The idea of a pager structure has escentially been eliminated. Objects now have types, and this type is used to index the appropriate pager. In most cases, items in the pager structure were duplicated in the object data structure and thus were unnecessary. In the few cases that remained, a un_pager structure union was created in the object to contain these items. 3) Because of the cleanup of #1 & #2, a lot of unnecessary layering can now be removed. For instance, vm_object_enter(), vm_object_lookup(), vm_object_remove(), and the associated object hash list were some of the things that were removed. 4) simple_lock's removed. Discussion with several people reveals that the SMP locking primitives used in the VM system aren't likely the mechanism that we'll be adopting. Even if it were, the locking that was in the code was very inadequate and would have to be mostly re-done anyway. The locking in a uni-processor kernel was a no-op but went a long way toward making the code difficult to read and debug. 5) Places that attempted to kludge-up the fact that we don't have kernel thread support have been fixed to reflect the reality that we are really dealing with processes, not threads. The VM system didn't have complete thread support, so the comments and mis-named routines were just wrong. We now use tsleep and wakeup directly in the lock routines, for instance. 6) Where appropriate, the pagers have been improved, especially in the pager_alloc routines. Most of the pager_allocs have been rewritten and are now faster and easier to maintain. 7) The pagedaemon pageout clustering algorithm has been rewritten and now tries harder to output an even number of pages before and after the requested page. This is sort of the reverse of the ideal pagein algorithm and should provide better overall performance. 8) Unnecessary (incorrect) casts to caddr_t in calls to tsleep & wakeup have been removed. Some other unnecessary casts have also been removed. 9) Some almost useless debugging code removed. 10) Terminology of shadow objects vs. backing objects straightened out. The fact that the vm_object data structure escentially had this backwards really confused things. The use of "shadow" and "backing object" throughout the code is now internally consistent and correct in the Mach terminology. 11) Several minor bug fixes, including one in the vm daemon that caused 0 RSS objects to not get purged as intended. 12) A "default pager" has now been created which cleans up the transition of objects to the "swap" type. The previous checks throughout the code for swp->pg_data != NULL were really ugly. This change also provides the rudiments for future backing of "anonymous" memory by something other than the swap pager (via the vnode pager, for example), and it allows the decision about which of these pagers to use to be made dynamically (although will need some additional decision code to do this, of course). 13) (dyson) MAP_COPY has been deprecated and the corresponding "copy object" code has been removed. MAP_COPY was undocumented and non- standard. It was furthermore broken in several ways which caused its behavior to degrade to MAP_PRIVATE. Binaries that use MAP_COPY will continue to work correctly, but via the slightly different semantics of MAP_PRIVATE. 14) (dyson) Sharing maps have been removed. It's marginal usefulness in a threads design can be worked around in other ways. Both #12 and #13 were done to simplify the code and improve readability and maintain- ability. (As were most all of these changes) TODO: 1) Rewrite most of the vnode pager to use VOP_GETPAGES/PUTPAGES. Doing this will reduce the vnode pager to a mere fraction of its current size. 2) Rewrite vm_fault and the swap/vnode pagers to use the clustering information provided by the new haspage pager interface. This will substantially reduce the overhead by eliminating a large number of VOP_BMAP() calls. The VOP_BMAP() filesystem interface should be improved to provide both a "behind" and "ahead" indication of contiguousness. 3) Implement the extended features of pager_haspage in swap_pager_haspage(). It currently just says 0 pages ahead/behind. 4) Re-implement the swap device (swstrategy) in a more elegant way, perhaps via a much more general mechanism that could also be used for disk striping of regular filesystems. 5) Do something to improve the architecture of vm_object_collapse(). The fact that it makes calls into the swap pager and knows too much about how the swap pager operates really bothers me. It also doesn't allow for collapsing of non-swap pager objects ("unnamed" objects backed by other pagers).
1995-07-13 08:48:48 +00:00
if (fmode & FWRITE)
vp->v_writecount++;
1994-05-24 10:09:53 +00:00
return (0);
bad:
vput(vp);
return (error);
}
/*
* Check for write permissions on the specified vnode.
* Prototype text segments cannot be written.
1994-05-24 10:09:53 +00:00
*/
int
1994-05-24 10:09:53 +00:00
vn_writechk(vp)
register struct vnode *vp;
{
/*
* If there's shared text associated with
* the vnode, try to free it up once. If
* we fail, we can't allow writing.
*/
if (vp->v_flag & VTEXT)
1994-05-24 10:09:53 +00:00
return (ETXTBSY);
return (0);
}
/*
* Vnode close call
*/
int
1994-05-24 10:09:53 +00:00
vn_close(vp, flags, cred, p)
register struct vnode *vp;
int flags;
struct ucred *cred;
struct proc *p;
{
int error;
if (flags & FWRITE)
vp->v_writecount--;
error = VOP_CLOSE(vp, flags, cred, p);
vrele(vp);
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* Package up an I/O request on a vnode into a uio and do it.
*/
int
1994-05-24 10:09:53 +00:00
vn_rdwr(rw, vp, base, len, offset, segflg, ioflg, cred, aresid, p)
enum uio_rw rw;
struct vnode *vp;
caddr_t base;
int len;
off_t offset;
enum uio_seg segflg;
int ioflg;
struct ucred *cred;
int *aresid;
struct proc *p;
{
struct uio auio;
struct iovec aiov;
int error;
if ((ioflg & IO_NODELOCKED) == 0)
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p);
1994-05-24 10:09:53 +00:00
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
aiov.iov_base = base;
aiov.iov_len = len;
auio.uio_resid = len;
auio.uio_offset = offset;
auio.uio_segflg = segflg;
auio.uio_rw = rw;
auio.uio_procp = p;
if (rw == UIO_READ) {
error = VOP_READ(vp, &auio, ioflg, cred);
} else {
error = VOP_WRITE(vp, &auio, ioflg, cred);
}
if (aresid)
*aresid = auio.uio_resid;
else
if (auio.uio_resid && error == 0)
error = EIO;
if ((ioflg & IO_NODELOCKED) == 0)
VOP_UNLOCK(vp, 0, p);
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* File table vnode read routine.
*/
static int
1994-05-24 10:09:53 +00:00
vn_read(fp, uio, cred)
struct file *fp;
struct uio *uio;
struct ucred *cred;
{
struct vnode *vp = (struct vnode *)fp->f_data;
struct proc *p = uio->uio_procp;
1994-05-24 10:09:53 +00:00
int count, error;
int flag, seq;
1994-05-24 10:09:53 +00:00
VOP_LEASE(vp, p, cred, LEASE_READ);
vn_lock(vp, LK_SHARED | LK_NOPAUSE | LK_RETRY, p);
1994-05-24 10:09:53 +00:00
uio->uio_offset = fp->f_offset;
count = uio->uio_resid;
flag = 0;
if (fp->f_flag & FNONBLOCK)
flag |= IO_NDELAY;
/*
* Sequential read heuristic.
* If we have been doing sequential input,
* a rewind operation doesn't turn off
* sequential input mode.
*/
if (((fp->f_offset == 0) && (fp->f_seqcount > 0)) ||
(fp->f_offset == fp->f_nextread)) {
int tmpseq = fp->f_seqcount;
/*
* XXX we assume that the filesystem block size is
* the default. Not true, but still gives us a pretty
* good indicator of how sequential the read operations
* are.
*/
tmpseq += ((count + BKVASIZE - 1) / BKVASIZE);
if (tmpseq >= CHAR_MAX)
tmpseq = CHAR_MAX;
fp->f_seqcount = tmpseq;
flag |= (fp->f_seqcount << 16);
} else {
if (fp->f_seqcount > 1)
fp->f_seqcount = 1;
else
fp->f_seqcount = 0;
}
error = VOP_READ(vp, uio, flag, cred);
1994-05-24 10:09:53 +00:00
fp->f_offset += count - uio->uio_resid;
fp->f_nextread = fp->f_offset;
VOP_UNLOCK(vp, 0, p);
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* File table vnode write routine.
*/
static int
1994-05-24 10:09:53 +00:00
vn_write(fp, uio, cred)
struct file *fp;
struct uio *uio;
struct ucred *cred;
{
struct vnode *vp = (struct vnode *)fp->f_data;
struct proc *p = uio->uio_procp;
int count, error, ioflag = IO_UNIT;
1994-05-24 10:09:53 +00:00
if (vp->v_type == VREG && (fp->f_flag & O_APPEND))
ioflag |= IO_APPEND;
if (fp->f_flag & FNONBLOCK)
ioflag |= IO_NDELAY;
if ((fp->f_flag & O_FSYNC) ||
(vp->v_mount && (vp->v_mount->mnt_flag & MNT_SYNCHRONOUS)))
ioflag |= IO_SYNC;
VOP_LEASE(vp, p, cred, LEASE_WRITE);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p);
1994-05-24 10:09:53 +00:00
uio->uio_offset = fp->f_offset;
count = uio->uio_resid;
error = VOP_WRITE(vp, uio, ioflag, cred);
if (ioflag & IO_APPEND)
fp->f_offset = uio->uio_offset;
else
fp->f_offset += count - uio->uio_resid;
VOP_UNLOCK(vp, 0, p);
1994-05-24 10:09:53 +00:00
return (error);
}
/*
* File table vnode stat routine.
*/
int
1994-05-24 10:09:53 +00:00
vn_stat(vp, sb, p)
struct vnode *vp;
register struct stat *sb;
struct proc *p;
{
struct vattr vattr;
register struct vattr *vap;
int error;
u_short mode;
vap = &vattr;
error = VOP_GETATTR(vp, vap, p->p_ucred, p);
if (error)
return (error);
/*
* Copy from vattr table
*/
sb->st_dev = vap->va_fsid;
sb->st_ino = vap->va_fileid;
mode = vap->va_mode;
switch (vp->v_type) {
case VREG:
mode |= S_IFREG;
break;
case VDIR:
mode |= S_IFDIR;
break;
case VBLK:
mode |= S_IFBLK;
break;
case VCHR:
mode |= S_IFCHR;
break;
case VLNK:
mode |= S_IFLNK;
break;
case VSOCK:
mode |= S_IFSOCK;
break;
case VFIFO:
mode |= S_IFIFO;
break;
default:
return (EBADF);
};
sb->st_mode = mode;
sb->st_nlink = vap->va_nlink;
sb->st_uid = vap->va_uid;
sb->st_gid = vap->va_gid;
sb->st_rdev = vap->va_rdev;
sb->st_size = vap->va_size;
sb->st_atimespec = vap->va_atime;
sb->st_mtimespec = vap->va_mtime;
1994-05-24 10:09:53 +00:00
sb->st_ctimespec = vap->va_ctime;
sb->st_blksize = vap->va_blocksize;
sb->st_flags = vap->va_flags;
if (p->p_ucred->cr_uid != 0)
sb->st_gen = 0;
else
sb->st_gen = vap->va_gen;
#if (S_BLKSIZE == 512)
/* Optimize this case */
sb->st_blocks = vap->va_bytes >> 9;
#else
1994-05-24 10:09:53 +00:00
sb->st_blocks = vap->va_bytes / S_BLKSIZE;
#endif
1994-05-24 10:09:53 +00:00
return (0);
}
/*
* File table vnode ioctl routine.
*/
static int
1994-05-24 10:09:53 +00:00
vn_ioctl(fp, com, data, p)
struct file *fp;
int com;
caddr_t data;
struct proc *p;
{
register struct vnode *vp = ((struct vnode *)fp->f_data);
struct vattr vattr;
int error;
switch (vp->v_type) {
case VREG:
case VDIR:
if (com == FIONREAD) {
error = VOP_GETATTR(vp, &vattr, p->p_ucred, p);
if (error)
1994-05-24 10:09:53 +00:00
return (error);
*(int *)data = vattr.va_size - fp->f_offset;
return (0);
}
if (com == FIONBIO || com == FIOASYNC) /* XXX */
return (0); /* XXX */
/* fall into ... */
default:
return (ENOTTY);
case VFIFO:
case VCHR:
case VBLK:
error = VOP_IOCTL(vp, com, data, fp->f_flag, p->p_ucred, p);
if (error == 0 && com == TIOCSCTTY) {
/* Do nothing if reassigning same control tty */
if (p->p_session->s_ttyvp == vp)
return (0);
/* Get rid of reference to old control tty */
if (p->p_session->s_ttyvp)
vrele(p->p_session->s_ttyvp);
1994-05-24 10:09:53 +00:00
p->p_session->s_ttyvp = vp;
VREF(vp);
}
return (error);
}
}
/*
1997-09-14 02:51:16 +00:00
* File table vnode poll routine.
1994-05-24 10:09:53 +00:00
*/
static int
1997-09-14 02:51:16 +00:00
vn_poll(fp, events, cred, p)
1994-05-24 10:09:53 +00:00
struct file *fp;
1997-09-14 02:51:16 +00:00
int events;
struct ucred *cred;
1994-05-24 10:09:53 +00:00
struct proc *p;
{
1997-09-14 02:51:16 +00:00
return (VOP_POLL(((struct vnode *)fp->f_data), events, cred, p));
1994-05-24 10:09:53 +00:00
}
/*
* File table vnode close routine.
*/
static int
1994-05-24 10:09:53 +00:00
vn_closefile(fp, p)
struct file *fp;
struct proc *p;
{
return (vn_close(((struct vnode *)fp->f_data), fp->f_flag,
fp->f_cred, p));
}
/*
* Check that the vnode is still valid, and if so
* acquire requested lock.
*/
int
vn_lock(vp, flags, p)
struct vnode *vp;
int flags;
struct proc *p;
{
int error;
do {
if ((flags & LK_INTERLOCK) == 0) {
simple_lock(&vp->v_interlock);
}
if (vp->v_flag & VXLOCK) {
vp->v_flag |= VXWANT;
simple_unlock(&vp->v_interlock);
tsleep((caddr_t)vp, PINOD, "vn_lock", 0);
error = ENOENT;
} else {
error = VOP_LOCK(vp, flags | LK_NOPAUSE | LK_INTERLOCK, p);
if (error == 0)
return (error);
}
flags &= ~LK_INTERLOCK;
} while (flags & LK_RETRY);
return (error);
}